diff --git a/Documentation/ABI/testing/sysfs-driver-intel-i915-hwmon b/Documentation/ABI/testing/sysfs-driver-intel-i915-hwmon index be4141a7522f..a885e5316d02 100644 --- a/Documentation/ABI/testing/sysfs-driver-intel-i915-hwmon +++ b/Documentation/ABI/testing/sysfs-driver-intel-i915-hwmon @@ -83,3 +83,11 @@ Contact: intel-gfx@lists.freedesktop.org Description: RO. Fan speed of device in RPM. Only supported for particular Intel i915 graphics platforms. + +What: /sys/bus/pci/drivers/i915/.../hwmon/hwmon/temp1_input +Date: November 2024 +KernelVersion: 6.12 +Contact: intel-gfx@lists.freedesktop.org +Description: RO. GPU package temperature in millidegree Celsius. + + Only supported for particular Intel i915 graphics platforms. diff --git a/Documentation/ABI/testing/sysfs-driver-panthor-profiling b/Documentation/ABI/testing/sysfs-driver-panthor-profiling new file mode 100644 index 000000000000..af05fccedc15 --- /dev/null +++ b/Documentation/ABI/testing/sysfs-driver-panthor-profiling @@ -0,0 +1,10 @@ +What: /sys/bus/platform/drivers/panthor/.../profiling +Date: September 2024 +KernelVersion: 6.11.0 +Contact: Adrian Larumbe +Description: + Bitmask to enable drm fdinfo's job profiling measurements. + Valid values are: + 0: Don't enable fdinfo job profiling sources. + 1: Enable GPU cycle measurements for running jobs. + 2: Enable GPU timestamp sampling for running jobs. diff --git a/Documentation/accel/qaic/aic080.rst b/Documentation/accel/qaic/aic080.rst new file mode 100644 index 000000000000..d563771ea6ce --- /dev/null +++ b/Documentation/accel/qaic/aic080.rst @@ -0,0 +1,14 @@ +.. SPDX-License-Identifier: GPL-2.0-only + +=============================== + Qualcomm Cloud AI 80 (AIC080) +=============================== + +Overview +======== + +The Qualcomm Cloud AI 80/AIC080 family of products are a derivative of AIC100. +The number of NSPs and clock rates are reduced to fit within resource +constrained solutions. The PCIe Product ID is 0xa080. + +As a derivative product, all AIC100 documentation applies. diff --git a/Documentation/accel/qaic/aic100.rst b/Documentation/accel/qaic/aic100.rst index 590dae77ea12..273da6192fb3 100644 --- a/Documentation/accel/qaic/aic100.rst +++ b/Documentation/accel/qaic/aic100.rst @@ -229,6 +229,8 @@ of the defined channels, and their uses. | _PERIODIC | | | timestamps in the device side logs with| | | | | the host time source. | +----------------+---------+----------+----------------------------------------+ +| IPCR | 24 & 25 | AMSS | AF_QIPCRTR clients and servers. | ++----------------+---------+----------+----------------------------------------+ DMA Bridge ========== diff --git a/Documentation/accel/qaic/index.rst b/Documentation/accel/qaic/index.rst index ad19b88d1a66..967b9dd8bace 100644 --- a/Documentation/accel/qaic/index.rst +++ b/Documentation/accel/qaic/index.rst @@ -10,4 +10,5 @@ accelerator cards. .. toctree:: qaic + aic080 aic100 diff --git a/Documentation/devicetree/bindings/display/bridge/ite,it6263.yaml b/Documentation/devicetree/bindings/display/bridge/ite,it6263.yaml new file mode 100644 index 000000000000..0a10e10d80ff --- /dev/null +++ b/Documentation/devicetree/bindings/display/bridge/ite,it6263.yaml @@ -0,0 +1,250 @@ +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/display/bridge/ite,it6263.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: ITE IT6263 LVDS to HDMI converter + +maintainers: + - Liu Ying + +description: | + The IT6263 is a high-performance single-chip De-SSC(De-Spread Spectrum) LVDS + to HDMI converter. Combined with LVDS receiver and HDMI 1.4a transmitter, + the IT6263 supports LVDS input and HDMI 1.4 output by conversion function. + The built-in LVDS receiver can support single-link and dual-link LVDS inputs, + and the built-in HDMI transmitter is fully compliant with HDMI 1.4a/3D, HDCP + 1.2 and backward compatible with DVI 1.0 specification. + + The IT6263 also encodes and transmits up to 8 channels of I2S digital audio, + with sampling rate up to 192KHz and sample size up to 24 bits. In addition, + an S/PDIF input port takes in compressed audio of up to 192KHz frame rate. + + The newly supported High-Bit Rate(HBR) audio by HDMI specifications v1.3 is + provided by the IT6263 in two interfaces: the four I2S input ports or the + S/PDIF input port. With both interfaces the highest possible HBR frame rate + is supported at up to 768KHz. + +allOf: + - $ref: /schemas/display/lvds-dual-ports.yaml# + +properties: + compatible: + const: ite,it6263 + + reg: + maxItems: 1 + + clocks: + maxItems: 1 + description: audio master clock + + clock-names: + const: mclk + + data-mapping: + enum: + - jeida-18 + - jeida-24 + - jeida-30 + - vesa-24 + - vesa-30 + + reset-gpios: + maxItems: 1 + + ivdd-supply: + description: 1.8V digital logic power + + ovdd-supply: + description: 3.3V I/O pin power + + txavcc18-supply: + description: 1.8V HDMI analog frontend power + + txavcc33-supply: + description: 3.3V HDMI analog frontend power + + pvcc1-supply: + description: 1.8V HDMI frontend core PLL power + + pvcc2-supply: + description: 1.8V HDMI frontend filter PLL power + + avcc-supply: + description: 3.3V LVDS frontend power + + anvdd-supply: + description: 1.8V LVDS frontend analog power + + apvdd-supply: + description: 1.8V LVDS frontend PLL power + + "#sound-dai-cells": + const: 0 + + ite,i2s-audio-fifo-sources: + $ref: /schemas/types.yaml#/definitions/uint32-array + minItems: 1 + maxItems: 4 + items: + enum: [0, 1, 2, 3] + description: + Each array element indicates the pin number of an I2S serial data input + line which is connected to an audio FIFO, from audio FIFO0 to FIFO3. + + ite,rl-channel-swap-audio-sources: + $ref: /schemas/types.yaml#/definitions/uint32-array + minItems: 1 + maxItems: 4 + uniqueItems: true + items: + enum: [0, 1, 2, 3] + description: + Each array element indicates an audio source whose right channel and left + channel are swapped by this converter. For I2S, the element is the pin + number of an I2S serial data input line. For S/PDIF, the element is always + 0. + + ports: + $ref: /schemas/graph.yaml#/properties/ports + + properties: + port@0: true + + port@1: + oneOf: + - required: [dual-lvds-odd-pixels] + - required: [dual-lvds-even-pixels] + + port@2: + $ref: /schemas/graph.yaml#/properties/port + description: video port for the HDMI output + + port@3: + $ref: /schemas/graph.yaml#/properties/port + description: sound input port + + required: + - port@0 + - port@2 + +required: + - compatible + - reg + - data-mapping + - ivdd-supply + - ovdd-supply + - txavcc18-supply + - txavcc33-supply + - pvcc1-supply + - pvcc2-supply + - avcc-supply + - anvdd-supply + - apvdd-supply + +unevaluatedProperties: false + +examples: + - | + /* single-link LVDS input */ + #include + + i2c { + #address-cells = <1>; + #size-cells = <0>; + + hdmi@4c { + compatible = "ite,it6263"; + reg = <0x4c>; + data-mapping = "jeida-24"; + reset-gpios = <&gpio1 10 GPIO_ACTIVE_LOW>; + ivdd-supply = <®_buck5>; + ovdd-supply = <®_vext_3v3>; + txavcc18-supply = <®_buck5>; + txavcc33-supply = <®_vext_3v3>; + pvcc1-supply = <®_buck5>; + pvcc2-supply = <®_buck5>; + avcc-supply = <®_vext_3v3>; + anvdd-supply = <®_buck5>; + apvdd-supply = <®_buck5>; + + ports { + #address-cells = <1>; + #size-cells = <0>; + + port@0 { + reg = <0>; + + it6263_lvds_link1: endpoint { + remote-endpoint = <&ldb_lvds_ch0>; + }; + }; + + port@2 { + reg = <2>; + + it6263_out: endpoint { + remote-endpoint = <&hdmi_in>; + }; + }; + }; + }; + }; + + - | + /* dual-link LVDS input */ + #include + + i2c { + #address-cells = <1>; + #size-cells = <0>; + + hdmi@4c { + compatible = "ite,it6263"; + reg = <0x4c>; + data-mapping = "jeida-24"; + reset-gpios = <&gpio1 10 GPIO_ACTIVE_LOW>; + ivdd-supply = <®_buck5>; + ovdd-supply = <®_vext_3v3>; + txavcc18-supply = <®_buck5>; + txavcc33-supply = <®_vext_3v3>; + pvcc1-supply = <®_buck5>; + pvcc2-supply = <®_buck5>; + avcc-supply = <®_vext_3v3>; + anvdd-supply = <®_buck5>; + apvdd-supply = <®_buck5>; + + ports { + #address-cells = <1>; + #size-cells = <0>; + + port@0 { + reg = <0>; + dual-lvds-odd-pixels; + + it6263_lvds_link1_dual: endpoint { + remote-endpoint = <&ldb_lvds_ch0>; + }; + }; + + port@1 { + reg = <1>; + dual-lvds-even-pixels; + + it6263_lvds_link2_dual: endpoint { + remote-endpoint = <&ldb_lvds_ch1>; + }; + }; + + port@2 { + reg = <2>; + + it6263_out_dual: endpoint { + remote-endpoint = <&hdmi_in>; + }; + }; + }; + }; + }; diff --git a/Documentation/devicetree/bindings/display/bridge/sil,sii9022.yaml b/Documentation/devicetree/bindings/display/bridge/sil,sii9022.yaml index 5a69547ad3d7..1509c4535e53 100644 --- a/Documentation/devicetree/bindings/display/bridge/sil,sii9022.yaml +++ b/Documentation/devicetree/bindings/display/bridge/sil,sii9022.yaml @@ -81,9 +81,22 @@ properties: properties: port@0: - $ref: /schemas/graph.yaml#/properties/port + unevaluatedProperties: false + $ref: /schemas/graph.yaml#/$defs/port-base description: Parallel RGB input port + properties: + endpoint: + $ref: /schemas/graph.yaml#/$defs/endpoint-base + unevaluatedProperties: false + + properties: + bus-width: + description: + Endpoint bus width. + enum: [ 16, 18, 24 ] + default: 24 + port@1: $ref: /schemas/graph.yaml#/properties/port description: HDMI output port diff --git a/Documentation/devicetree/bindings/display/bridge/ti,tdp158.yaml b/Documentation/devicetree/bindings/display/bridge/ti,tdp158.yaml new file mode 100644 index 000000000000..1c522f72c4ba --- /dev/null +++ b/Documentation/devicetree/bindings/display/bridge/ti,tdp158.yaml @@ -0,0 +1,57 @@ +# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/display/bridge/ti,tdp158.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: TI TDP158 HDMI to TMDS Redriver + +maintainers: + - Arnaud Vrac + - Pierre-Hugues Husson + +properties: + compatible: + const: ti,tdp158 + +# The reg property is required if and only if the device is connected +# to an I2C bus. In pin strap mode, reg must not be specified. + reg: + description: I2C address of the device + +# Pin 36 = Operation Enable / Reset Pin +# OE = L: Power Down Mode +# OE = H: Normal Operation +# Internal weak pullup - device resets on H to L transitions + enable-gpios: + description: GPIO controlling bridge enable + + vcc-supply: + description: Power supply 3.3V + + vdd-supply: + description: Power supply 1.1V + + ports: + $ref: /schemas/graph.yaml#/properties/ports + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: Bridge input + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: Bridge output + + required: + - port@0 + - port@1 + +required: + - compatible + - vcc-supply + - vdd-supply + - ports + +additionalProperties: false diff --git a/Documentation/devicetree/bindings/display/bridge/toshiba,tc358768.yaml b/Documentation/devicetree/bindings/display/bridge/toshiba,tc358768.yaml index 779d8c57f854..bb5d3b543800 100644 --- a/Documentation/devicetree/bindings/display/bridge/toshiba,tc358768.yaml +++ b/Documentation/devicetree/bindings/display/bridge/toshiba,tc358768.yaml @@ -60,6 +60,10 @@ properties: data-lines: $ref: /schemas/types.yaml#/definitions/uint32 enum: [ 16, 18, 24 ] + deprecated: true + + bus-width: + enum: [ 16, 18, 24 ] port@1: $ref: /schemas/graph.yaml#/properties/port diff --git a/Documentation/devicetree/bindings/display/imx/fsl-imx-drm.txt b/Documentation/devicetree/bindings/display/imx/fsl-imx-drm.txt index 3c35338a2867..269b1ae2fca9 100644 --- a/Documentation/devicetree/bindings/display/imx/fsl-imx-drm.txt +++ b/Documentation/devicetree/bindings/display/imx/fsl-imx-drm.txt @@ -119,7 +119,6 @@ Optional properties: - interface-pix-fmt: How this display is connected to the display interface. Currently supported types: "rgb24", "rgb565", "bgr666" and "lvds666". -- edid: verbatim EDID data block describing attached display. - ddc: phandle describing the i2c bus handling the display data channel - port@[0-1]: Port nodes with endpoint definitions as defined in @@ -131,7 +130,6 @@ example: disp0 { compatible = "fsl,imx-parallel-display"; - edid = [edid-data]; interface-pix-fmt = "rgb24"; port@0 { diff --git a/Documentation/devicetree/bindings/display/imx/ldb.txt b/Documentation/devicetree/bindings/display/imx/ldb.txt index 8e6e7d797943..03653a291b54 100644 --- a/Documentation/devicetree/bindings/display/imx/ldb.txt +++ b/Documentation/devicetree/bindings/display/imx/ldb.txt @@ -62,7 +62,6 @@ Required properties: display-timings are used instead. Optional properties (required if display-timings are used): - - ddc-i2c-bus: phandle of an I2C controller used for DDC EDID probing - display-timings : A node that describes the display timings as defined in Documentation/devicetree/bindings/display/panel/display-timing.txt. - fsl,data-mapping : should be "spwg" or "jeida" diff --git a/Documentation/devicetree/bindings/display/lvds-data-mapping.yaml b/Documentation/devicetree/bindings/display/lvds-data-mapping.yaml index d68982fe2e9b..ab842594feb9 100644 --- a/Documentation/devicetree/bindings/display/lvds-data-mapping.yaml +++ b/Documentation/devicetree/bindings/display/lvds-data-mapping.yaml @@ -26,12 +26,17 @@ description: | Device compatible with those specifications have been marketed under the FPD-Link and FlatLink brands. + This bindings also supports 30-bit data mapping compatible with JEIDA and + VESA. + properties: data-mapping: enum: - jeida-18 - jeida-24 + - jeida-30 - vesa-24 + - vesa-30 description: | The color signals mapping order. @@ -60,6 +65,19 @@ properties: DATA2 ><_CTL2_><_CTL1_><_CTL0_><__B7__><__B6__><__B5__><__B4__>< DATA3 ><_CTL3_><__B1__><__B0__><__G1__><__G0__><__R1__><__R0__>< + - "jeida-30" - 30-bit data mapping compatible with JEIDA and VESA. Data + are transferred as follows on 5 LVDS lanes. + + Slot 0 1 2 3 4 5 6 + ________________ _________________ + Clock \_______________________/ + ______ ______ ______ ______ ______ ______ ______ + DATA0 ><__G4__><__R9__><__R8__><__R7__><__R6__><__R5__><__R4__>< + DATA1 ><__B5__><__B4__><__G9__><__G8__><__G7__><__G6__><__G5__>< + DATA2 ><_CTL2_><_CTL1_><_CTL0_><__B9__><__B8__><__B7__><__B6__>< + DATA3 ><_CTL3_><__B3__><__B2__><__G3__><__G2__><__R3__><__R2__>< + DATA4 ><_CTL3_><__B1__><__B0__><__G1__><__G0__><__R1__><__R0__>< + - "vesa-24" - 24-bit data mapping compatible with the [VESA] specification. Data are transferred as follows on 4 LVDS lanes. @@ -72,6 +90,19 @@ properties: DATA2 ><_CTL2_><_CTL1_><_CTL0_><__B5__><__B4__><__B3__><__B2__>< DATA3 ><_CTL3_><__B7__><__B6__><__G7__><__G6__><__R7__><__R6__>< + - "vesa-30" - 30-bit data mapping compatible with VESA. Data are + transferred as follows on 5 LVDS lanes. + + Slot 0 1 2 3 4 5 6 + ________________ _________________ + Clock \_______________________/ + ______ ______ ______ ______ ______ ______ ______ + DATA0 ><__G0__><__R5__><__R4__><__R3__><__R2__><__R1__><__R0__>< + DATA1 ><__B1__><__B0__><__G5__><__G4__><__G3__><__G2__><__G1__>< + DATA2 ><_CTL2_><_CTL1_><_CTL0_><__B5__><__B4__><__B3__><__B2__>< + DATA3 ><_CTL3_><__B7__><__B6__><__G7__><__G6__><__R7__><__R6__>< + DATA4 ><_CTL3_><__B9__><__B8__><__G9__><__G8__><__R9__><__R8__>< + Control signals are mapped as follows. CTL0: HSync diff --git a/Documentation/devicetree/bindings/display/lvds-dual-ports.yaml b/Documentation/devicetree/bindings/display/lvds-dual-ports.yaml new file mode 100644 index 000000000000..785701fe1590 --- /dev/null +++ b/Documentation/devicetree/bindings/display/lvds-dual-ports.yaml @@ -0,0 +1,63 @@ +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/display/lvds-dual-ports.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: Dual-link LVDS Display Common Properties + +maintainers: + - Liu Ying + +description: | + Common properties for LVDS displays with dual LVDS links. Extend LVDS display + common properties defined in lvds.yaml. + + Dual-link LVDS displays receive odd pixels and even pixels separately from + the dual LVDS links. One link receives odd pixels and the other receives + even pixels. Some of those displays may also use only one LVDS link to + receive all pixels, being odd and even agnostic. + +allOf: + - $ref: lvds.yaml# + +properties: + ports: + $ref: /schemas/graph.yaml#/properties/ports + + patternProperties: + '^port@[01]$': + $ref: /schemas/graph.yaml#/$defs/port-base + unevaluatedProperties: false + description: | + port@0 represents the first LVDS input link. + port@1 represents the second LVDS input link. + + properties: + dual-lvds-odd-pixels: + type: boolean + description: LVDS input link for odd pixels + + dual-lvds-even-pixels: + type: boolean + description: LVDS input link for even pixels + + oneOf: + - required: [dual-lvds-odd-pixels] + - required: [dual-lvds-even-pixels] + - properties: + dual-lvds-odd-pixels: false + dual-lvds-even-pixels: false + + anyOf: + - required: + - port@0 + - required: + - port@1 + +required: + - ports + +additionalProperties: true + +... diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,aal.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,aal.yaml index cf24434854ff..47ddba5c41af 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,aal.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,aal.yaml @@ -62,6 +62,27 @@ properties: $ref: /schemas/types.yaml#/definitions/phandle-array maxItems: 1 + ports: + $ref: /schemas/graph.yaml#/properties/ports + description: + Input and output ports can have multiple endpoints, each of those + connects to either the primary, secondary, etc, display pipeline. + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: AAL input port + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + AAL output to the next component's input, for example could be one + of many gamma, overdrive or other blocks. + + required: + - port@0 + - port@1 + required: - compatible - reg @@ -89,5 +110,24 @@ examples: power-domains = <&scpsys MT8173_POWER_DOMAIN_MM>; clocks = <&mmsys CLK_MM_DISP_AAL>; mediatek,gce-client-reg = <&gce SUBSYS_1401XXXX 0x5000 0x1000>; + + ports { + #address-cells = <1>; + #size-cells = <0>; + + port@0 { + reg = <0>; + aal0_in: endpoint { + remote-endpoint = <&ccorr0_out>; + }; + }; + + port@1 { + reg = <1>; + aal0_out: endpoint { + remote-endpoint = <&gamma0_in>; + }; + }; + }; }; }; diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,ccorr.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,ccorr.yaml index 9f8366763831..fca8e7bb0cbc 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,ccorr.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,ccorr.yaml @@ -57,6 +57,27 @@ properties: $ref: /schemas/types.yaml#/definitions/phandle-array maxItems: 1 + ports: + $ref: /schemas/graph.yaml#/properties/ports + description: + Input and output ports can have multiple endpoints, each of those + connects to either the primary, secondary, etc, display pipeline. + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: CCORR input port + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + CCORR output to the input of the next desired component in the + display pipeline, usually only one of the available AAL blocks. + + required: + - port@0 + - port@1 + required: - compatible - reg diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,color.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,color.yaml index 7df786bbad20..6160439ce4d7 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,color.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,color.yaml @@ -65,6 +65,28 @@ properties: $ref: /schemas/types.yaml#/definitions/phandle-array maxItems: 1 + ports: + $ref: /schemas/graph.yaml#/properties/ports + description: + Input and output ports can have multiple endpoints, each of those + connects to either the primary, secondary, etc, display pipeline. + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: COLOR input port + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + COLOR output to the input of the next desired component in the + display pipeline, for example one of the available CCORR or AAL + blocks. + + required: + - port@0 + - port@1 + required: - compatible - reg diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,dither.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,dither.yaml index 6fceb1f95d2a..abaf27916d13 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,dither.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,dither.yaml @@ -56,6 +56,28 @@ properties: $ref: /schemas/types.yaml#/definitions/phandle-array maxItems: 1 + ports: + $ref: /schemas/graph.yaml#/properties/ports + description: + Input and output ports can have multiple endpoints, each of those + connects to either the primary, secondary, etc, display pipeline. + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: DITHER input, usually from a POSTMASK or GAMMA block. + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + DITHER output to the input of the next desired component in the + display pipeline, for example one of the available DSC compressors, + DP_INTF, DSI, LVDS or others. + + required: + - port@0 + - port@1 + required: - compatible - reg diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,dpi.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,dpi.yaml index 497c0eb4ed0b..0f1e556dc8ef 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,dpi.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,dpi.yaml @@ -81,13 +81,34 @@ properties: Output port node. This port should be connected to the input port of an attached HDMI, LVDS or DisplayPort encoder chip. + ports: + $ref: /schemas/graph.yaml#/properties/ports + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: DPI input port + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: DPI output to an HDMI, LVDS or DisplayPort encoder input + + required: + - port@0 + - port@1 + required: - compatible - reg - interrupts - clocks - clock-names - - port + +oneOf: + - required: + - port + - required: + - ports additionalProperties: false @@ -96,7 +117,7 @@ examples: #include #include - dpi0: dpi@1401d000 { + dpi: dpi@1401d000 { compatible = "mediatek,mt8173-dpi"; reg = <0x1401d000 0x1000>; interrupts = ; diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,dsc.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,dsc.yaml index 2cbdd9ee449d..846de6c17d93 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,dsc.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,dsc.yaml @@ -49,6 +49,30 @@ properties: $ref: /schemas/types.yaml#/definitions/phandle-array maxItems: 1 + ports: + $ref: /schemas/graph.yaml#/properties/ports + description: + Input and output ports can have multiple endpoints, each of those + connects to either the primary, secondary, etc, display pipeline. + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: + Display Stream Compression input, usually from one of the DITHER + or MERGE blocks. + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + Display Stream Compression output to the input of the next desired + component in the display pipeline, for example to MERGE, DP_INTF, + DPI or DSI. + + required: + - port@0 + - port@1 + required: - compatible - reg diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,dsi.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,dsi.yaml index a7aa8fcb0dd1..27ffbccc2a08 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,dsi.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,dsi.yaml @@ -77,6 +77,26 @@ properties: Output port node. This port should be connected to the input port of an attached DSI panel or DSI-to-eDP encoder chip. + ports: + $ref: /schemas/graph.yaml#/properties/ports + description: + Input ports can have multiple endpoints, each of those connects + to either the primary, secondary, etc, display pipeline. + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: DSI input port, usually from DITHER, DSC or MERGE + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + DSI output to an attached DSI panel, or a DSI-to-X encoder chip + + required: + - port@0 + - port@1 + required: - compatible - reg @@ -86,7 +106,12 @@ required: - clock-names - phys - phy-names - - port + +oneOf: + - required: + - port + - required: + - ports unevaluatedProperties: false diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,ethdr.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,ethdr.yaml index 677882348ede..98db47894eeb 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,ethdr.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,ethdr.yaml @@ -110,6 +110,28 @@ properties: include/dt-bindings/gce/-gce.h, mapping to the register of display function block. + ports: + $ref: /schemas/graph.yaml#/properties/ports + description: + Input and output ports can have multiple endpoints, each of those + connects to either the primary, secondary, etc, display pipeline. + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: ETHDR input, usually from one of the MERGE blocks. + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + ETHDR output to the input of the next desired component in the + display pipeline, for example one of the available MERGE blocks, + or others. + + required: + - port@0 + - port@1 + required: - compatible - reg diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,gamma.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,gamma.yaml index 6823d3ce5049..48542dc7e784 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,gamma.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,gamma.yaml @@ -65,6 +65,25 @@ properties: $ref: /schemas/types.yaml#/definitions/phandle-array maxItems: 1 + ports: + $ref: /schemas/graph.yaml#/properties/ports + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: GAMMA input, usually from one of the AAL blocks. + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + GAMMA output to the input of the next desired component in the + display pipeline, for example one of the available DITHER or + POSTMASK blocks. + + required: + - port@0 + - port@1 + required: - compatible - reg diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,merge.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,merge.yaml index dae839279950..0de9f64f3f84 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,merge.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,merge.yaml @@ -77,6 +77,29 @@ properties: $ref: /schemas/types.yaml#/definitions/phandle-array maxItems: 1 + ports: + $ref: /schemas/graph.yaml#/properties/ports + description: + Input and output ports can have multiple endpoints, each of those + connects to either the primary, secondary, etc, display pipeline. + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: + MERGE input port, usually from DITHER, DPI, DSC, DSI, MDP_RDMA, + ETHDR or even from a different MERGE block + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + MERGE output to a DSC, DPI, DP_INTF, DSI, ETHDR, Write DMA, or + a different MERGE block, or others. + + required: + - port@0 + - port@1 + resets: description: reset controller See Documentation/devicetree/bindings/reset/reset.txt for details. diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,od.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,od.yaml index 831c653caffd..71534febd49c 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,od.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,od.yaml @@ -38,6 +38,28 @@ properties: items: - description: OD Clock + ports: + $ref: /schemas/graph.yaml#/properties/ports + description: + Input and output ports can have multiple endpoints, each of those + connects to either the primary, secondary, etc, display pipeline. + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: OD input port, usually from an AAL block + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + OD output to the input of the next desired component in the + display pipeline, for example one of the available RDMA or + other blocks. + + required: + - port@0 + - port@1 + required: - compatible - reg diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,ovl-2l.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,ovl-2l.yaml index c7dd0ef02dcf..bacdfe7d08a6 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,ovl-2l.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,ovl-2l.yaml @@ -57,6 +57,28 @@ properties: $ref: /schemas/types.yaml#/definitions/phandle-array maxItems: 1 + ports: + $ref: /schemas/graph.yaml#/properties/ports + description: + Input and output ports can have multiple endpoints, each of those + connects to either the primary, secondary, etc, display pipeline. + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: OVL input port from MMSYS, VDOSYS or other OVLs + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + OVL output to the input of the next desired component in the + display pipeline, for example one of the available COLOR, RDMA + or WDMA blocks. + + required: + - port@0 + - port@1 + required: - compatible - reg diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,ovl.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,ovl.yaml index d55611c7ce5e..9ea796a033b2 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,ovl.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,ovl.yaml @@ -75,6 +75,28 @@ properties: $ref: /schemas/types.yaml#/definitions/phandle-array maxItems: 1 + ports: + $ref: /schemas/graph.yaml#/properties/ports + description: + Input and output ports can have multiple endpoints, each of those + connects to either the primary, secondary, etc, display pipeline. + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: OVL input port from MMSYS or one of multiple VDOSYS + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + OVL output to the input of the next desired component in the + display pipeline, for example one of the available COLOR, RDMA + or WDMA blocks. + + required: + - port@0 + - port@1 + required: - compatible - reg diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,postmask.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,postmask.yaml index 11fe32e50a59..fb6fe4742624 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,postmask.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,postmask.yaml @@ -52,6 +52,27 @@ properties: $ref: /schemas/types.yaml#/definitions/phandle-array maxItems: 1 + ports: + $ref: /schemas/graph.yaml#/properties/ports + description: + Input and output ports can have multiple endpoints, each of those + connects to either the primary, secondary, etc, display pipeline. + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: POSTMASK input port, usually from GAMMA + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + POSTMASK output to the input of the next desired component in the + display pipeline, for example one of the available DITHER blocks. + + required: + - port@0 + - port@1 + required: - compatible - reg diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,rdma.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,rdma.yaml index 4cadb245d028..878f676b581f 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,rdma.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,rdma.yaml @@ -87,6 +87,28 @@ properties: $ref: /schemas/types.yaml#/definitions/phandle-array maxItems: 1 + ports: + $ref: /schemas/graph.yaml#/properties/ports + description: + Input and output ports can have multiple endpoints, each of those + connects to either the primary, secondary, etc, display pipeline. + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: RDMA input port, usually from MMSYS, OD or OVL + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + RDMA output to the input of the next desired component in the + display pipeline, for example one of the available COLOR, DPI, + DSI, MERGE or UFOE blocks. + + required: + - port@0 + - port@1 + required: - compatible - reg diff --git a/Documentation/devicetree/bindings/display/mediatek/mediatek,ufoe.yaml b/Documentation/devicetree/bindings/display/mediatek/mediatek,ufoe.yaml index 39e3e2d4a0db..61a5e22effbf 100644 --- a/Documentation/devicetree/bindings/display/mediatek/mediatek,ufoe.yaml +++ b/Documentation/devicetree/bindings/display/mediatek/mediatek,ufoe.yaml @@ -43,6 +43,27 @@ properties: items: - description: UFOe Clock + ports: + $ref: /schemas/graph.yaml#/properties/ports + description: + Input and output ports can have multiple endpoints, each of those + connects to either the primary, secondary, etc, display pipeline. + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: UFOE input, usually from one of the RDMA blocks. + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: + UFOE output to the input of the next desired component in the + display pipeline, usually one of the available DSI blocks. + + required: + - port@0 + - port@1 + required: - compatible - reg diff --git a/Documentation/devicetree/bindings/display/msm/dp-controller.yaml b/Documentation/devicetree/bindings/display/msm/dp-controller.yaml index 97993feda193..a212f335d5ff 100644 --- a/Documentation/devicetree/bindings/display/msm/dp-controller.yaml +++ b/Documentation/devicetree/bindings/display/msm/dp-controller.yaml @@ -17,6 +17,7 @@ properties: compatible: oneOf: - enum: + - qcom,sa8775p-dp - qcom,sc7180-dp - qcom,sc7280-dp - qcom,sc7280-edp diff --git a/Documentation/devicetree/bindings/display/msm/gmu.yaml b/Documentation/devicetree/bindings/display/msm/gmu.yaml index b1bd372996d5..ab884e236429 100644 --- a/Documentation/devicetree/bindings/display/msm/gmu.yaml +++ b/Documentation/devicetree/bindings/display/msm/gmu.yaml @@ -125,6 +125,7 @@ allOf: enum: - qcom,adreno-gmu-635.0 - qcom,adreno-gmu-660.1 + - qcom,adreno-gmu-663.0 then: properties: reg: diff --git a/Documentation/devicetree/bindings/display/msm/qcom,sa8775p-mdss.yaml b/Documentation/devicetree/bindings/display/msm/qcom,sa8775p-mdss.yaml new file mode 100644 index 000000000000..58f8a01f29c7 --- /dev/null +++ b/Documentation/devicetree/bindings/display/msm/qcom,sa8775p-mdss.yaml @@ -0,0 +1,241 @@ +# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/display/msm/qcom,sa8775p-mdss.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: Qualcomm Technologies, Inc. SA87755P Display MDSS + +maintainers: + - Mahadevan + +description: + SA8775P MSM Mobile Display Subsystem(MDSS), which encapsulates sub-blocks like + DPU display controller, DP interfaces and EDP etc. + +$ref: /schemas/display/msm/mdss-common.yaml# + +properties: + compatible: + const: qcom,sa8775p-mdss + + clocks: + items: + - description: Display AHB + - description: Display hf AXI + - description: Display core + + iommus: + maxItems: 1 + + interconnects: + maxItems: 3 + + interconnect-names: + maxItems: 3 + +patternProperties: + "^display-controller@[0-9a-f]+$": + type: object + additionalProperties: true + + properties: + compatible: + const: qcom,sa8775p-dpu + + "^displayport-controller@[0-9a-f]+$": + type: object + additionalProperties: true + + properties: + compatible: + items: + - const: qcom,sa8775p-dp + +required: + - compatible + +unevaluatedProperties: false + +examples: + - | + #include + #include + #include + #include + #include + #include + + display-subsystem@ae00000 { + compatible = "qcom,sa8775p-mdss"; + reg = <0x0ae00000 0x1000>; + reg-names = "mdss"; + + interconnects = <&mmss_noc MASTER_MDP0 &mc_virt SLAVE_EBI1>, + <&mmss_noc MASTER_MDP1 &mc_virt SLAVE_EBI1>, + <&gem_noc MASTER_APPSS_PROC &config_noc SLAVE_DISPLAY_CFG>; + interconnect-names = "mdp0-mem", + "mdp1-mem", + "cpu-cfg"; + + + resets = <&dispcc_core_bcr>; + power-domains = <&dispcc_gdsc>; + + clocks = <&dispcc_ahb_clk>, + <&gcc GCC_DISP_HF_AXI_CLK>, + <&dispcc_mdp_clk>; + + interrupts = ; + interrupt-controller; + #interrupt-cells = <1>; + + iommus = <&apps_smmu 0x1000 0x402>; + + #address-cells = <1>; + #size-cells = <1>; + ranges; + + display-controller@ae01000 { + compatible = "qcom,sa8775p-dpu"; + reg = <0x0ae01000 0x8f000>, + <0x0aeb0000 0x2008>; + reg-names = "mdp", "vbif"; + + clocks = <&gcc GCC_DISP_HF_AXI_CLK>, + <&dispcc_ahb_clk>, + <&dispcc_mdp_lut_clk>, + <&dispcc_mdp_clk>, + <&dispcc_mdp_vsync_clk>; + clock-names = "nrt_bus", + "iface", + "lut", + "core", + "vsync"; + + assigned-clocks = <&dispcc_mdp_vsync_clk>; + assigned-clock-rates = <19200000>; + + operating-points-v2 = <&mdss0_mdp_opp_table>; + power-domains = <&rpmhpd RPMHPD_MMCX>; + + interrupt-parent = <&mdss0>; + interrupts = <0>; + + ports { + #address-cells = <1>; + #size-cells = <0>; + + port@0 { + reg = <0>; + dpu_intf0_out: endpoint { + remote-endpoint = <&mdss0_dp0_in>; + }; + }; + }; + + mdss0_mdp_opp_table: opp-table { + compatible = "operating-points-v2"; + + opp-375000000 { + opp-hz = /bits/ 64 <375000000>; + required-opps = <&rpmhpd_opp_svs_l1>; + }; + + opp-500000000 { + opp-hz = /bits/ 64 <500000000>; + required-opps = <&rpmhpd_opp_nom>; + }; + + opp-575000000 { + opp-hz = /bits/ 64 <575000000>; + required-opps = <&rpmhpd_opp_turbo>; + }; + + opp-650000000 { + opp-hz = /bits/ 64 <650000000>; + required-opps = <&rpmhpd_opp_turbo_l1>; + }; + }; + }; + + displayport-controller@af54000 { + compatible = "qcom,sa8775p-dp"; + + pinctrl-0 = <&dp_hot_plug_det>; + pinctrl-names = "default"; + + reg = <0xaf54000 0x104>, + <0xaf54200 0x0c0>, + <0xaf55000 0x770>, + <0xaf56000 0x09c>; + + interrupt-parent = <&mdss0>; + interrupts = <12>; + + clocks = <&dispcc_mdss_ahb_clk>, + <&dispcc_dptx0_aux_clk>, + <&dispcc_dptx0_link_clk>, + <&dispcc_dptx0_link_intf_clk>, + <&dispcc_dptx0_pixel0_clk>; + clock-names = "core_iface", + "core_aux", + "ctrl_link", + "ctrl_link_iface", + "stream_pixel"; + + assigned-clocks = <&dispcc_mdss_dptx0_link_clk_src>, + <&dispcc_mdss_dptx0_pixel0_clk_src>; + assigned-clock-parents = <&mdss0_edp_phy 0>, <&mdss0_edp_phy 1>; + + phys = <&mdss0_edp_phy>; + phy-names = "dp"; + + operating-points-v2 = <&dp_opp_table>; + power-domains = <&rpmhpd SA8775P_MMCX>; + + #sound-dai-cells = <0>; + + ports { + #address-cells = <1>; + #size-cells = <0>; + + port@0 { + reg = <0>; + mdss0_dp0_in: endpoint { + remote-endpoint = <&dpu_intf0_out>; + }; + }; + + port@1 { + reg = <1>; + mdss0_dp_out: endpoint { }; + }; + }; + + dp_opp_table: opp-table { + compatible = "operating-points-v2"; + + opp-160000000 { + opp-hz = /bits/ 64 <160000000>; + required-opps = <&rpmhpd_opp_low_svs>; + }; + + opp-270000000 { + opp-hz = /bits/ 64 <270000000>; + required-opps = <&rpmhpd_opp_svs>; + }; + + opp-540000000 { + opp-hz = /bits/ 64 <540000000>; + required-opps = <&rpmhpd_opp_svs_l1>; + }; + + opp-810000000 { + opp-hz = /bits/ 64 <810000000>; + required-opps = <&rpmhpd_opp_nom>; + }; + }; + }; + }; +... diff --git a/Documentation/devicetree/bindings/display/msm/qcom,sc7280-dpu.yaml b/Documentation/devicetree/bindings/display/msm/qcom,sc7280-dpu.yaml index b0fbe86219d1..6902795b4e2c 100644 --- a/Documentation/devicetree/bindings/display/msm/qcom,sc7280-dpu.yaml +++ b/Documentation/devicetree/bindings/display/msm/qcom,sc7280-dpu.yaml @@ -7,13 +7,21 @@ $schema: http://devicetree.org/meta-schemas/core.yaml# title: Qualcomm Display DPU on SC7280 maintainers: + - Bjorn Andersson + - Neil Armstrong + - Dmitry Baryshkov - Krishna Manikandan $ref: /schemas/display/msm/dpu-common.yaml# properties: compatible: - const: qcom,sc7280-dpu + enum: + - qcom,sc7280-dpu + - qcom,sc8280xp-dpu + - qcom,sm8350-dpu + - qcom,sm8450-dpu + - qcom,sm8550-dpu reg: items: diff --git a/Documentation/devicetree/bindings/display/msm/qcom,sc8280xp-dpu.yaml b/Documentation/devicetree/bindings/display/msm/qcom,sc8280xp-dpu.yaml deleted file mode 100644 index d19e3bec4600..000000000000 --- a/Documentation/devicetree/bindings/display/msm/qcom,sc8280xp-dpu.yaml +++ /dev/null @@ -1,122 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause -%YAML 1.2 ---- -$id: http://devicetree.org/schemas/display/msm/qcom,sc8280xp-dpu.yaml# -$schema: http://devicetree.org/meta-schemas/core.yaml# - -title: Qualcomm SC8280XP Display Processing Unit - -maintainers: - - Bjorn Andersson - -description: - Device tree bindings for SC8280XP Display Processing Unit. - -$ref: /schemas/display/msm/dpu-common.yaml# - -properties: - compatible: - const: qcom,sc8280xp-dpu - - reg: - items: - - description: Address offset and size for mdp register set - - description: Address offset and size for vbif register set - - reg-names: - items: - - const: mdp - - const: vbif - - clocks: - items: - - description: Display hf axi clock - - description: Display sf axi clock - - description: Display ahb clock - - description: Display lut clock - - description: Display core clock - - description: Display vsync clock - - clock-names: - items: - - const: bus - - const: nrt_bus - - const: iface - - const: lut - - const: core - - const: vsync - -unevaluatedProperties: false - -examples: - - | - #include - #include - #include - #include - #include - - display-controller@ae01000 { - compatible = "qcom,sc8280xp-dpu"; - reg = <0x0ae01000 0x8f000>, - <0x0aeb0000 0x2008>; - reg-names = "mdp", "vbif"; - - clocks = <&gcc GCC_DISP_HF_AXI_CLK>, - <&gcc GCC_DISP_SF_AXI_CLK>, - <&dispcc0 DISP_CC_MDSS_AHB_CLK>, - <&dispcc0 DISP_CC_MDSS_MDP_LUT_CLK>, - <&dispcc0 DISP_CC_MDSS_MDP_CLK>, - <&dispcc0 DISP_CC_MDSS_VSYNC_CLK>; - clock-names = "bus", - "nrt_bus", - "iface", - "lut", - "core", - "vsync"; - - assigned-clocks = <&dispcc0 DISP_CC_MDSS_MDP_CLK>, - <&dispcc0 DISP_CC_MDSS_VSYNC_CLK>; - assigned-clock-rates = <460000000>, - <19200000>; - - operating-points-v2 = <&mdp_opp_table>; - power-domains = <&rpmhpd SC8280XP_MMCX>; - - interrupt-parent = <&mdss0>; - interrupts = <0>; - - ports { - #address-cells = <1>; - #size-cells = <0>; - - port@0 { - reg = <0>; - endpoint { - remote-endpoint = <&mdss0_dp0_in>; - }; - }; - - port@4 { - reg = <4>; - endpoint { - remote-endpoint = <&mdss0_dp1_in>; - }; - }; - - port@5 { - reg = <5>; - endpoint { - remote-endpoint = <&mdss0_dp3_in>; - }; - }; - - port@6 { - reg = <6>; - endpoint { - remote-endpoint = <&mdss0_dp2_in>; - }; - }; - }; - }; -... diff --git a/Documentation/devicetree/bindings/display/msm/qcom,sm8150-dpu.yaml b/Documentation/devicetree/bindings/display/msm/qcom,sm8150-dpu.yaml index 13146b3f053c..a88d22f30a60 100644 --- a/Documentation/devicetree/bindings/display/msm/qcom,sm8150-dpu.yaml +++ b/Documentation/devicetree/bindings/display/msm/qcom,sm8150-dpu.yaml @@ -13,7 +13,9 @@ $ref: /schemas/display/msm/dpu-common.yaml# properties: compatible: - const: qcom,sm8150-dpu + enum: + - qcom,sm8150-dpu + - qcom,sm8250-dpu reg: items: diff --git a/Documentation/devicetree/bindings/display/msm/qcom,sm8250-dpu.yaml b/Documentation/devicetree/bindings/display/msm/qcom,sm8250-dpu.yaml deleted file mode 100644 index ffa5047e901f..000000000000 --- a/Documentation/devicetree/bindings/display/msm/qcom,sm8250-dpu.yaml +++ /dev/null @@ -1,99 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause -%YAML 1.2 ---- -$id: http://devicetree.org/schemas/display/msm/qcom,sm8250-dpu.yaml# -$schema: http://devicetree.org/meta-schemas/core.yaml# - -title: Qualcomm SM8250 Display DPU - -maintainers: - - Dmitry Baryshkov - -$ref: /schemas/display/msm/dpu-common.yaml# - -properties: - compatible: - const: qcom,sm8250-dpu - - reg: - items: - - description: Address offset and size for mdp register set - - description: Address offset and size for vbif register set - - reg-names: - items: - - const: mdp - - const: vbif - - clocks: - items: - - description: Display ahb clock - - description: Display hf axi clock - - description: Display core clock - - description: Display vsync clock - - clock-names: - items: - - const: iface - - const: bus - - const: core - - const: vsync - -required: - - compatible - - reg - - reg-names - - clocks - - clock-names - -unevaluatedProperties: false - -examples: - - | - #include - #include - #include - #include - #include - - display-controller@ae01000 { - compatible = "qcom,sm8250-dpu"; - reg = <0x0ae01000 0x8f000>, - <0x0aeb0000 0x2008>; - reg-names = "mdp", "vbif"; - - clocks = <&dispcc DISP_CC_MDSS_AHB_CLK>, - <&gcc GCC_DISP_HF_AXI_CLK>, - <&dispcc DISP_CC_MDSS_MDP_CLK>, - <&dispcc DISP_CC_MDSS_VSYNC_CLK>; - clock-names = "iface", "bus", "core", "vsync"; - - assigned-clocks = <&dispcc DISP_CC_MDSS_VSYNC_CLK>; - assigned-clock-rates = <19200000>; - - operating-points-v2 = <&mdp_opp_table>; - power-domains = <&rpmhpd RPMHPD_MMCX>; - - interrupt-parent = <&mdss>; - interrupts = <0>; - - ports { - #address-cells = <1>; - #size-cells = <0>; - - port@0 { - reg = <0>; - endpoint { - remote-endpoint = <&dsi0_in>; - }; - }; - - port@1 { - reg = <1>; - endpoint { - remote-endpoint = <&dsi1_in>; - }; - }; - }; - }; -... diff --git a/Documentation/devicetree/bindings/display/msm/qcom,sm8350-dpu.yaml b/Documentation/devicetree/bindings/display/msm/qcom,sm8350-dpu.yaml deleted file mode 100644 index 96ef2d9c3512..000000000000 --- a/Documentation/devicetree/bindings/display/msm/qcom,sm8350-dpu.yaml +++ /dev/null @@ -1,120 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause -%YAML 1.2 ---- -$id: http://devicetree.org/schemas/display/msm/qcom,sm8350-dpu.yaml# -$schema: http://devicetree.org/meta-schemas/core.yaml# - -title: Qualcomm SM8350 Display DPU - -maintainers: - - Robert Foss - -$ref: /schemas/display/msm/dpu-common.yaml# - -properties: - compatible: - const: qcom,sm8350-dpu - - reg: - items: - - description: Address offset and size for mdp register set - - description: Address offset and size for vbif register set - - reg-names: - items: - - const: mdp - - const: vbif - - clocks: - items: - - description: Display hf axi clock - - description: Display sf axi clock - - description: Display ahb clock - - description: Display lut clock - - description: Display core clock - - description: Display vsync clock - - clock-names: - items: - - const: bus - - const: nrt_bus - - const: iface - - const: lut - - const: core - - const: vsync - -unevaluatedProperties: false - -examples: - - | - #include - #include - #include - #include - #include - - display-controller@ae01000 { - compatible = "qcom,sm8350-dpu"; - reg = <0x0ae01000 0x8f000>, - <0x0aeb0000 0x2008>; - reg-names = "mdp", "vbif"; - - clocks = <&gcc GCC_DISP_HF_AXI_CLK>, - <&gcc GCC_DISP_SF_AXI_CLK>, - <&dispcc DISP_CC_MDSS_AHB_CLK>, - <&dispcc DISP_CC_MDSS_MDP_LUT_CLK>, - <&dispcc DISP_CC_MDSS_MDP_CLK>, - <&dispcc DISP_CC_MDSS_VSYNC_CLK>; - clock-names = "bus", - "nrt_bus", - "iface", - "lut", - "core", - "vsync"; - - assigned-clocks = <&dispcc DISP_CC_MDSS_VSYNC_CLK>; - assigned-clock-rates = <19200000>; - - operating-points-v2 = <&mdp_opp_table>; - power-domains = <&rpmhpd RPMHPD_MMCX>; - - interrupt-parent = <&mdss>; - interrupts = <0>; - - ports { - #address-cells = <1>; - #size-cells = <0>; - - port@0 { - reg = <0>; - dpu_intf1_out: endpoint { - remote-endpoint = <&dsi0_in>; - }; - }; - }; - - mdp_opp_table: opp-table { - compatible = "operating-points-v2"; - - opp-200000000 { - opp-hz = /bits/ 64 <200000000>; - required-opps = <&rpmhpd_opp_low_svs>; - }; - - opp-300000000 { - opp-hz = /bits/ 64 <300000000>; - required-opps = <&rpmhpd_opp_svs>; - }; - - opp-345000000 { - opp-hz = /bits/ 64 <345000000>; - required-opps = <&rpmhpd_opp_svs_l1>; - }; - - opp-460000000 { - opp-hz = /bits/ 64 <460000000>; - required-opps = <&rpmhpd_opp_nom>; - }; - }; - }; -... diff --git a/Documentation/devicetree/bindings/display/msm/qcom,sm8450-dpu.yaml b/Documentation/devicetree/bindings/display/msm/qcom,sm8450-dpu.yaml deleted file mode 100644 index 2a5d3daed0e1..000000000000 --- a/Documentation/devicetree/bindings/display/msm/qcom,sm8450-dpu.yaml +++ /dev/null @@ -1,139 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause -%YAML 1.2 ---- -$id: http://devicetree.org/schemas/display/msm/qcom,sm8450-dpu.yaml# -$schema: http://devicetree.org/meta-schemas/core.yaml# - -title: Qualcomm SM8450 Display DPU - -maintainers: - - Dmitry Baryshkov - -$ref: /schemas/display/msm/dpu-common.yaml# - -properties: - compatible: - const: qcom,sm8450-dpu - - reg: - items: - - description: Address offset and size for mdp register set - - description: Address offset and size for vbif register set - - reg-names: - items: - - const: mdp - - const: vbif - - clocks: - items: - - description: Display hf axi - - description: Display sf axi - - description: Display ahb - - description: Display lut - - description: Display core - - description: Display vsync - - clock-names: - items: - - const: bus - - const: nrt_bus - - const: iface - - const: lut - - const: core - - const: vsync - -required: - - compatible - - reg - - reg-names - - clocks - - clock-names - -unevaluatedProperties: false - -examples: - - | - #include - #include - #include - #include - #include - - display-controller@ae01000 { - compatible = "qcom,sm8450-dpu"; - reg = <0x0ae01000 0x8f000>, - <0x0aeb0000 0x2008>; - reg-names = "mdp", "vbif"; - - clocks = <&gcc GCC_DISP_HF_AXI_CLK>, - <&gcc GCC_DISP_SF_AXI_CLK>, - <&dispcc DISP_CC_MDSS_AHB_CLK>, - <&dispcc DISP_CC_MDSS_MDP_LUT_CLK>, - <&dispcc DISP_CC_MDSS_MDP_CLK>, - <&dispcc DISP_CC_MDSS_VSYNC_CLK>; - clock-names = "bus", - "nrt_bus", - "iface", - "lut", - "core", - "vsync"; - - assigned-clocks = <&dispcc DISP_CC_MDSS_VSYNC_CLK>; - assigned-clock-rates = <19200000>; - - operating-points-v2 = <&mdp_opp_table>; - power-domains = <&rpmhpd RPMHPD_MMCX>; - - interrupt-parent = <&mdss>; - interrupts = <0>; - - ports { - #address-cells = <1>; - #size-cells = <0>; - - port@0 { - reg = <0>; - dpu_intf1_out: endpoint { - remote-endpoint = <&dsi0_in>; - }; - }; - - port@1 { - reg = <1>; - dpu_intf2_out: endpoint { - remote-endpoint = <&dsi1_in>; - }; - }; - }; - - mdp_opp_table: opp-table { - compatible = "operating-points-v2"; - - opp-172000000{ - opp-hz = /bits/ 64 <172000000>; - required-opps = <&rpmhpd_opp_low_svs_d1>; - }; - - opp-200000000 { - opp-hz = /bits/ 64 <200000000>; - required-opps = <&rpmhpd_opp_low_svs>; - }; - - opp-325000000 { - opp-hz = /bits/ 64 <325000000>; - required-opps = <&rpmhpd_opp_svs>; - }; - - opp-375000000 { - opp-hz = /bits/ 64 <375000000>; - required-opps = <&rpmhpd_opp_svs_l1>; - }; - - opp-500000000 { - opp-hz = /bits/ 64 <500000000>; - required-opps = <&rpmhpd_opp_nom>; - }; - }; - }; -... diff --git a/Documentation/devicetree/bindings/display/msm/qcom,sm8550-dpu.yaml b/Documentation/devicetree/bindings/display/msm/qcom,sm8550-dpu.yaml deleted file mode 100644 index 16a541fca66f..000000000000 --- a/Documentation/devicetree/bindings/display/msm/qcom,sm8550-dpu.yaml +++ /dev/null @@ -1,133 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause -%YAML 1.2 ---- -$id: http://devicetree.org/schemas/display/msm/qcom,sm8550-dpu.yaml# -$schema: http://devicetree.org/meta-schemas/core.yaml# - -title: Qualcomm SM8550 Display DPU - -maintainers: - - Neil Armstrong - -$ref: /schemas/display/msm/dpu-common.yaml# - -properties: - compatible: - const: qcom,sm8550-dpu - - reg: - items: - - description: Address offset and size for mdp register set - - description: Address offset and size for vbif register set - - reg-names: - items: - - const: mdp - - const: vbif - - clocks: - items: - - description: Display AHB - - description: Display hf axi - - description: Display MDSS ahb - - description: Display lut - - description: Display core - - description: Display vsync - - clock-names: - items: - - const: bus - - const: nrt_bus - - const: iface - - const: lut - - const: core - - const: vsync - -required: - - compatible - - reg - - reg-names - - clocks - - clock-names - -unevaluatedProperties: false - -examples: - - | - #include - #include - #include - #include - - display-controller@ae01000 { - compatible = "qcom,sm8550-dpu"; - reg = <0x0ae01000 0x8f000>, - <0x0aeb0000 0x2008>; - reg-names = "mdp", "vbif"; - - clocks = <&gcc GCC_DISP_AHB_CLK>, - <&gcc GCC_DISP_HF_AXI_CLK>, - <&dispcc DISP_CC_MDSS_AHB_CLK>, - <&dispcc DISP_CC_MDSS_MDP_LUT_CLK>, - <&dispcc DISP_CC_MDSS_MDP_CLK>, - <&dispcc DISP_CC_MDSS_VSYNC_CLK>; - clock-names = "bus", - "nrt_bus", - "iface", - "lut", - "core", - "vsync"; - - assigned-clocks = <&dispcc DISP_CC_MDSS_VSYNC_CLK>; - assigned-clock-rates = <19200000>; - - operating-points-v2 = <&mdp_opp_table>; - power-domains = <&rpmhpd RPMHPD_MMCX>; - - interrupt-parent = <&mdss>; - interrupts = <0>; - - ports { - #address-cells = <1>; - #size-cells = <0>; - - port@0 { - reg = <0>; - dpu_intf1_out: endpoint { - remote-endpoint = <&dsi0_in>; - }; - }; - - port@1 { - reg = <1>; - dpu_intf2_out: endpoint { - remote-endpoint = <&dsi1_in>; - }; - }; - }; - - mdp_opp_table: opp-table { - compatible = "operating-points-v2"; - - opp-200000000 { - opp-hz = /bits/ 64 <200000000>; - required-opps = <&rpmhpd_opp_low_svs>; - }; - - opp-325000000 { - opp-hz = /bits/ 64 <325000000>; - required-opps = <&rpmhpd_opp_svs>; - }; - - opp-375000000 { - opp-hz = /bits/ 64 <375000000>; - required-opps = <&rpmhpd_opp_svs_l1>; - }; - - opp-514000000 { - opp-hz = /bits/ 64 <514000000>; - required-opps = <&rpmhpd_opp_nom>; - }; - }; - }; -... diff --git a/Documentation/devicetree/bindings/display/msm/qcom,sm8650-dpu.yaml b/Documentation/devicetree/bindings/display/msm/qcom,sm8650-dpu.yaml index c4087cc5abbd..01cf79bd754b 100644 --- a/Documentation/devicetree/bindings/display/msm/qcom,sm8650-dpu.yaml +++ b/Documentation/devicetree/bindings/display/msm/qcom,sm8650-dpu.yaml @@ -14,6 +14,7 @@ $ref: /schemas/display/msm/dpu-common.yaml# properties: compatible: enum: + - qcom,sa8775p-dpu - qcom,sm8650-dpu - qcom,x1e80100-dpu diff --git a/Documentation/devicetree/bindings/display/panel/advantech,idk-2121wr.yaml b/Documentation/devicetree/bindings/display/panel/advantech,idk-2121wr.yaml index 2e8dbdb5a3d5..05ca3b2385f8 100644 --- a/Documentation/devicetree/bindings/display/panel/advantech,idk-2121wr.yaml +++ b/Documentation/devicetree/bindings/display/panel/advantech,idk-2121wr.yaml @@ -20,6 +20,7 @@ description: | dual-lvds-odd-pixels or dual-lvds-even-pixels). allOf: + - $ref: /schemas/display/lvds-dual-ports.yaml# - $ref: panel-common.yaml# properties: @@ -44,22 +45,10 @@ properties: properties: port@0: - $ref: /schemas/graph.yaml#/$defs/port-base - unevaluatedProperties: false - description: The sink for odd pixels. - properties: - dual-lvds-odd-pixels: true - required: - dual-lvds-odd-pixels port@1: - $ref: /schemas/graph.yaml#/$defs/port-base - unevaluatedProperties: false - description: The sink for even pixels. - properties: - dual-lvds-even-pixels: true - required: - dual-lvds-even-pixels @@ -75,7 +64,6 @@ required: - height-mm - data-mapping - panel-timing - - ports examples: - |+ diff --git a/Documentation/devicetree/bindings/display/panel/panel-common.yaml b/Documentation/devicetree/bindings/display/panel/panel-common.yaml index 0a57a31f4f3d..087415753d60 100644 --- a/Documentation/devicetree/bindings/display/panel/panel-common.yaml +++ b/Documentation/devicetree/bindings/display/panel/panel-common.yaml @@ -51,6 +51,14 @@ properties: $ref: /schemas/types.yaml#/definitions/uint32 enum: [0, 90, 180, 270] + flip-horizontal: + description: boolean to flip image horizontally + type: boolean + + flip-vertical: + description: boolean to flip image vertically + type: boolean + # Display Timings panel-timing: description: diff --git a/Documentation/devicetree/bindings/display/panel/panel-lvds.yaml b/Documentation/devicetree/bindings/display/panel/panel-lvds.yaml index 155d8ffa8f6e..5af2d6930075 100644 --- a/Documentation/devicetree/bindings/display/panel/panel-lvds.yaml +++ b/Documentation/devicetree/bindings/display/panel/panel-lvds.yaml @@ -50,6 +50,8 @@ properties: - hannstar,hsd101pww2 # Hydis Technologies 7" WXGA (800x1280) TFT LCD LVDS panel - hydis,hv070wx2-1e0 + # Jenson Display BL-JT60050-01A 7" WSVGA (1024x600) color TFT LCD LVDS panel + - jenson,bl-jt60050-01a - tbs,a711-panel - const: panel-lvds diff --git a/Documentation/devicetree/bindings/display/panel/panel-simple-lvds-dual-ports.yaml b/Documentation/devicetree/bindings/display/panel/panel-simple-lvds-dual-ports.yaml index 10ed4b57232b..e80fc7006984 100644 --- a/Documentation/devicetree/bindings/display/panel/panel-simple-lvds-dual-ports.yaml +++ b/Documentation/devicetree/bindings/display/panel/panel-simple-lvds-dual-ports.yaml @@ -22,6 +22,7 @@ description: | If the panel is more advanced a dedicated binding file is required. allOf: + - $ref: /schemas/display/lvds-dual-ports.yaml# - $ref: panel-common.yaml# properties: @@ -55,28 +56,10 @@ properties: properties: port@0: - $ref: /schemas/graph.yaml#/$defs/port-base - unevaluatedProperties: false - description: The first sink port. - - properties: - dual-lvds-odd-pixels: - type: boolean - description: The first sink port for odd pixels. - required: - dual-lvds-odd-pixels port@1: - $ref: /schemas/graph.yaml#/$defs/port-base - unevaluatedProperties: false - description: The second sink port. - - properties: - dual-lvds-even-pixels: - type: boolean - description: The second sink port for even pixels. - required: - dual-lvds-even-pixels @@ -88,7 +71,6 @@ unevaluatedProperties: false required: - compatible - - ports - power-supply examples: diff --git a/Documentation/devicetree/bindings/display/panel/panel-simple.yaml b/Documentation/devicetree/bindings/display/panel/panel-simple.yaml index b89e39790579..18b63f356bb4 100644 --- a/Documentation/devicetree/bindings/display/panel/panel-simple.yaml +++ b/Documentation/devicetree/bindings/display/panel/panel-simple.yaml @@ -200,6 +200,8 @@ properties: - logictechno,lttd800480070-l2rt # Logic Technologies LTTD800480070-L6WH-RT 7” 800x480 TFT Resistive Touch Module - logictechno,lttd800480070-l6wh-rt + # Microchip AC69T88A 5" 800X480 LVDS interface TFT LCD Panel + - microchip,ac69t88a # Mitsubishi "AA070MC01 7.0" WVGA TFT LCD panel - mitsubishi,aa070mc01-ca1 # Mitsubishi AA084XE01 8.4" XGA TFT LCD panel diff --git a/Documentation/devicetree/bindings/display/panel/samsung,ams581vf01.yaml b/Documentation/devicetree/bindings/display/panel/samsung,ams581vf01.yaml new file mode 100644 index 000000000000..70dff9c0ef2b --- /dev/null +++ b/Documentation/devicetree/bindings/display/panel/samsung,ams581vf01.yaml @@ -0,0 +1,79 @@ +# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/display/panel/samsung,ams581vf01.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: Samsung AMS581VF01 SOFEF01-based 5.81" 1080x2340 MIPI-DSI Panel + +maintainers: + - Danila Tikhonov + +description: + The Samsung AMS581VF01 is a 5.81 inch 1080x2340 MIPI-DSI CMD mode OLED panel. + +allOf: + - $ref: panel-common.yaml# + +properties: + compatible: + const: samsung,ams581vf01 + + reg: + maxItems: 1 + + vdd3p3-supply: + description: 3.3V source voltage rail + + vddio-supply: + description: I/O source voltage rail + + vsn-supply: + description: Negative source voltage rail + + vsp-supply: + description: Positive source voltage rail + + reset-gpios: true + port: true + +required: + - compatible + - reg + - vdd3p3-supply + - vddio-supply + - vsn-supply + - vsp-supply + - reset-gpios + - port + +additionalProperties: false + +examples: + - | + #include + + dsi { + #address-cells = <1>; + #size-cells = <0>; + + panel@0 { + compatible = "samsung,ams581vf01"; + reg = <0>; + + vdd3p3-supply = <&vreg_l7c_3p0>; + vddio-supply = <&vreg_l13a_1p8>; + vsn-supply = <&vreg_ibb>; + vsp-supply = <&vreg_lab>; + + reset-gpios = <&pm6150l_gpios 9 GPIO_ACTIVE_LOW>; + + port { + panel_in: endpoint { + remote-endpoint = <&mdss_dsi0_out>; + }; + }; + }; + }; + +... diff --git a/Documentation/devicetree/bindings/display/panel/samsung,ams639rq08.yaml b/Documentation/devicetree/bindings/display/panel/samsung,ams639rq08.yaml new file mode 100644 index 000000000000..f5b6ecb96f99 --- /dev/null +++ b/Documentation/devicetree/bindings/display/panel/samsung,ams639rq08.yaml @@ -0,0 +1,80 @@ +# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/display/panel/samsung,ams639rq08.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: Samsung AMS639RQ08 EA8076-based 6.39" 1080x2340 MIPI-DSI Panel + +maintainers: + - Danila Tikhonov + - Jens Reidel + +description: + The Samsung AMS639RQ08 is a 6.39 inch 1080x2340 MIPI-DSI CMD mode AMOLED panel. + +allOf: + - $ref: panel-common.yaml# + +properties: + compatible: + const: samsung,ams639rq08 + + reg: + maxItems: 1 + + vdd3p3-supply: + description: 3.3V source voltage rail + + vddio-supply: + description: I/O source voltage rail + + vsn-supply: + description: Negative source voltage rail + + vsp-supply: + description: Positive source voltage rail + + reset-gpios: true + port: true + +required: + - compatible + - reg + - vdd3p3-supply + - vddio-supply + - vsn-supply + - vsp-supply + - reset-gpios + - port + +additionalProperties: false + +examples: + - | + #include + + dsi { + #address-cells = <1>; + #size-cells = <0>; + + panel@0 { + compatible = "samsung,ams639rq08"; + reg = <0>; + + vdd3p3-supply = <&vreg_l18a_2p8>; + vddio-supply = <&vreg_l13a_1p8>; + vsn-supply = <&vreg_ibb>; + vsp-supply = <&vreg_lab>; + + reset-gpios = <&pm6150l_gpios 9 GPIO_ACTIVE_LOW>; + + port { + panel_in: endpoint { + remote-endpoint = <&mdss_dsi0_out>; + }; + }; + }; + }; + +... diff --git a/Documentation/devicetree/bindings/display/panel/samsung,s6e3ha8.yaml b/Documentation/devicetree/bindings/display/panel/samsung,s6e3ha8.yaml new file mode 100644 index 000000000000..05a78429aaea --- /dev/null +++ b/Documentation/devicetree/bindings/display/panel/samsung,s6e3ha8.yaml @@ -0,0 +1,75 @@ +# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/display/panel/samsung,s6e3ha8.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: Samsung s6e3ha8 AMOLED DSI panel + +description: The s6e3ha8 is a 1440x2960 DPI display panel from Samsung Mobile + Displays (SMD). + +maintainers: + - Dzmitry Sankouski + +allOf: + - $ref: panel-common.yaml# + +properties: + compatible: + const: samsung,s6e3ha8 + + reg: + maxItems: 1 + + reset-gpios: true + + port: true + + vdd3-supply: + description: VDD regulator + + vci-supply: + description: VCI regulator + + vddr-supply: + description: VDDR regulator + +required: + - compatible + - reset-gpios + - vdd3-supply + - vci-supply + - vddr-supply + +unevaluatedProperties: false + +examples: + - | + #include + + dsi { + #address-cells = <1>; + #size-cells = <0>; + + panel@0 { + compatible = "samsung,s6e3ha8"; + reg = <0>; + vci-supply = <&s2dos05_ldo4>; + vddr-supply = <&s2dos05_buck1>; + vdd3-supply = <&s2dos05_ldo1>; + te-gpios = <&tlmm 10 GPIO_ACTIVE_HIGH>; + reset-gpios = <&tlmm 6 GPIO_ACTIVE_HIGH>; + pinctrl-0 = <&sde_dsi_active &sde_te_active_sleep>; + pinctrl-1 = <&sde_dsi_suspend &sde_te_active_sleep>; + pinctrl-names = "default", "sleep"; + + port { + panel_in: endpoint { + remote-endpoint = <&mdss_dsi0_out>; + }; + }; + }; + }; + +... diff --git a/Documentation/devicetree/bindings/display/panel/samsung,s6e88a0-ams427ap24.yaml b/Documentation/devicetree/bindings/display/panel/samsung,s6e88a0-ams427ap24.yaml new file mode 100644 index 000000000000..db284ba5be20 --- /dev/null +++ b/Documentation/devicetree/bindings/display/panel/samsung,s6e88a0-ams427ap24.yaml @@ -0,0 +1,65 @@ +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/display/panel/samsung,s6e88a0-ams427ap24.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: Samsung AMS427AP24 panel with S6E88A0 controller + +maintainers: + - Jakob Hauser + +allOf: + - $ref: panel-common.yaml# + +properties: + compatible: + const: samsung,s6e88a0-ams427ap24 + + reg: + maxItems: 1 + + port: true + reset-gpios: true + flip-horizontal: true + + vdd3-supply: + description: core voltage supply + + vci-supply: + description: voltage supply for analog circuits + +required: + - compatible + - reg + - port + - reset-gpios + - vdd3-supply + - vci-supply + +additionalProperties: false + +examples: + - | + #include + + dsi { + #address-cells = <1>; + #size-cells = <0>; + + panel@0 { + compatible = "samsung,s6e88a0-ams427ap24"; + reg = <0>; + + vdd3-supply = <&pm8916_l17>; + vci-supply = <&pm8916_l6>; + reset-gpios = <&tlmm 25 GPIO_ACTIVE_LOW>; + flip-horizontal; + + port { + panel_in: endpoint { + remote-endpoint = <&mdss_dsi0_out>; + }; + }; + }; + }; diff --git a/Documentation/devicetree/bindings/display/panel/samsung,s6e8aa0.yaml b/Documentation/devicetree/bindings/display/panel/samsung,s6e8aa0.yaml index 4601fa460680..19c8cc83db97 100644 --- a/Documentation/devicetree/bindings/display/panel/samsung,s6e8aa0.yaml +++ b/Documentation/devicetree/bindings/display/panel/samsung,s6e8aa0.yaml @@ -21,6 +21,8 @@ properties: reset-gpios: true display-timings: true + flip-horizontal: true + flip-vertical: true vdd3-supply: description: core voltage supply @@ -46,14 +48,6 @@ properties: panel-height-mm: description: physical panel height [mm] - flip-horizontal: - description: boolean to flip image horizontally - type: boolean - - flip-vertical: - description: boolean to flip image vertically - type: boolean - required: - compatible - reg diff --git a/Documentation/devicetree/bindings/display/rockchip/rockchip,rk3588-dw-hdmi-qp.yaml b/Documentation/devicetree/bindings/display/rockchip/rockchip,rk3588-dw-hdmi-qp.yaml new file mode 100644 index 000000000000..d8e761865f27 --- /dev/null +++ b/Documentation/devicetree/bindings/display/rockchip/rockchip,rk3588-dw-hdmi-qp.yaml @@ -0,0 +1,188 @@ +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/display/rockchip/rockchip,rk3588-dw-hdmi-qp.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: Rockchip DW HDMI QP TX Encoder + +maintainers: + - Cristian Ciocaltea + +description: | + Rockchip RK3588 SoC integrates the Synopsys DesignWare HDMI QP TX controller + IP and a HDMI/eDP TX Combo PHY based on a Samsung IP block, providing the + following features, among others: + + * Fixed Rate Link (FRL) + * Display Stream Compression (DSC) + * 4K@120Hz and 8K@60Hz video modes + * Variable Refresh Rate (VRR) including Quick Media Switching (QMS) + * Fast Vactive (FVA) + * SCDC I2C DDC access + * Multi-stream audio + * Enhanced Audio Return Channel (EARC) + +allOf: + - $ref: /schemas/sound/dai-common.yaml# + +properties: + compatible: + enum: + - rockchip,rk3588-dw-hdmi-qp + + reg: + maxItems: 1 + + clocks: + items: + - description: Peripheral/APB bus clock + - description: EARC RX biphase clock + - description: Reference clock + - description: Audio interface clock + - description: TMDS/FRL link clock + - description: Video datapath clock + + clock-names: + items: + - const: pclk + - const: earc + - const: ref + - const: aud + - const: hdp + - const: hclk_vo1 + + interrupts: + items: + - description: AVP Unit interrupt + - description: CEC interrupt + - description: eARC RX interrupt + - description: Main Unit interrupt + - description: HPD interrupt + + interrupt-names: + items: + - const: avp + - const: cec + - const: earc + - const: main + - const: hpd + + phys: + maxItems: 1 + description: The HDMI/eDP PHY + + ports: + $ref: /schemas/graph.yaml#/properties/ports + + properties: + port@0: + $ref: /schemas/graph.yaml#/properties/port + description: Video port for RGB/YUV input. + + port@1: + $ref: /schemas/graph.yaml#/properties/port + description: Video port for HDMI/eDP output. + + required: + - port@0 + - port@1 + + power-domains: + maxItems: 1 + + resets: + maxItems: 2 + + reset-names: + items: + - const: ref + - const: hdp + + "#sound-dai-cells": + const: 0 + + rockchip,grf: + $ref: /schemas/types.yaml#/definitions/phandle + description: + Some HDMI QP related data is accessed through SYS GRF regs. + + rockchip,vo-grf: + $ref: /schemas/types.yaml#/definitions/phandle + description: + Additional HDMI QP related data is accessed through VO GRF regs. + +required: + - compatible + - reg + - clocks + - clock-names + - interrupts + - interrupt-names + - phys + - ports + - resets + - reset-names + - rockchip,grf + - rockchip,vo-grf + +unevaluatedProperties: false + +examples: + - | + #include + #include + #include + #include + #include + + soc { + #address-cells = <2>; + #size-cells = <2>; + + hdmi@fde80000 { + compatible = "rockchip,rk3588-dw-hdmi-qp"; + reg = <0x0 0xfde80000 0x0 0x20000>; + clocks = <&cru PCLK_HDMITX0>, + <&cru CLK_HDMITX0_EARC>, + <&cru CLK_HDMITX0_REF>, + <&cru MCLK_I2S5_8CH_TX>, + <&cru CLK_HDMIHDP0>, + <&cru HCLK_VO1>; + clock-names = "pclk", "earc", "ref", "aud", "hdp", "hclk_vo1"; + interrupts = , + , + , + , + ; + interrupt-names = "avp", "cec", "earc", "main", "hpd"; + phys = <&hdptxphy_hdmi0>; + power-domains = <&power RK3588_PD_VO1>; + resets = <&cru SRST_HDMITX0_REF>, <&cru SRST_HDMIHDP0>; + reset-names = "ref", "hdp"; + rockchip,grf = <&sys_grf>; + rockchip,vo-grf = <&vo1_grf>; + #sound-dai-cells = <0>; + + ports { + #address-cells = <1>; + #size-cells = <0>; + + port@0 { + reg = <0>; + + hdmi0_in_vp0: endpoint { + remote-endpoint = <&vp0_out_hdmi0>; + }; + }; + + port@1 { + reg = <1>; + + hdmi0_out_con0: endpoint { + remote-endpoint = <&hdmi_con0_in>; + }; + }; + }; + }; + }; diff --git a/Documentation/devicetree/bindings/display/samsung/samsung,exynos7-decon.yaml b/Documentation/devicetree/bindings/display/samsung/samsung,exynos7-decon.yaml index 992c23ca7a4e..53916e4c95d8 100644 --- a/Documentation/devicetree/bindings/display/samsung/samsung,exynos7-decon.yaml +++ b/Documentation/devicetree/bindings/display/samsung/samsung,exynos7-decon.yaml @@ -19,7 +19,9 @@ description: | properties: compatible: - const: samsung,exynos7-decon + enum: + - samsung,exynos7-decon + - samsung,exynos7870-decon clocks: maxItems: 4 diff --git a/Documentation/devicetree/bindings/display/sharp,ls010b7dh04.yaml b/Documentation/devicetree/bindings/display/sharp,ls010b7dh04.yaml new file mode 100644 index 000000000000..8097f091c2a5 --- /dev/null +++ b/Documentation/devicetree/bindings/display/sharp,ls010b7dh04.yaml @@ -0,0 +1,92 @@ +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/display/sharp,ls010b7dh04.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: Sharp Memory LCD panels + +maintainers: + - Alex Lanzano + +description: + Sharp Memory LCDs are a series of monochrome displays that operate over + a SPI bus. The displays require a signal (VCOM) to be generated to prevent + DC bias build up resulting in pixels being unable to change. Three modes + can be used to provide the VCOM signal ("software", "external", "pwm"). + +properties: + compatible: + enum: + - sharp,ls010b7dh04 + - sharp,ls011b7dh03 + - sharp,ls012b7dd01 + - sharp,ls013b7dh03 + - sharp,ls013b7dh05 + - sharp,ls018b7dh02 + - sharp,ls027b7dh01 + - sharp,ls027b7dh01a + - sharp,ls032b7dd02 + - sharp,ls044q7dh01 + + reg: + maxItems: 1 + + spi-max-frequency: + maximum: 2000000 + + sharp,vcom-mode: + $ref: /schemas/types.yaml#/definitions/string + description: | + software - This mode relies on a software operation to send a + "maintain display" message to the display, toggling the vcom + bit on and off with each message + + external - This mode relies on an external clock to generate + the signal on the EXTCOMM pin + + pwm - This mode relies on a pwm device to generate the signal + on the EXTCOMM pin + + enum: [software, external, pwm] + + enable-gpios: true + + pwms: + maxItems: 1 + description: External VCOM signal + +required: + - compatible + - reg + - sharp,vcom-mode + +allOf: + - $ref: panel/panel-common.yaml# + - $ref: /schemas/spi/spi-peripheral-props.yaml# + + - if: + properties: + sharp,vcom-mode: + const: pwm + then: + required: + - pwms + +unevaluatedProperties: false + +examples: + - | + spi { + #address-cells = <1>; + #size-cells = <0>; + + display@0 { + compatible = "sharp,ls013b7dh03"; + reg = <0>; + spi-cs-high; + spi-max-frequency = <1000000>; + sharp,vcom-mode = "software"; + }; + }; +... diff --git a/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.yaml b/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.yaml index 278399adc550..735c7f06c24e 100644 --- a/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.yaml +++ b/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.yaml @@ -26,6 +26,7 @@ properties: - renesas,r9a07g054-mali - rockchip,px30-mali - rockchip,rk3568-mali + - rockchip,rk3576-mali - const: arm,mali-bifrost # Mali Bifrost GPU model/revision is fully discoverable - items: - enum: diff --git a/Documentation/devicetree/bindings/vendor-prefixes.yaml b/Documentation/devicetree/bindings/vendor-prefixes.yaml index b0531b935b3e..715663b450f8 100644 --- a/Documentation/devicetree/bindings/vendor-prefixes.yaml +++ b/Documentation/devicetree/bindings/vendor-prefixes.yaml @@ -756,6 +756,8 @@ patternProperties: description: Japan Display Inc. "^jedec,.*": description: JEDEC Solid State Technology Association + "^jenson,.*": + description: Jenson Display Co. Ltd. "^jesurun,.*": description: Shenzhen Jesurun Electronics Business Dept. "^jethome,.*": diff --git a/Documentation/gpu/amdgpu/display/dc-arch-overview.svg b/Documentation/gpu/amdgpu/display/dc-arch-overview.svg new file mode 100644 index 000000000000..23394931cf26 --- /dev/null +++ b/Documentation/gpu/amdgpu/display/dc-arch-overview.svg @@ -0,0 +1,731 @@ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + image/svg+xml + + + + + + + + + + Board/Platform + SoC + Component + DRAM + + dc_plane + + + + dc_plane + + + + + DC + + + + + dc_link + + + + + dc_link + + + + + + + dc_link + + + + + + + + + + + DCN + SoC + Board/Platform + Display + Connector + Connector + + diff --git a/Documentation/gpu/amdgpu/display/dc-components.svg b/Documentation/gpu/amdgpu/display/dc-components.svg new file mode 100644 index 000000000000..f84bb2a57c05 --- /dev/null +++ b/Documentation/gpu/amdgpu/display/dc-components.svg @@ -0,0 +1,732 @@ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + image/svg+xml + + + + + + + + + Core + + + + + + + + + + Display Core API (dc/dc.h) + Link + Hardware Sequencer API(dc/inc/hw_sequence.h) + Hardware Sequencer + Block Level API (dc/inc/hw) + + + DCHUB + + + + HUBP + + + + DPP + + + + MPC + + + + ... + + Hardware Registers + DMUBBlock + DMUB Hardware API(dmub/dmub_srv.h) + DMUB Service + DMUB Service API(dc/dc_dmub_srv.h) + + diff --git a/Documentation/gpu/amdgpu/display/dc-debug.rst b/Documentation/gpu/amdgpu/display/dc-debug.rst index 817631b1dbf3..013f63b271f3 100644 --- a/Documentation/gpu/amdgpu/display/dc-debug.rst +++ b/Documentation/gpu/amdgpu/display/dc-debug.rst @@ -2,6 +2,181 @@ Display Core Debug tools ======================== +In this section, you will find helpful information on debugging the amdgpu +driver from the display perspective. This page introduces debug mechanisms and +procedures to help you identify if some issues are related to display code. + +Narrow down display issues +========================== + +Since the display is the driver's visual component, it is common to see users +reporting issues as a display when another component causes the problem. This +section equips users to determine if a specific issue was caused by the display +component or another part of the driver. + +DC dmesg important messages +--------------------------- + +The dmesg log is the first source of information to be checked, and amdgpu +takes advantage of this feature by logging some valuable information. When +looking for the issues associated with amdgpu, remember that each component of +the driver (e.g., smu, PSP, dm, etc.) is loaded one by one, and this +information can be found in the dmesg log. In this sense, look for the part of +the log that looks like the below log snippet:: + + [ 4.254295] [drm] initializing kernel modesetting (IP DISCOVERY 0x1002:0x744C 0x1002:0x0E3B 0xC8). + [ 4.254718] [drm] register mmio base: 0xFCB00000 + [ 4.254918] [drm] register mmio size: 1048576 + [ 4.260095] [drm] add ip block number 0 + [ 4.260318] [drm] add ip block number 1 + [ 4.260510] [drm] add ip block number 2 + [ 4.260696] [drm] add ip block number 3 + [ 4.260878] [drm] add ip block number 4 + [ 4.261057] [drm] add ip block number 5 + [ 4.261231] [drm] add ip block number 6 + [ 4.261402] [drm] add ip block number 7 + [ 4.261568] [drm] add ip block number 8 + [ 4.261729] [drm] add ip block number 9 + [ 4.261887] [drm] add ip block number 10 + +From the above example, you can see the line that reports that ``, +(**Display Manager**), was loaded, which means that display can be part of the +issue. If you do not see that line, something else might have failed before +amdgpu loads the display component, indicating that we don't have a +display issue. + +After you identified that the DM was loaded correctly, you can check for the +display version of the hardware in use, which can be retrieved from the dmesg +log with the command:: + + dmesg | grep -i 'display core' + +This command shows a message that looks like this:: + + [ 4.655828] [drm] Display Core v3.2.285 initialized on DCN 3.2 + +This message has two key pieces of information: + +* **The DC version (e.g., v3.2.285)**: Display developers release a new DC version + every week, and this information can be advantageous in a situation where a + user/developer must find a good point versus a bad point based on a tested + version of the display code. Remember from page :ref:`Display Core `, + that every week the new patches for display are heavily tested with IGT and + manual tests. +* **The DCN version (e.g., DCN 3.2)**: The DCN block is associated with the + hardware generation, and the DCN version conveys the hardware generation that + the driver is currently running. This information helps to narrow down the + code debug area since each DCN version has its files in the DC folder per DCN + component (from the example, the developer might want to focus on + files/folders/functions/structs with the dcn32 label might be executed). + However, keep in mind that DC reuses code across different DCN versions; for + example, it is expected to have some callbacks set in one DCN that are the same + as those from another DCN. In summary, use the DCN version just as a guide. + +From the dmesg file, it is also possible to get the ATOM bios code by using:: + + dmesg | grep -i 'ATOM BIOS' + +Which generates an output that looks like this:: + + [ 4.274534] amdgpu: ATOM BIOS: 113-D7020100-102 + +This type of information is useful to be reported. + +Avoid loading display core +-------------------------- + +Sometimes, it might be hard to figure out which part of the driver is causing +the issue; if you suspect that the display is not part of the problem and your +bug scenario is simple (e.g., some desktop configuration) you can try to remove +the display component from the equation. First, you need to identify `dm` ID +from the dmesg log; for example, search for the following log:: + + [ 4.254295] [drm] initializing kernel modesetting (IP DISCOVERY 0x1002:0x744C 0x1002:0x0E3B 0xC8). + [..] + [ 4.260095] [drm] add ip block number 0 + [ 4.260318] [drm] add ip block number 1 + [..] + [ 4.261057] [drm] add ip block number 5 + +Notice from the above example that the `dm` id is 5 for this specific hardware. +Next, you need to run the following binary operation to identify the IP block +mask:: + + 0xffffffff & ~(1 << [DM ID]) + +From our example the IP mask is:: + + 0xffffffff & ~(1 << 5) = 0xffffffdf + +Finally, to disable DC, you just need to set the below parameter in your +bootloader:: + + amdgpu.ip_block_mask = 0xffffffdf + +If you can boot your system with the DC disabled and still see the issue, it +means you can rule DC out of the equation. However, if the bug disappears, you +still need to consider the DC part of the problem and keep narrowing down the +issue. In some scenarios, disabling DC is impossible since it might be +necessary to use the display component to reproduce the issue (e.g., play a +game). + +**Note: This will probably lead to the absence of a display output.** + +Display flickering +------------------ + +Display flickering might have multiple causes; one is the lack of proper power +to the GPU or problems in the DPM switches. A good first generic verification +is to set the GPU to use high voltage:: + + bash -c "echo high > /sys/class/drm/card0/device/power_dpm_force_performance_level" + +The above command sets the GPU/APU to use the maximum power allowed which +disables DPM switches. If forcing DPM levels high does not fix the issue, it +is less likely that the issue is related to power management. If the issue +disappears, there is a good chance that other components might be involved, and +the display should not be ignored since this could be a DPM issues. From the +display side, if the power increase fixes the issue, it is worth debugging the +clock configuration and the pipe split police used in the specific +configuration. + +Display artifacts +----------------- + +Users may see some screen artifacts that can be categorized into two different +types: localized artifacts and general artifacts. The localized artifacts +happen in some specific areas, such as around the UI window corners; if you see +this type of issue, there is a considerable chance that you have a userspace +problem, likely Mesa or similar. The general artifacts usually happen on the +entire screen. They might be caused by a misconfiguration at the driver level +of the display parameters, but the userspace might also cause this issue. One +way to identify the source of the problem is to take a screenshot or make a +desktop video capture when the problem happens; after checking the +screenshot/video recording, if you don't see any of the artifacts, it means +that the issue is likely on the the driver side. If you can still see the +problem in the data collected, it is an issue that probably happened during +rendering, and the display code just got the framebuffer already corrupted. + +Disabling/Enabling specific features +==================================== + +DC has a struct named `dc_debug_options`, which is statically initialized by +all DCE/DCN components based on the specific hardware characteristic. This +structure usually facilitates the bring-up phase since developers can start +with many disabled features and enable them individually. This is also an +important debug feature since users can change it when debugging specific +issues. + +For example, dGPU users sometimes see a problem where a horizontal fillet of +flickering happens in some specific part of the screen. This could be an +indication of Sub-Viewport issues; after the users identified the target DCN, +they can set the `force_disable_subvp` field to true in the statically +initialized version of `dc_debug_options` to see if the issue gets fixed. Along +the same lines, users/developers can also try to turn off `fams2_config` and +`enable_single_display_2to1_odm_policy`. In summary, the `dc_debug_options` is +an interesting form for identifying the problem. + DC Visual Confirmation ====================== @@ -76,6 +251,18 @@ change in real-time by using something like:: When reporting a bug related to DC, consider attaching this log before and after you reproduce the bug. +Collect Firmware information +============================ + +When reporting issues, it is important to have the firmware information since +it can be helpful for debugging purposes. To get all the firmware information, +use the command:: + + cat /sys/kernel/debug/dri/0/amdgpu_firmware_info + +From the display perspective, pay attention to the firmware of the DMCU and +DMCUB. + DMUB Firmware Debug =================== diff --git a/Documentation/gpu/amdgpu/display/dcn-blocks.rst b/Documentation/gpu/amdgpu/display/dcn-blocks.rst index 5e34366f6dbe..756957128dad 100644 --- a/Documentation/gpu/amdgpu/display/dcn-blocks.rst +++ b/Documentation/gpu/amdgpu/display/dcn-blocks.rst @@ -1,3 +1,5 @@ +.. _dcn_blocks: + ========== DCN Blocks ========== diff --git a/Documentation/gpu/amdgpu/display/dcn-overview.rst b/Documentation/gpu/amdgpu/display/dcn-overview.rst index 9fea6500448b..eb54a6802e04 100644 --- a/Documentation/gpu/amdgpu/display/dcn-overview.rst +++ b/Documentation/gpu/amdgpu/display/dcn-overview.rst @@ -1,3 +1,5 @@ +.. _dcn_overview: + ======================= Display Core Next (DCN) ======================= diff --git a/Documentation/gpu/amdgpu/display/index.rst b/Documentation/gpu/amdgpu/display/index.rst index f0c342e00a39..bd2d797c123e 100644 --- a/Documentation/gpu/amdgpu/display/index.rst +++ b/Documentation/gpu/amdgpu/display/index.rst @@ -90,6 +90,7 @@ table of content: display-manager.rst dcn-overview.rst dcn-blocks.rst + programming-model-dcn.rst mpo-overview.rst dc-debug.rst display-contributing.rst diff --git a/Documentation/gpu/amdgpu/display/programming-model-dcn.rst b/Documentation/gpu/amdgpu/display/programming-model-dcn.rst new file mode 100644 index 000000000000..c1b48d49fb0b --- /dev/null +++ b/Documentation/gpu/amdgpu/display/programming-model-dcn.rst @@ -0,0 +1,162 @@ +==================== +DC Programming Model +==================== + +In the :ref:`Display Core Next (DCN) ` and :ref:`DCN Block +` pages, you learned about the hardware components and how they +interact with each other. On this page, the focus is shifted to the display +code architecture. Hence, it is reasonable to remind the reader that the code +in DC is shared with other OSes; for this reason, DC provides a set of +abstractions and operations to connect different APIs with the hardware +configuration. See DC as a service available for a Display Manager (amdgpu_dm) +to access and configure DCN/DCE hardware (DCE is also part of DC, but for +simplicity's sake, this documentation only examines DCN). + +.. note:: + For this page, we will use the term GPU to refers to dGPU and APU. + +Overview +======== + +From the display hardware perspective, it is plausible to expect that if a +problem is well-defined, it will probably be implemented at the hardware level. +On the other hand, when there are multiple ways of achieving something without +a very well-defined scope, the solution is usually implemented as a policy at +the DC level. In other words, some policies are defined in the DC core +(resource management, power optimization, image quality, etc.), and the others +implemented in hardware are enabled via DC configuration. + +In terms of hardware management, DCN has multiple instances of the same block +(e.g., HUBP, DPP, MPC, etc), and during the driver execution, it might be +necessary to use some of these instances. The core has policies in place for +handling those instances. Regarding resource management, the DC objective is +quite simple: minimize the hardware shuffle when the driver performs some +actions. When the state changes from A to B, the transition is considered +easier to maneuver if the hardware resource is still used for the same set of +driver objects. Usually, adding and removing a resource to a `pipe_ctx` (more +details below) is not a problem; however, moving a resource from one `pipe_ctx` +to another should be avoided. + +Another area of influence for DC is power optimization, which has a myriad of +arrangement possibilities. In some way, just displaying an image via DCN should +be relatively straightforward; however, showing it with the best power +footprint is more desirable, but it has many associated challenges. +Unfortunately, there is no straight-forward analytic way to determine if a +configuration is the best one for the context due to the enormous variety of +variables related to this problem (e.g., many different DCN/DCE hardware +versions, different displays configurations, etc.) for this reason DC +implements a dedicated library for trying some configuration and verify if it +is possible to support it or not. This type of policy is extremely complex to +create and maintain, and amdgpu driver relies on Display Mode Library (DML) to +generate the best decisions. + +In summary, DC must deal with the complexity of handling multiple scenarios and +determine policies to manage them. All of the above information is conveyed to +give the reader some idea about the complexity of driving a display from the +driver's perspective. This page hopes to allow the reader to better navigate +over the amdgpu display code. + +Display Driver Architecture Overview +==================================== + +The diagram below provides an overview of the display driver architecture; +notice it illustrates the software layers adopted by DC: + +.. kernel-figure:: dc-components.svg + +The first layer of the diagram is the high-level DC API represented by the +`dc.h` file; below it are two big blocks represented by Core and Link. Next is +the hardware configuration block; the main file describing it is +the`hw_sequencer.h`, where the implementation of the callbacks can be found in +the hardware sequencer folder. Almost at the end, you can see the block level +API (`dc/inc/hw`), which represents each DCN low-level block, such as HUBP, +DPP, MPC, OPTC, etc. Notice on the left side of the diagram that we have a +different set of layers representing the interaction with the DMUB +microcontroller. + +Basic Objects +------------- + +The below diagram outlines the basic display objects. In particular, pay +attention to the names in the boxes since they represent a data structure in +the driver: + +.. kernel-figure:: dc-arch-overview.svg + +Let's start with the central block in the image, `dc`. The `dc` struct is +initialized per GPU; for example, one GPU has one `dc` instance, two GPUs have +two `dc` instances, and so forth. In other words we have one 'dc' per 'amdgpu' +instance. In some ways, this object behaves like the `Singleton` pattern. + +After the `dc` block in the diagram, you can see the `dc_link` component, which +is a low-level abstraction for the connector. One interesting aspect of the +image is that connectors are not part of the DCN block; they are defined by the +platform/board and not by the SoC. The `dc_link` struct is the high-level data +container with information such as connected sinks, connection status, signal +types, etc. After `dc_link`, there is the `dc_sink`, which is the object that +represents the connected display. + +.. note:: + For historical reasons, we used the name `dc_link`, which gives the + wrong impression that this abstraction only deals with physical connections + that the developer can easily manipulate. However, this also covers + conections like eDP or cases where the output is connected to other devices. + +There are two structs that are not represented in the diagram since they were +elaborated in the DCN overview page (check the DCN block diagram :ref:`Display +Core Next (DCN) `); still, it is worth bringing back for this +overview which is `dc_stream` and `dc_state`. The `dc_stream` stores many +properties associated with the data transmission, but most importantly, it +represents the data flow from the connector to the display. Next we have +`dc_state`, which represents the logic state within the hardware at the moment; +`dc_state` is composed of `dc_stream` and `dc_plane`. The `dc_stream` is the DC +version of `drm_crtc` and represents the post-blending pipeline. + +Speaking of the `dc_plane` data structure (first part of the diagram), you can +think about it as an abstraction similar to `drm_plane` that represents the +pre-blending portion of the pipeline. This image was probably processed by GFX +and is ready to be composited under a `dc_stream`. Normally, the driver may +have one or more `dc_plane` connected to the same `dc_stream`, which defines a +composition at the DC level. + +Basic Operations +---------------- + +Now that we have covered the basic objects, it is time to examine some of the +basic hardware/software operations. Let's start with the `dc_create()` +function, which directly works with the `dc` data struct; this function behaves +like a constructor responsible for the basic software initialization and +preparing for enabling other parts of the API. It is important to highlight +that this operation does not touch any hardware configuration; it is only a +software initialization. + +Next, we have the `dc_hardware_init()`, which also relies on the `dc` data +struct. Its main function is to put the hardware in a valid state. It is worth +highlighting that the hardware might initialize in an unknown state, and it is +a requirement to put it in a valid state; this function has multiple callbacks +for the hardware-specific initialization, whereas `dc_hardware_init` does the +hardware initialization and is the first point where we touch hardware. + +The `dc_get_link_at_index` is an operation that depends on the `dc_link` data +structure. This function retrieves and enumerates all the `dc_links` available +on the device; this is required since this information is not part of the SoC +definition but depends on the board configuration. As soon as the `dc_link` is +initialized, it is useful to figure out if any of them are already connected to +the display by using the `dc_link_detect()` function. After the driver figures +out if any display is connected to the device, the challenging phase starts: +configuring the monitor to show something. Nonetheless, dealing with the ideal +configuration is not a DC task since this is the Display Manager (`amdgpu_dm`) +responsibility which in turn is responsible for dealing with the atomic +commits. The only interface DC provides to the configuration phase is the +function `dc_validate_with_context` that receives the configuration information +and, based on that, validates whether the hardware can support it or not. It is +important to add that even if the display supports some specific configuration, +it does not mean the DCN hardware can support it. + +After the DM and DC agree upon the configuration, the stream configuration +phase starts. This task activates one or more `dc_stream` at this phase, and in +the best-case scenario, you might be able to turn the display on with a black +screen (it does not show anything yet since it does not have any plane +associated with it). The final step would be to call the +`dc_update_planes_and_stream,` which will add or remove planes. + diff --git a/Documentation/gpu/amdgpu/index.rst b/Documentation/gpu/amdgpu/index.rst index 847e04924030..302d039928ee 100644 --- a/Documentation/gpu/amdgpu/index.rst +++ b/Documentation/gpu/amdgpu/index.rst @@ -16,4 +16,5 @@ Next (GCN), Radeon DNA (RDNA), and Compute DNA (CDNA) architectures. thermal driver-misc debugging + process-isolation amdgpu-glossary diff --git a/Documentation/gpu/amdgpu/process-isolation.rst b/Documentation/gpu/amdgpu/process-isolation.rst new file mode 100644 index 000000000000..6b6d70e357a7 --- /dev/null +++ b/Documentation/gpu/amdgpu/process-isolation.rst @@ -0,0 +1,59 @@ +.. SPDX-License-Identifier: GPL-2.0 + +========================= + AMDGPU Process Isolation +========================= + +The AMDGPU driver includes a feature that enables automatic process isolation on the graphics engine. This feature serializes access to the graphics engine and adds a cleaner shader which clears the Local Data Store (LDS) and General Purpose Registers (GPRs) between jobs. All processes using the GPU, including both graphics and compute workloads, are serialized when this feature is enabled. On GPUs that support partitionable graphics engines, this feature can be enabled on a per-partition basis. + +In addition, there is an interface to manually run the cleaner shader when the use of the GPU is complete. This may be preferable in some use cases, such as a single-user system where the login manager triggers the cleaner shader when the user logs out. + +Process Isolation +================= + +The `run_cleaner_shader` and `enforce_isolation` sysfs interfaces allow users to manually execute the cleaner shader and control the process isolation feature, respectively. + +Partition Handling +------------------ + +The `enforce_isolation` file in sysfs can be used to enable process isolation and automatic shader cleanup between processes. On GPUs that support graphics engine partitioning, this can be enabled per partition. The partition and its current setting (0 disabled, 1 enabled) can be read from sysfs. On GPUs that do not support graphics engine partitioning, only a single partition will be present. Writing 1 to the partition position enables enforce isolation, writing 0 disables it. + +Example of enabling enforce isolation on a GPU with multiple partitions: + +.. code-block:: console + + $ echo 1 0 1 0 > /sys/class/drm/card0/device/enforce_isolation + $ cat /sys/class/drm/card0/device/enforce_isolation + 1 0 1 0 + +The output indicates that enforce isolation is enabled on zeroth and second parition and disabled on first and fourth parition. + +For devices with a single partition or those that do not support partitions, there will be only one element: + +.. code-block:: console + + $ echo 1 > /sys/class/drm/card0/device/enforce_isolation + $ cat /sys/class/drm/card0/device/enforce_isolation + 1 + +Cleaner Shader Execution +======================== + +The driver can trigger a cleaner shader to clean up the LDS and GPR state on the graphics engine. When process isolation is enabled, this happens automatically between processes. In addition, there is a sysfs file to manually trigger cleaner shader execution. + +To manually trigger the execution of the cleaner shader, write `0` to the `run_cleaner_shader` sysfs file: + +.. code-block:: console + + $ echo 0 > /sys/class/drm/card0/device/run_cleaner_shader + +For multi-partition devices, you can specify the partition index when triggering the cleaner shader: + +.. code-block:: console + + $ echo 0 > /sys/class/drm/card0/device/run_cleaner_shader # For partition 0 + $ echo 1 > /sys/class/drm/card0/device/run_cleaner_shader # For partition 1 + $ echo 2 > /sys/class/drm/card0/device/run_cleaner_shader # For partition 2 + # ... and so on for each partition + +This command initiates the cleaner shader, which will run and complete before any new tasks are scheduled on the GPU. diff --git a/Documentation/gpu/amdgpu/thermal.rst b/Documentation/gpu/amdgpu/thermal.rst index 6d942b5c58f0..1768a106aab1 100644 --- a/Documentation/gpu/amdgpu/thermal.rst +++ b/Documentation/gpu/amdgpu/thermal.rst @@ -100,6 +100,18 @@ fan_minimum_pwm .. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c :doc: fan_minimum_pwm +fan_zero_rpm_enable +---------------------- + +.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c + :doc: fan_zero_rpm_enable + +fan_zero_rpm_stop_temperature +----------------------------- + +.. kernel-doc:: drivers/gpu/drm/amd/pm/amdgpu_pm.c + :doc: fan_zero_rpm_stop_temperature + GFXOFF ====== diff --git a/Documentation/gpu/automated_testing.rst b/Documentation/gpu/automated_testing.rst index 2d5a28866afe..6d7c6086034d 100644 --- a/Documentation/gpu/automated_testing.rst +++ b/Documentation/gpu/automated_testing.rst @@ -68,19 +68,25 @@ known to behave unreliably. These tests won't cause a job to fail regardless of the result. They will still be run. Each new flake entry must be associated with a link to the email reporting the -bug to the author of the affected driver, the board name or Device Tree name of -the board, the first kernel version affected, the IGT version used for tests, -and an approximation of the failure rate. +bug to the author of the affected driver or the relevant GitLab issue. The entry +must also include the board name or Device Tree name, the first kernel version +affected, the IGT version used for tests, and an approximation of the failure rate. They should be provided under the following format:: - # Bug Report: $LORE_OR_PATCHWORK_URL + # Bug Report: $LORE_URL_OR_GITLAB_ISSUE # Board Name: broken-board.dtb # Linux Version: 6.6-rc1 # IGT Version: 1.28-gd2af13d9f # Failure Rate: 100 flaky-test +Use the appropriate link below to create a GitLab issue: +amdgpu driver: https://gitlab.freedesktop.org/drm/amd/-/issues +i915 driver: https://gitlab.freedesktop.org/drm/i915/kernel/-/issues +msm driver: https://gitlab.freedesktop.org/drm/msm/-/issues +xe driver: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues + drivers/gpu/drm/ci/${DRIVER_NAME}-${HW_REVISION}-skips.txt ----------------------------------------------------------- diff --git a/Documentation/gpu/drivers.rst b/Documentation/gpu/drivers.rst index b899cbc5c2b4..1f17ad0790d7 100644 --- a/Documentation/gpu/drivers.rst +++ b/Documentation/gpu/drivers.rst @@ -22,6 +22,8 @@ GPU Driver Documentation afbc komeda-kms panfrost + panthor + zynqmp .. only:: subproject and html diff --git a/Documentation/gpu/drm-client.rst b/Documentation/gpu/drm-client.rst index 58b5a1d1219d..cbcfe30de777 100644 --- a/Documentation/gpu/drm-client.rst +++ b/Documentation/gpu/drm-client.rst @@ -13,3 +13,6 @@ Kernel clients .. kernel-doc:: drivers/gpu/drm/drm_client_modeset.c :export: + +.. kernel-doc:: drivers/gpu/drm/drm_client_event.c + :export: diff --git a/Documentation/gpu/drm-internals.rst b/Documentation/gpu/drm-internals.rst index 11d9a5730fb2..cb9ae282771c 100644 --- a/Documentation/gpu/drm-internals.rst +++ b/Documentation/gpu/drm-internals.rst @@ -75,18 +75,6 @@ Module Initialization .. kernel-doc:: include/drm/drm_module.h :doc: overview -Managing Ownership of the Framebuffer Aperture ----------------------------------------------- - -.. kernel-doc:: drivers/gpu/drm/drm_aperture.c - :doc: overview - -.. kernel-doc:: include/drm/drm_aperture.h - :internal: - -.. kernel-doc:: drivers/gpu/drm/drm_aperture.c - :export: - Device Instance and Driver Handling ----------------------------------- diff --git a/Documentation/gpu/drm-kms-helpers.rst b/Documentation/gpu/drm-kms-helpers.rst index c3e58856f75b..8cf2f041af47 100644 --- a/Documentation/gpu/drm-kms-helpers.rst +++ b/Documentation/gpu/drm-kms-helpers.rst @@ -110,15 +110,6 @@ fbdev Helper Functions Reference .. kernel-doc:: drivers/gpu/drm/drm_fb_helper.c :doc: fbdev helpers -.. kernel-doc:: drivers/gpu/drm/drm_fbdev_dma.c - :export: - -.. kernel-doc:: drivers/gpu/drm/drm_fbdev_shmem.c - :export: - -.. kernel-doc:: drivers/gpu/drm/drm_fbdev_ttm.c - :export: - .. kernel-doc:: include/drm/drm_fb_helper.h :internal: diff --git a/Documentation/gpu/drm-uapi.rst b/Documentation/gpu/drm-uapi.rst index 370d820be248..b75cc9a70d1f 100644 --- a/Documentation/gpu/drm-uapi.rst +++ b/Documentation/gpu/drm-uapi.rst @@ -305,13 +305,26 @@ Kernel Mode Driver ------------------ The KMD is responsible for checking if the device needs a reset, and to perform -it as needed. Usually a hang is detected when a job gets stuck executing. KMD -should keep track of resets, because userspace can query any time about the -reset status for a specific context. This is needed to propagate to the rest of -the stack that a reset has happened. Currently, this is implemented by each -driver separately, with no common DRM interface. Ideally this should be properly -integrated at DRM scheduler to provide a common ground for all drivers. After a -reset, KMD should reject new command submissions for affected contexts. +it as needed. Usually a hang is detected when a job gets stuck executing. + +Propagation of errors to userspace has proven to be tricky since it goes in +the opposite direction of the usual flow of commands. Because of this vendor +independent error handling was added to the &dma_fence object, this way drivers +can add an error code to their fences before signaling them. See function +dma_fence_set_error() on how to do this and for examples of error codes to use. + +The DRM scheduler also allows setting error codes on all pending fences when +hardware submissions are restarted after an reset. Error codes are also +forwarded from the hardware fence to the scheduler fence to bubble up errors +to the higher levels of the stack and eventually userspace. + +Fence errors can be queried by userspace through the generic SYNC_IOC_FILE_INFO +IOCTL as well as through driver specific interfaces. + +Additional to setting fence errors drivers should also keep track of resets per +context, the DRM scheduler provides the drm_sched_entity_error() function as +helper for this use case. After a reset, KMD should reject new command +submissions for affected contexts. User Mode Driver ---------------- diff --git a/Documentation/gpu/drm-usage-stats.rst b/Documentation/gpu/drm-usage-stats.rst index a80f95ca1b2f..2717cb2a597e 100644 --- a/Documentation/gpu/drm-usage-stats.rst +++ b/Documentation/gpu/drm-usage-stats.rst @@ -73,6 +73,11 @@ scope of each device, in which case `drm-pdev` shall be present as well. Userspace should make sure to not double account any usage statistics by using the above described criteria in order to associate data to individual clients. +- drm-client-name: + +String optionally set by userspace using DRM_IOCTL_SET_CLIENT_NAME. + + Utilization ^^^^^^^^^^^ @@ -144,7 +149,9 @@ Memory Each possible memory type which can be used to store buffer objects by the GPU in question shall be given a stable and unique name to be returned as the -string here. The name "memory" is reserved to refer to normal system memory. +string here. + +The region name "memory" is reserved to refer to normal system memory. Value shall reflect the amount of storage currently consumed by the buffer objects belong to this client, in the respective memory region. @@ -152,6 +159,9 @@ objects belong to this client, in the respective memory region. Default unit shall be bytes with optional unit specifiers of 'KiB' or 'MiB' indicating kibi- or mebi-bytes. +This key is deprecated and is an alias for drm-resident-. Only one of +the two should be present in the output. + - drm-shared-: [KiB|MiB] The total size of buffers that are shared with another file (e.g., have more @@ -159,20 +169,34 @@ than a single handle). - drm-total-: [KiB|MiB] -The total size of buffers that including shared and private memory. +The total size of all created buffers including shared and private memory. The +backing store for the buffers does not have to be currently instantiated to be +counted under this category. - drm-resident-: [KiB|MiB] -The total size of buffers that are resident in the specified region. +The total size of buffers that are resident (have their backing store present or +instantiated) in the specified region. + +This is an alias for drm-memory- and only one of the two should be +present in the output. - drm-purgeable-: [KiB|MiB] The total size of buffers that are purgeable. +For example drivers which implement a form of 'madvise' like functionality can +here count buffers which have instantiated backing store, but have been marked +with an equivalent of MADV_DONTNEED. + - drm-active-: [KiB|MiB] The total size of buffers that are active on one or more engines. +One practical example of this can be presence of unsignaled fences in an GEM +buffer reservation object. Therefore the active category is a subset of +resident. + Implementation Details ====================== @@ -186,4 +210,5 @@ Driver specific implementations * :ref:`i915-usage-stats` * :ref:`panfrost-usage-stats` +* :ref:`panthor-usage-stats` * :ref:`xe-usage-stats` diff --git a/Documentation/gpu/i915.rst b/Documentation/gpu/i915.rst index ad59ae579237..7a469df675d8 100644 --- a/Documentation/gpu/i915.rst +++ b/Documentation/gpu/i915.rst @@ -35,10 +35,10 @@ Interrupt Handling :functions: intel_irq_init intel_irq_init_hw intel_hpd_init .. kernel-doc:: drivers/gpu/drm/i915/i915_irq.c - :functions: intel_runtime_pm_disable_interrupts + :functions: intel_irq_suspend .. kernel-doc:: drivers/gpu/drm/i915/i915_irq.c - :functions: intel_runtime_pm_enable_interrupts + :functions: intel_irq_resume Intel GVT-g Guest Support(vGPU) ------------------------------- diff --git a/Documentation/gpu/msm-preemption.rst b/Documentation/gpu/msm-preemption.rst new file mode 100644 index 000000000000..d768ca09fdec --- /dev/null +++ b/Documentation/gpu/msm-preemption.rst @@ -0,0 +1,99 @@ +.. SPDX-License-Identifier: GPL-2.0 + +:orphan: + +============== +MSM Preemption +============== + +Preemption allows Adreno GPUs to switch to a higher priority ring when work is +pushed to it, reducing latency for high priority submissions. + +When preemption is enabled 4 rings are initialized, corresponding to different +priority levels. Having multiple rings is purely a software concept as the GPU +only has registers to keep track of one graphics ring. +The kernel is able to switch which ring is currently being processed by +requesting preemption. When certain conditions are met, depending on the +priority level, the GPU will save its current state in a series of buffers, +then restores state from a similar set of buffers specified by the kernel. It +then resumes execution and fires an IRQ to let the kernel know the context +switch has completed. + +This mechanism can be used by the kernel to switch between rings. Whenever a +submission occurs the kernel finds the highest priority ring which isn't empty +and preempts to it if said ring is not the one being currently executed. This is +also done whenever a submission completes to make sure execution resumes on a +lower priority ring when a higher priority ring is done. + +Preemption levels +----------------- + +Preemption can only occur at certain boundaries. The exact conditions can be +configured by changing the preemption level, this allows to compromise between +latency (ie. the time that passes between when the kernel requests preemption +and when the SQE begins saving state) and overhead (the amount of state that +needs to be saved). + +The GPU offers 3 levels: + +Level 0 + Preemption only occurs at the submission level. This requires the least amount + of state to be saved as the execution of userspace submitted IBs is never + interrupted, however it offers very little benefit compared to not enabling + preemption of any kind. + +Level 1 + Preemption occurs at either bin level, if using GMEM rendering, or draw level + in the sysmem rendering case. + +Level 2 + Preemption occurs at draw level. + +Level 1 is the mode that is used by the msm driver. + +Additionally the GPU allows to specify a `skip_save_restore` option. This +disables the saving and restoring of all registers except those relating to the +operation of the SQE itself, reducing overhead. Saving and restoring is only +skipped when using GMEM with Level 1 preemption. When enabling this userspace is +expected to set the state that isn't preserved whenever preemption occurs which +is done by specifying preamble and postambles. Those are IBs that are executed +before and after preemption. + +Preemption buffers +------------------ + +A series of buffers are necessary to store the state of rings while they are not +being executed. There are different kinds of preemption records and most of +those require one buffer per ring. This is because preemption never occurs +between submissions on the same ring, which always run in sequence when the ring +is active. This means that only one context per ring is effectively active. + +SMMU_INFO + This buffer contains info about the current SMMU configuration such as the + ttbr0 register. The SQE firmware isn't actually able to save this record. + As a result SMMU info must be saved manually from the CP to a buffer and the + SMMU record updated with info from said buffer before triggering + preemption. + +NON_SECURE + This is the main preemption record where most state is saved. It is mostly + opaque to the kernel except for the first few words that must be initialized + by the kernel. + +SECURE + This saves state related to the GPU's secure mode. + +NON_PRIV + The intended purpose of this record is unknown. The SQE firmware actually + ignores it and therefore msm doesn't handle it. + +COUNTER + This record is used to save and restore performance counters. + +Handling the permissions of those buffers is critical for security. All but the +NON_PRIV records need to be inaccessible from userspace, so they must be mapped +in the kernel address space with the MSM_BO_MAP_PRIV flag. +For example, making the NON_SECURE record accessible from userspace would allow +any process to manipulate a saved ring's RPTR which can be used to skip the +execution of some packets in a ring and execute user commands with higher +privileges. diff --git a/Documentation/gpu/panthor.rst b/Documentation/gpu/panthor.rst new file mode 100644 index 000000000000..3f8979fa2b86 --- /dev/null +++ b/Documentation/gpu/panthor.rst @@ -0,0 +1,46 @@ +.. SPDX-License-Identifier: GPL-2.0+ + +========================= + drm/Panthor CSF driver +========================= + +.. _panthor-usage-stats: + +Panthor DRM client usage stats implementation +============================================== + +The drm/Panthor driver implements the DRM client usage stats specification as +documented in :ref:`drm-client-usage-stats`. + +Example of the output showing the implemented key value pairs and entirety of +the currently possible format options: + +:: + pos: 0 + flags: 02400002 + mnt_id: 29 + ino: 491 + drm-driver: panthor + drm-client-id: 10 + drm-engine-panthor: 111110952750 ns + drm-cycles-panthor: 94439687187 + drm-maxfreq-panthor: 1000000000 Hz + drm-curfreq-panthor: 1000000000 Hz + drm-total-memory: 16480 KiB + drm-shared-memory: 0 + drm-active-memory: 16200 KiB + drm-resident-memory: 16480 KiB + drm-purgeable-memory: 0 + +Possible `drm-engine-` key names are: `panthor`. +`drm-curfreq-` values convey the current operating frequency for that engine. + +Users must bear in mind that engine and cycle sampling are disabled by default, +because of power saving concerns. `fdinfo` users and benchmark applications which +query the fdinfo file must make sure to toggle the job profiling status of the +driver by writing into the appropriate sysfs node:: + + echo > /sys/bus/platform/drivers/panthor/[a-f0-9]*.gpu/profiling + +Where `N` is a bit mask where cycle and timestamp sampling are respectively +enabled by the first and second bits. diff --git a/Documentation/gpu/todo.rst b/Documentation/gpu/todo.rst index 2b281e3c75a4..256d0d1cb216 100644 --- a/Documentation/gpu/todo.rst +++ b/Documentation/gpu/todo.rst @@ -834,6 +834,22 @@ Contact: Javier Martinez Canillas Level: Advanced +Querying errors from drm_syncobj +================================ + +The drm_syncobj container can be used by driver independent code to signal +complection of submission. + +One minor feature still missing is a generic DRM IOCTL to query the error +status of binary and timeline drm_syncobj. + +This should probably be improved by implementing the necessary kernel interface +and adding support for that in the userspace stack. + +Contact: Christian König + +Level: Starter + Outside DRM =========== diff --git a/Documentation/gpu/zynqmp.rst b/Documentation/gpu/zynqmp.rst new file mode 100644 index 000000000000..f57bfa0ad6ec --- /dev/null +++ b/Documentation/gpu/zynqmp.rst @@ -0,0 +1,149 @@ +.. SPDX-License-Identifier: GPL-2.0+ + +=============================================== +Xilinx ZynqMP Ultrascale+ DisplayPort Subsystem +=============================================== + +This subsystem handles DisplayPort video and audio output on the ZynqMP. It +supports in-memory framebuffers with the DisplayPort DMA controller +(xilinx-dpdma), as well as "live" video and audio from the programmable logic +(PL). This subsystem can perform several transformations, including color space +conversion, alpha blending, and audio mixing, although not all features are +currently supported. + +debugfs +------- + +To support debugging and compliance testing, several test modes can be enabled +though debugfs. The following files in /sys/kernel/debug/dri/X/DP-1/test/ +control the DisplayPort test modes: + +active: + Writing a 1 to this file will activate test mode, and writing a 0 will + deactivate test mode. Writing a 1 or 0 when the test mode is already + active/inactive will re-activate/re-deactivate test mode. When test + mode is inactive, changes made to other files will have no (immediate) + effect, although the settings will be saved for when test mode is + activated. When test mode is active, changes made to other files will + apply immediately. + +custom: + Custom test pattern value + +downspread: + Enable/disable clock downspreading (spread-spectrum clocking) by + writing 1/0 + +enhanced: + Enable/disable enhanced framing + +ignore_aux_errors: + Ignore AUX errors when set to 1. Writes to this file take effect + immediately (regardless of whether test mode is active) and affect all + AUX transfers. + +ignore_hpd: + Ignore hotplug events (such as cable removals or monitor link + retraining requests) when set to 1. Writes to this file take effect + immediately (regardless of whether test mode is active). + +laneX_preemphasis: + Preemphasis from 0 (lowest) to 2 (highest) for lane X + +laneX_swing: + Voltage swing from 0 (lowest) to 3 (highest) for lane X + +lanes: + Number of lanes to use (1, 2, or 4) + +pattern: + Test pattern. May be one of: + + video + Use regular video input + + symbol-error + Symbol error measurement pattern + + prbs7 + Output of the PRBS7 (x^7 + x^6 + 1) polynomial + + 80bit-custom + A custom 80-bit pattern + + cp2520 + HBR2 compliance eye pattern + + tps1 + Link training symbol pattern TPS1 (/D10.2/) + + tps2 + Link training symbol pattern TPS2 + + tps3 + Link training symbol pattern TPS3 (for HBR2) + +rate: + Rate in hertz. One of + + * 5400000000 (HBR2) + * 2700000000 (HBR) + * 1620000000 (RBR) + +You can dump the displayport test settings with the following command:: + + for prop in /sys/kernel/debug/dri/1/DP-1/test/*; do + printf '%-17s ' ${prop##*/} + if [ ${prop##*/} = custom ]; then + hexdump -C $prop | head -1 + else + cat $prop + fi + done + +The output could look something like:: + + active 1 + custom 00000000 00 00 00 00 00 00 00 00 00 00 |..........| + downspread 0 + enhanced 1 + ignore_aux_errors 1 + ignore_hpd 1 + lane0_preemphasis 0 + lane0_swing 3 + lane1_preemphasis 0 + lane1_swing 3 + lanes 2 + pattern prbs7 + rate 1620000000 + +The recommended test procedure is to connect the board to a monitor, +configure test mode, activate test mode, and then disconnect the cable +and connect it to your test equipment of choice. For example, one +sequence of commands could be:: + + echo 1 > /sys/kernel/debug/dri/1/DP-1/test/enhanced + echo tps1 > /sys/kernel/debug/dri/1/DP-1/test/pattern + echo 1620000000 > /sys/kernel/debug/dri/1/DP-1/test/rate + echo 1 > /sys/kernel/debug/dri/1/DP-1/test/ignore_aux_errors + echo 1 > /sys/kernel/debug/dri/1/DP-1/test/ignore_hpd + echo 1 > /sys/kernel/debug/dri/1/DP-1/test/active + +at which point the cable could be disconnected from the monitor. + +Internals +--------- + +.. kernel-doc:: drivers/gpu/drm/xlnx/zynqmp_disp.h + +.. kernel-doc:: drivers/gpu/drm/xlnx/zynqmp_dpsub.h + +.. kernel-doc:: drivers/gpu/drm/xlnx/zynqmp_kms.h + +.. kernel-doc:: drivers/gpu/drm/xlnx/zynqmp_disp.c + +.. kernel-doc:: drivers/gpu/drm/xlnx/zynqmp_dp.c + +.. kernel-doc:: drivers/gpu/drm/xlnx/zynqmp_dpsub.c + +.. kernel-doc:: drivers/gpu/drm/xlnx/zynqmp_kms.c diff --git a/Documentation/userspace-api/media/v4l/subdev-formats.rst b/Documentation/userspace-api/media/v4l/subdev-formats.rst index d2a6cd2e1eb2..2a94371448dc 100644 --- a/Documentation/userspace-api/media/v4l/subdev-formats.rst +++ b/Documentation/userspace-api/media/v4l/subdev-formats.rst @@ -2225,7 +2225,7 @@ The following table list existing packed 48bit wide RGB formats. \endgroup On LVDS buses, usually each sample is transferred serialized in seven -time slots per pixel clock, on three (18-bit) or four (24-bit) +time slots per pixel clock, on three (18-bit) or four (24-bit) or five (30-bit) differential data pairs at the same time. The remaining bits are used for control signals as defined by SPWG/PSWG/VESA or JEIDA standards. The 24-bit RGB format serialized in seven time slots on four lanes using @@ -2246,11 +2246,12 @@ JEIDA defined bit mapping will be named - Code - - - - :cspan:`3` Data organization + - :cspan:`4` Data organization * - - - Timeslot - Lane + - 4 - 3 - 2 - 1 @@ -2262,6 +2263,7 @@ JEIDA defined bit mapping will be named - 0 - - + - - d - b\ :sub:`1` - g\ :sub:`0` @@ -2270,6 +2272,7 @@ JEIDA defined bit mapping will be named - 1 - - + - - d - b\ :sub:`0` - r\ :sub:`5` @@ -2278,6 +2281,7 @@ JEIDA defined bit mapping will be named - 2 - - + - - d - g\ :sub:`5` - r\ :sub:`4` @@ -2286,6 +2290,7 @@ JEIDA defined bit mapping will be named - 3 - - + - - b\ :sub:`5` - g\ :sub:`4` - r\ :sub:`3` @@ -2294,6 +2299,7 @@ JEIDA defined bit mapping will be named - 4 - - + - - b\ :sub:`4` - g\ :sub:`3` - r\ :sub:`2` @@ -2302,6 +2308,7 @@ JEIDA defined bit mapping will be named - 5 - - + - - b\ :sub:`3` - g\ :sub:`2` - r\ :sub:`1` @@ -2310,6 +2317,7 @@ JEIDA defined bit mapping will be named - 6 - - + - - b\ :sub:`2` - g\ :sub:`1` - r\ :sub:`0` @@ -2319,6 +2327,7 @@ JEIDA defined bit mapping will be named - 0x1011 - 0 - + - - d - d - b\ :sub:`1` @@ -2327,6 +2336,7 @@ JEIDA defined bit mapping will be named - - 1 - + - - b\ :sub:`7` - d - b\ :sub:`0` @@ -2335,6 +2345,7 @@ JEIDA defined bit mapping will be named - - 2 - + - - b\ :sub:`6` - d - g\ :sub:`5` @@ -2343,6 +2354,7 @@ JEIDA defined bit mapping will be named - - 3 - + - - g\ :sub:`7` - b\ :sub:`5` - g\ :sub:`4` @@ -2351,6 +2363,7 @@ JEIDA defined bit mapping will be named - - 4 - + - - g\ :sub:`6` - b\ :sub:`4` - g\ :sub:`3` @@ -2359,6 +2372,7 @@ JEIDA defined bit mapping will be named - - 5 - + - - r\ :sub:`7` - b\ :sub:`3` - g\ :sub:`2` @@ -2367,6 +2381,7 @@ JEIDA defined bit mapping will be named - - 6 - + - - r\ :sub:`6` - b\ :sub:`2` - g\ :sub:`1` @@ -2377,6 +2392,7 @@ JEIDA defined bit mapping will be named - 0x1012 - 0 - + - - d - d - b\ :sub:`3` @@ -2385,6 +2401,7 @@ JEIDA defined bit mapping will be named - - 1 - + - - b\ :sub:`1` - d - b\ :sub:`2` @@ -2393,6 +2410,7 @@ JEIDA defined bit mapping will be named - - 2 - + - - b\ :sub:`0` - d - g\ :sub:`7` @@ -2401,6 +2419,7 @@ JEIDA defined bit mapping will be named - - 3 - + - - g\ :sub:`1` - b\ :sub:`7` - g\ :sub:`6` @@ -2409,6 +2428,7 @@ JEIDA defined bit mapping will be named - - 4 - + - - g\ :sub:`0` - b\ :sub:`6` - g\ :sub:`5` @@ -2417,6 +2437,7 @@ JEIDA defined bit mapping will be named - - 5 - + - - r\ :sub:`1` - b\ :sub:`5` - g\ :sub:`4` @@ -2425,10 +2446,141 @@ JEIDA defined bit mapping will be named - - 6 - + - - r\ :sub:`0` - b\ :sub:`4` - g\ :sub:`3` - r\ :sub:`2` + * .. _MEDIA-BUS-FMT-RGB101010-1X7X5-SPWG: + + - MEDIA_BUS_FMT_RGB101010_1X7X5_SPWG + - 0x1026 + - 0 + - + - d + - d + - d + - b\ :sub:`1` + - g\ :sub:`0` + * - + - + - 1 + - + - b\ :sub:`9` + - b\ :sub:`7` + - d + - b\ :sub:`0` + - r\ :sub:`5` + * - + - + - 2 + - + - b\ :sub:`8` + - b\ :sub:`6` + - d + - g\ :sub:`5` + - r\ :sub:`4` + * - + - + - 3 + - + - g\ :sub:`9` + - g\ :sub:`7` + - b\ :sub:`5` + - g\ :sub:`4` + - r\ :sub:`3` + * - + - + - 4 + - + - g\ :sub:`8` + - g\ :sub:`6` + - b\ :sub:`4` + - g\ :sub:`3` + - r\ :sub:`2` + * - + - + - 5 + - + - r\ :sub:`9` + - r\ :sub:`7` + - b\ :sub:`3` + - g\ :sub:`2` + - r\ :sub:`1` + * - + - + - 6 + - + - r\ :sub:`8` + - r\ :sub:`6` + - b\ :sub:`2` + - g\ :sub:`1` + - r\ :sub:`0` + * .. _MEDIA-BUS-FMT-RGB101010-1X7X5-JEIDA: + + - MEDIA_BUS_FMT_RGB101010_1X7X5_JEIDA + - 0x1027 + - 0 + - + - d + - d + - d + - b\ :sub:`5` + - g\ :sub:`4` + * - + - + - 1 + - + - b\ :sub:`1` + - b\ :sub:`3` + - d + - b\ :sub:`4` + - r\ :sub:`9` + * - + - + - 2 + - + - b\ :sub:`0` + - b\ :sub:`2` + - d + - g\ :sub:`9` + - r\ :sub:`8` + * - + - + - 3 + - + - g\ :sub:`1` + - g\ :sub:`3` + - b\ :sub:`9` + - g\ :sub:`8` + - r\ :sub:`7` + * - + - + - 4 + - + - g\ :sub:`0` + - g\ :sub:`2` + - b\ :sub:`8` + - g\ :sub:`7` + - r\ :sub:`6` + * - + - + - 5 + - + - r\ :sub:`1` + - r\ :sub:`3` + - b\ :sub:`7` + - g\ :sub:`6` + - r\ :sub:`5` + * - + - + - 6 + - + - r\ :sub:`0` + - r\ :sub:`2` + - b\ :sub:`6` + - g\ :sub:`5` + - r\ :sub:`4` .. raw:: latex diff --git a/MAINTAINERS b/MAINTAINERS index 48688a2f3b7e..bebd3436b590 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -7103,12 +7103,10 @@ M: Javier Martinez Canillas L: dri-devel@lists.freedesktop.org S: Maintained T: git https://gitlab.freedesktop.org/drm/misc/kernel.git -F: drivers/gpu/drm/drm_aperture.c F: drivers/gpu/drm/tiny/ofdrm.c F: drivers/gpu/drm/tiny/simpledrm.c F: drivers/video/aperture.c F: drivers/video/nomodeset.c -F: include/drm/drm_aperture.h F: include/linux/aperture.h F: include/video/nomodeset.h @@ -7389,6 +7387,18 @@ S: Maintained F: Documentation/devicetree/bindings/display/panel/samsung,s6d7aa0.yaml F: drivers/gpu/drm/panel/panel-samsung-s6d7aa0.c +DRM DRIVER FOR SAMSUNG S6E3HA8 PANELS +M: Dzmitry Sankouski +S: Maintained +F: Documentation/devicetree/bindings/display/panel/samsung,s6e3ha8.yaml +F: drivers/gpu/drm/panel/panel-samsung-s6e3ha8.c + +DRM DRIVER FOR SHARP MEMORY LCD +M: Alex Lanzano +S: Maintained +F: Documentation/devicetree/bindings/display/sharp,ls010b7dh04.yaml +F: drivers/gpu/drm/tiny/sharp-memory.c + DRM DRIVER FOR SITRONIX ST7586 PANELS M: David Lechner S: Maintained @@ -7466,8 +7476,7 @@ T: git https://gitlab.freedesktop.org/drm/misc/kernel.git F: drivers/gpu/drm/udl/ DRM DRIVER FOR VIRTUAL KERNEL MODESETTING (VKMS) -M: Rodrigo Siqueira -M: Maíra Canal +M: Louis Chauvet R: Haneen Mohammed R: Simona Vetter R: Melissa Wen @@ -7799,6 +7808,7 @@ F: include/uapi/drm/v3d_drm.h DRM DRIVERS FOR VC4 M: Maxime Ripard M: Dave Stevenson +R: Maíra Canal R: Raspberry Pi Kernel Maintenance S: Supported T: git https://gitlab.freedesktop.org/drm/misc/kernel.git @@ -7833,6 +7843,7 @@ L: dri-devel@lists.freedesktop.org S: Maintained T: git https://gitlab.freedesktop.org/drm/misc/kernel.git F: Documentation/devicetree/bindings/display/xlnx/ +F: Documentation/gpu/zynqmp.rst F: drivers/gpu/drm/xlnx/ DRM GPU SCHEDULER @@ -12174,6 +12185,14 @@ W: https://linuxtv.org Q: http://patchwork.linuxtv.org/project/linux-media/list/ F: drivers/media/tuners/it913x* +ITE IT6263 LVDS TO HDMI BRIDGE DRIVER +M: Liu Ying +L: dri-devel@lists.freedesktop.org +S: Maintained +T: git https://gitlab.freedesktop.org/drm/misc/kernel.git +F: Documentation/devicetree/bindings/display/bridge/ite,it6263.yaml +F: drivers/gpu/drm/bridge/ite-it6263.c + ITE IT66121 HDMI BRIDGE DRIVER M: Phong LE M: Neil Armstrong @@ -24301,6 +24320,7 @@ F: lib/iov_iter.c USERSPACE DMA BUFFER DRIVER M: Gerd Hoffmann +M: Vivek Kasireddy L: dri-devel@lists.freedesktop.org S: Maintained T: git https://gitlab.freedesktop.org/drm/misc/kernel.git diff --git a/arch/x86/kernel/early-quirks.c b/arch/x86/kernel/early-quirks.c index 29d1f9104e94..6b6f32f40cbe 100644 --- a/arch/x86/kernel/early-quirks.c +++ b/arch/x86/kernel/early-quirks.c @@ -18,7 +18,7 @@ #include #include #include -#include +#include #include #include #include diff --git a/drivers/accel/ivpu/Kconfig b/drivers/accel/ivpu/Kconfig index 682c53245286..9e055b5ce03d 100644 --- a/drivers/accel/ivpu/Kconfig +++ b/drivers/accel/ivpu/Kconfig @@ -8,6 +8,7 @@ config DRM_ACCEL_IVPU select FW_LOADER select DRM_GEM_SHMEM_HELPER select GENERIC_ALLOCATOR + select WANT_DEV_COREDUMP help Choose this option if you have a system with an 14th generation Intel CPU (Meteor Lake) or newer. Intel NPU (formerly called Intel VPU) @@ -15,3 +16,12 @@ config DRM_ACCEL_IVPU and Deep Learning applications. If "M" is selected, the module will be called intel_vpu. + +config DRM_ACCEL_IVPU_DEBUG + bool "Intel NPU debug mode" + depends on DRM_ACCEL_IVPU + help + Choose this option to enable additional + debug features for the Intel NPU driver: + - Always print debug messages regardless of dyndbg config, + - Enable unsafe module params. diff --git a/drivers/accel/ivpu/Makefile b/drivers/accel/ivpu/Makefile index ebd682a42eb1..1029e0bab061 100644 --- a/drivers/accel/ivpu/Makefile +++ b/drivers/accel/ivpu/Makefile @@ -16,8 +16,14 @@ intel_vpu-y := \ ivpu_mmu_context.o \ ivpu_ms.o \ ivpu_pm.o \ - ivpu_sysfs.o + ivpu_sysfs.o \ + ivpu_trace_points.o intel_vpu-$(CONFIG_DEBUG_FS) += ivpu_debugfs.o +intel_vpu-$(CONFIG_DEV_COREDUMP) += ivpu_coredump.o obj-$(CONFIG_DRM_ACCEL_IVPU) += intel_vpu.o + +subdir-ccflags-$(CONFIG_DRM_ACCEL_IVPU_DEBUG) += -DDEBUG + +CFLAGS_ivpu_trace_points.o = -I$(src) diff --git a/drivers/accel/ivpu/ivpu_coredump.c b/drivers/accel/ivpu/ivpu_coredump.c new file mode 100644 index 000000000000..16ad0c30818c --- /dev/null +++ b/drivers/accel/ivpu/ivpu_coredump.c @@ -0,0 +1,39 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (C) 2020-2024 Intel Corporation + */ + +#include +#include + +#include "ivpu_coredump.h" +#include "ivpu_fw.h" +#include "ivpu_gem.h" +#include "vpu_boot_api.h" + +#define CRASH_DUMP_HEADER "Intel NPU crash dump" +#define CRASH_DUMP_HEADERS_SIZE SZ_4K + +void ivpu_dev_coredump(struct ivpu_device *vdev) +{ + struct drm_print_iterator pi = {}; + struct drm_printer p; + size_t coredump_size; + char *coredump; + + coredump_size = CRASH_DUMP_HEADERS_SIZE + FW_VERSION_HEADER_SIZE + + ivpu_bo_size(vdev->fw->mem_log_crit) + ivpu_bo_size(vdev->fw->mem_log_verb); + coredump = vmalloc(coredump_size); + if (!coredump) + return; + + pi.data = coredump; + pi.remain = coredump_size; + p = drm_coredump_printer(&pi); + + drm_printf(&p, "%s\n", CRASH_DUMP_HEADER); + drm_printf(&p, "FW version: %s\n", vdev->fw->version); + ivpu_fw_log_print(vdev, false, &p); + + dev_coredumpv(vdev->drm.dev, coredump, pi.offset, GFP_KERNEL); +} diff --git a/drivers/accel/ivpu/ivpu_coredump.h b/drivers/accel/ivpu/ivpu_coredump.h new file mode 100644 index 000000000000..8efb09d02441 --- /dev/null +++ b/drivers/accel/ivpu/ivpu_coredump.h @@ -0,0 +1,25 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* + * Copyright (C) 2020-2024 Intel Corporation + */ + +#ifndef __IVPU_COREDUMP_H__ +#define __IVPU_COREDUMP_H__ + +#include + +#include "ivpu_drv.h" +#include "ivpu_fw_log.h" + +#ifdef CONFIG_DEV_COREDUMP +void ivpu_dev_coredump(struct ivpu_device *vdev); +#else +static inline void ivpu_dev_coredump(struct ivpu_device *vdev) +{ + struct drm_printer p = drm_info_printer(vdev->drm.dev); + + ivpu_fw_log_print(vdev, false, &p); +} +#endif + +#endif /* __IVPU_COREDUMP_H__ */ diff --git a/drivers/accel/ivpu/ivpu_debugfs.c b/drivers/accel/ivpu/ivpu_debugfs.c index 8d50981594d1..8180b95ed69d 100644 --- a/drivers/accel/ivpu/ivpu_debugfs.c +++ b/drivers/accel/ivpu/ivpu_debugfs.c @@ -45,6 +45,14 @@ static int fw_name_show(struct seq_file *s, void *v) return 0; } +static int fw_version_show(struct seq_file *s, void *v) +{ + struct ivpu_device *vdev = seq_to_ivpu(s); + + seq_printf(s, "%s\n", vdev->fw->version); + return 0; +} + static int fw_trace_capability_show(struct seq_file *s, void *v) { struct ivpu_device *vdev = seq_to_ivpu(s); @@ -119,6 +127,7 @@ static int firewall_irq_counter_show(struct seq_file *s, void *v) static const struct drm_debugfs_info vdev_debugfs_list[] = { {"bo_list", bo_list_show, 0}, {"fw_name", fw_name_show, 0}, + {"fw_version", fw_version_show, 0}, {"fw_trace_capability", fw_trace_capability_show, 0}, {"fw_trace_config", fw_trace_config_show, 0}, {"last_bootmode", last_bootmode_show, 0}, @@ -127,32 +136,23 @@ static const struct drm_debugfs_info vdev_debugfs_list[] = { {"firewall_irq_counter", firewall_irq_counter_show, 0}, }; -static ssize_t -dvfs_mode_fops_write(struct file *file, const char __user *user_buf, size_t size, loff_t *pos) +static int dvfs_mode_get(void *data, u64 *dvfs_mode) { - struct ivpu_device *vdev = file->private_data; - struct ivpu_fw_info *fw = vdev->fw; - u32 dvfs_mode; - int ret; + struct ivpu_device *vdev = (struct ivpu_device *)data; - ret = kstrtou32_from_user(user_buf, size, 0, &dvfs_mode); - if (ret < 0) - return ret; - - fw->dvfs_mode = dvfs_mode; - - ret = pci_try_reset_function(to_pci_dev(vdev->drm.dev)); - if (ret) - return ret; - - return size; + *dvfs_mode = vdev->fw->dvfs_mode; + return 0; } -static const struct file_operations dvfs_mode_fops = { - .owner = THIS_MODULE, - .open = simple_open, - .write = dvfs_mode_fops_write, -}; +static int dvfs_mode_set(void *data, u64 dvfs_mode) +{ + struct ivpu_device *vdev = (struct ivpu_device *)data; + + vdev->fw->dvfs_mode = (u32)dvfs_mode; + return pci_try_reset_function(to_pci_dev(vdev->drm.dev)); +} + +DEFINE_DEBUGFS_ATTRIBUTE(dvfs_mode_fops, dvfs_mode_get, dvfs_mode_set, "%llu\n"); static ssize_t fw_dyndbg_fops_write(struct file *file, const char __user *user_buf, size_t size, loff_t *pos) @@ -201,7 +201,7 @@ fw_log_fops_write(struct file *file, const char __user *user_buf, size_t size, l if (!size) return -EINVAL; - ivpu_fw_log_clear(vdev); + ivpu_fw_log_mark_read(vdev); return size; } @@ -346,49 +346,23 @@ static const struct file_operations ivpu_force_recovery_fops = { .write = ivpu_force_recovery_fn, }; -static ssize_t -ivpu_reset_engine_fn(struct file *file, const char __user *user_buf, size_t size, loff_t *pos) +static int ivpu_reset_engine_fn(void *data, u64 val) { - struct ivpu_device *vdev = file->private_data; + struct ivpu_device *vdev = (struct ivpu_device *)data; - if (!size) - return -EINVAL; - - if (ivpu_jsm_reset_engine(vdev, DRM_IVPU_ENGINE_COMPUTE)) - return -ENODEV; - if (ivpu_jsm_reset_engine(vdev, DRM_IVPU_ENGINE_COPY)) - return -ENODEV; - - return size; + return ivpu_jsm_reset_engine(vdev, (u32)val); } -static const struct file_operations ivpu_reset_engine_fops = { - .owner = THIS_MODULE, - .open = simple_open, - .write = ivpu_reset_engine_fn, -}; +DEFINE_DEBUGFS_ATTRIBUTE(ivpu_reset_engine_fops, NULL, ivpu_reset_engine_fn, "0x%02llx\n"); -static ssize_t -ivpu_resume_engine_fn(struct file *file, const char __user *user_buf, size_t size, loff_t *pos) +static int ivpu_resume_engine_fn(void *data, u64 val) { - struct ivpu_device *vdev = file->private_data; + struct ivpu_device *vdev = (struct ivpu_device *)data; - if (!size) - return -EINVAL; - - if (ivpu_jsm_hws_resume_engine(vdev, DRM_IVPU_ENGINE_COMPUTE)) - return -ENODEV; - if (ivpu_jsm_hws_resume_engine(vdev, DRM_IVPU_ENGINE_COPY)) - return -ENODEV; - - return size; + return ivpu_jsm_hws_resume_engine(vdev, (u32)val); } -static const struct file_operations ivpu_resume_engine_fops = { - .owner = THIS_MODULE, - .open = simple_open, - .write = ivpu_resume_engine_fn, -}; +DEFINE_DEBUGFS_ATTRIBUTE(ivpu_resume_engine_fops, NULL, ivpu_resume_engine_fn, "0x%02llx\n"); static int dct_active_get(void *data, u64 *active_percent) { @@ -432,7 +406,7 @@ void ivpu_debugfs_init(struct ivpu_device *vdev) debugfs_create_file("force_recovery", 0200, debugfs_root, vdev, &ivpu_force_recovery_fops); - debugfs_create_file("dvfs_mode", 0200, debugfs_root, vdev, + debugfs_create_file("dvfs_mode", 0644, debugfs_root, vdev, &dvfs_mode_fops); debugfs_create_file("fw_dyndbg", 0200, debugfs_root, vdev, diff --git a/drivers/accel/ivpu/ivpu_drv.c b/drivers/accel/ivpu/ivpu_drv.c index c91400ecf926..ca2bf47ce248 100644 --- a/drivers/accel/ivpu/ivpu_drv.c +++ b/drivers/accel/ivpu/ivpu_drv.c @@ -7,6 +7,7 @@ #include #include #include +#include #include #include @@ -14,7 +15,7 @@ #include #include -#include "vpu_boot_api.h" +#include "ivpu_coredump.h" #include "ivpu_debugfs.h" #include "ivpu_drv.h" #include "ivpu_fw.h" @@ -29,10 +30,10 @@ #include "ivpu_ms.h" #include "ivpu_pm.h" #include "ivpu_sysfs.h" +#include "vpu_boot_api.h" #ifndef DRIVER_VERSION_STR -#define DRIVER_VERSION_STR __stringify(DRM_IVPU_DRIVER_MAJOR) "." \ - __stringify(DRM_IVPU_DRIVER_MINOR) "." +#define DRIVER_VERSION_STR "1.0.0 " UTS_RELEASE #endif static struct lock_class_key submitted_jobs_xa_lock_class_key; @@ -42,8 +43,10 @@ module_param_named(dbg_mask, ivpu_dbg_mask, int, 0644); MODULE_PARM_DESC(dbg_mask, "Driver debug mask. See IVPU_DBG_* macros."); int ivpu_test_mode; +#if IS_ENABLED(CONFIG_DRM_ACCEL_IVPU_DEBUG) module_param_named_unsafe(test_mode, ivpu_test_mode, int, 0644); MODULE_PARM_DESC(test_mode, "Test mode mask. See IVPU_TEST_MODE_* macros."); +#endif u8 ivpu_pll_min_ratio; module_param_named(pll_min_ratio, ivpu_pll_min_ratio, byte, 0644); @@ -53,9 +56,9 @@ u8 ivpu_pll_max_ratio = U8_MAX; module_param_named(pll_max_ratio, ivpu_pll_max_ratio, byte, 0644); MODULE_PARM_DESC(pll_max_ratio, "Maximum PLL ratio used to set NPU frequency"); -int ivpu_sched_mode; +int ivpu_sched_mode = IVPU_SCHED_MODE_AUTO; module_param_named(sched_mode, ivpu_sched_mode, int, 0444); -MODULE_PARM_DESC(sched_mode, "Scheduler mode: 0 - Default scheduler, 1 - Force HW scheduler"); +MODULE_PARM_DESC(sched_mode, "Scheduler mode: -1 - Use default scheduler, 0 - Use OS scheduler, 1 - Use HW scheduler"); bool ivpu_disable_mmu_cont_pages; module_param_named(disable_mmu_cont_pages, ivpu_disable_mmu_cont_pages, bool, 0444); @@ -85,7 +88,7 @@ static void file_priv_unbind(struct ivpu_device *vdev, struct ivpu_file_priv *fi ivpu_cmdq_release_all_locked(file_priv); ivpu_bo_unbind_all_bos_from_context(vdev, &file_priv->ctx); - ivpu_mmu_user_context_fini(vdev, &file_priv->ctx); + ivpu_mmu_context_fini(vdev, &file_priv->ctx); file_priv->bound = false; drm_WARN_ON(&vdev->drm, !xa_erase_irq(&vdev->context_xa, file_priv->ctx.id)); } @@ -103,6 +106,8 @@ static void file_priv_release(struct kref *ref) pm_runtime_get_sync(vdev->drm.dev); mutex_lock(&vdev->context_list_lock); file_priv_unbind(vdev, file_priv); + drm_WARN_ON(&vdev->drm, !xa_empty(&file_priv->cmdq_xa)); + xa_destroy(&file_priv->cmdq_xa); mutex_unlock(&vdev->context_list_lock); pm_runtime_put_autosuspend(vdev->drm.dev); @@ -116,8 +121,6 @@ void ivpu_file_priv_put(struct ivpu_file_priv **link) struct ivpu_file_priv *file_priv = *link; struct ivpu_device *vdev = file_priv->vdev; - drm_WARN_ON(&vdev->drm, !file_priv); - ivpu_dbg(vdev, KREF, "file_priv put: ctx %u refcount %u\n", file_priv->ctx.id, kref_read(&file_priv->ref)); @@ -255,9 +258,14 @@ static int ivpu_open(struct drm_device *dev, struct drm_file *file) goto err_unlock; } - ret = ivpu_mmu_user_context_init(vdev, &file_priv->ctx, ctx_id); - if (ret) - goto err_xa_erase; + ivpu_mmu_context_init(vdev, &file_priv->ctx, ctx_id); + + file_priv->job_limit.min = FIELD_PREP(IVPU_JOB_ID_CONTEXT_MASK, (file_priv->ctx.id - 1)); + file_priv->job_limit.max = file_priv->job_limit.min | IVPU_JOB_ID_JOB_MASK; + + xa_init_flags(&file_priv->cmdq_xa, XA_FLAGS_ALLOC1); + file_priv->cmdq_limit.min = IVPU_CMDQ_MIN_ID; + file_priv->cmdq_limit.max = IVPU_CMDQ_MAX_ID; mutex_unlock(&vdev->context_list_lock); drm_dev_exit(idx); @@ -269,8 +277,6 @@ static int ivpu_open(struct drm_device *dev, struct drm_file *file) return 0; -err_xa_erase: - xa_erase_irq(&vdev->context_xa, ctx_id); err_unlock: mutex_unlock(&vdev->context_list_lock); mutex_destroy(&file_priv->ms_lock); @@ -346,7 +352,7 @@ static int ivpu_hw_sched_init(struct ivpu_device *vdev) { int ret = 0; - if (vdev->hw->sched_mode == VPU_SCHEDULING_MODE_HW) { + if (vdev->fw->sched_mode == VPU_SCHEDULING_MODE_HW) { ret = ivpu_jsm_hws_setup_priority_bands(vdev); if (ret) { ivpu_err(vdev, "Failed to enable hw scheduler: %d", ret); @@ -380,10 +386,7 @@ int ivpu_boot(struct ivpu_device *vdev) ret = ivpu_wait_for_ready(vdev); if (ret) { ivpu_err(vdev, "Failed to boot the firmware: %d\n", ret); - ivpu_hw_diagnose_failure(vdev); - ivpu_mmu_evtq_dump(vdev); - ivpu_fw_log_dump(vdev); - return ret; + goto err_diagnose_failure; } ivpu_hw_irq_clear(vdev); @@ -394,12 +397,20 @@ int ivpu_boot(struct ivpu_device *vdev) if (ivpu_fw_is_cold_boot(vdev)) { ret = ivpu_pm_dct_init(vdev); if (ret) - return ret; + goto err_diagnose_failure; - return ivpu_hw_sched_init(vdev); + ret = ivpu_hw_sched_init(vdev); + if (ret) + goto err_diagnose_failure; } return 0; + +err_diagnose_failure: + ivpu_hw_diagnose_failure(vdev); + ivpu_mmu_evtq_dump(vdev); + ivpu_dev_coredump(vdev); + return ret; } void ivpu_prepare_for_reset(struct ivpu_device *vdev) @@ -446,9 +457,16 @@ static const struct drm_driver driver = { .name = DRIVER_NAME, .desc = DRIVER_DESC, + +#ifdef DRIVER_DATE .date = DRIVER_DATE, - .major = DRM_IVPU_DRIVER_MAJOR, - .minor = DRM_IVPU_DRIVER_MINOR, + .major = DRIVER_MAJOR, + .minor = DRIVER_MINOR, + .patchlevel = DRIVER_PATCHLEVEL, +#else + .date = UTS_RELEASE, + .major = 1, +#endif }; static void ivpu_context_abort_invalid(struct ivpu_device *vdev) @@ -606,6 +624,9 @@ static int ivpu_dev_init(struct ivpu_device *vdev) lockdep_set_class(&vdev->submitted_jobs_xa.xa_lock, &submitted_jobs_xa_lock_class_key); INIT_LIST_HEAD(&vdev->bo_list); + vdev->db_limit.min = IVPU_MIN_DB; + vdev->db_limit.max = IVPU_MAX_DB; + ret = drmm_mutex_init(&vdev->drm, &vdev->context_list_lock); if (ret) goto err_xa_destroy; @@ -632,9 +653,7 @@ static int ivpu_dev_init(struct ivpu_device *vdev) if (ret) goto err_shutdown; - ret = ivpu_mmu_global_context_init(vdev); - if (ret) - goto err_shutdown; + ivpu_mmu_global_context_init(vdev); ret = ivpu_mmu_init(vdev); if (ret) @@ -722,6 +741,7 @@ static struct pci_device_id ivpu_pci_ids[] = { { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_MTL) }, { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_ARL) }, { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_LNL) }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_PTL_P) }, { } }; MODULE_DEVICE_TABLE(pci, ivpu_pci_ids); diff --git a/drivers/accel/ivpu/ivpu_drv.h b/drivers/accel/ivpu/ivpu_drv.h index 63f13b697eed..3fdff3f6cffd 100644 --- a/drivers/accel/ivpu/ivpu_drv.h +++ b/drivers/accel/ivpu/ivpu_drv.h @@ -21,11 +21,11 @@ #define DRIVER_NAME "intel_vpu" #define DRIVER_DESC "Driver for Intel NPU (Neural Processing Unit)" -#define DRIVER_DATE "20230117" -#define PCI_DEVICE_ID_MTL 0x7d1d -#define PCI_DEVICE_ID_ARL 0xad1d -#define PCI_DEVICE_ID_LNL 0x643e +#define PCI_DEVICE_ID_MTL 0x7d1d +#define PCI_DEVICE_ID_ARL 0xad1d +#define PCI_DEVICE_ID_LNL 0x643e +#define PCI_DEVICE_ID_PTL_P 0xb03e #define IVPU_HW_IP_37XX 37 #define IVPU_HW_IP_40XX 40 @@ -46,17 +46,22 @@ #define IVPU_MIN_DB 1 #define IVPU_MAX_DB 255 -#define IVPU_NUM_ENGINES 2 -#define IVPU_NUM_PRIORITIES 4 -#define IVPU_NUM_CMDQS_PER_CTX (IVPU_NUM_ENGINES * IVPU_NUM_PRIORITIES) +#define IVPU_JOB_ID_JOB_MASK GENMASK(7, 0) +#define IVPU_JOB_ID_CONTEXT_MASK GENMASK(31, 8) -#define IVPU_CMDQ_INDEX(engine, priority) ((engine) * IVPU_NUM_PRIORITIES + (priority)) +#define IVPU_NUM_PRIORITIES 4 +#define IVPU_NUM_CMDQS_PER_CTX (IVPU_NUM_PRIORITIES) + +#define IVPU_CMDQ_MIN_ID 1 +#define IVPU_CMDQ_MAX_ID 255 #define IVPU_PLATFORM_SILICON 0 #define IVPU_PLATFORM_SIMICS 2 #define IVPU_PLATFORM_FPGA 3 #define IVPU_PLATFORM_INVALID 8 +#define IVPU_SCHED_MODE_AUTO -1 + #define IVPU_DBG_REG BIT(0) #define IVPU_DBG_IRQ BIT(1) #define IVPU_DBG_MMU BIT(2) @@ -134,6 +139,8 @@ struct ivpu_device { struct xa_limit context_xa_limit; struct xarray db_xa; + struct xa_limit db_limit; + u32 db_next; struct mutex bo_list_lock; /* Protects bo_list */ struct list_head bo_list; @@ -152,6 +159,7 @@ struct ivpu_device { int tdr; int autosuspend; int d0i3_entry_msg; + int state_dump_msg; } timeout; }; @@ -163,11 +171,15 @@ struct ivpu_file_priv { struct kref ref; struct ivpu_device *vdev; struct mutex lock; /* Protects cmdq */ - struct ivpu_cmdq *cmdq[IVPU_NUM_CMDQS_PER_CTX]; + struct xarray cmdq_xa; struct ivpu_mmu_context ctx; struct mutex ms_lock; /* Protects ms_instance_list, ms_info_bo */ struct list_head ms_instance_list; struct ivpu_bo *ms_info_bo; + struct xa_limit job_limit; + u32 job_id_next; + struct xa_limit cmdq_limit; + u32 cmdq_id_next; bool has_mmu_faults; bool bound; bool aborted; @@ -185,9 +197,9 @@ extern bool ivpu_force_snoop; #define IVPU_TEST_MODE_NULL_SUBMISSION BIT(2) #define IVPU_TEST_MODE_D0I3_MSG_DISABLE BIT(4) #define IVPU_TEST_MODE_D0I3_MSG_ENABLE BIT(5) -#define IVPU_TEST_MODE_PREEMPTION_DISABLE BIT(6) -#define IVPU_TEST_MODE_HWS_EXTRA_EVENTS BIT(7) +#define IVPU_TEST_MODE_MIP_DISABLE BIT(6) #define IVPU_TEST_MODE_DISABLE_TIMEOUTS BIT(8) +#define IVPU_TEST_MODE_TURBO BIT(9) extern int ivpu_test_mode; struct ivpu_file_priv *ivpu_file_priv_get(struct ivpu_file_priv *file_priv); @@ -215,6 +227,8 @@ static inline int ivpu_hw_ip_gen(struct ivpu_device *vdev) return IVPU_HW_IP_37XX; case PCI_DEVICE_ID_LNL: return IVPU_HW_IP_40XX; + case PCI_DEVICE_ID_PTL_P: + return IVPU_HW_IP_50XX; default: dump_stack(); ivpu_err(vdev, "Unknown NPU IP generation\n"); @@ -229,6 +243,7 @@ static inline int ivpu_hw_btrs_gen(struct ivpu_device *vdev) case PCI_DEVICE_ID_ARL: return IVPU_HW_BTRS_MTL; case PCI_DEVICE_ID_LNL: + case PCI_DEVICE_ID_PTL_P: return IVPU_HW_BTRS_LNL; default: dump_stack(); diff --git a/drivers/accel/ivpu/ivpu_fw.c b/drivers/accel/ivpu/ivpu_fw.c index ede6165e09d9..6037ec0b3096 100644 --- a/drivers/accel/ivpu/ivpu_fw.c +++ b/drivers/accel/ivpu/ivpu_fw.c @@ -25,7 +25,6 @@ #define FW_SHAVE_NN_MAX_SIZE SZ_2M #define FW_RUNTIME_MIN_ADDR (FW_GLOBAL_MEM_START) #define FW_RUNTIME_MAX_ADDR (FW_GLOBAL_MEM_END - FW_SHARED_MEM_SIZE) -#define FW_VERSION_HEADER_SIZE SZ_4K #define FW_FILE_IMAGE_OFFSET (VPU_FW_HEADER_SIZE + FW_VERSION_HEADER_SIZE) #define WATCHDOG_MSS_REDIRECT 32 @@ -47,8 +46,10 @@ #define IVPU_FOCUS_PRESENT_TIMER_MS 1000 static char *ivpu_firmware; +#if IS_ENABLED(CONFIG_DRM_ACCEL_IVPU_DEBUG) module_param_named_unsafe(firmware, ivpu_firmware, charp, 0644); MODULE_PARM_DESC(firmware, "NPU firmware binary in /lib/firmware/.."); +#endif static struct { int gen; @@ -58,11 +59,14 @@ static struct { { IVPU_HW_IP_37XX, "intel/vpu/vpu_37xx_v0.0.bin" }, { IVPU_HW_IP_40XX, "vpu_40xx.bin" }, { IVPU_HW_IP_40XX, "intel/vpu/vpu_40xx_v0.0.bin" }, + { IVPU_HW_IP_50XX, "vpu_50xx.bin" }, + { IVPU_HW_IP_50XX, "intel/vpu/vpu_50xx_v0.0.bin" }, }; /* Production fw_names from the table above */ MODULE_FIRMWARE("intel/vpu/vpu_37xx_v0.0.bin"); MODULE_FIRMWARE("intel/vpu/vpu_40xx_v0.0.bin"); +MODULE_FIRMWARE("intel/vpu/vpu_50xx_v0.0.bin"); static int ivpu_fw_request(struct ivpu_device *vdev) { @@ -135,6 +139,15 @@ static bool is_within_range(u64 addr, size_t size, u64 range_start, size_t range return true; } +static u32 +ivpu_fw_sched_mode_select(struct ivpu_device *vdev, const struct vpu_firmware_header *fw_hdr) +{ + if (ivpu_sched_mode != IVPU_SCHED_MODE_AUTO) + return ivpu_sched_mode; + + return VPU_SCHEDULING_MODE_OS; +} + static int ivpu_fw_parse(struct ivpu_device *vdev) { struct ivpu_fw_info *fw = vdev->fw; @@ -191,8 +204,10 @@ static int ivpu_fw_parse(struct ivpu_device *vdev) ivpu_dbg(vdev, FW_BOOT, "Header version: 0x%x, format 0x%x\n", fw_hdr->header_version, fw_hdr->image_format); - ivpu_info(vdev, "Firmware: %s, version: %s", fw->name, - (const char *)fw_hdr + VPU_FW_HEADER_SIZE); + if (!scnprintf(fw->version, sizeof(fw->version), "%s", fw->file->data + VPU_FW_HEADER_SIZE)) + ivpu_warn(vdev, "Missing firmware version\n"); + + ivpu_info(vdev, "Firmware: %s, version: %s\n", fw->name, fw->version); if (IVPU_FW_CHECK_API_COMPAT(vdev, fw_hdr, BOOT, 3)) return -EINVAL; @@ -208,14 +223,16 @@ static int ivpu_fw_parse(struct ivpu_device *vdev) fw->cold_boot_entry_point = fw_hdr->entry_point; fw->entry_point = fw->cold_boot_entry_point; - fw->trace_level = min_t(u32, ivpu_log_level, IVPU_FW_LOG_FATAL); + fw->trace_level = min_t(u32, ivpu_fw_log_level, IVPU_FW_LOG_FATAL); fw->trace_destination_mask = VPU_TRACE_DESTINATION_VERBOSE_TRACING; fw->trace_hw_component_mask = -1; fw->dvfs_mode = 0; + fw->sched_mode = ivpu_fw_sched_mode_select(vdev, fw_hdr); fw->primary_preempt_buf_size = fw_hdr->preemption_buffer_1_size; fw->secondary_preempt_buf_size = fw_hdr->preemption_buffer_2_size; + ivpu_info(vdev, "Scheduler mode: %s\n", fw->sched_mode ? "HW" : "OS"); if (fw_hdr->ro_section_start_address && !is_within_range(fw_hdr->ro_section_start_address, fw_hdr->ro_section_size, @@ -311,7 +328,7 @@ static int ivpu_fw_mem_init(struct ivpu_device *vdev) goto err_free_fw_mem; } - if (ivpu_log_level <= IVPU_FW_LOG_INFO) + if (ivpu_fw_log_level <= IVPU_FW_LOG_INFO) log_verb_size = IVPU_FW_VERBOSE_BUFFER_LARGE_SIZE; else log_verb_size = IVPU_FW_VERBOSE_BUFFER_SMALL_SIZE; @@ -567,8 +584,10 @@ void ivpu_fw_boot_params_setup(struct ivpu_device *vdev, struct vpu_boot_params boot_params->ipc_payload_area_start = ipc_mem_rx->vpu_addr + ivpu_bo_size(ipc_mem_rx) / 2; boot_params->ipc_payload_area_size = ivpu_bo_size(ipc_mem_rx) / 2; - boot_params->global_aliased_pio_base = vdev->hw->ranges.user.start; - boot_params->global_aliased_pio_size = ivpu_hw_range_size(&vdev->hw->ranges.user); + if (ivpu_hw_ip_gen(vdev) == IVPU_HW_IP_37XX) { + boot_params->global_aliased_pio_base = vdev->hw->ranges.user.start; + boot_params->global_aliased_pio_size = ivpu_hw_range_size(&vdev->hw->ranges.user); + } /* Allow configuration for L2C_PAGE_TABLE with boot param value */ boot_params->autoconfig = 1; @@ -604,8 +623,8 @@ void ivpu_fw_boot_params_setup(struct ivpu_device *vdev, struct vpu_boot_params boot_params->punit_telemetry_sram_base = ivpu_hw_telemetry_offset_get(vdev); boot_params->punit_telemetry_sram_size = ivpu_hw_telemetry_size_get(vdev); boot_params->vpu_telemetry_enable = ivpu_hw_telemetry_enable_get(vdev); - boot_params->vpu_scheduling_mode = vdev->hw->sched_mode; - if (vdev->hw->sched_mode == VPU_SCHEDULING_MODE_HW) + boot_params->vpu_scheduling_mode = vdev->fw->sched_mode; + if (vdev->fw->sched_mode == VPU_SCHEDULING_MODE_HW) boot_params->vpu_focus_present_timer_ms = IVPU_FOCUS_PRESENT_TIMER_MS; boot_params->dvfs_mode = vdev->fw->dvfs_mode; if (!IVPU_WA(disable_d0i3_msg)) diff --git a/drivers/accel/ivpu/ivpu_fw.h b/drivers/accel/ivpu/ivpu_fw.h index 40d9d17be3f5..1d0b2bd9d65c 100644 --- a/drivers/accel/ivpu/ivpu_fw.h +++ b/drivers/accel/ivpu/ivpu_fw.h @@ -1,11 +1,16 @@ /* SPDX-License-Identifier: GPL-2.0-only */ /* - * Copyright (C) 2020-2023 Intel Corporation + * Copyright (C) 2020-2024 Intel Corporation */ #ifndef __IVPU_FW_H__ #define __IVPU_FW_H__ +#include "vpu_jsm_api.h" + +#define FW_VERSION_HEADER_SIZE SZ_4K +#define FW_VERSION_STR_SIZE SZ_256 + struct ivpu_device; struct ivpu_bo; struct vpu_boot_params; @@ -13,6 +18,7 @@ struct vpu_boot_params; struct ivpu_fw_info { const struct firmware *file; const char *name; + char version[FW_VERSION_STR_SIZE]; struct ivpu_bo *mem; struct ivpu_bo *mem_shave_nn; struct ivpu_bo *mem_log_crit; @@ -32,6 +38,7 @@ struct ivpu_fw_info { u32 secondary_preempt_buf_size; u64 read_only_addr; u32 read_only_size; + u32 sched_mode; }; int ivpu_fw_init(struct ivpu_device *vdev); diff --git a/drivers/accel/ivpu/ivpu_fw_log.c b/drivers/accel/ivpu/ivpu_fw_log.c index ef0adb5e0fbe..337c906b0210 100644 --- a/drivers/accel/ivpu/ivpu_fw_log.c +++ b/drivers/accel/ivpu/ivpu_fw_log.c @@ -1,6 +1,6 @@ // SPDX-License-Identifier: GPL-2.0-only /* - * Copyright (C) 2020-2023 Intel Corporation + * Copyright (C) 2020-2024 Intel Corporation */ #include @@ -15,19 +15,19 @@ #include "ivpu_fw_log.h" #include "ivpu_gem.h" -#define IVPU_FW_LOG_LINE_LENGTH 256 +#define IVPU_FW_LOG_LINE_LENGTH 256 -unsigned int ivpu_log_level = IVPU_FW_LOG_ERROR; -module_param(ivpu_log_level, uint, 0444); -MODULE_PARM_DESC(ivpu_log_level, - "NPU firmware default trace level: debug=" __stringify(IVPU_FW_LOG_DEBUG) +unsigned int ivpu_fw_log_level = IVPU_FW_LOG_ERROR; +module_param_named(fw_log_level, ivpu_fw_log_level, uint, 0444); +MODULE_PARM_DESC(fw_log_level, + "NPU firmware default log level: debug=" __stringify(IVPU_FW_LOG_DEBUG) " info=" __stringify(IVPU_FW_LOG_INFO) " warn=" __stringify(IVPU_FW_LOG_WARN) " error=" __stringify(IVPU_FW_LOG_ERROR) " fatal=" __stringify(IVPU_FW_LOG_FATAL)); -static int fw_log_ptr(struct ivpu_device *vdev, struct ivpu_bo *bo, u32 *offset, - struct vpu_tracing_buffer_header **log_header) +static int fw_log_from_bo(struct ivpu_device *vdev, struct ivpu_bo *bo, u32 *offset, + struct vpu_tracing_buffer_header **out_log) { struct vpu_tracing_buffer_header *log; @@ -48,7 +48,7 @@ static int fw_log_ptr(struct ivpu_device *vdev, struct ivpu_bo *bo, u32 *offset, return -EINVAL; } - *log_header = log; + *out_log = log; *offset += log->size; ivpu_dbg(vdev, FW_BOOT, @@ -59,7 +59,7 @@ static int fw_log_ptr(struct ivpu_device *vdev, struct ivpu_bo *bo, u32 *offset, return 0; } -static void buffer_print(char *buffer, u32 size, struct drm_printer *p) +static void fw_log_print_lines(char *buffer, u32 size, struct drm_printer *p) { char line[IVPU_FW_LOG_LINE_LENGTH]; u32 index = 0; @@ -87,56 +87,89 @@ static void buffer_print(char *buffer, u32 size, struct drm_printer *p) } line[index] = 0; if (index != 0) - drm_printf(p, "%s\n", line); + drm_printf(p, "%s", line); } -static void fw_log_print_buffer(struct ivpu_device *vdev, struct vpu_tracing_buffer_header *log, - const char *prefix, bool only_new_msgs, struct drm_printer *p) +static void fw_log_print_buffer(struct vpu_tracing_buffer_header *log, const char *prefix, + bool only_new_msgs, struct drm_printer *p) { - char *log_buffer = (void *)log + log->header_size; - u32 log_size = log->size - log->header_size; - u32 log_start = log->read_index; - u32 log_end = log->write_index; + char *log_data = (void *)log + log->header_size; + u32 data_size = log->size - log->header_size; + u32 log_start = only_new_msgs ? READ_ONCE(log->read_index) : 0; + u32 log_end = READ_ONCE(log->write_index); - if (!(log->write_index || log->wrap_count) || - (log->write_index == log->read_index && only_new_msgs)) { - drm_printf(p, "==== %s \"%s\" log empty ====\n", prefix, log->name); - return; + if (log->wrap_count == log->read_wrap_count) { + if (log_end <= log_start) { + drm_printf(p, "==== %s \"%s\" log empty ====\n", prefix, log->name); + return; + } + } else if (log->wrap_count == log->read_wrap_count + 1) { + if (log_end > log_start) + log_start = log_end; + } else { + log_start = log_end; } drm_printf(p, "==== %s \"%s\" log start ====\n", prefix, log->name); - if (log->write_index > log->read_index) { - buffer_print(log_buffer + log_start, log_end - log_start, p); + if (log_end > log_start) { + fw_log_print_lines(log_data + log_start, log_end - log_start, p); } else { - buffer_print(log_buffer + log_end, log_size - log_end, p); - buffer_print(log_buffer, log_end, p); + fw_log_print_lines(log_data + log_start, data_size - log_start, p); + fw_log_print_lines(log_data, log_end, p); } - drm_printf(p, "\x1b[0m"); + drm_printf(p, "\n\x1b[0m"); /* add new line and clear formatting */ drm_printf(p, "==== %s \"%s\" log end ====\n", prefix, log->name); } +static void +fw_log_print_all_in_bo(struct ivpu_device *vdev, const char *name, + struct ivpu_bo *bo, bool only_new_msgs, struct drm_printer *p) +{ + struct vpu_tracing_buffer_header *log; + u32 next = 0; + + while (fw_log_from_bo(vdev, bo, &next, &log) == 0) + fw_log_print_buffer(log, name, only_new_msgs, p); +} + void ivpu_fw_log_print(struct ivpu_device *vdev, bool only_new_msgs, struct drm_printer *p) { - struct vpu_tracing_buffer_header *log_header; - u32 next = 0; - - while (fw_log_ptr(vdev, vdev->fw->mem_log_crit, &next, &log_header) == 0) - fw_log_print_buffer(vdev, log_header, "NPU critical", only_new_msgs, p); - - next = 0; - while (fw_log_ptr(vdev, vdev->fw->mem_log_verb, &next, &log_header) == 0) - fw_log_print_buffer(vdev, log_header, "NPU verbose", only_new_msgs, p); + fw_log_print_all_in_bo(vdev, "NPU critical", vdev->fw->mem_log_crit, only_new_msgs, p); + fw_log_print_all_in_bo(vdev, "NPU verbose", vdev->fw->mem_log_verb, only_new_msgs, p); } -void ivpu_fw_log_clear(struct ivpu_device *vdev) +void ivpu_fw_log_mark_read(struct ivpu_device *vdev) { - struct vpu_tracing_buffer_header *log_header; - u32 next = 0; - - while (fw_log_ptr(vdev, vdev->fw->mem_log_crit, &next, &log_header) == 0) - log_header->read_index = log_header->write_index; + struct vpu_tracing_buffer_header *log; + u32 next; next = 0; - while (fw_log_ptr(vdev, vdev->fw->mem_log_verb, &next, &log_header) == 0) - log_header->read_index = log_header->write_index; + while (fw_log_from_bo(vdev, vdev->fw->mem_log_crit, &next, &log) == 0) { + log->read_index = READ_ONCE(log->write_index); + log->read_wrap_count = READ_ONCE(log->wrap_count); + } + + next = 0; + while (fw_log_from_bo(vdev, vdev->fw->mem_log_verb, &next, &log) == 0) { + log->read_index = READ_ONCE(log->write_index); + log->read_wrap_count = READ_ONCE(log->wrap_count); + } +} + +void ivpu_fw_log_reset(struct ivpu_device *vdev) +{ + struct vpu_tracing_buffer_header *log; + u32 next; + + next = 0; + while (fw_log_from_bo(vdev, vdev->fw->mem_log_crit, &next, &log) == 0) { + log->read_index = 0; + log->read_wrap_count = 0; + } + + next = 0; + while (fw_log_from_bo(vdev, vdev->fw->mem_log_verb, &next, &log) == 0) { + log->read_index = 0; + log->read_wrap_count = 0; + } } diff --git a/drivers/accel/ivpu/ivpu_fw_log.h b/drivers/accel/ivpu/ivpu_fw_log.h index 0b2573f6f315..8bb528a73cb7 100644 --- a/drivers/accel/ivpu/ivpu_fw_log.h +++ b/drivers/accel/ivpu/ivpu_fw_log.h @@ -1,6 +1,6 @@ /* SPDX-License-Identifier: GPL-2.0-only */ /* - * Copyright (C) 2020-2023 Intel Corporation + * Copyright (C) 2020-2024 Intel Corporation */ #ifndef __IVPU_FW_LOG_H__ @@ -8,8 +8,6 @@ #include -#include - #include "ivpu_drv.h" #define IVPU_FW_LOG_DEFAULT 0 @@ -19,20 +17,15 @@ #define IVPU_FW_LOG_ERROR 4 #define IVPU_FW_LOG_FATAL 5 -extern unsigned int ivpu_log_level; - #define IVPU_FW_VERBOSE_BUFFER_SMALL_SIZE SZ_1M #define IVPU_FW_VERBOSE_BUFFER_LARGE_SIZE SZ_8M #define IVPU_FW_CRITICAL_BUFFER_SIZE SZ_512K +extern unsigned int ivpu_fw_log_level; + void ivpu_fw_log_print(struct ivpu_device *vdev, bool only_new_msgs, struct drm_printer *p); -void ivpu_fw_log_clear(struct ivpu_device *vdev); +void ivpu_fw_log_mark_read(struct ivpu_device *vdev); +void ivpu_fw_log_reset(struct ivpu_device *vdev); -static inline void ivpu_fw_log_dump(struct ivpu_device *vdev) -{ - struct drm_printer p = drm_info_printer(vdev->drm.dev); - - ivpu_fw_log_print(vdev, false, &p); -} #endif /* __IVPU_FW_LOG_H__ */ diff --git a/drivers/accel/ivpu/ivpu_gem.c b/drivers/accel/ivpu/ivpu_gem.c index 1b409dbd332d..d8e97a760fbc 100644 --- a/drivers/accel/ivpu/ivpu_gem.c +++ b/drivers/accel/ivpu/ivpu_gem.c @@ -384,6 +384,9 @@ int ivpu_bo_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file timeout = drm_timeout_abs_to_jiffies(args->timeout_ns); + /* Add 1 jiffy to ensure the wait function never times out before intended timeout_ns */ + timeout += 1; + obj = drm_gem_object_lookup(file, args->handle); if (!obj) return -EINVAL; diff --git a/drivers/accel/ivpu/ivpu_hw.c b/drivers/accel/ivpu/ivpu_hw.c index e69c0613513f..4e1054f3466e 100644 --- a/drivers/accel/ivpu/ivpu_hw.c +++ b/drivers/accel/ivpu/ivpu_hw.c @@ -89,12 +89,14 @@ static void timeouts_init(struct ivpu_device *vdev) vdev->timeout.tdr = 2000000; vdev->timeout.autosuspend = -1; vdev->timeout.d0i3_entry_msg = 500; + vdev->timeout.state_dump_msg = 10; } else if (ivpu_is_simics(vdev)) { vdev->timeout.boot = 50; vdev->timeout.jsm = 500; vdev->timeout.tdr = 10000; - vdev->timeout.autosuspend = -1; + vdev->timeout.autosuspend = 100; vdev->timeout.d0i3_entry_msg = 100; + vdev->timeout.state_dump_msg = 10; } else { vdev->timeout.boot = 1000; vdev->timeout.jsm = 500; @@ -104,6 +106,7 @@ static void timeouts_init(struct ivpu_device *vdev) else vdev->timeout.autosuspend = 100; vdev->timeout.d0i3_entry_msg = 5; + vdev->timeout.state_dump_msg = 10; } } @@ -111,14 +114,14 @@ static void memory_ranges_init(struct ivpu_device *vdev) { if (ivpu_hw_ip_gen(vdev) == IVPU_HW_IP_37XX) { ivpu_hw_range_init(&vdev->hw->ranges.global, 0x80000000, SZ_512M); - ivpu_hw_range_init(&vdev->hw->ranges.user, 0xc0000000, 255 * SZ_1M); + ivpu_hw_range_init(&vdev->hw->ranges.user, 0x88000000, 511 * SZ_1M); ivpu_hw_range_init(&vdev->hw->ranges.shave, 0x180000000, SZ_2G); - ivpu_hw_range_init(&vdev->hw->ranges.dma, 0x200000000, SZ_8G); + ivpu_hw_range_init(&vdev->hw->ranges.dma, 0x200000000, SZ_128G); } else { ivpu_hw_range_init(&vdev->hw->ranges.global, 0x80000000, SZ_512M); - ivpu_hw_range_init(&vdev->hw->ranges.user, 0x80000000, SZ_256M); - ivpu_hw_range_init(&vdev->hw->ranges.shave, 0x80000000 + SZ_256M, SZ_2G - SZ_256M); - ivpu_hw_range_init(&vdev->hw->ranges.dma, 0x200000000, SZ_8G); + ivpu_hw_range_init(&vdev->hw->ranges.shave, 0x80000000, SZ_2G); + ivpu_hw_range_init(&vdev->hw->ranges.user, 0x100000000, SZ_256G); + vdev->hw->ranges.dma = vdev->hw->ranges.user; } } diff --git a/drivers/accel/ivpu/ivpu_hw.h b/drivers/accel/ivpu/ivpu_hw.h index a96a05b2acda..fc4dbfc980c8 100644 --- a/drivers/accel/ivpu/ivpu_hw.h +++ b/drivers/accel/ivpu/ivpu_hw.h @@ -46,7 +46,6 @@ struct ivpu_hw_info { u32 profiling_freq; } pll; u32 tile_fuse; - u32 sched_mode; u32 sku; u16 config; int dma_bits; diff --git a/drivers/accel/ivpu/ivpu_hw_40xx_reg.h b/drivers/accel/ivpu/ivpu_hw_40xx_reg.h index d0b795b344c7..fc0ee8d637f9 100644 --- a/drivers/accel/ivpu/ivpu_hw_40xx_reg.h +++ b/drivers/accel/ivpu/ivpu_hw_40xx_reg.h @@ -115,6 +115,8 @@ #define VPU_50XX_HOST_SS_AON_PWR_ISLAND_EN_POST_DLY 0x00030068u #define VPU_50XX_HOST_SS_AON_PWR_ISLAND_EN_POST_DLY_POST_DLY_MASK GENMASK(7, 0) +#define VPU_50XX_HOST_SS_AON_PWR_ISLAND_EN_POST_DLY_POST1_DLY_MASK GENMASK(15, 8) +#define VPU_50XX_HOST_SS_AON_PWR_ISLAND_EN_POST_DLY_POST2_DLY_MASK GENMASK(23, 16) #define VPU_50XX_HOST_SS_AON_PWR_ISLAND_STATUS_DLY 0x0003006cu #define VPU_50XX_HOST_SS_AON_PWR_ISLAND_STATUS_DLY_STATUS_DLY_MASK GENMASK(7, 0) diff --git a/drivers/accel/ivpu/ivpu_hw_btrs.c b/drivers/accel/ivpu/ivpu_hw_btrs.c index 745e5248803d..3212c99f3682 100644 --- a/drivers/accel/ivpu/ivpu_hw_btrs.c +++ b/drivers/accel/ivpu/ivpu_hw_btrs.c @@ -141,16 +141,10 @@ static int read_tile_config_fuse(struct ivpu_device *vdev, u32 *tile_fuse_config } config = REG_GET_FLD(VPU_HW_BTRS_LNL_TILE_FUSE, CONFIG, fuse); - if (!tile_disable_check(config)) { - ivpu_err(vdev, "Fuse: Invalid tile disable config (0x%x)\n", config); - return -EIO; - } + if (!tile_disable_check(config)) + ivpu_warn(vdev, "More than 1 tile disabled, tile fuse config mask: 0x%x\n", config); - if (config) - ivpu_dbg(vdev, MISC, "Fuse: %d tiles enabled. Tile number %d disabled\n", - BTRS_LNL_TILE_MAX_NUM - 1, ffs(config) - 1); - else - ivpu_dbg(vdev, MISC, "Fuse: All %d tiles enabled\n", BTRS_LNL_TILE_MAX_NUM); + ivpu_dbg(vdev, MISC, "Tile disable config mask: 0x%x\n", config); *tile_fuse_config = config; return 0; @@ -163,7 +157,6 @@ static int info_init_mtl(struct ivpu_device *vdev) hw->tile_fuse = BTRS_MTL_TILE_FUSE_ENABLE_BOTH; hw->sku = BTRS_MTL_TILE_SKU_BOTH; hw->config = BTRS_MTL_WP_CONFIG_2_TILE_4_3_RATIO; - hw->sched_mode = ivpu_sched_mode; return 0; } @@ -178,7 +171,6 @@ static int info_init_lnl(struct ivpu_device *vdev) if (ret) return ret; - hw->sched_mode = ivpu_sched_mode; hw->tile_fuse = tile_fuse_config; hw->pll.profiling_freq = PLL_PROFILING_FREQ_DEFAULT; @@ -315,10 +307,6 @@ static void prepare_wp_request(struct ivpu_device *vdev, struct wp_request *wp, wp->cdyn = enable ? PLL_CDYN_DEFAULT : 0; wp->epp = enable ? PLL_EPP_DEFAULT : 0; } - - /* Simics cannot start without at least one tile */ - if (enable && ivpu_is_simics(vdev)) - wp->cfg = 1; } static int wait_for_pll_lock(struct ivpu_device *vdev, bool enable) @@ -465,9 +453,6 @@ int ivpu_hw_btrs_wait_for_clock_res_own_ack(struct ivpu_device *vdev) if (ivpu_hw_btrs_gen(vdev) == IVPU_HW_BTRS_MTL) return 0; - if (ivpu_is_simics(vdev)) - return 0; - return REGB_POLL_FLD(VPU_HW_BTRS_LNL_VPU_STATUS, CLOCK_RESOURCE_OWN_ACK, 1, TIMEOUT_US); } diff --git a/drivers/accel/ivpu/ivpu_hw_ip.c b/drivers/accel/ivpu/ivpu_hw_ip.c index 60b33fc59d96..029dd065614b 100644 --- a/drivers/accel/ivpu/ivpu_hw_ip.c +++ b/drivers/accel/ivpu/ivpu_hw_ip.c @@ -8,15 +8,12 @@ #include "ivpu_hw.h" #include "ivpu_hw_37xx_reg.h" #include "ivpu_hw_40xx_reg.h" +#include "ivpu_hw_btrs.h" #include "ivpu_hw_ip.h" #include "ivpu_hw_reg_io.h" #include "ivpu_mmu.h" #include "ivpu_pm.h" -#define PWR_ISLAND_EN_POST_DLY_FREQ_DEFAULT 0 -#define PWR_ISLAND_EN_POST_DLY_FREQ_HIGH 18 -#define PWR_ISLAND_STATUS_DLY_FREQ_DEFAULT 3 -#define PWR_ISLAND_STATUS_DLY_FREQ_HIGH 46 #define PWR_ISLAND_STATUS_TIMEOUT_US (5 * USEC_PER_MSEC) #define TIM_SAFE_ENABLE 0xf1d0dead @@ -268,20 +265,15 @@ void ivpu_hw_ip_idle_gen_disable(struct ivpu_device *vdev) idle_gen_drive_40xx(vdev, false); } -static void pwr_island_delay_set_50xx(struct ivpu_device *vdev) +static void +pwr_island_delay_set_50xx(struct ivpu_device *vdev, u32 post, u32 post1, u32 post2, u32 status) { - u32 val, post, status; - - if (vdev->hw->pll.profiling_freq == PLL_PROFILING_FREQ_DEFAULT) { - post = PWR_ISLAND_EN_POST_DLY_FREQ_DEFAULT; - status = PWR_ISLAND_STATUS_DLY_FREQ_DEFAULT; - } else { - post = PWR_ISLAND_EN_POST_DLY_FREQ_HIGH; - status = PWR_ISLAND_STATUS_DLY_FREQ_HIGH; - } + u32 val; val = REGV_RD32(VPU_50XX_HOST_SS_AON_PWR_ISLAND_EN_POST_DLY); val = REG_SET_FLD_NUM(VPU_50XX_HOST_SS_AON_PWR_ISLAND_EN_POST_DLY, POST_DLY, post, val); + val = REG_SET_FLD_NUM(VPU_50XX_HOST_SS_AON_PWR_ISLAND_EN_POST_DLY, POST1_DLY, post1, val); + val = REG_SET_FLD_NUM(VPU_50XX_HOST_SS_AON_PWR_ISLAND_EN_POST_DLY, POST2_DLY, post2, val); REGV_WR32(VPU_50XX_HOST_SS_AON_PWR_ISLAND_EN_POST_DLY, val); val = REGV_RD32(VPU_50XX_HOST_SS_AON_PWR_ISLAND_STATUS_DLY); @@ -311,9 +303,6 @@ static void pwr_island_trickle_drive_40xx(struct ivpu_device *vdev, bool enable) val = REG_CLR_FLD(VPU_40XX_HOST_SS_AON_PWR_ISLAND_TRICKLE_EN0, CSS_CPU, val); REGV_WR32(VPU_40XX_HOST_SS_AON_PWR_ISLAND_TRICKLE_EN0, val); - - if (enable) - ndelay(500); } static void pwr_island_drive_37xx(struct ivpu_device *vdev, bool enable) @@ -326,9 +315,6 @@ static void pwr_island_drive_37xx(struct ivpu_device *vdev, bool enable) val = REG_CLR_FLD(VPU_40XX_HOST_SS_AON_PWR_ISLAND_EN0, CSS_CPU, val); REGV_WR32(VPU_40XX_HOST_SS_AON_PWR_ISLAND_EN0, val); - - if (!enable) - ndelay(500); } static void pwr_island_drive_40xx(struct ivpu_device *vdev, bool enable) @@ -347,9 +333,11 @@ static void pwr_island_enable(struct ivpu_device *vdev) { if (ivpu_hw_ip_gen(vdev) == IVPU_HW_IP_37XX) { pwr_island_trickle_drive_37xx(vdev, true); + ndelay(500); pwr_island_drive_37xx(vdev, true); } else { pwr_island_trickle_drive_40xx(vdev, true); + ndelay(500); pwr_island_drive_40xx(vdev, true); } } @@ -686,13 +674,36 @@ static void dpu_active_drive_37xx(struct ivpu_device *vdev, bool enable) REGV_WR32(VPU_37XX_HOST_SS_AON_DPU_ACTIVE, val); } +static void pwr_island_delay_set(struct ivpu_device *vdev) +{ + bool high = vdev->hw->pll.profiling_freq == PLL_PROFILING_FREQ_HIGH; + u32 post, post1, post2, status; + + if (ivpu_hw_ip_gen(vdev) < IVPU_HW_IP_50XX) + return; + + switch (ivpu_device_id(vdev)) { + case PCI_DEVICE_ID_PTL_P: + post = high ? 18 : 0; + post1 = 0; + post2 = 0; + status = high ? 46 : 3; + break; + + default: + dump_stack(); + ivpu_err(vdev, "Unknown device ID\n"); + return; + } + + pwr_island_delay_set_50xx(vdev, post, post1, post2, status); +} + int ivpu_hw_ip_pwr_domain_enable(struct ivpu_device *vdev) { int ret; - if (ivpu_hw_ip_gen(vdev) == IVPU_HW_IP_50XX) - pwr_island_delay_set_50xx(vdev); - + pwr_island_delay_set(vdev); pwr_island_enable(vdev); ret = wait_for_pwr_island_status(vdev, 0x1); diff --git a/drivers/accel/ivpu/ivpu_ipc.c b/drivers/accel/ivpu/ivpu_ipc.c index 78b32a823241..01ebf88fe6ef 100644 --- a/drivers/accel/ivpu/ivpu_ipc.c +++ b/drivers/accel/ivpu/ivpu_ipc.c @@ -15,6 +15,7 @@ #include "ivpu_ipc.h" #include "ivpu_jsm_msg.h" #include "ivpu_pm.h" +#include "ivpu_trace.h" #define IPC_MAX_RX_MSG 128 @@ -227,6 +228,7 @@ int ivpu_ipc_send(struct ivpu_device *vdev, struct ivpu_ipc_consumer *cons, stru goto unlock; ivpu_ipc_tx(vdev, cons->tx_vpu_addr); + trace_jsm("[tx]", req); unlock: mutex_unlock(&ipc->lock); @@ -278,12 +280,13 @@ int ivpu_ipc_receive(struct ivpu_device *vdev, struct ivpu_ipc_consumer *cons, u32 size = min_t(int, rx_msg->ipc_hdr->data_size, sizeof(*jsm_msg)); if (rx_msg->jsm_msg->result != VPU_JSM_STATUS_SUCCESS) { - ivpu_dbg(vdev, IPC, "IPC resp result error: %d\n", rx_msg->jsm_msg->result); + ivpu_err(vdev, "IPC resp result error: %d\n", rx_msg->jsm_msg->result); ret = -EBADMSG; } if (jsm_msg) memcpy(jsm_msg, rx_msg->jsm_msg, size); + trace_jsm("[rx]", rx_msg->jsm_msg); } ivpu_ipc_rx_msg_del(vdev, rx_msg); @@ -291,15 +294,16 @@ int ivpu_ipc_receive(struct ivpu_device *vdev, struct ivpu_ipc_consumer *cons, return ret; } -static int +int ivpu_ipc_send_receive_internal(struct ivpu_device *vdev, struct vpu_jsm_msg *req, enum vpu_ipc_msg_type expected_resp_type, - struct vpu_jsm_msg *resp, u32 channel, - unsigned long timeout_ms) + struct vpu_jsm_msg *resp, u32 channel, unsigned long timeout_ms) { struct ivpu_ipc_consumer cons; int ret; + drm_WARN_ON(&vdev->drm, pm_runtime_status_suspended(vdev->drm.dev)); + ivpu_ipc_consumer_add(vdev, &cons, channel, NULL); ret = ivpu_ipc_send(vdev, &cons, req); @@ -325,19 +329,21 @@ consumer_del: return ret; } -int ivpu_ipc_send_receive_active(struct ivpu_device *vdev, struct vpu_jsm_msg *req, - enum vpu_ipc_msg_type expected_resp, struct vpu_jsm_msg *resp, - u32 channel, unsigned long timeout_ms) +int ivpu_ipc_send_receive(struct ivpu_device *vdev, struct vpu_jsm_msg *req, + enum vpu_ipc_msg_type expected_resp, struct vpu_jsm_msg *resp, + u32 channel, unsigned long timeout_ms) { struct vpu_jsm_msg hb_req = { .type = VPU_JSM_MSG_QUERY_ENGINE_HB }; struct vpu_jsm_msg hb_resp; int ret, hb_ret; - drm_WARN_ON(&vdev->drm, pm_runtime_status_suspended(vdev->drm.dev)); + ret = ivpu_rpm_get(vdev); + if (ret < 0) + return ret; ret = ivpu_ipc_send_receive_internal(vdev, req, expected_resp, resp, channel, timeout_ms); if (ret != -ETIMEDOUT) - return ret; + goto rpm_put; hb_ret = ivpu_ipc_send_receive_internal(vdev, &hb_req, VPU_JSM_MSG_QUERY_ENGINE_HB_DONE, &hb_resp, VPU_IPC_CHAN_ASYNC_CMD, @@ -345,21 +351,33 @@ int ivpu_ipc_send_receive_active(struct ivpu_device *vdev, struct vpu_jsm_msg *r if (hb_ret == -ETIMEDOUT) ivpu_pm_trigger_recovery(vdev, "IPC timeout"); +rpm_put: + ivpu_rpm_put(vdev); return ret; } -int ivpu_ipc_send_receive(struct ivpu_device *vdev, struct vpu_jsm_msg *req, - enum vpu_ipc_msg_type expected_resp, struct vpu_jsm_msg *resp, - u32 channel, unsigned long timeout_ms) +int ivpu_ipc_send_and_wait(struct ivpu_device *vdev, struct vpu_jsm_msg *req, + u32 channel, unsigned long timeout_ms) { + struct ivpu_ipc_consumer cons; int ret; ret = ivpu_rpm_get(vdev); if (ret < 0) return ret; - ret = ivpu_ipc_send_receive_active(vdev, req, expected_resp, resp, channel, timeout_ms); + ivpu_ipc_consumer_add(vdev, &cons, channel, NULL); + ret = ivpu_ipc_send(vdev, &cons, req); + if (ret) { + ivpu_warn_ratelimited(vdev, "IPC send failed: %d\n", ret); + goto consumer_del; + } + + msleep(timeout_ms); + +consumer_del: + ivpu_ipc_consumer_del(vdev, &cons); ivpu_rpm_put(vdev); return ret; } @@ -518,7 +536,6 @@ void ivpu_ipc_fini(struct ivpu_device *vdev) { struct ivpu_ipc_info *ipc = vdev->ipc; - drm_WARN_ON(&vdev->drm, ipc->on); drm_WARN_ON(&vdev->drm, !list_empty(&ipc->cons_list)); drm_WARN_ON(&vdev->drm, !list_empty(&ipc->cb_msg_list)); drm_WARN_ON(&vdev->drm, atomic_read(&ipc->rx_msg_count) > 0); diff --git a/drivers/accel/ivpu/ivpu_ipc.h b/drivers/accel/ivpu/ivpu_ipc.h index 4fe38141045e..b4dfb504679b 100644 --- a/drivers/accel/ivpu/ivpu_ipc.h +++ b/drivers/accel/ivpu/ivpu_ipc.h @@ -101,12 +101,13 @@ int ivpu_ipc_send(struct ivpu_device *vdev, struct ivpu_ipc_consumer *cons, int ivpu_ipc_receive(struct ivpu_device *vdev, struct ivpu_ipc_consumer *cons, struct ivpu_ipc_hdr *ipc_buf, struct vpu_jsm_msg *jsm_msg, unsigned long timeout_ms); - -int ivpu_ipc_send_receive_active(struct ivpu_device *vdev, struct vpu_jsm_msg *req, - enum vpu_ipc_msg_type expected_resp, struct vpu_jsm_msg *resp, - u32 channel, unsigned long timeout_ms); +int ivpu_ipc_send_receive_internal(struct ivpu_device *vdev, struct vpu_jsm_msg *req, + enum vpu_ipc_msg_type expected_resp_type, + struct vpu_jsm_msg *resp, u32 channel, unsigned long timeout_ms); int ivpu_ipc_send_receive(struct ivpu_device *vdev, struct vpu_jsm_msg *req, enum vpu_ipc_msg_type expected_resp, struct vpu_jsm_msg *resp, u32 channel, unsigned long timeout_ms); +int ivpu_ipc_send_and_wait(struct ivpu_device *vdev, struct vpu_jsm_msg *req, + u32 channel, unsigned long timeout_ms); #endif /* __IVPU_IPC_H__ */ diff --git a/drivers/accel/ivpu/ivpu_job.c b/drivers/accel/ivpu/ivpu_job.c index be2e2bf0f43f..7149312f16e1 100644 --- a/drivers/accel/ivpu/ivpu_job.c +++ b/drivers/accel/ivpu/ivpu_job.c @@ -18,11 +18,10 @@ #include "ivpu_job.h" #include "ivpu_jsm_msg.h" #include "ivpu_pm.h" +#include "ivpu_trace.h" #include "vpu_boot_api.h" #define CMD_BUF_IDX 0 -#define JOB_ID_JOB_MASK GENMASK(7, 0) -#define JOB_ID_CONTEXT_MASK GENMASK(31, 8) #define JOB_MAX_BUFFER_COUNT 65535 static void ivpu_cmdq_ring_db(struct ivpu_device *vdev, struct ivpu_cmdq *cmdq) @@ -35,24 +34,20 @@ static int ivpu_preemption_buffers_create(struct ivpu_device *vdev, { u64 primary_size = ALIGN(vdev->fw->primary_preempt_buf_size, PAGE_SIZE); u64 secondary_size = ALIGN(vdev->fw->secondary_preempt_buf_size, PAGE_SIZE); - struct ivpu_addr_range range; - if (vdev->hw->sched_mode != VPU_SCHEDULING_MODE_HW) + if (vdev->fw->sched_mode != VPU_SCHEDULING_MODE_HW || + ivpu_test_mode & IVPU_TEST_MODE_MIP_DISABLE) return 0; - range.start = vdev->hw->ranges.user.end - (primary_size * IVPU_NUM_CMDQS_PER_CTX); - range.end = vdev->hw->ranges.user.end; - cmdq->primary_preempt_buf = ivpu_bo_create(vdev, &file_priv->ctx, &range, primary_size, - DRM_IVPU_BO_WC); + cmdq->primary_preempt_buf = ivpu_bo_create(vdev, &file_priv->ctx, &vdev->hw->ranges.user, + primary_size, DRM_IVPU_BO_WC); if (!cmdq->primary_preempt_buf) { ivpu_err(vdev, "Failed to create primary preemption buffer\n"); return -ENOMEM; } - range.start = vdev->hw->ranges.shave.end - (secondary_size * IVPU_NUM_CMDQS_PER_CTX); - range.end = vdev->hw->ranges.shave.end; - cmdq->secondary_preempt_buf = ivpu_bo_create(vdev, &file_priv->ctx, &range, secondary_size, - DRM_IVPU_BO_WC); + cmdq->secondary_preempt_buf = ivpu_bo_create(vdev, &file_priv->ctx, &vdev->hw->ranges.dma, + secondary_size, DRM_IVPU_BO_WC); if (!cmdq->secondary_preempt_buf) { ivpu_err(vdev, "Failed to create secondary preemption buffer\n"); goto err_free_primary; @@ -62,24 +57,24 @@ static int ivpu_preemption_buffers_create(struct ivpu_device *vdev, err_free_primary: ivpu_bo_free(cmdq->primary_preempt_buf); + cmdq->primary_preempt_buf = NULL; return -ENOMEM; } static void ivpu_preemption_buffers_free(struct ivpu_device *vdev, struct ivpu_file_priv *file_priv, struct ivpu_cmdq *cmdq) { - if (vdev->hw->sched_mode != VPU_SCHEDULING_MODE_HW) + if (vdev->fw->sched_mode != VPU_SCHEDULING_MODE_HW) return; - drm_WARN_ON(&vdev->drm, !cmdq->primary_preempt_buf); - drm_WARN_ON(&vdev->drm, !cmdq->secondary_preempt_buf); - ivpu_bo_free(cmdq->primary_preempt_buf); - ivpu_bo_free(cmdq->secondary_preempt_buf); + if (cmdq->primary_preempt_buf) + ivpu_bo_free(cmdq->primary_preempt_buf); + if (cmdq->secondary_preempt_buf) + ivpu_bo_free(cmdq->secondary_preempt_buf); } static struct ivpu_cmdq *ivpu_cmdq_alloc(struct ivpu_file_priv *file_priv) { - struct xa_limit db_xa_limit = {.max = IVPU_MAX_DB, .min = IVPU_MIN_DB}; struct ivpu_device *vdev = file_priv->vdev; struct ivpu_cmdq *cmdq; int ret; @@ -88,25 +83,33 @@ static struct ivpu_cmdq *ivpu_cmdq_alloc(struct ivpu_file_priv *file_priv) if (!cmdq) return NULL; - ret = xa_alloc(&vdev->db_xa, &cmdq->db_id, NULL, db_xa_limit, GFP_KERNEL); - if (ret) { + ret = xa_alloc_cyclic(&vdev->db_xa, &cmdq->db_id, NULL, vdev->db_limit, &vdev->db_next, + GFP_KERNEL); + if (ret < 0) { ivpu_err(vdev, "Failed to allocate doorbell id: %d\n", ret); goto err_free_cmdq; } + ret = xa_alloc_cyclic(&file_priv->cmdq_xa, &cmdq->id, cmdq, file_priv->cmdq_limit, + &file_priv->cmdq_id_next, GFP_KERNEL); + if (ret < 0) { + ivpu_err(vdev, "Failed to allocate command queue id: %d\n", ret); + goto err_erase_db_xa; + } + cmdq->mem = ivpu_bo_create_global(vdev, SZ_4K, DRM_IVPU_BO_WC | DRM_IVPU_BO_MAPPABLE); if (!cmdq->mem) - goto err_erase_xa; + goto err_erase_cmdq_xa; ret = ivpu_preemption_buffers_create(vdev, file_priv, cmdq); if (ret) - goto err_free_cmdq_mem; + ivpu_warn(vdev, "Failed to allocate preemption buffers, preemption limited\n"); return cmdq; -err_free_cmdq_mem: - ivpu_bo_free(cmdq->mem); -err_erase_xa: +err_erase_cmdq_xa: + xa_erase(&file_priv->cmdq_xa, cmdq->id); +err_erase_db_xa: xa_erase(&vdev->db_xa, cmdq->db_id); err_free_cmdq: kfree(cmdq); @@ -130,13 +133,13 @@ static int ivpu_hws_cmdq_init(struct ivpu_file_priv *file_priv, struct ivpu_cmdq struct ivpu_device *vdev = file_priv->vdev; int ret; - ret = ivpu_jsm_hws_create_cmdq(vdev, file_priv->ctx.id, file_priv->ctx.id, cmdq->db_id, + ret = ivpu_jsm_hws_create_cmdq(vdev, file_priv->ctx.id, file_priv->ctx.id, cmdq->id, task_pid_nr(current), engine, cmdq->mem->vpu_addr, ivpu_bo_size(cmdq->mem)); if (ret) return ret; - ret = ivpu_jsm_hws_set_context_sched_properties(vdev, file_priv->ctx.id, cmdq->db_id, + ret = ivpu_jsm_hws_set_context_sched_properties(vdev, file_priv->ctx.id, cmdq->id, priority); if (ret) return ret; @@ -149,21 +152,22 @@ static int ivpu_register_db(struct ivpu_file_priv *file_priv, struct ivpu_cmdq * struct ivpu_device *vdev = file_priv->vdev; int ret; - if (vdev->hw->sched_mode == VPU_SCHEDULING_MODE_HW) - ret = ivpu_jsm_hws_register_db(vdev, file_priv->ctx.id, cmdq->db_id, cmdq->db_id, + if (vdev->fw->sched_mode == VPU_SCHEDULING_MODE_HW) + ret = ivpu_jsm_hws_register_db(vdev, file_priv->ctx.id, cmdq->id, cmdq->db_id, cmdq->mem->vpu_addr, ivpu_bo_size(cmdq->mem)); else ret = ivpu_jsm_register_db(vdev, file_priv->ctx.id, cmdq->db_id, cmdq->mem->vpu_addr, ivpu_bo_size(cmdq->mem)); if (!ret) - ivpu_dbg(vdev, JOB, "DB %d registered to ctx %d\n", cmdq->db_id, file_priv->ctx.id); + ivpu_dbg(vdev, JOB, "DB %d registered to cmdq %d ctx %d\n", + cmdq->db_id, cmdq->id, file_priv->ctx.id); return ret; } static int -ivpu_cmdq_init(struct ivpu_file_priv *file_priv, struct ivpu_cmdq *cmdq, u16 engine, u8 priority) +ivpu_cmdq_init(struct ivpu_file_priv *file_priv, struct ivpu_cmdq *cmdq, u8 priority) { struct ivpu_device *vdev = file_priv->vdev; struct vpu_job_queue_header *jobq_header; @@ -179,13 +183,18 @@ ivpu_cmdq_init(struct ivpu_file_priv *file_priv, struct ivpu_cmdq *cmdq, u16 eng cmdq->jobq = (struct vpu_job_queue *)ivpu_bo_vaddr(cmdq->mem); jobq_header = &cmdq->jobq->header; - jobq_header->engine_idx = engine; + jobq_header->engine_idx = VPU_ENGINE_COMPUTE; jobq_header->head = 0; jobq_header->tail = 0; + if (ivpu_test_mode & IVPU_TEST_MODE_TURBO) { + ivpu_dbg(vdev, JOB, "Turbo mode enabled"); + jobq_header->flags = VPU_JOB_QUEUE_FLAGS_TURBO_MODE; + } + wmb(); /* Flush WC buffer for jobq->header */ - if (vdev->hw->sched_mode == VPU_SCHEDULING_MODE_HW) { - ret = ivpu_hws_cmdq_init(file_priv, cmdq, engine, priority); + if (vdev->fw->sched_mode == VPU_SCHEDULING_MODE_HW) { + ret = ivpu_hws_cmdq_init(file_priv, cmdq, VPU_ENGINE_COMPUTE, priority); if (ret) return ret; } @@ -211,10 +220,10 @@ static int ivpu_cmdq_fini(struct ivpu_file_priv *file_priv, struct ivpu_cmdq *cm cmdq->db_registered = false; - if (vdev->hw->sched_mode == VPU_SCHEDULING_MODE_HW) { - ret = ivpu_jsm_hws_destroy_cmdq(vdev, file_priv->ctx.id, cmdq->db_id); + if (vdev->fw->sched_mode == VPU_SCHEDULING_MODE_HW) { + ret = ivpu_jsm_hws_destroy_cmdq(vdev, file_priv->ctx.id, cmdq->id); if (!ret) - ivpu_dbg(vdev, JOB, "Command queue %d destroyed\n", cmdq->db_id); + ivpu_dbg(vdev, JOB, "Command queue %d destroyed\n", cmdq->id); } ret = ivpu_jsm_unregister_db(vdev, cmdq->db_id); @@ -224,55 +233,46 @@ static int ivpu_cmdq_fini(struct ivpu_file_priv *file_priv, struct ivpu_cmdq *cm return 0; } -static struct ivpu_cmdq *ivpu_cmdq_acquire(struct ivpu_file_priv *file_priv, u16 engine, - u8 priority) +static struct ivpu_cmdq *ivpu_cmdq_acquire(struct ivpu_file_priv *file_priv, u8 priority) { - int cmdq_idx = IVPU_CMDQ_INDEX(engine, priority); - struct ivpu_cmdq *cmdq = file_priv->cmdq[cmdq_idx]; + struct ivpu_cmdq *cmdq; + unsigned long cmdq_id; int ret; lockdep_assert_held(&file_priv->lock); + xa_for_each(&file_priv->cmdq_xa, cmdq_id, cmdq) + if (cmdq->priority == priority) + break; + if (!cmdq) { cmdq = ivpu_cmdq_alloc(file_priv); if (!cmdq) return NULL; - file_priv->cmdq[cmdq_idx] = cmdq; + cmdq->priority = priority; } - ret = ivpu_cmdq_init(file_priv, cmdq, engine, priority); + ret = ivpu_cmdq_init(file_priv, cmdq, priority); if (ret) return NULL; return cmdq; } -static void ivpu_cmdq_release_locked(struct ivpu_file_priv *file_priv, u16 engine, u8 priority) +void ivpu_cmdq_release_all_locked(struct ivpu_file_priv *file_priv) { - int cmdq_idx = IVPU_CMDQ_INDEX(engine, priority); - struct ivpu_cmdq *cmdq = file_priv->cmdq[cmdq_idx]; + struct ivpu_cmdq *cmdq; + unsigned long cmdq_id; lockdep_assert_held(&file_priv->lock); - if (cmdq) { - file_priv->cmdq[cmdq_idx] = NULL; + xa_for_each(&file_priv->cmdq_xa, cmdq_id, cmdq) { + xa_erase(&file_priv->cmdq_xa, cmdq_id); ivpu_cmdq_fini(file_priv, cmdq); ivpu_cmdq_free(file_priv, cmdq); } } -void ivpu_cmdq_release_all_locked(struct ivpu_file_priv *file_priv) -{ - u16 engine; - u8 priority; - - lockdep_assert_held(&file_priv->lock); - - for (engine = 0; engine < IVPU_NUM_ENGINES; engine++) - for (priority = 0; priority < IVPU_NUM_PRIORITIES; priority++) - ivpu_cmdq_release_locked(file_priv, engine, priority); -} - /* * Mark the doorbell as unregistered * This function needs to be called when the VPU hardware is restarted @@ -281,20 +281,13 @@ void ivpu_cmdq_release_all_locked(struct ivpu_file_priv *file_priv) */ static void ivpu_cmdq_reset(struct ivpu_file_priv *file_priv) { - u16 engine; - u8 priority; + struct ivpu_cmdq *cmdq; + unsigned long cmdq_id; mutex_lock(&file_priv->lock); - for (engine = 0; engine < IVPU_NUM_ENGINES; engine++) { - for (priority = 0; priority < IVPU_NUM_PRIORITIES; priority++) { - int cmdq_idx = IVPU_CMDQ_INDEX(engine, priority); - struct ivpu_cmdq *cmdq = file_priv->cmdq[cmdq_idx]; - - if (cmdq) - cmdq->db_registered = false; - } - } + xa_for_each(&file_priv->cmdq_xa, cmdq_id, cmdq) + cmdq->db_registered = false; mutex_unlock(&file_priv->lock); } @@ -314,17 +307,11 @@ void ivpu_cmdq_reset_all_contexts(struct ivpu_device *vdev) static void ivpu_cmdq_fini_all(struct ivpu_file_priv *file_priv) { - u16 engine; - u8 priority; + struct ivpu_cmdq *cmdq; + unsigned long cmdq_id; - for (engine = 0; engine < IVPU_NUM_ENGINES; engine++) { - for (priority = 0; priority < IVPU_NUM_PRIORITIES; priority++) { - int cmdq_idx = IVPU_CMDQ_INDEX(engine, priority); - - if (file_priv->cmdq[cmdq_idx]) - ivpu_cmdq_fini(file_priv, file_priv->cmdq[cmdq_idx]); - } - } + xa_for_each(&file_priv->cmdq_xa, cmdq_id, cmdq) + ivpu_cmdq_fini(file_priv, cmdq); } void ivpu_context_abort_locked(struct ivpu_file_priv *file_priv) @@ -335,7 +322,7 @@ void ivpu_context_abort_locked(struct ivpu_file_priv *file_priv) ivpu_cmdq_fini_all(file_priv); - if (vdev->hw->sched_mode == VPU_SCHEDULING_MODE_OS) + if (vdev->fw->sched_mode == VPU_SCHEDULING_MODE_OS) ivpu_jsm_context_release(vdev, file_priv->ctx.id); } @@ -349,24 +336,29 @@ static int ivpu_cmdq_push_job(struct ivpu_cmdq *cmdq, struct ivpu_job *job) /* Check if there is space left in job queue */ if (next_entry == header->head) { - ivpu_dbg(vdev, JOB, "Job queue full: ctx %d engine %d db %d head %d tail %d\n", - job->file_priv->ctx.id, job->engine_idx, cmdq->db_id, header->head, tail); + ivpu_dbg(vdev, JOB, "Job queue full: ctx %d cmdq %d db %d head %d tail %d\n", + job->file_priv->ctx.id, cmdq->id, cmdq->db_id, header->head, tail); return -EBUSY; } - entry = &cmdq->jobq->job[tail]; + entry = &cmdq->jobq->slot[tail].job; entry->batch_buf_addr = job->cmd_buf_vpu_addr; entry->job_id = job->job_id; entry->flags = 0; if (unlikely(ivpu_test_mode & IVPU_TEST_MODE_NULL_SUBMISSION)) entry->flags = VPU_JOB_FLAGS_NULL_SUBMISSION_MASK; - if (vdev->hw->sched_mode == VPU_SCHEDULING_MODE_HW && - (unlikely(!(ivpu_test_mode & IVPU_TEST_MODE_PREEMPTION_DISABLE)))) { - entry->primary_preempt_buf_addr = cmdq->primary_preempt_buf->vpu_addr; - entry->primary_preempt_buf_size = ivpu_bo_size(cmdq->primary_preempt_buf); - entry->secondary_preempt_buf_addr = cmdq->secondary_preempt_buf->vpu_addr; - entry->secondary_preempt_buf_size = ivpu_bo_size(cmdq->secondary_preempt_buf); + if (vdev->fw->sched_mode == VPU_SCHEDULING_MODE_HW) { + if (cmdq->primary_preempt_buf) { + entry->primary_preempt_buf_addr = cmdq->primary_preempt_buf->vpu_addr; + entry->primary_preempt_buf_size = ivpu_bo_size(cmdq->primary_preempt_buf); + } + + if (cmdq->secondary_preempt_buf) { + entry->secondary_preempt_buf_addr = cmdq->secondary_preempt_buf->vpu_addr; + entry->secondary_preempt_buf_size = + ivpu_bo_size(cmdq->secondary_preempt_buf); + } } wmb(); /* Ensure that tail is updated after filling entry */ @@ -457,6 +449,7 @@ ivpu_job_create(struct ivpu_file_priv *file_priv, u32 engine_idx, u32 bo_count) job->file_priv = ivpu_file_priv_get(file_priv); + trace_job("create", job); ivpu_dbg(vdev, JOB, "Job created: ctx %2d engine %d", file_priv->ctx.id, job->engine_idx); return job; @@ -496,6 +489,7 @@ static int ivpu_job_signal_and_destroy(struct ivpu_device *vdev, u32 job_id, u32 job->bos[CMD_BUF_IDX]->job_status = job_status; dma_fence_signal(job->done_fence); + trace_job("done", job); ivpu_dbg(vdev, JOB, "Job complete: id %3u ctx %2d engine %d status 0x%x\n", job->job_id, job->file_priv->ctx.id, job->engine_idx, job_status); @@ -519,7 +513,6 @@ static int ivpu_job_submit(struct ivpu_job *job, u8 priority) { struct ivpu_file_priv *file_priv = job->file_priv; struct ivpu_device *vdev = job->vdev; - struct xa_limit job_id_range; struct ivpu_cmdq *cmdq; bool is_first_job; int ret; @@ -530,7 +523,7 @@ static int ivpu_job_submit(struct ivpu_job *job, u8 priority) mutex_lock(&file_priv->lock); - cmdq = ivpu_cmdq_acquire(job->file_priv, job->engine_idx, priority); + cmdq = ivpu_cmdq_acquire(file_priv, priority); if (!cmdq) { ivpu_warn_ratelimited(vdev, "Failed to get job queue, ctx %d engine %d prio %d\n", file_priv->ctx.id, job->engine_idx, priority); @@ -538,13 +531,11 @@ static int ivpu_job_submit(struct ivpu_job *job, u8 priority) goto err_unlock_file_priv; } - job_id_range.min = FIELD_PREP(JOB_ID_CONTEXT_MASK, (file_priv->ctx.id - 1)); - job_id_range.max = job_id_range.min | JOB_ID_JOB_MASK; - xa_lock(&vdev->submitted_jobs_xa); is_first_job = xa_empty(&vdev->submitted_jobs_xa); - ret = __xa_alloc(&vdev->submitted_jobs_xa, &job->job_id, job, job_id_range, GFP_KERNEL); - if (ret) { + ret = __xa_alloc_cyclic(&vdev->submitted_jobs_xa, &job->job_id, job, file_priv->job_limit, + &file_priv->job_id_next, GFP_KERNEL); + if (ret < 0) { ivpu_dbg(vdev, JOB, "Too many active jobs in ctx %d\n", file_priv->ctx.id); ret = -EBUSY; @@ -566,6 +557,7 @@ static int ivpu_job_submit(struct ivpu_job *job, u8 priority) vdev->busy_start_ts = ktime_get(); } + trace_job("submit", job); ivpu_dbg(vdev, JOB, "Job submitted: id %3u ctx %2d engine %d prio %d addr 0x%llx next %d\n", job->job_id, file_priv->ctx.id, job->engine_idx, priority, job->cmd_buf_vpu_addr, cmdq->jobq->header.tail); @@ -673,7 +665,7 @@ int ivpu_submit_ioctl(struct drm_device *dev, void *data, struct drm_file *file) int idx, ret; u8 priority; - if (params->engine > DRM_IVPU_ENGINE_COPY) + if (params->engine != DRM_IVPU_ENGINE_COMPUTE) return -EINVAL; if (params->priority > DRM_IVPU_JOB_PRIORITY_REALTIME) diff --git a/drivers/accel/ivpu/ivpu_job.h b/drivers/accel/ivpu/ivpu_job.h index 6accb94028c7..8b19e3f8b4cf 100644 --- a/drivers/accel/ivpu/ivpu_job.h +++ b/drivers/accel/ivpu/ivpu_job.h @@ -28,8 +28,10 @@ struct ivpu_cmdq { struct ivpu_bo *secondary_preempt_buf; struct ivpu_bo *mem; u32 entry_count; + u32 id; u32 db_id; bool db_registered; + u8 priority; }; /** diff --git a/drivers/accel/ivpu/ivpu_jsm_msg.c b/drivers/accel/ivpu/ivpu_jsm_msg.c index 46ef16c3c069..30a40be76930 100644 --- a/drivers/accel/ivpu/ivpu_jsm_msg.c +++ b/drivers/accel/ivpu/ivpu_jsm_msg.c @@ -48,9 +48,10 @@ const char *ivpu_jsm_msg_type_to_str(enum vpu_ipc_msg_type type) IVPU_CASE_TO_STR(VPU_JSM_MSG_HWS_RESUME_ENGINE_DONE); IVPU_CASE_TO_STR(VPU_JSM_MSG_STATE_DUMP); IVPU_CASE_TO_STR(VPU_JSM_MSG_STATE_DUMP_RSP); - IVPU_CASE_TO_STR(VPU_JSM_MSG_BLOB_DEINIT); + IVPU_CASE_TO_STR(VPU_JSM_MSG_BLOB_DEINIT_DEPRECATED); IVPU_CASE_TO_STR(VPU_JSM_MSG_DYNDBG_CONTROL); IVPU_CASE_TO_STR(VPU_JSM_MSG_JOB_DONE); + IVPU_CASE_TO_STR(VPU_JSM_MSG_NATIVE_FENCE_SIGNALLED); IVPU_CASE_TO_STR(VPU_JSM_MSG_ENGINE_RESET_DONE); IVPU_CASE_TO_STR(VPU_JSM_MSG_ENGINE_PREEMPT_DONE); IVPU_CASE_TO_STR(VPU_JSM_MSG_REGISTER_DB_DONE); @@ -131,7 +132,7 @@ int ivpu_jsm_get_heartbeat(struct ivpu_device *vdev, u32 engine, u64 *heartbeat) struct vpu_jsm_msg resp; int ret; - if (engine > VPU_ENGINE_COPY) + if (engine != VPU_ENGINE_COMPUTE) return -EINVAL; req.payload.query_engine_hb.engine_idx = engine; @@ -154,7 +155,7 @@ int ivpu_jsm_reset_engine(struct ivpu_device *vdev, u32 engine) struct vpu_jsm_msg resp; int ret; - if (engine > VPU_ENGINE_COPY) + if (engine != VPU_ENGINE_COMPUTE) return -EINVAL; req.payload.engine_reset.engine_idx = engine; @@ -173,7 +174,7 @@ int ivpu_jsm_preempt_engine(struct ivpu_device *vdev, u32 engine, u32 preempt_id struct vpu_jsm_msg resp; int ret; - if (engine > VPU_ENGINE_COPY) + if (engine != VPU_ENGINE_COMPUTE) return -EINVAL; req.payload.engine_preempt.engine_idx = engine; @@ -196,7 +197,7 @@ int ivpu_jsm_dyndbg_control(struct ivpu_device *vdev, char *command, size_t size strscpy(req.payload.dyndbg_control.dyndbg_cmd, command, VPU_DYNDBG_CMD_MAX_LEN); ret = ivpu_ipc_send_receive(vdev, &req, VPU_JSM_MSG_DYNDBG_CONTROL_RSP, &resp, - VPU_IPC_CHAN_ASYNC_CMD, vdev->timeout.jsm); + VPU_IPC_CHAN_GEN_CMD, vdev->timeout.jsm); if (ret) ivpu_warn_ratelimited(vdev, "Failed to send command \"%s\": ret %d\n", command, ret); @@ -270,9 +271,8 @@ int ivpu_jsm_pwr_d0i3_enter(struct ivpu_device *vdev) req.payload.pwr_d0i3_enter.send_response = 1; - ret = ivpu_ipc_send_receive_active(vdev, &req, VPU_JSM_MSG_PWR_D0I3_ENTER_DONE, - &resp, VPU_IPC_CHAN_GEN_CMD, - vdev->timeout.d0i3_entry_msg); + ret = ivpu_ipc_send_receive_internal(vdev, &req, VPU_JSM_MSG_PWR_D0I3_ENTER_DONE, &resp, + VPU_IPC_CHAN_GEN_CMD, vdev->timeout.d0i3_entry_msg); if (ret) return ret; @@ -346,7 +346,7 @@ int ivpu_jsm_hws_resume_engine(struct ivpu_device *vdev, u32 engine) struct vpu_jsm_msg resp; int ret; - if (engine >= VPU_ENGINE_NB) + if (engine != VPU_ENGINE_COMPUTE) return -EINVAL; req.payload.hws_resume_engine.engine_idx = engine; @@ -394,8 +394,6 @@ int ivpu_jsm_hws_set_scheduling_log(struct ivpu_device *vdev, u32 engine_idx, u3 req.payload.hws_set_scheduling_log.host_ssid = host_ssid; req.payload.hws_set_scheduling_log.vpu_log_buffer_va = vpu_log_buffer_va; req.payload.hws_set_scheduling_log.notify_index = 0; - req.payload.hws_set_scheduling_log.enable_extra_events = - ivpu_test_mode & IVPU_TEST_MODE_HWS_EXTRA_EVENTS; ret = ivpu_ipc_send_receive(vdev, &req, VPU_JSM_MSG_HWS_SET_SCHEDULING_LOG_RSP, &resp, VPU_IPC_CHAN_ASYNC_CMD, vdev->timeout.jsm); @@ -430,8 +428,8 @@ int ivpu_jsm_hws_setup_priority_bands(struct ivpu_device *vdev) req.payload.hws_priority_band_setup.normal_band_percentage = 10; - ret = ivpu_ipc_send_receive_active(vdev, &req, VPU_JSM_MSG_SET_PRIORITY_BAND_SETUP_RSP, - &resp, VPU_IPC_CHAN_ASYNC_CMD, vdev->timeout.jsm); + ret = ivpu_ipc_send_receive_internal(vdev, &req, VPU_JSM_MSG_SET_PRIORITY_BAND_SETUP_RSP, + &resp, VPU_IPC_CHAN_ASYNC_CMD, vdev->timeout.jsm); if (ret) ivpu_warn_ratelimited(vdev, "Failed to set priority bands: %d\n", ret); @@ -544,9 +542,8 @@ int ivpu_jsm_dct_enable(struct ivpu_device *vdev, u32 active_us, u32 inactive_us req.payload.pwr_dct_control.dct_active_us = active_us; req.payload.pwr_dct_control.dct_inactive_us = inactive_us; - return ivpu_ipc_send_receive_active(vdev, &req, VPU_JSM_MSG_DCT_ENABLE_DONE, - &resp, VPU_IPC_CHAN_ASYNC_CMD, - vdev->timeout.jsm); + return ivpu_ipc_send_receive_internal(vdev, &req, VPU_JSM_MSG_DCT_ENABLE_DONE, &resp, + VPU_IPC_CHAN_ASYNC_CMD, vdev->timeout.jsm); } int ivpu_jsm_dct_disable(struct ivpu_device *vdev) @@ -554,7 +551,14 @@ int ivpu_jsm_dct_disable(struct ivpu_device *vdev) struct vpu_jsm_msg req = { .type = VPU_JSM_MSG_DCT_DISABLE }; struct vpu_jsm_msg resp; - return ivpu_ipc_send_receive_active(vdev, &req, VPU_JSM_MSG_DCT_DISABLE_DONE, - &resp, VPU_IPC_CHAN_ASYNC_CMD, - vdev->timeout.jsm); + return ivpu_ipc_send_receive_internal(vdev, &req, VPU_JSM_MSG_DCT_DISABLE_DONE, &resp, + VPU_IPC_CHAN_ASYNC_CMD, vdev->timeout.jsm); +} + +int ivpu_jsm_state_dump(struct ivpu_device *vdev) +{ + struct vpu_jsm_msg req = { .type = VPU_JSM_MSG_STATE_DUMP }; + + return ivpu_ipc_send_and_wait(vdev, &req, VPU_IPC_CHAN_ASYNC_CMD, + vdev->timeout.state_dump_msg); } diff --git a/drivers/accel/ivpu/ivpu_jsm_msg.h b/drivers/accel/ivpu/ivpu_jsm_msg.h index e4e42c0ff6e6..9e84d3526a14 100644 --- a/drivers/accel/ivpu/ivpu_jsm_msg.h +++ b/drivers/accel/ivpu/ivpu_jsm_msg.h @@ -43,4 +43,6 @@ int ivpu_jsm_metric_streamer_info(struct ivpu_device *vdev, u64 metric_group_mas u64 buffer_size, u32 *sample_size, u64 *info_size); int ivpu_jsm_dct_enable(struct ivpu_device *vdev, u32 active_us, u32 inactive_us); int ivpu_jsm_dct_disable(struct ivpu_device *vdev); +int ivpu_jsm_state_dump(struct ivpu_device *vdev); + #endif diff --git a/drivers/accel/ivpu/ivpu_mmu.c b/drivers/accel/ivpu/ivpu_mmu.c index c078e214b221..26ef52fbb93e 100644 --- a/drivers/accel/ivpu/ivpu_mmu.c +++ b/drivers/accel/ivpu/ivpu_mmu.c @@ -696,7 +696,7 @@ unlock: return ret; } -static int ivpu_mmu_cd_add(struct ivpu_device *vdev, u32 ssid, u64 cd_dma) +static int ivpu_mmu_cdtab_entry_set(struct ivpu_device *vdev, u32 ssid, u64 cd_dma, bool valid) { struct ivpu_mmu_info *mmu = vdev->mmu; struct ivpu_mmu_cdtab *cdtab = &mmu->cdtab; @@ -708,30 +708,29 @@ static int ivpu_mmu_cd_add(struct ivpu_device *vdev, u32 ssid, u64 cd_dma) return -EINVAL; entry = cdtab->base + (ssid * IVPU_MMU_CDTAB_ENT_SIZE); + drm_WARN_ON(&vdev->drm, (entry[0] & IVPU_MMU_CD_0_V) == valid); - if (cd_dma != 0) { - cd[0] = FIELD_PREP(IVPU_MMU_CD_0_TCR_T0SZ, IVPU_MMU_T0SZ_48BIT) | - FIELD_PREP(IVPU_MMU_CD_0_TCR_TG0, 0) | - FIELD_PREP(IVPU_MMU_CD_0_TCR_IRGN0, 0) | - FIELD_PREP(IVPU_MMU_CD_0_TCR_ORGN0, 0) | - FIELD_PREP(IVPU_MMU_CD_0_TCR_SH0, 0) | - FIELD_PREP(IVPU_MMU_CD_0_TCR_IPS, IVPU_MMU_IPS_48BIT) | - FIELD_PREP(IVPU_MMU_CD_0_ASID, ssid) | - IVPU_MMU_CD_0_TCR_EPD1 | - IVPU_MMU_CD_0_AA64 | - IVPU_MMU_CD_0_R | - IVPU_MMU_CD_0_ASET | - IVPU_MMU_CD_0_V; - cd[1] = cd_dma & IVPU_MMU_CD_1_TTB0_MASK; - cd[2] = 0; - cd[3] = 0x0000000000007444; + cd[0] = FIELD_PREP(IVPU_MMU_CD_0_TCR_T0SZ, IVPU_MMU_T0SZ_48BIT) | + FIELD_PREP(IVPU_MMU_CD_0_TCR_TG0, 0) | + FIELD_PREP(IVPU_MMU_CD_0_TCR_IRGN0, 0) | + FIELD_PREP(IVPU_MMU_CD_0_TCR_ORGN0, 0) | + FIELD_PREP(IVPU_MMU_CD_0_TCR_SH0, 0) | + FIELD_PREP(IVPU_MMU_CD_0_TCR_IPS, IVPU_MMU_IPS_48BIT) | + FIELD_PREP(IVPU_MMU_CD_0_ASID, ssid) | + IVPU_MMU_CD_0_TCR_EPD1 | + IVPU_MMU_CD_0_AA64 | + IVPU_MMU_CD_0_R | + IVPU_MMU_CD_0_ASET; + cd[1] = cd_dma & IVPU_MMU_CD_1_TTB0_MASK; + cd[2] = 0; + cd[3] = 0x0000000000007444; - /* For global context generate memory fault on VPU */ - if (ssid == IVPU_GLOBAL_CONTEXT_MMU_SSID) - cd[0] |= IVPU_MMU_CD_0_A; - } else { - memset(cd, 0, sizeof(cd)); - } + /* For global context generate memory fault on VPU */ + if (ssid == IVPU_GLOBAL_CONTEXT_MMU_SSID) + cd[0] |= IVPU_MMU_CD_0_A; + + if (valid) + cd[0] |= IVPU_MMU_CD_0_V; WRITE_ONCE(entry[1], cd[1]); WRITE_ONCE(entry[2], cd[2]); @@ -741,8 +740,8 @@ static int ivpu_mmu_cd_add(struct ivpu_device *vdev, u32 ssid, u64 cd_dma) if (!ivpu_is_force_snoop_enabled(vdev)) clflush_cache_range(entry, IVPU_MMU_CDTAB_ENT_SIZE); - ivpu_dbg(vdev, MMU, "CDTAB %s entry (SSID=%u, dma=%pad): 0x%llx, 0x%llx, 0x%llx, 0x%llx\n", - cd_dma ? "write" : "clear", ssid, &cd_dma, cd[0], cd[1], cd[2], cd[3]); + ivpu_dbg(vdev, MMU, "CDTAB set %s entry (SSID=%u, dma=%pad): 0x%llx, 0x%llx, 0x%llx, 0x%llx\n", + valid ? "valid" : "invalid", ssid, &cd_dma, cd[0], cd[1], cd[2], cd[3]); mutex_lock(&mmu->lock); if (!mmu->on) @@ -750,38 +749,18 @@ static int ivpu_mmu_cd_add(struct ivpu_device *vdev, u32 ssid, u64 cd_dma) ret = ivpu_mmu_cmdq_write_cfgi_all(vdev); if (ret) - goto unlock; + goto err_invalidate; ret = ivpu_mmu_cmdq_sync(vdev); + if (ret) + goto err_invalidate; unlock: mutex_unlock(&mmu->lock); - return ret; -} - -static int ivpu_mmu_cd_add_gbl(struct ivpu_device *vdev) -{ - int ret; - - ret = ivpu_mmu_cd_add(vdev, 0, vdev->gctx.pgtable.pgd_dma); - if (ret) - ivpu_err(vdev, "Failed to add global CD entry: %d\n", ret); - - return ret; -} - -static int ivpu_mmu_cd_add_user(struct ivpu_device *vdev, u32 ssid, dma_addr_t cd_dma) -{ - int ret; - - if (ssid == 0) { - ivpu_err(vdev, "Invalid SSID: %u\n", ssid); - return -EINVAL; - } - - ret = ivpu_mmu_cd_add(vdev, ssid, cd_dma); - if (ret) - ivpu_err(vdev, "Failed to add CD entry SSID=%u: %d\n", ssid, ret); + return 0; +err_invalidate: + WRITE_ONCE(entry[0], 0); + mutex_unlock(&mmu->lock); return ret; } @@ -808,12 +787,6 @@ int ivpu_mmu_init(struct ivpu_device *vdev) return ret; } - ret = ivpu_mmu_cd_add_gbl(vdev); - if (ret) { - ivpu_err(vdev, "Failed to initialize strtab: %d\n", ret); - return ret; - } - ret = ivpu_mmu_enable(vdev); if (ret) { ivpu_err(vdev, "Failed to resume MMU: %d\n", ret); @@ -966,12 +939,12 @@ void ivpu_mmu_irq_gerr_handler(struct ivpu_device *vdev) REGV_WR32(IVPU_MMU_REG_GERRORN, gerror_val); } -int ivpu_mmu_set_pgtable(struct ivpu_device *vdev, int ssid, struct ivpu_mmu_pgtable *pgtable) +int ivpu_mmu_cd_set(struct ivpu_device *vdev, int ssid, struct ivpu_mmu_pgtable *pgtable) { - return ivpu_mmu_cd_add_user(vdev, ssid, pgtable->pgd_dma); + return ivpu_mmu_cdtab_entry_set(vdev, ssid, pgtable->pgd_dma, true); } -void ivpu_mmu_clear_pgtable(struct ivpu_device *vdev, int ssid) +void ivpu_mmu_cd_clear(struct ivpu_device *vdev, int ssid) { - ivpu_mmu_cd_add_user(vdev, ssid, 0); /* 0 will clear CD entry */ + ivpu_mmu_cdtab_entry_set(vdev, ssid, 0, false); } diff --git a/drivers/accel/ivpu/ivpu_mmu.h b/drivers/accel/ivpu/ivpu_mmu.h index 6fa35c240710..7afea9cd8731 100644 --- a/drivers/accel/ivpu/ivpu_mmu.h +++ b/drivers/accel/ivpu/ivpu_mmu.h @@ -40,8 +40,8 @@ struct ivpu_mmu_info { int ivpu_mmu_init(struct ivpu_device *vdev); void ivpu_mmu_disable(struct ivpu_device *vdev); int ivpu_mmu_enable(struct ivpu_device *vdev); -int ivpu_mmu_set_pgtable(struct ivpu_device *vdev, int ssid, struct ivpu_mmu_pgtable *pgtable); -void ivpu_mmu_clear_pgtable(struct ivpu_device *vdev, int ssid); +int ivpu_mmu_cd_set(struct ivpu_device *vdev, int ssid, struct ivpu_mmu_pgtable *pgtable); +void ivpu_mmu_cd_clear(struct ivpu_device *vdev, int ssid); int ivpu_mmu_invalidate_tlb(struct ivpu_device *vdev, u16 ssid); void ivpu_mmu_irq_evtq_handler(struct ivpu_device *vdev); diff --git a/drivers/accel/ivpu/ivpu_mmu_context.c b/drivers/accel/ivpu/ivpu_mmu_context.c index bbe652a7019d..891967a95bc3 100644 --- a/drivers/accel/ivpu/ivpu_mmu_context.c +++ b/drivers/accel/ivpu/ivpu_mmu_context.c @@ -90,19 +90,6 @@ static void ivpu_pgtable_free_page(struct ivpu_device *vdev, u64 *cpu_addr, dma_ } } -static int ivpu_mmu_pgtable_init(struct ivpu_device *vdev, struct ivpu_mmu_pgtable *pgtable) -{ - dma_addr_t pgd_dma; - - pgtable->pgd_dma_ptr = ivpu_pgtable_alloc_page(vdev, &pgd_dma); - if (!pgtable->pgd_dma_ptr) - return -ENOMEM; - - pgtable->pgd_dma = pgd_dma; - - return 0; -} - static void ivpu_mmu_pgtables_free(struct ivpu_device *vdev, struct ivpu_mmu_pgtable *pgtable) { int pgd_idx, pud_idx, pmd_idx; @@ -140,6 +127,27 @@ static void ivpu_mmu_pgtables_free(struct ivpu_device *vdev, struct ivpu_mmu_pgt } ivpu_pgtable_free_page(vdev, pgtable->pgd_dma_ptr, pgtable->pgd_dma); + pgtable->pgd_dma_ptr = NULL; + pgtable->pgd_dma = 0; +} + +static u64* +ivpu_mmu_ensure_pgd(struct ivpu_device *vdev, struct ivpu_mmu_pgtable *pgtable) +{ + u64 *pgd_dma_ptr = pgtable->pgd_dma_ptr; + dma_addr_t pgd_dma; + + if (pgd_dma_ptr) + return pgd_dma_ptr; + + pgd_dma_ptr = ivpu_pgtable_alloc_page(vdev, &pgd_dma); + if (!pgd_dma_ptr) + return NULL; + + pgtable->pgd_dma_ptr = pgd_dma_ptr; + pgtable->pgd_dma = pgd_dma; + + return pgd_dma_ptr; } static u64* @@ -237,6 +245,12 @@ ivpu_mmu_context_map_page(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx int pmd_idx = FIELD_GET(IVPU_MMU_PMD_INDEX_MASK, vpu_addr); int pte_idx = FIELD_GET(IVPU_MMU_PTE_INDEX_MASK, vpu_addr); + drm_WARN_ON(&vdev->drm, ctx->id == IVPU_RESERVED_CONTEXT_MMU_SSID); + + /* Allocate PGD - first level page table if needed */ + if (!ivpu_mmu_ensure_pgd(vdev, &ctx->pgtable)) + return -ENOMEM; + /* Allocate PUD - second level page table if needed */ if (!ivpu_mmu_ensure_pud(vdev, &ctx->pgtable, pgd_idx)) return -ENOMEM; @@ -418,6 +432,7 @@ int ivpu_mmu_context_map_sgt(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx, u64 vpu_addr, struct sg_table *sgt, bool llc_coherent) { + size_t start_vpu_addr = vpu_addr; struct scatterlist *sg; int ret; u64 prot; @@ -448,20 +463,36 @@ ivpu_mmu_context_map_sgt(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx, ret = ivpu_mmu_context_map_pages(vdev, ctx, vpu_addr, dma_addr, size, prot); if (ret) { ivpu_err(vdev, "Failed to map context pages\n"); - mutex_unlock(&ctx->lock); - return ret; + goto err_unmap_pages; } vpu_addr += size; } + if (!ctx->is_cd_valid) { + ret = ivpu_mmu_cd_set(vdev, ctx->id, &ctx->pgtable); + if (ret) { + ivpu_err(vdev, "Failed to set context descriptor for context %u: %d\n", + ctx->id, ret); + goto err_unmap_pages; + } + ctx->is_cd_valid = true; + } + /* Ensure page table modifications are flushed from wc buffers to memory */ wmb(); - mutex_unlock(&ctx->lock); - ret = ivpu_mmu_invalidate_tlb(vdev, ctx->id); - if (ret) + if (ret) { ivpu_err(vdev, "Failed to invalidate TLB for ctx %u: %d\n", ctx->id, ret); + goto err_unmap_pages; + } + + mutex_unlock(&ctx->lock); + return 0; + +err_unmap_pages: + ivpu_mmu_context_unmap_pages(ctx, start_vpu_addr, vpu_addr - start_vpu_addr); + mutex_unlock(&ctx->lock); return ret; } @@ -530,65 +561,75 @@ ivpu_mmu_context_remove_node(struct ivpu_mmu_context *ctx, struct drm_mm_node *n mutex_unlock(&ctx->lock); } -static int -ivpu_mmu_context_init(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx, u32 context_id) +void ivpu_mmu_context_init(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx, u32 context_id) { u64 start, end; - int ret; mutex_init(&ctx->lock); - ret = ivpu_mmu_pgtable_init(vdev, &ctx->pgtable); - if (ret) { - ivpu_err(vdev, "Failed to initialize pgtable for ctx %u: %d\n", context_id, ret); - return ret; - } - if (!context_id) { start = vdev->hw->ranges.global.start; end = vdev->hw->ranges.shave.end; } else { - start = vdev->hw->ranges.user.start; - end = vdev->hw->ranges.dma.end; + start = min_t(u64, vdev->hw->ranges.user.start, vdev->hw->ranges.shave.start); + end = max_t(u64, vdev->hw->ranges.user.end, vdev->hw->ranges.dma.end); } drm_mm_init(&ctx->mm, start, end - start); ctx->id = context_id; - - return 0; } -static void ivpu_mmu_context_fini(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx) +void ivpu_mmu_context_fini(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx) { - if (drm_WARN_ON(&vdev->drm, !ctx->pgtable.pgd_dma_ptr)) - return; + if (ctx->is_cd_valid) { + ivpu_mmu_cd_clear(vdev, ctx->id); + ctx->is_cd_valid = false; + } mutex_destroy(&ctx->lock); ivpu_mmu_pgtables_free(vdev, &ctx->pgtable); drm_mm_takedown(&ctx->mm); - - ctx->pgtable.pgd_dma_ptr = NULL; - ctx->pgtable.pgd_dma = 0; } -int ivpu_mmu_global_context_init(struct ivpu_device *vdev) +void ivpu_mmu_global_context_init(struct ivpu_device *vdev) { - return ivpu_mmu_context_init(vdev, &vdev->gctx, IVPU_GLOBAL_CONTEXT_MMU_SSID); + ivpu_mmu_context_init(vdev, &vdev->gctx, IVPU_GLOBAL_CONTEXT_MMU_SSID); } void ivpu_mmu_global_context_fini(struct ivpu_device *vdev) { - return ivpu_mmu_context_fini(vdev, &vdev->gctx); + ivpu_mmu_context_fini(vdev, &vdev->gctx); } int ivpu_mmu_reserved_context_init(struct ivpu_device *vdev) { - return ivpu_mmu_user_context_init(vdev, &vdev->rctx, IVPU_RESERVED_CONTEXT_MMU_SSID); + int ret; + + ivpu_mmu_context_init(vdev, &vdev->rctx, IVPU_RESERVED_CONTEXT_MMU_SSID); + + mutex_lock(&vdev->rctx.lock); + + if (!ivpu_mmu_ensure_pgd(vdev, &vdev->rctx.pgtable)) { + ivpu_err(vdev, "Failed to allocate root page table for reserved context\n"); + ret = -ENOMEM; + goto unlock; + } + + ret = ivpu_mmu_cd_set(vdev, vdev->rctx.id, &vdev->rctx.pgtable); + if (ret) { + ivpu_err(vdev, "Failed to set context descriptor for reserved context\n"); + goto unlock; + } + +unlock: + mutex_unlock(&vdev->rctx.lock); + return ret; } void ivpu_mmu_reserved_context_fini(struct ivpu_device *vdev) { - return ivpu_mmu_user_context_fini(vdev, &vdev->rctx); + ivpu_mmu_cd_clear(vdev, vdev->rctx.id); + ivpu_mmu_context_fini(vdev, &vdev->rctx); } void ivpu_mmu_user_context_mark_invalid(struct ivpu_device *vdev, u32 ssid) @@ -603,36 +644,3 @@ void ivpu_mmu_user_context_mark_invalid(struct ivpu_device *vdev, u32 ssid) xa_unlock(&vdev->context_xa); } - -int ivpu_mmu_user_context_init(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx, u32 ctx_id) -{ - int ret; - - drm_WARN_ON(&vdev->drm, !ctx_id); - - ret = ivpu_mmu_context_init(vdev, ctx, ctx_id); - if (ret) { - ivpu_err(vdev, "Failed to initialize context %u: %d\n", ctx_id, ret); - return ret; - } - - ret = ivpu_mmu_set_pgtable(vdev, ctx_id, &ctx->pgtable); - if (ret) { - ivpu_err(vdev, "Failed to set page table for context %u: %d\n", ctx_id, ret); - goto err_context_fini; - } - - return 0; - -err_context_fini: - ivpu_mmu_context_fini(vdev, ctx); - return ret; -} - -void ivpu_mmu_user_context_fini(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx) -{ - drm_WARN_ON(&vdev->drm, !ctx->id); - - ivpu_mmu_clear_pgtable(vdev, ctx->id); - ivpu_mmu_context_fini(vdev, ctx); -} diff --git a/drivers/accel/ivpu/ivpu_mmu_context.h b/drivers/accel/ivpu/ivpu_mmu_context.h index 7f9aaf3d10c2..8042fc067062 100644 --- a/drivers/accel/ivpu/ivpu_mmu_context.h +++ b/drivers/accel/ivpu/ivpu_mmu_context.h @@ -23,19 +23,20 @@ struct ivpu_mmu_pgtable { }; struct ivpu_mmu_context { - struct mutex lock; /* Protects: mm, pgtable */ + struct mutex lock; /* Protects: mm, pgtable, is_cd_valid */ struct drm_mm mm; struct ivpu_mmu_pgtable pgtable; + bool is_cd_valid; u32 id; }; -int ivpu_mmu_global_context_init(struct ivpu_device *vdev); +void ivpu_mmu_context_init(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx, u32 context_id); +void ivpu_mmu_context_fini(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx); +void ivpu_mmu_global_context_init(struct ivpu_device *vdev); void ivpu_mmu_global_context_fini(struct ivpu_device *vdev); int ivpu_mmu_reserved_context_init(struct ivpu_device *vdev); void ivpu_mmu_reserved_context_fini(struct ivpu_device *vdev); -int ivpu_mmu_user_context_init(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx, u32 ctx_id); -void ivpu_mmu_user_context_fini(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx); void ivpu_mmu_user_context_mark_invalid(struct ivpu_device *vdev, u32 ssid); int ivpu_mmu_context_insert_node(struct ivpu_mmu_context *ctx, const struct ivpu_addr_range *range, diff --git a/drivers/accel/ivpu/ivpu_ms.c b/drivers/accel/ivpu/ivpu_ms.c index 2f9d37f5c208..ffe7b10f8a76 100644 --- a/drivers/accel/ivpu/ivpu_ms.c +++ b/drivers/accel/ivpu/ivpu_ms.c @@ -11,7 +11,7 @@ #include "ivpu_ms.h" #include "ivpu_pm.h" -#define MS_INFO_BUFFER_SIZE SZ_16K +#define MS_INFO_BUFFER_SIZE SZ_64K #define MS_NUM_BUFFERS 2 #define MS_READ_PERIOD_MULTIPLIER 2 #define MS_MIN_SAMPLE_PERIOD_NS 1000000 diff --git a/drivers/accel/ivpu/ivpu_pm.c b/drivers/accel/ivpu/ivpu_pm.c index 59d3170f5e35..dbc0711e28d1 100644 --- a/drivers/accel/ivpu/ivpu_pm.c +++ b/drivers/accel/ivpu/ivpu_pm.c @@ -9,21 +9,25 @@ #include #include -#include "vpu_boot_api.h" +#include "ivpu_coredump.h" #include "ivpu_drv.h" -#include "ivpu_hw.h" #include "ivpu_fw.h" #include "ivpu_fw_log.h" +#include "ivpu_hw.h" #include "ivpu_ipc.h" #include "ivpu_job.h" #include "ivpu_jsm_msg.h" #include "ivpu_mmu.h" #include "ivpu_ms.h" #include "ivpu_pm.h" +#include "ivpu_trace.h" +#include "vpu_boot_api.h" static bool ivpu_disable_recovery; +#if IS_ENABLED(CONFIG_DRM_ACCEL_IVPU_DEBUG) module_param_named_unsafe(disable_recovery, ivpu_disable_recovery, bool, 0644); MODULE_PARM_DESC(disable_recovery, "Disables recovery when NPU hang is detected"); +#endif static unsigned long ivpu_tdr_timeout_ms; module_param_named(tdr_timeout_ms, ivpu_tdr_timeout_ms, ulong, 0644); @@ -37,6 +41,7 @@ static void ivpu_pm_prepare_cold_boot(struct ivpu_device *vdev) ivpu_cmdq_reset_all_contexts(vdev); ivpu_ipc_reset(vdev); + ivpu_fw_log_reset(vdev); ivpu_fw_load(vdev); fw->entry_point = fw->cold_boot_entry_point; } @@ -123,7 +128,8 @@ static void ivpu_pm_recovery_work(struct work_struct *work) if (ret) ivpu_err(vdev, "Failed to resume NPU: %d\n", ret); - ivpu_fw_log_dump(vdev); + ivpu_jsm_state_dump(vdev); + ivpu_dev_coredump(vdev); atomic_inc(&vdev->pm->reset_counter); atomic_set(&vdev->pm->reset_pending, 1); @@ -195,6 +201,7 @@ int ivpu_pm_suspend_cb(struct device *dev) struct ivpu_device *vdev = to_ivpu_device(drm); unsigned long timeout; + trace_pm("suspend"); ivpu_dbg(vdev, PM, "Suspend..\n"); timeout = jiffies + msecs_to_jiffies(vdev->timeout.tdr); @@ -212,6 +219,7 @@ int ivpu_pm_suspend_cb(struct device *dev) ivpu_pm_prepare_warm_boot(vdev); ivpu_dbg(vdev, PM, "Suspend done.\n"); + trace_pm("suspend done"); return 0; } @@ -222,6 +230,7 @@ int ivpu_pm_resume_cb(struct device *dev) struct ivpu_device *vdev = to_ivpu_device(drm); int ret; + trace_pm("resume"); ivpu_dbg(vdev, PM, "Resume..\n"); ret = ivpu_resume(vdev); @@ -229,6 +238,7 @@ int ivpu_pm_resume_cb(struct device *dev) ivpu_err(vdev, "Failed to resume: %d\n", ret); ivpu_dbg(vdev, PM, "Resume done.\n"); + trace_pm("resume done"); return ret; } @@ -243,6 +253,7 @@ int ivpu_pm_runtime_suspend_cb(struct device *dev) drm_WARN_ON(&vdev->drm, !xa_empty(&vdev->submitted_jobs_xa)); drm_WARN_ON(&vdev->drm, work_pending(&vdev->pm->recovery_work)); + trace_pm("runtime suspend"); ivpu_dbg(vdev, PM, "Runtime suspend..\n"); ivpu_mmu_disable(vdev); @@ -262,13 +273,14 @@ int ivpu_pm_runtime_suspend_cb(struct device *dev) if (!is_idle || ret_d0i3) { ivpu_err(vdev, "Forcing cold boot due to previous errors\n"); atomic_inc(&vdev->pm->reset_counter); - ivpu_fw_log_dump(vdev); + ivpu_dev_coredump(vdev); ivpu_pm_prepare_cold_boot(vdev); } else { ivpu_pm_prepare_warm_boot(vdev); } ivpu_dbg(vdev, PM, "Runtime suspend done.\n"); + trace_pm("runtime suspend done"); return 0; } @@ -279,6 +291,7 @@ int ivpu_pm_runtime_resume_cb(struct device *dev) struct ivpu_device *vdev = to_ivpu_device(drm); int ret; + trace_pm("runtime resume"); ivpu_dbg(vdev, PM, "Runtime resume..\n"); ret = ivpu_resume(vdev); @@ -286,6 +299,7 @@ int ivpu_pm_runtime_resume_cb(struct device *dev) ivpu_err(vdev, "Failed to set RESUME state: %d\n", ret); ivpu_dbg(vdev, PM, "Runtime resume done.\n"); + trace_pm("runtime resume done"); return ret; } @@ -411,7 +425,7 @@ int ivpu_pm_dct_enable(struct ivpu_device *vdev, u8 active_percent) ret = ivpu_jsm_dct_enable(vdev, active_us, inactive_us); if (ret) { - ivpu_err_ratelimited(vdev, "Filed to enable DCT: %d\n", ret); + ivpu_err_ratelimited(vdev, "Failed to enable DCT: %d\n", ret); return ret; } @@ -428,7 +442,7 @@ int ivpu_pm_dct_disable(struct ivpu_device *vdev) ret = ivpu_jsm_dct_disable(vdev); if (ret) { - ivpu_err_ratelimited(vdev, "Filed to disable DCT: %d\n", ret); + ivpu_err_ratelimited(vdev, "Failed to disable DCT: %d\n", ret); return ret; } diff --git a/drivers/accel/ivpu/ivpu_sysfs.c b/drivers/accel/ivpu/ivpu_sysfs.c index 913669f1786e..616477fc17fa 100644 --- a/drivers/accel/ivpu/ivpu_sysfs.c +++ b/drivers/accel/ivpu/ivpu_sysfs.c @@ -6,6 +6,8 @@ #include #include +#include "ivpu_drv.h" +#include "ivpu_fw.h" #include "ivpu_hw.h" #include "ivpu_sysfs.h" @@ -39,8 +41,30 @@ npu_busy_time_us_show(struct device *dev, struct device_attribute *attr, char *b static DEVICE_ATTR_RO(npu_busy_time_us); +/** + * DOC: sched_mode + * + * The sched_mode is used to report current NPU scheduling mode. + * + * It returns following strings: + * - "HW" - Hardware Scheduler mode + * - "OS" - Operating System Scheduler mode + * + */ +static ssize_t +sched_mode_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct drm_device *drm = dev_get_drvdata(dev); + struct ivpu_device *vdev = to_ivpu_device(drm); + + return sysfs_emit(buf, "%s\n", vdev->fw->sched_mode ? "HW" : "OS"); +} + +static DEVICE_ATTR_RO(sched_mode); + static struct attribute *ivpu_dev_attrs[] = { &dev_attr_npu_busy_time_us.attr, + &dev_attr_sched_mode.attr, NULL, }; diff --git a/drivers/accel/ivpu/ivpu_trace.h b/drivers/accel/ivpu/ivpu_trace.h new file mode 100644 index 000000000000..eb792038e701 --- /dev/null +++ b/drivers/accel/ivpu/ivpu_trace.h @@ -0,0 +1,73 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* + * Copyright (C) 2020-2024 Intel Corporation + */ + +#if !defined(__IVPU_TRACE_H__) || defined(TRACE_HEADER_MULTI_READ) +#define __IVPU_TRACE_H__ + +#include +#include "ivpu_drv.h" +#include "ivpu_job.h" +#include "vpu_jsm_api.h" +#include "ivpu_jsm_msg.h" +#include "ivpu_ipc.h" + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM vpu +#define TRACE_INCLUDE_FILE ivpu_trace + +TRACE_EVENT(pm, + TP_PROTO(const char *event), + TP_ARGS(event), + TP_STRUCT__entry(__field(const char *, event)), + TP_fast_assign(__entry->event = event;), + TP_printk("%s", __entry->event) +); + +TRACE_EVENT(job, + TP_PROTO(const char *event, struct ivpu_job *job), + TP_ARGS(event, job), + TP_STRUCT__entry(__field(const char *, event) + __field(u32, ctx_id) + __field(u32, engine_id) + __field(u32, job_id) + ), + TP_fast_assign(__entry->event = event; + __entry->ctx_id = job->file_priv->ctx.id; + __entry->engine_id = job->engine_idx; + __entry->job_id = job->job_id;), + TP_printk("%s context:%d engine:%d job:%d", + __entry->event, + __entry->ctx_id, + __entry->engine_id, + __entry->job_id) +); + +TRACE_EVENT(jsm, + TP_PROTO(const char *event, struct vpu_jsm_msg *msg), + TP_ARGS(event, msg), + TP_STRUCT__entry(__field(const char *, event) + __field(const char *, type) + __field(enum vpu_ipc_msg_status, status) + __field(u32, request_id) + __field(u32, result) + ), + TP_fast_assign(__entry->event = event; + __entry->type = ivpu_jsm_msg_type_to_str(msg->type); + __entry->status = msg->status; + __entry->request_id = msg->request_id; + __entry->result = msg->result;), + TP_printk("%s type:%s, status:%#x, id:%#x, result:%#x", + __entry->event, + __entry->type, + __entry->status, + __entry->request_id, + __entry->result) +); + +#endif /* __IVPU_TRACE_H__ */ + +#undef TRACE_INCLUDE_PATH +#define TRACE_INCLUDE_PATH . +#include diff --git a/drivers/accel/ivpu/ivpu_trace_points.c b/drivers/accel/ivpu/ivpu_trace_points.c new file mode 100644 index 000000000000..f8fb99de0de3 --- /dev/null +++ b/drivers/accel/ivpu/ivpu_trace_points.c @@ -0,0 +1,9 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (C) 2020-2024 Intel Corporation + */ + +#ifndef __CHECKER__ +#define CREATE_TRACE_POINTS +#include "ivpu_trace.h" +#endif diff --git a/drivers/accel/ivpu/vpu_boot_api.h b/drivers/accel/ivpu/vpu_boot_api.h index 82954b91b748..908e68ea1c39 100644 --- a/drivers/accel/ivpu/vpu_boot_api.h +++ b/drivers/accel/ivpu/vpu_boot_api.h @@ -1,14 +1,13 @@ /* SPDX-License-Identifier: MIT */ /* - * Copyright (c) 2020-2023, Intel Corporation. + * Copyright (c) 2020-2024, Intel Corporation. */ #ifndef VPU_BOOT_API_H #define VPU_BOOT_API_H /* - * =========== FW API version information beginning ================ - * The bellow values will be used to construct the version info this way: + * The below values will be used to construct the version info this way: * fw_bin_header->api_version[VPU_BOOT_API_VER_ID] = (VPU_BOOT_API_VER_MAJOR << 16) | * VPU_BOOT_API_VER_MINOR; * VPU_BOOT_API_VER_PATCH will be ignored. KMD and compatibility is not affected if this changes @@ -27,19 +26,18 @@ * Minor version changes when API backward compatibility is preserved. * Resets to 0 if Major version is incremented. */ -#define VPU_BOOT_API_VER_MINOR 24 +#define VPU_BOOT_API_VER_MINOR 26 /* * API header changed (field names, documentation, formatting) but API itself has not been changed */ -#define VPU_BOOT_API_VER_PATCH 0 +#define VPU_BOOT_API_VER_PATCH 3 /* * Index in the API version table * Must be unique for each API */ #define VPU_BOOT_API_VER_INDEX 0 -/* ------------ FW API version information end ---------------------*/ #pragma pack(push, 4) @@ -164,8 +162,6 @@ enum vpu_trace_destination { /* VPU 30xx HW component IDs are sequential, so define first and last IDs. */ #define VPU_TRACE_PROC_BIT_30XX_FIRST VPU_TRACE_PROC_BIT_LRT #define VPU_TRACE_PROC_BIT_30XX_LAST VPU_TRACE_PROC_BIT_SHV_15 -#define VPU_TRACE_PROC_BIT_KMB_FIRST VPU_TRACE_PROC_BIT_30XX_FIRST -#define VPU_TRACE_PROC_BIT_KMB_LAST VPU_TRACE_PROC_BIT_30XX_LAST struct vpu_boot_l2_cache_config { u8 use; @@ -199,6 +195,17 @@ struct vpu_warm_boot_section { */ #define POWER_PROFILE_SURVIVABILITY 0x1 +/** + * Enum for dvfs_mode boot param. + */ +enum vpu_governor { + VPU_GOV_DEFAULT = 0, /* Default Governor for the system */ + VPU_GOV_MAX_PERFORMANCE = 1, /* Maximum performance governor */ + VPU_GOV_ON_DEMAND = 2, /* On Demand frequency control governor */ + VPU_GOV_POWER_SAVE = 3, /* Power save governor */ + VPU_GOV_ON_DEMAND_PRIORITY_AWARE = 4 /* On Demand priority based governor */ +}; + struct vpu_boot_params { u32 magic; u32 vpu_id; @@ -301,7 +308,14 @@ struct vpu_boot_params { u32 temp_sensor_period_ms; /** PLL ratio for efficient clock frequency */ u32 pn_freq_pll_ratio; - /** DVFS Mode: Default: 0, Max Performance: 1, On Demand: 2, Power Save: 3 */ + /** + * DVFS Mode: + * 0 - Default, DVFS mode selected by the firmware + * 1 - Max Performance + * 2 - On Demand + * 3 - Power Save + * 4 - On Demand Priority Aware + */ u32 dvfs_mode; /** * Depending on DVFS Mode: @@ -332,8 +346,8 @@ struct vpu_boot_params { u64 d0i3_entry_vpu_ts; /* * The system time of the host operating system in microseconds. - * E.g the number of microseconds since 1st of January 1970, or whatever date the - * host operating system uses to maintain system time. + * E.g the number of microseconds since 1st of January 1970, or whatever + * date the host operating system uses to maintain system time. * This value will be used to track system time on the VPU. * The KMD is required to update this value on every VPU reset. */ @@ -382,10 +396,7 @@ struct vpu_boot_params { u32 pad6[734]; }; -/* - * Magic numbers set between host and vpu to detect corruptio of tracing init - */ - +/* Magic numbers set between host and vpu to detect corruption of tracing init */ #define VPU_TRACING_BUFFER_CANARY (0xCAFECAFE) /* Tracing buffer message format definitions */ @@ -405,7 +416,9 @@ struct vpu_tracing_buffer_header { u32 host_canary_start; /* offset from start of buffer for trace entries */ u32 read_index; - u32 pad_to_cache_line_size_0[14]; + /* keeps track of wrapping on the reader side */ + u32 read_wrap_count; + u32 pad_to_cache_line_size_0[13]; /* End of first cache line */ /** diff --git a/drivers/accel/ivpu/vpu_jsm_api.h b/drivers/accel/ivpu/vpu_jsm_api.h index 33f462b1a25d..7215c144158c 100644 --- a/drivers/accel/ivpu/vpu_jsm_api.h +++ b/drivers/accel/ivpu/vpu_jsm_api.h @@ -22,7 +22,7 @@ /* * Minor version changes when API backward compatibility is preserved. */ -#define VPU_JSM_API_VER_MINOR 16 +#define VPU_JSM_API_VER_MINOR 25 /* * API header changed (field names, documentation, formatting) but API itself has not been changed @@ -36,7 +36,7 @@ /* * Number of Priority Bands for Hardware Scheduling - * Bands: RealTime, Focus, Normal, Idle + * Bands: Idle(0), Normal(1), Focus(2), RealTime(3) */ #define VPU_HWS_NUM_PRIORITY_BANDS 4 @@ -74,6 +74,7 @@ #define VPU_JSM_STATUS_MVNCI_INTERNAL_ERROR 0xCU /* Job status returned when the job was preempted mid-inference */ #define VPU_JSM_STATUS_PREEMPTED_MID_INFERENCE 0xDU +#define VPU_JSM_STATUS_MVNCI_CONTEXT_VIOLATION_HW 0xEU /* * Host <-> VPU IPC channels. @@ -86,18 +87,58 @@ /* * Job flags bit masks. */ -#define VPU_JOB_FLAGS_NULL_SUBMISSION_MASK 0x00000001 -#define VPU_JOB_FLAGS_PRIVATE_DATA_MASK 0xFF000000 +enum { + /* + * Null submission mask. + * When set, batch buffer's commands are not processed but returned as + * successful immediately, except fences and timestamps. + * When cleared, batch buffer's commands are processed normally. + * Used for testing and profiling purposes. + */ + VPU_JOB_FLAGS_NULL_SUBMISSION_MASK = (1 << 0U), + /* + * Inline command mask. + * When set, the object in job queue is an inline command (see struct vpu_inline_cmd below). + * When cleared, the object in job queue is a job (see struct vpu_job_queue_entry below). + */ + VPU_JOB_FLAGS_INLINE_CMD_MASK = (1 << 1U), + /* + * VPU private data mask. + * Reserved for the VPU to store private data about the job (or inline command) + * while being processed. + */ + VPU_JOB_FLAGS_PRIVATE_DATA_MASK = 0xFFFF0000U +}; /* - * Sizes of the reserved areas in jobs, in bytes. + * Job queue flags bit masks. */ -#define VPU_JOB_RESERVED_BYTES 8 +enum { + /* + * No job done notification mask. + * When set, indicates that no job done notification should be sent for any + * job from this queue. When cleared, indicates that job done notification + * should be sent for every job completed from this queue. + */ + VPU_JOB_QUEUE_FLAGS_NO_JOB_DONE_MASK = (1 << 0U), + /* + * Native fence usage mask. + * When set, indicates that job queue uses native fences (as inline commands + * in job queue). Such queues may also use legacy fences (as commands in batch buffers). + * When cleared, indicates the job queue only uses legacy fences. + * NOTE: For queues using native fences, VPU expects that all jobs in the queue + * are immediately followed by an inline command object. This object is expected + * to be a fence signal command in most cases, but can also be a NOP in case the host + * does not need per-job fence signalling. Other inline commands objects can be + * inserted between "job and inline command" pairs. + */ + VPU_JOB_QUEUE_FLAGS_USE_NATIVE_FENCE_MASK = (1 << 1U), -/* - * Sizes of the reserved areas in job queues, in bytes. - */ -#define VPU_JOB_QUEUE_RESERVED_BYTES 52 + /* + * Enable turbo mode for testing NPU performance; not recommended for regular usage. + */ + VPU_JOB_QUEUE_FLAGS_TURBO_MODE = (1 << 2U) +}; /* * Max length (including trailing NULL char) of trace entity name (e.g., the @@ -140,24 +181,113 @@ */ #define VPU_HWS_INVALID_CMDQ_HANDLE 0ULL +/* + * Inline commands types. + */ +/* + * NOP. + * VPU does nothing other than consuming the inline command object. + */ +#define VPU_INLINE_CMD_TYPE_NOP 0x0 +/* + * Fence wait. + * VPU waits for the fence current value to reach monitored value. + * Fence wait operations are executed upon job dispatching. While waiting for + * the fence to be satisfied, VPU blocks fetching of the next objects in the queue. + * Jobs present in the queue prior to the fence wait object may be processed + * concurrently. + */ +#define VPU_INLINE_CMD_TYPE_FENCE_WAIT 0x1 +/* + * Fence signal. + * VPU sets the fence current value to the provided value. If new current value + * is equal to or higher than monitored value, VPU sends fence signalled notification + * to the host. Fence signal operations are executed upon completion of all the jobs + * present in the queue prior to them, and in-order relative to each other in the queue. + * But jobs in-between them may be processed concurrently and may complete out-of-order. + */ +#define VPU_INLINE_CMD_TYPE_FENCE_SIGNAL 0x2 + +/* + * Job scheduling priority bands for both hardware scheduling and OS scheduling. + */ +enum vpu_job_scheduling_priority_band { + VPU_JOB_SCHEDULING_PRIORITY_BAND_IDLE = 0, + VPU_JOB_SCHEDULING_PRIORITY_BAND_NORMAL = 1, + VPU_JOB_SCHEDULING_PRIORITY_BAND_FOCUS = 2, + VPU_JOB_SCHEDULING_PRIORITY_BAND_REALTIME = 3, + VPU_JOB_SCHEDULING_PRIORITY_BAND_COUNT = 4, +}; + /* * Job format. + * Jobs defines the actual workloads to be executed by a given engine. */ struct vpu_job_queue_entry { - u64 batch_buf_addr; /**< Address of VPU commands batch buffer */ - u32 job_id; /**< Job ID */ - u32 flags; /**< Flags bit field, see VPU_JOB_FLAGS_* above */ - u64 root_page_table_addr; /**< Address of root page table to use for this job */ - u64 root_page_table_update_counter; /**< Page tables update events counter */ - u64 primary_preempt_buf_addr; + /**< Address of VPU commands batch buffer */ + u64 batch_buf_addr; + /**< Job ID */ + u32 job_id; + /**< Flags bit field, see VPU_JOB_FLAGS_* above */ + u32 flags; + /** + * Doorbell ring timestamp taken by KMD from SoC's global system clock, in + * microseconds. NPU can convert this value to its own fixed clock's timebase, + * to match other profiling timestamps. + */ + u64 doorbell_timestamp; + /**< Extra id for job tracking, used only in the firmware perf traces */ + u64 host_tracking_id; /**< Address of the primary preemption buffer to use for this job */ - u32 primary_preempt_buf_size; + u64 primary_preempt_buf_addr; /**< Size of the primary preemption buffer to use for this job */ - u32 secondary_preempt_buf_size; + u32 primary_preempt_buf_size; /**< Size of secondary preemption buffer to use for this job */ - u64 secondary_preempt_buf_addr; + u32 secondary_preempt_buf_size; /**< Address of secondary preemption buffer to use for this job */ - u8 reserved_0[VPU_JOB_RESERVED_BYTES]; + u64 secondary_preempt_buf_addr; + u64 reserved_0; +}; + +/* + * Inline command format. + * Inline commands are the commands executed at scheduler level (typically, + * synchronization directives). Inline command and job objects must be of + * the same size and have flags field at same offset. + */ +struct vpu_inline_cmd { + u64 reserved_0; + /* Inline command type, see VPU_INLINE_CMD_TYPE_* defines. */ + u32 type; + /* Flags bit field, see VPU_JOB_FLAGS_* above. */ + u32 flags; + /* Inline command payload. Depends on inline command type. */ + union { + /* Fence (wait and signal) commands' payload. */ + struct { + /* Fence object handle. */ + u64 fence_handle; + /* User VA of the current fence value. */ + u64 current_value_va; + /* User VA of the monitored fence value (read-only). */ + u64 monitored_value_va; + /* Value to wait for or write in fence location. */ + u64 value; + /* User VA of the log buffer in which to add log entry on completion. */ + u64 log_buffer_va; + } fence; + /* Other commands do not have a payload. */ + /* Payload definition for future inline commands can be inserted here. */ + u64 reserved_1[6]; + } payload; +}; + +/* + * Job queue slots can be populated either with job objects or inline command objects. + */ +union vpu_jobq_slot { + struct vpu_job_queue_entry job; + struct vpu_inline_cmd inline_cmd; }; /* @@ -167,7 +297,21 @@ struct vpu_job_queue_header { u32 engine_idx; u32 head; u32 tail; - u8 reserved_0[VPU_JOB_QUEUE_RESERVED_BYTES]; + u32 flags; + /* Set to 1 to indicate priority_band field is valid */ + u32 priority_band_valid; + /* + * Priority for the work of this job queue, valid only if the HWS is NOT used + * and the `priority_band_valid` is set to 1. It is applied only during + * the VPU_JSM_MSG_REGISTER_DB message processing. + * The device firmware might use the `priority_band` to optimize the power + * management logic, but it will not affect the order of jobs. + * Available priority bands: @see enum vpu_job_scheduling_priority_band + */ + u32 priority_band; + /* Inside realtime band assigns a further priority, limited to 0..31 range */ + u32 realtime_priority_level; + u32 reserved_0[9]; }; /* @@ -175,7 +319,7 @@ struct vpu_job_queue_header { */ struct vpu_job_queue { struct vpu_job_queue_header header; - struct vpu_job_queue_entry job[]; + union vpu_jobq_slot slot[]; }; /** @@ -197,9 +341,7 @@ enum vpu_trace_entity_type { struct vpu_hws_log_buffer_header { /* Written by VPU after adding a log entry. Initialised by host to 0. */ u32 first_free_entry_index; - /* Incremented by VPU every time the VPU overwrites the 0th entry; - * initialised by host to 0. - */ + /* Incremented by VPU every time the VPU writes the 0th entry; initialised by host to 0. */ u32 wraparound_count; /* * This is the number of buffers that can be stored in the log buffer provided by the host. @@ -230,14 +372,80 @@ struct vpu_hws_log_buffer_entry { u64 operation_data[2]; }; +/* Native fence log buffer types. */ +enum vpu_hws_native_fence_log_type { + VPU_HWS_NATIVE_FENCE_LOG_TYPE_WAITS = 1, + VPU_HWS_NATIVE_FENCE_LOG_TYPE_SIGNALS = 2 +}; + +/* HWS native fence log buffer header. */ +struct vpu_hws_native_fence_log_header { + union { + struct { + /* Index of the first free entry in buffer. */ + u32 first_free_entry_idx; + /* Incremented each time NPU wraps around the buffer to write next entry. */ + u32 wraparound_count; + }; + /* Field allowing atomic update of both fields above. */ + u64 atomic_wraparound_and_entry_idx; + }; + /* Log buffer type, see enum vpu_hws_native_fence_log_type. */ + u64 type; + /* Allocated number of entries in the log buffer. */ + u64 entry_nb; + u64 reserved[2]; +}; + +/* Native fence log operation types. */ +enum vpu_hws_native_fence_log_op { + VPU_HWS_NATIVE_FENCE_LOG_OP_SIGNAL_EXECUTED = 0, + VPU_HWS_NATIVE_FENCE_LOG_OP_WAIT_UNBLOCKED = 1 +}; + +/* HWS native fence log entry. */ +struct vpu_hws_native_fence_log_entry { + /* Newly signaled/unblocked fence value. */ + u64 fence_value; + /* Native fence object handle to which this operation belongs. */ + u64 fence_handle; + /* Operation type, see enum vpu_hws_native_fence_log_op. */ + u64 op_type; + u64 reserved_0; + /* + * VPU_HWS_NATIVE_FENCE_LOG_OP_WAIT_UNBLOCKED only: Timestamp at which fence + * wait was started (in NPU SysTime). + */ + u64 fence_wait_start_ts; + u64 reserved_1; + /* Timestamp at which fence operation was completed (in NPU SysTime). */ + u64 fence_end_ts; +}; + +/* Native fence log buffer. */ +struct vpu_hws_native_fence_log_buffer { + struct vpu_hws_native_fence_log_header header; + struct vpu_hws_native_fence_log_entry entry[]; +}; + /* * Host <-> VPU IPC messages types. */ enum vpu_ipc_msg_type { VPU_JSM_MSG_UNKNOWN = 0xFFFFFFFF, + /* IPC Host -> Device, Async commands */ VPU_JSM_MSG_ASYNC_CMD = 0x1100, VPU_JSM_MSG_ENGINE_RESET = VPU_JSM_MSG_ASYNC_CMD, + /** + * Preempt engine. The NPU stops (preempts) all the jobs currently + * executing on the target engine making the engine become idle and ready to + * execute new jobs. + * NOTE: The NPU does not remove unstarted jobs (if any) from job queues of + * the target engine, but it stops processing them (until the queue doorbell + * is rung again); the host is responsible to reset the job queue, either + * after preemption or when resubmitting jobs to the queue. + */ VPU_JSM_MSG_ENGINE_PREEMPT = 0x1101, VPU_JSM_MSG_REGISTER_DB = 0x1102, VPU_JSM_MSG_UNREGISTER_DB = 0x1103, @@ -323,9 +531,10 @@ enum vpu_ipc_msg_type { * NOTE: Please introduce new ASYNC commands before this one. * */ VPU_JSM_MSG_STATE_DUMP = 0x11FF, + /* IPC Host -> Device, General commands */ VPU_JSM_MSG_GENERAL_CMD = 0x1200, - VPU_JSM_MSG_BLOB_DEINIT = VPU_JSM_MSG_GENERAL_CMD, + VPU_JSM_MSG_BLOB_DEINIT_DEPRECATED = VPU_JSM_MSG_GENERAL_CMD, /** * Control dyndbg behavior by executing a dyndbg command; equivalent to * Linux command: `echo '' > /dynamic_debug/control`. @@ -335,8 +544,12 @@ enum vpu_ipc_msg_type { * Perform the save procedure for the D0i3 entry */ VPU_JSM_MSG_PWR_D0I3_ENTER = 0x1202, + /* IPC Device -> Host, Job completion */ VPU_JSM_MSG_JOB_DONE = 0x2100, + /* IPC Device -> Host, Fence signalled */ + VPU_JSM_MSG_NATIVE_FENCE_SIGNALLED = 0x2101, + /* IPC Device -> Host, Async command completion */ VPU_JSM_MSG_ASYNC_CMD_DONE = 0x2200, VPU_JSM_MSG_ENGINE_RESET_DONE = VPU_JSM_MSG_ASYNC_CMD_DONE, @@ -422,6 +635,7 @@ enum vpu_ipc_msg_type { * NOTE: Please introduce new ASYNC responses before this one. * */ VPU_JSM_MSG_STATE_DUMP_RSP = 0x22FF, + /* IPC Device -> Host, General command completion */ VPU_JSM_MSG_GENERAL_CMD_DONE = 0x2300, VPU_JSM_MSG_BLOB_DEINIT_DONE = VPU_JSM_MSG_GENERAL_CMD_DONE, @@ -600,11 +814,6 @@ struct vpu_jsm_metric_streamer_update { u64 next_buffer_size; }; -struct vpu_ipc_msg_payload_blob_deinit { - /* 64-bit unique ID for the blob to be de-initialized. */ - u64 blob_id; -}; - struct vpu_ipc_msg_payload_job_done { /* Engine to which the job was submitted. */ u32 engine_idx; @@ -622,6 +831,21 @@ struct vpu_ipc_msg_payload_job_done { u64 cmdq_id; }; +/* + * Notification message upon native fence signalling. + * @see VPU_JSM_MSG_NATIVE_FENCE_SIGNALLED + */ +struct vpu_ipc_msg_payload_native_fence_signalled { + /* Engine ID. */ + u32 engine_idx; + /* Host SSID. */ + u32 host_ssid; + /* CMDQ ID */ + u64 cmdq_id; + /* Fence object handle. */ + u64 fence_handle; +}; + struct vpu_jsm_engine_reset_context { /* Host SSID */ u32 host_ssid; @@ -700,11 +924,6 @@ struct vpu_ipc_msg_payload_get_power_level_count_done { u8 power_limit[16]; }; -struct vpu_ipc_msg_payload_blob_deinit_done { - /* 64-bit unique ID for the blob de-initialized. */ - u64 blob_id; -}; - /* HWS priority band setup request / response */ struct vpu_ipc_msg_payload_hws_priority_band_setup { /* @@ -794,7 +1013,10 @@ struct vpu_ipc_msg_payload_hws_set_context_sched_properties { u32 reserved_0; /* Command queue id */ u64 cmdq_id; - /* Priority band to assign to work of this context */ + /* + * Priority band to assign to work of this context. + * Available priority bands: @see enum vpu_job_scheduling_priority_band + */ u32 priority_band; /* Inside realtime band assigns a further priority */ u32 realtime_priority_level; @@ -869,9 +1091,7 @@ struct vpu_ipc_msg_payload_hws_set_scheduling_log { */ u64 notify_index; /* - * Enable extra events to be output to log for debug of scheduling algorithm. - * Interpreted by VPU as a boolean to enable or disable, expected values are - * 0 and 1. + * Field is now deprecated, will be removed when KMD is updated to support removal */ u32 enable_extra_events; /* Zero Padding */ @@ -1243,10 +1463,10 @@ union vpu_ipc_msg_payload { struct vpu_jsm_metric_streamer_start metric_streamer_start; struct vpu_jsm_metric_streamer_stop metric_streamer_stop; struct vpu_jsm_metric_streamer_update metric_streamer_update; - struct vpu_ipc_msg_payload_blob_deinit blob_deinit; struct vpu_ipc_msg_payload_ssid_release ssid_release; struct vpu_jsm_hws_register_db hws_register_db; struct vpu_ipc_msg_payload_job_done job_done; + struct vpu_ipc_msg_payload_native_fence_signalled native_fence_signalled; struct vpu_ipc_msg_payload_engine_reset_done engine_reset_done; struct vpu_ipc_msg_payload_engine_preempt_done engine_preempt_done; struct vpu_ipc_msg_payload_register_db_done register_db_done; @@ -1254,7 +1474,6 @@ union vpu_ipc_msg_payload { struct vpu_ipc_msg_payload_query_engine_hb_done query_engine_hb_done; struct vpu_ipc_msg_payload_get_power_level_count_done get_power_level_count_done; struct vpu_jsm_metric_streamer_done metric_streamer_done; - struct vpu_ipc_msg_payload_blob_deinit_done blob_deinit_done; struct vpu_ipc_msg_payload_trace_config trace_config; struct vpu_ipc_msg_payload_trace_capability_rsp trace_capability; struct vpu_ipc_msg_payload_trace_get_name trace_get_name; diff --git a/drivers/accel/qaic/mhi_controller.c b/drivers/accel/qaic/mhi_controller.c index ada9b1eb0787..8ab82e78dd94 100644 --- a/drivers/accel/qaic/mhi_controller.c +++ b/drivers/accel/qaic/mhi_controller.c @@ -405,6 +405,38 @@ static const struct mhi_channel_config aic100_channels[] = { .auto_queue = false, .wake_capable = false, }, + { + .name = "IPCR", + .num = 24, + .num_elements = 32, + .local_elements = 0, + .event_ring = 0, + .dir = DMA_TO_DEVICE, + .ee_mask = MHI_CH_EE_AMSS, + .pollcfg = 0, + .doorbell = MHI_DB_BRST_DISABLE, + .lpm_notify = false, + .offload_channel = false, + .doorbell_mode_switch = false, + .auto_queue = false, + .wake_capable = false, + }, + { + .name = "IPCR", + .num = 25, + .num_elements = 32, + .local_elements = 0, + .event_ring = 0, + .dir = DMA_FROM_DEVICE, + .ee_mask = MHI_CH_EE_AMSS, + .pollcfg = 0, + .doorbell = MHI_DB_BRST_DISABLE, + .lpm_notify = false, + .offload_channel = false, + .doorbell_mode_switch = false, + .auto_queue = true, + .wake_capable = false, + }, }; static struct mhi_event_config aic100_events[] = { diff --git a/drivers/accel/qaic/qaic_debugfs.c b/drivers/accel/qaic/qaic_debugfs.c index 20b653d99e52..ba0cf2f94732 100644 --- a/drivers/accel/qaic/qaic_debugfs.c +++ b/drivers/accel/qaic/qaic_debugfs.c @@ -64,20 +64,9 @@ static int bootlog_show(struct seq_file *s, void *unused) return 0; } -static int bootlog_fops_open(struct inode *inode, struct file *file) -{ - return single_open(file, bootlog_show, inode->i_private); -} +DEFINE_SHOW_ATTRIBUTE(bootlog); -static const struct file_operations bootlog_fops = { - .owner = THIS_MODULE, - .open = bootlog_fops_open, - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, -}; - -static int read_dbc_fifo_size(struct seq_file *s, void *unused) +static int fifo_size_show(struct seq_file *s, void *unused) { struct dma_bridge_chan *dbc = s->private; @@ -85,20 +74,9 @@ static int read_dbc_fifo_size(struct seq_file *s, void *unused) return 0; } -static int fifo_size_open(struct inode *inode, struct file *file) -{ - return single_open(file, read_dbc_fifo_size, inode->i_private); -} +DEFINE_SHOW_ATTRIBUTE(fifo_size); -static const struct file_operations fifo_size_fops = { - .owner = THIS_MODULE, - .open = fifo_size_open, - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, -}; - -static int read_dbc_queued(struct seq_file *s, void *unused) +static int queued_show(struct seq_file *s, void *unused) { struct dma_bridge_chan *dbc = s->private; u32 tail = 0, head = 0; @@ -115,18 +93,7 @@ static int read_dbc_queued(struct seq_file *s, void *unused) return 0; } -static int queued_open(struct inode *inode, struct file *file) -{ - return single_open(file, read_dbc_queued, inode->i_private); -} - -static const struct file_operations queued_fops = { - .owner = THIS_MODULE, - .open = queued_open, - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, -}; +DEFINE_SHOW_ATTRIBUTE(queued); void qaic_debugfs_init(struct qaic_drm_device *qddev) { diff --git a/drivers/accel/qaic/qaic_drv.c b/drivers/accel/qaic/qaic_drv.c index bf10156c334e..3575e0c984d6 100644 --- a/drivers/accel/qaic/qaic_drv.c +++ b/drivers/accel/qaic/qaic_drv.c @@ -34,6 +34,7 @@ MODULE_IMPORT_NS(DMA_BUF); +#define PCI_DEV_AIC080 0xa080 #define PCI_DEV_AIC100 0xa100 #define QAIC_NAME "qaic" #define QAIC_DESC "Qualcomm Cloud AI Accelerators" @@ -53,12 +54,12 @@ static void qaicm_wq_release(struct drm_device *dev, void *res) destroy_workqueue(wq); } -static struct workqueue_struct *qaicm_wq_init(struct drm_device *dev, const char *fmt) +static struct workqueue_struct *qaicm_wq_init(struct drm_device *dev, const char *name) { struct workqueue_struct *wq; int ret; - wq = alloc_workqueue(fmt, WQ_UNBOUND, 0); + wq = alloc_workqueue("%s", WQ_UNBOUND, 0, name); if (!wq) return ERR_PTR(-ENOMEM); ret = drmm_add_action_or_reset(dev, qaicm_wq_release, wq); @@ -365,7 +366,7 @@ static struct qaic_device *create_qdev(struct pci_dev *pdev, const struct pci_de return NULL; qdev->dev_state = QAIC_OFFLINE; - if (id->device == PCI_DEV_AIC100) { + if (id->device == PCI_DEV_AIC080 || id->device == PCI_DEV_AIC100) { qdev->num_dbc = 16; qdev->dbc = devm_kcalloc(dev, qdev->num_dbc, sizeof(*qdev->dbc), GFP_KERNEL); if (!qdev->dbc) @@ -607,6 +608,7 @@ static struct mhi_driver qaic_mhi_driver = { }; static const struct pci_device_id qaic_ids[] = { + { PCI_DEVICE(PCI_VENDOR_ID_QCOM, PCI_DEV_AIC080), }, { PCI_DEVICE(PCI_VENDOR_ID_QCOM, PCI_DEV_AIC100), }, { } }; diff --git a/drivers/accel/qaic/sahara.c b/drivers/accel/qaic/sahara.c index bf94bbab6be5..6d772143d612 100644 --- a/drivers/accel/qaic/sahara.c +++ b/drivers/accel/qaic/sahara.c @@ -2,6 +2,7 @@ /* Copyright (c) 2024 Qualcomm Innovation Center, Inc. All rights reserved. */ +#include #include #include #include @@ -9,6 +10,7 @@ #include #include #include +#include #include #include "sahara.h" @@ -36,12 +38,14 @@ #define SAHARA_PACKET_MAX_SIZE 0xffffU /* MHI_MAX_MTU */ #define SAHARA_TRANSFER_MAX_SIZE 0x80000 +#define SAHARA_READ_MAX_SIZE 0xfff0U /* Avoid unaligned requests */ #define SAHARA_NUM_TX_BUF DIV_ROUND_UP(SAHARA_TRANSFER_MAX_SIZE,\ SAHARA_PACKET_MAX_SIZE) #define SAHARA_IMAGE_ID_NONE U32_MAX #define SAHARA_VERSION 2 #define SAHARA_SUCCESS 0 +#define SAHARA_TABLE_ENTRY_STR_LEN 20 #define SAHARA_MODE_IMAGE_TX_PENDING 0x0 #define SAHARA_MODE_IMAGE_TX_COMPLETE 0x1 @@ -53,6 +57,8 @@ #define SAHARA_END_OF_IMAGE_LENGTH 0x10 #define SAHARA_DONE_LENGTH 0x8 #define SAHARA_RESET_LENGTH 0x8 +#define SAHARA_MEM_DEBUG64_LENGTH 0x18 +#define SAHARA_MEM_READ64_LENGTH 0x18 struct sahara_packet { __le32 cmd; @@ -80,18 +86,95 @@ struct sahara_packet { __le32 image; __le32 status; } end_of_image; + struct { + __le64 table_address; + __le64 table_length; + } memory_debug64; + struct { + __le64 memory_address; + __le64 memory_length; + } memory_read64; }; }; +struct sahara_debug_table_entry64 { + __le64 type; + __le64 address; + __le64 length; + char description[SAHARA_TABLE_ENTRY_STR_LEN]; + char filename[SAHARA_TABLE_ENTRY_STR_LEN]; +}; + +struct sahara_dump_table_entry { + u64 type; + u64 address; + u64 length; + char description[SAHARA_TABLE_ENTRY_STR_LEN]; + char filename[SAHARA_TABLE_ENTRY_STR_LEN]; +}; + +#define SAHARA_DUMP_V1_MAGIC 0x1234567890abcdef +#define SAHARA_DUMP_V1_VER 1 +struct sahara_memory_dump_meta_v1 { + u64 magic; + u64 version; + u64 dump_size; + u64 table_size; +}; + +/* + * Layout of crashdump provided to user via devcoredump + * +------------------------------------------+ + * | Crashdump Meta structure | + * | type: struct sahara_memory_dump_meta_v1 | + * +------------------------------------------+ + * | Crashdump Table | + * | type: array of struct | + * | sahara_dump_table_entry | + * | | + * | | + * +------------------------------------------+ + * | Crashdump | + * | | + * | | + * | | + * | | + * | | + * +------------------------------------------+ + * + * First is the metadata header. Userspace can use the magic number to verify + * the content type, and then check the version for the rest of the format. + * New versions should keep the magic number location/value, and version + * location, but increment the version value. + * + * For v1, the metadata lists the size of the entire dump (header + table + + * dump) and the size of the table. Then the dump image table, which describes + * the contents of the dump. Finally all the images are listed in order, with + * no deadspace in between. Userspace can use the sizes listed in the image + * table to reconstruct the individual images. + */ + struct sahara_context { struct sahara_packet *tx[SAHARA_NUM_TX_BUF]; struct sahara_packet *rx; - struct work_struct work; + struct work_struct fw_work; + struct work_struct dump_work; struct mhi_device *mhi_dev; const char **image_table; u32 table_size; u32 active_image_id; const struct firmware *firmware; + u64 dump_table_address; + u64 dump_table_length; + size_t rx_size; + size_t rx_size_requested; + void *mem_dump; + size_t mem_dump_sz; + struct sahara_dump_table_entry *dump_image; + u64 dump_image_offset; + void *mem_dump_freespace; + u64 dump_images_left; + bool is_mem_dump_mode; }; static const char *aic100_image_table[] = { @@ -153,6 +236,8 @@ static void sahara_send_reset(struct sahara_context *context) { int ret; + context->is_mem_dump_mode = false; + context->tx[0]->cmd = cpu_to_le32(SAHARA_RESET_CMD); context->tx[0]->length = cpu_to_le32(SAHARA_RESET_LENGTH); @@ -186,7 +271,8 @@ static void sahara_hello(struct sahara_context *context) } if (le32_to_cpu(context->rx->hello.mode) != SAHARA_MODE_IMAGE_TX_PENDING && - le32_to_cpu(context->rx->hello.mode) != SAHARA_MODE_IMAGE_TX_COMPLETE) { + le32_to_cpu(context->rx->hello.mode) != SAHARA_MODE_IMAGE_TX_COMPLETE && + le32_to_cpu(context->rx->hello.mode) != SAHARA_MODE_MEMORY_DEBUG) { dev_err(&context->mhi_dev->dev, "Unsupported hello packet - mode %d\n", le32_to_cpu(context->rx->hello.mode)); return; @@ -320,9 +406,70 @@ static void sahara_end_of_image(struct sahara_context *context) dev_dbg(&context->mhi_dev->dev, "Unable to send done response %d\n", ret); } +static void sahara_memory_debug64(struct sahara_context *context) +{ + int ret; + + dev_dbg(&context->mhi_dev->dev, + "MEMORY DEBUG64 cmd received. length:%d table_address:%#llx table_length:%#llx\n", + le32_to_cpu(context->rx->length), + le64_to_cpu(context->rx->memory_debug64.table_address), + le64_to_cpu(context->rx->memory_debug64.table_length)); + + if (le32_to_cpu(context->rx->length) != SAHARA_MEM_DEBUG64_LENGTH) { + dev_err(&context->mhi_dev->dev, "Malformed memory debug64 packet - length %d\n", + le32_to_cpu(context->rx->length)); + return; + } + + context->dump_table_address = le64_to_cpu(context->rx->memory_debug64.table_address); + context->dump_table_length = le64_to_cpu(context->rx->memory_debug64.table_length); + + if (context->dump_table_length % sizeof(struct sahara_debug_table_entry64) != 0 || + !context->dump_table_length) { + dev_err(&context->mhi_dev->dev, "Malformed memory debug64 packet - table length %lld\n", + context->dump_table_length); + return; + } + + /* + * From this point, the protocol flips. We make memory_read requests to + * the device, and the device responds with the raw data. If the device + * has an error, it will send an End of Image command. First we need to + * request the memory dump table so that we know where all the pieces + * of the dump are that we can consume. + */ + + context->is_mem_dump_mode = true; + + /* + * Assume that the table is smaller than our MTU so that we can read it + * in one shot. The spec does not put an upper limit on the table, but + * no known device will exceed this. + */ + if (context->dump_table_length > SAHARA_PACKET_MAX_SIZE) { + dev_err(&context->mhi_dev->dev, "Memory dump table length %lld exceeds supported size. Discarding dump\n", + context->dump_table_length); + sahara_send_reset(context); + return; + } + + context->tx[0]->cmd = cpu_to_le32(SAHARA_MEM_READ64_CMD); + context->tx[0]->length = cpu_to_le32(SAHARA_MEM_READ64_LENGTH); + context->tx[0]->memory_read64.memory_address = cpu_to_le64(context->dump_table_address); + context->tx[0]->memory_read64.memory_length = cpu_to_le64(context->dump_table_length); + + context->rx_size_requested = context->dump_table_length; + + ret = mhi_queue_buf(context->mhi_dev, DMA_TO_DEVICE, context->tx[0], + SAHARA_MEM_READ64_LENGTH, MHI_EOT); + if (ret) + dev_err(&context->mhi_dev->dev, "Unable to send read for dump table %d\n", ret); +} + static void sahara_processing(struct work_struct *work) { - struct sahara_context *context = container_of(work, struct sahara_context, work); + struct sahara_context *context = container_of(work, struct sahara_context, fw_work); int ret; switch (le32_to_cpu(context->rx->cmd)) { @@ -338,6 +485,12 @@ static void sahara_processing(struct work_struct *work) case SAHARA_DONE_RESP_CMD: /* Intentional do nothing as we don't need to exit an app */ break; + case SAHARA_RESET_RESP_CMD: + /* Intentional do nothing as we don't need to exit an app */ + break; + case SAHARA_MEM_DEBUG64_CMD: + sahara_memory_debug64(context); + break; default: dev_err(&context->mhi_dev->dev, "Unknown command %d\n", le32_to_cpu(context->rx->cmd)); @@ -350,6 +503,217 @@ static void sahara_processing(struct work_struct *work) dev_err(&context->mhi_dev->dev, "Unable to requeue rx buf %d\n", ret); } +static void sahara_parse_dump_table(struct sahara_context *context) +{ + struct sahara_dump_table_entry *image_out_table; + struct sahara_debug_table_entry64 *dev_table; + struct sahara_memory_dump_meta_v1 *dump_meta; + u64 table_nents; + u64 dump_length; + int ret; + u64 i; + + table_nents = context->dump_table_length / sizeof(*dev_table); + context->dump_images_left = table_nents; + dump_length = 0; + + dev_table = (struct sahara_debug_table_entry64 *)(context->rx); + for (i = 0; i < table_nents; ++i) { + /* Do not trust the device, ensure the strings are terminated */ + dev_table[i].description[SAHARA_TABLE_ENTRY_STR_LEN - 1] = 0; + dev_table[i].filename[SAHARA_TABLE_ENTRY_STR_LEN - 1] = 0; + + dump_length = size_add(dump_length, le64_to_cpu(dev_table[i].length)); + if (dump_length == SIZE_MAX) { + /* Discard the dump */ + sahara_send_reset(context); + return; + } + + dev_dbg(&context->mhi_dev->dev, + "Memory dump table entry %lld type: %lld address: %#llx length: %#llx description: \"%s\" filename \"%s\"\n", + i, + le64_to_cpu(dev_table[i].type), + le64_to_cpu(dev_table[i].address), + le64_to_cpu(dev_table[i].length), + dev_table[i].description, + dev_table[i].filename); + } + + dump_length = size_add(dump_length, sizeof(*dump_meta)); + if (dump_length == SIZE_MAX) { + /* Discard the dump */ + sahara_send_reset(context); + return; + } + dump_length = size_add(dump_length, size_mul(sizeof(*image_out_table), table_nents)); + if (dump_length == SIZE_MAX) { + /* Discard the dump */ + sahara_send_reset(context); + return; + } + + context->mem_dump_sz = dump_length; + context->mem_dump = vzalloc(dump_length); + if (!context->mem_dump) { + /* Discard the dump */ + sahara_send_reset(context); + return; + } + + /* Populate the dump metadata and table for userspace */ + dump_meta = context->mem_dump; + dump_meta->magic = SAHARA_DUMP_V1_MAGIC; + dump_meta->version = SAHARA_DUMP_V1_VER; + dump_meta->dump_size = dump_length; + dump_meta->table_size = context->dump_table_length; + + image_out_table = context->mem_dump + sizeof(*dump_meta); + for (i = 0; i < table_nents; ++i) { + image_out_table[i].type = le64_to_cpu(dev_table[i].type); + image_out_table[i].address = le64_to_cpu(dev_table[i].address); + image_out_table[i].length = le64_to_cpu(dev_table[i].length); + strscpy(image_out_table[i].description, dev_table[i].description, + SAHARA_TABLE_ENTRY_STR_LEN); + strscpy(image_out_table[i].filename, + dev_table[i].filename, + SAHARA_TABLE_ENTRY_STR_LEN); + } + + context->mem_dump_freespace = &image_out_table[i]; + + /* Done parsing the table, switch to image dump mode */ + context->dump_table_length = 0; + + /* Request the first chunk of the first image */ + context->dump_image = &image_out_table[0]; + dump_length = min(context->dump_image->length, SAHARA_READ_MAX_SIZE); + /* Avoid requesting EOI sized data so that we can identify errors */ + if (dump_length == SAHARA_END_OF_IMAGE_LENGTH) + dump_length = SAHARA_END_OF_IMAGE_LENGTH / 2; + + context->dump_image_offset = dump_length; + + context->tx[0]->cmd = cpu_to_le32(SAHARA_MEM_READ64_CMD); + context->tx[0]->length = cpu_to_le32(SAHARA_MEM_READ64_LENGTH); + context->tx[0]->memory_read64.memory_address = cpu_to_le64(context->dump_image->address); + context->tx[0]->memory_read64.memory_length = cpu_to_le64(dump_length); + + context->rx_size_requested = dump_length; + + ret = mhi_queue_buf(context->mhi_dev, DMA_TO_DEVICE, context->tx[0], + SAHARA_MEM_READ64_LENGTH, MHI_EOT); + if (ret) + dev_err(&context->mhi_dev->dev, "Unable to send read for dump content %d\n", ret); +} + +static void sahara_parse_dump_image(struct sahara_context *context) +{ + u64 dump_length; + int ret; + + memcpy(context->mem_dump_freespace, context->rx, context->rx_size); + context->mem_dump_freespace += context->rx_size; + + if (context->dump_image_offset >= context->dump_image->length) { + /* Need to move to next image */ + context->dump_image++; + context->dump_images_left--; + context->dump_image_offset = 0; + + if (!context->dump_images_left) { + /* Dump done */ + dev_coredumpv(context->mhi_dev->mhi_cntrl->cntrl_dev, + context->mem_dump, + context->mem_dump_sz, + GFP_KERNEL); + context->mem_dump = NULL; + sahara_send_reset(context); + return; + } + } + + /* Get next image chunk */ + dump_length = context->dump_image->length - context->dump_image_offset; + dump_length = min(dump_length, SAHARA_READ_MAX_SIZE); + /* Avoid requesting EOI sized data so that we can identify errors */ + if (dump_length == SAHARA_END_OF_IMAGE_LENGTH) + dump_length = SAHARA_END_OF_IMAGE_LENGTH / 2; + + context->tx[0]->cmd = cpu_to_le32(SAHARA_MEM_READ64_CMD); + context->tx[0]->length = cpu_to_le32(SAHARA_MEM_READ64_LENGTH); + context->tx[0]->memory_read64.memory_address = + cpu_to_le64(context->dump_image->address + context->dump_image_offset); + context->tx[0]->memory_read64.memory_length = cpu_to_le64(dump_length); + + context->dump_image_offset += dump_length; + context->rx_size_requested = dump_length; + + ret = mhi_queue_buf(context->mhi_dev, DMA_TO_DEVICE, context->tx[0], + SAHARA_MEM_READ64_LENGTH, MHI_EOT); + if (ret) + dev_err(&context->mhi_dev->dev, + "Unable to send read for dump content %d\n", ret); +} + +static void sahara_dump_processing(struct work_struct *work) +{ + struct sahara_context *context = container_of(work, struct sahara_context, dump_work); + int ret; + + /* + * We should get the expected raw data, but if the device has an error + * it is supposed to send EOI with an error code. + */ + if (context->rx_size != context->rx_size_requested && + context->rx_size != SAHARA_END_OF_IMAGE_LENGTH) { + dev_err(&context->mhi_dev->dev, + "Unexpected response to read_data. Expected size: %#zx got: %#zx\n", + context->rx_size_requested, + context->rx_size); + goto error; + } + + if (context->rx_size == SAHARA_END_OF_IMAGE_LENGTH && + le32_to_cpu(context->rx->cmd) == SAHARA_END_OF_IMAGE_CMD) { + dev_err(&context->mhi_dev->dev, + "Unexpected EOI response to read_data. Status: %d\n", + le32_to_cpu(context->rx->end_of_image.status)); + goto error; + } + + if (context->rx_size == SAHARA_END_OF_IMAGE_LENGTH && + le32_to_cpu(context->rx->cmd) != SAHARA_END_OF_IMAGE_CMD) { + dev_err(&context->mhi_dev->dev, + "Invalid EOI response to read_data. CMD: %d\n", + le32_to_cpu(context->rx->cmd)); + goto error; + } + + /* + * Need to know if we received the dump table, or part of a dump image. + * Since we get raw data, we cannot tell from the data itself. Instead, + * we use the stored dump_table_length, which we zero after we read and + * process the entire table. + */ + if (context->dump_table_length) + sahara_parse_dump_table(context); + else + sahara_parse_dump_image(context); + + ret = mhi_queue_buf(context->mhi_dev, DMA_FROM_DEVICE, context->rx, + SAHARA_PACKET_MAX_SIZE, MHI_EOT); + if (ret) + dev_err(&context->mhi_dev->dev, "Unable to requeue rx buf %d\n", ret); + + return; + +error: + vfree(context->mem_dump); + context->mem_dump = NULL; + sahara_send_reset(context); +} + static int sahara_mhi_probe(struct mhi_device *mhi_dev, const struct mhi_device_id *id) { struct sahara_context *context; @@ -382,7 +746,8 @@ static int sahara_mhi_probe(struct mhi_device *mhi_dev, const struct mhi_device_ } context->mhi_dev = mhi_dev; - INIT_WORK(&context->work, sahara_processing); + INIT_WORK(&context->fw_work, sahara_processing); + INIT_WORK(&context->dump_work, sahara_dump_processing); context->image_table = aic100_image_table; context->table_size = ARRAY_SIZE(aic100_image_table); context->active_image_id = SAHARA_IMAGE_ID_NONE; @@ -405,7 +770,10 @@ static void sahara_mhi_remove(struct mhi_device *mhi_dev) { struct sahara_context *context = dev_get_drvdata(&mhi_dev->dev); - cancel_work_sync(&context->work); + cancel_work_sync(&context->fw_work); + cancel_work_sync(&context->dump_work); + if (context->mem_dump) + vfree(context->mem_dump); sahara_release_image(context); mhi_unprepare_from_transfer(mhi_dev); } @@ -418,8 +786,14 @@ static void sahara_mhi_dl_xfer_cb(struct mhi_device *mhi_dev, struct mhi_result { struct sahara_context *context = dev_get_drvdata(&mhi_dev->dev); - if (!mhi_result->transaction_status) - schedule_work(&context->work); + if (!mhi_result->transaction_status) { + context->rx_size = mhi_result->bytes_xferd; + if (context->is_mem_dump_mode) + schedule_work(&context->dump_work); + else + schedule_work(&context->fw_work); + } + } static const struct mhi_device_id sahara_mhi_match_table[] = { diff --git a/drivers/dma-buf/Kconfig b/drivers/dma-buf/Kconfig index b46eb8a552d7..fee04fdb0822 100644 --- a/drivers/dma-buf/Kconfig +++ b/drivers/dma-buf/Kconfig @@ -36,6 +36,7 @@ config UDMABUF depends on DMA_SHARED_BUFFER depends on MEMFD_CREATE || COMPILE_TEST depends on MMU + select VMAP_PFN help A driver to let userspace turn memfd regions into dma-bufs. Qemu can use this to create host dmabufs for guest framebuffers. diff --git a/drivers/dma-buf/dma-buf.c b/drivers/dma-buf/dma-buf.c index 8892bc701a66..5ad0e9e2e1b9 100644 --- a/drivers/dma-buf/dma-buf.c +++ b/drivers/dma-buf/dma-buf.c @@ -176,8 +176,9 @@ static loff_t dma_buf_llseek(struct file *file, loff_t offset, int whence) dmabuf = file->private_data; /* only support discovering the end of the buffer, - but also allow SEEK_SET to maintain the idiomatic - SEEK_END(0), SEEK_CUR(0) pattern */ + * but also allow SEEK_SET to maintain the idiomatic + * SEEK_END(0), SEEK_CUR(0) pattern. + */ if (whence == SEEK_END) base = dmabuf->size; else if (whence == SEEK_SET) @@ -558,7 +559,7 @@ static struct file *dma_buf_getfile(size_t size, int flags) * Override ->i_ino with the unique and dmabuffs specific * value. */ - inode->i_ino = atomic64_add_return(1, &dmabuf_inode); + inode->i_ino = atomic64_inc_return(&dmabuf_inode); flags &= O_ACCMODE | O_NONBLOCK; file = alloc_file_pseudo(inode, dma_buf_mnt, "dmabuf", flags, &dma_buf_fops); @@ -782,13 +783,14 @@ static void mangle_sg_table(struct sg_table *sg_table) /* To catch abuse of the underlying struct page by importers mix * up the bits, but take care to preserve the low SG_ bits to * not corrupt the sgt. The mixing is undone in __unmap_dma_buf - * before passing the sgt back to the exporter. */ + * before passing the sgt back to the exporter. + */ for_each_sgtable_sg(sg_table, sg, i) sg->page_link ^= ~0xffUL; #endif } -static struct sg_table * __map_dma_buf(struct dma_buf_attachment *attach, +static struct sg_table *__map_dma_buf(struct dma_buf_attachment *attach, enum dma_data_direction direction) { struct sg_table *sg_table; @@ -1296,10 +1298,12 @@ EXPORT_SYMBOL_NS_GPL(dma_buf_move_notify, DMA_BUF); * vmap interface is introduced. Note that on very old 32-bit architectures * vmalloc space might be limited and result in vmap calls failing. * - * Interfaces:: + * Interfaces: * - * void \*dma_buf_vmap(struct dma_buf \*dmabuf, struct iosys_map \*map) - * void dma_buf_vunmap(struct dma_buf \*dmabuf, struct iosys_map \*map) + * .. code-block:: c + * + * void *dma_buf_vmap(struct dma_buf *dmabuf, struct iosys_map *map) + * void dma_buf_vunmap(struct dma_buf *dmabuf, struct iosys_map *map) * * The vmap call can fail if there is no vmap support in the exporter, or if * it runs out of vmalloc space. Note that the dma-buf layer keeps a reference @@ -1356,10 +1360,11 @@ EXPORT_SYMBOL_NS_GPL(dma_buf_move_notify, DMA_BUF); * enough, since adding interfaces to intercept pagefaults and allow pte * shootdowns would increase the complexity quite a bit. * - * Interface:: + * Interface: * - * int dma_buf_mmap(struct dma_buf \*, struct vm_area_struct \*, - * unsigned long); + * .. code-block:: c + * + * int dma_buf_mmap(struct dma_buf *, struct vm_area_struct *, unsigned long); * * If the importing subsystem simply provides a special-purpose mmap call to * set up a mapping in userspace, calling do_mmap with &dma_buf.file will @@ -1694,7 +1699,7 @@ static int dma_buf_init_debugfs(void) dma_buf_debugfs_dir = d; - d = debugfs_create_file("bufinfo", S_IRUGO, dma_buf_debugfs_dir, + d = debugfs_create_file("bufinfo", 0444, dma_buf_debugfs_dir, NULL, &dma_buf_debug_fops); if (IS_ERR(d)) { pr_debug("dma_buf: debugfs: failed to create node bufinfo\n"); diff --git a/drivers/dma-buf/dma-fence.c b/drivers/dma-buf/dma-fence.c index 0393a9bba3a8..f0cdd3e99d36 100644 --- a/drivers/dma-buf/dma-fence.c +++ b/drivers/dma-buf/dma-fence.c @@ -309,8 +309,8 @@ bool dma_fence_begin_signalling(void) if (in_atomic()) return true; - /* ... and non-recursive readlock */ - lock_acquire(&dma_fence_lockdep_map, 0, 0, 1, 1, NULL, _RET_IP_); + /* ... and non-recursive successful read_trylock */ + lock_acquire(&dma_fence_lockdep_map, 0, 1, 1, 1, NULL, _RET_IP_); return false; } @@ -341,7 +341,7 @@ void __dma_fence_might_wait(void) lock_map_acquire(&dma_fence_lockdep_map); lock_map_release(&dma_fence_lockdep_map); if (tmp) - lock_acquire(&dma_fence_lockdep_map, 0, 0, 1, 1, NULL, _THIS_IP_); + lock_acquire(&dma_fence_lockdep_map, 0, 1, 1, 1, NULL, _THIS_IP_); } #endif @@ -412,7 +412,7 @@ int dma_fence_signal_timestamp(struct dma_fence *fence, ktime_t timestamp) unsigned long flags; int ret; - if (!fence) + if (WARN_ON(!fence)) return -EINVAL; spin_lock_irqsave(fence->lock, flags); @@ -464,7 +464,7 @@ int dma_fence_signal(struct dma_fence *fence) int ret; bool tmp; - if (!fence) + if (WARN_ON(!fence)) return -EINVAL; tmp = dma_fence_begin_signalling(); diff --git a/drivers/dma-buf/heaps/cma_heap.c b/drivers/dma-buf/heaps/cma_heap.c index 93be88b805fe..9512d050563a 100644 --- a/drivers/dma-buf/heaps/cma_heap.c +++ b/drivers/dma-buf/heaps/cma_heap.c @@ -309,13 +309,13 @@ static struct dma_buf *cma_heap_allocate(struct dma_heap *heap, struct page *page = cma_pages; while (nr_clear_pages > 0) { - void *vaddr = kmap_atomic(page); + void *vaddr = kmap_local_page(page); memset(vaddr, 0, PAGE_SIZE); - kunmap_atomic(vaddr); + kunmap_local(vaddr); /* * Avoid wasting time zeroing memory if the process - * has been killed by by SIGKILL + * has been killed by SIGKILL. */ if (fatal_signal_pending(current)) goto free_cma; @@ -366,7 +366,7 @@ static const struct dma_heap_ops cma_heap_ops = { .allocate = cma_heap_allocate, }; -static int __add_cma_heap(struct cma *cma, void *data) +static int __init __add_cma_heap(struct cma *cma, void *data) { struct cma_heap *cma_heap; struct dma_heap_export_info exp_info; @@ -391,7 +391,7 @@ static int __add_cma_heap(struct cma *cma, void *data) return 0; } -static int add_default_cma_heap(void) +static int __init add_default_cma_heap(void) { struct cma *default_cma = dev_get_cma_area(NULL); int ret = 0; diff --git a/drivers/dma-buf/heaps/system_heap.c b/drivers/dma-buf/heaps/system_heap.c index d78cdb9d01e5..26d5dc89ea16 100644 --- a/drivers/dma-buf/heaps/system_heap.c +++ b/drivers/dma-buf/heaps/system_heap.c @@ -421,7 +421,7 @@ static const struct dma_heap_ops system_heap_ops = { .allocate = system_heap_allocate, }; -static int system_heap_create(void) +static int __init system_heap_create(void) { struct dma_heap_export_info exp_info; diff --git a/drivers/dma-buf/sw_sync.c b/drivers/dma-buf/sw_sync.c index c353029789cf..f5905d67dedb 100644 --- a/drivers/dma-buf/sw_sync.c +++ b/drivers/dma-buf/sw_sync.c @@ -173,11 +173,6 @@ static bool timeline_fence_signaled(struct dma_fence *fence) return !__dma_fence_is_later(fence->seqno, parent->value, fence->ops); } -static bool timeline_fence_enable_signaling(struct dma_fence *fence) -{ - return true; -} - static void timeline_fence_value_str(struct dma_fence *fence, char *str, int size) { @@ -211,7 +206,6 @@ static void timeline_fence_set_deadline(struct dma_fence *fence, ktime_t deadlin static const struct dma_fence_ops timeline_fence_ops = { .get_driver_name = timeline_fence_get_driver_name, .get_timeline_name = timeline_fence_get_timeline_name, - .enable_signaling = timeline_fence_enable_signaling, .signaled = timeline_fence_signaled, .release = timeline_fence_release, .fence_value_str = timeline_fence_value_str, diff --git a/drivers/dma-buf/udmabuf.c b/drivers/dma-buf/udmabuf.c index 047c3cd2ceff..8ce1f074c2d3 100644 --- a/drivers/dma-buf/udmabuf.c +++ b/drivers/dma-buf/udmabuf.c @@ -27,15 +27,21 @@ MODULE_PARM_DESC(size_limit_mb, "Max size of a dmabuf, in megabytes. Default is struct udmabuf { pgoff_t pagecount; struct folio **folios; + + /** + * Unlike folios, pinned_folios is only used for unpin. + * So, nr_pinned is not the same to pagecount, the pinned_folios + * only set each folio which already pinned when udmabuf_create. + * Note that, since a folio may be pinned multiple times, each folio + * can be added to pinned_folios multiple times, depending on how many + * times the folio has been pinned when create. + */ + pgoff_t nr_pinned; + struct folio **pinned_folios; + struct sg_table *sg; struct miscdevice *device; pgoff_t *offsets; - struct list_head unpin_list; -}; - -struct udmabuf_folio { - struct folio *folio; - struct list_head list; }; static vm_fault_t udmabuf_vm_fault(struct vm_fault *vmf) @@ -43,7 +49,8 @@ static vm_fault_t udmabuf_vm_fault(struct vm_fault *vmf) struct vm_area_struct *vma = vmf->vma; struct udmabuf *ubuf = vma->vm_private_data; pgoff_t pgoff = vmf->pgoff; - unsigned long pfn; + unsigned long addr, pfn; + vm_fault_t ret; if (pgoff >= ubuf->pagecount) return VM_FAULT_SIGBUS; @@ -51,7 +58,35 @@ static vm_fault_t udmabuf_vm_fault(struct vm_fault *vmf) pfn = folio_pfn(ubuf->folios[pgoff]); pfn += ubuf->offsets[pgoff] >> PAGE_SHIFT; - return vmf_insert_pfn(vma, vmf->address, pfn); + ret = vmf_insert_pfn(vma, vmf->address, pfn); + if (ret & VM_FAULT_ERROR) + return ret; + + /* pre fault */ + pgoff = vma->vm_pgoff; + addr = vma->vm_start; + + for (; addr < vma->vm_end; pgoff++, addr += PAGE_SIZE) { + if (addr == vmf->address) + continue; + + if (WARN_ON(pgoff >= ubuf->pagecount)) + break; + + pfn = folio_pfn(ubuf->folios[pgoff]); + pfn += ubuf->offsets[pgoff] >> PAGE_SHIFT; + + /** + * If the below vmf_insert_pfn() fails, we do not return an + * error here during this pre-fault step. However, an error + * will be returned if the failure occurs when the addr is + * truly accessed. + */ + if (vmf_insert_pfn(vma, addr, pfn) & VM_FAULT_ERROR) + break; + } + + return ret; } static const struct vm_operations_struct udmabuf_vm_ops = { @@ -74,21 +109,29 @@ static int mmap_udmabuf(struct dma_buf *buf, struct vm_area_struct *vma) static int vmap_udmabuf(struct dma_buf *buf, struct iosys_map *map) { struct udmabuf *ubuf = buf->priv; - struct page **pages; + unsigned long *pfns; void *vaddr; pgoff_t pg; dma_resv_assert_held(buf->resv); - pages = kmalloc_array(ubuf->pagecount, sizeof(*pages), GFP_KERNEL); - if (!pages) + /** + * HVO may free tail pages, so just use pfn to map each folio + * into vmalloc area. + */ + pfns = kvmalloc_array(ubuf->pagecount, sizeof(*pfns), GFP_KERNEL); + if (!pfns) return -ENOMEM; - for (pg = 0; pg < ubuf->pagecount; pg++) - pages[pg] = &ubuf->folios[pg]->page; + for (pg = 0; pg < ubuf->pagecount; pg++) { + unsigned long pfn = folio_pfn(ubuf->folios[pg]); - vaddr = vm_map_ram(pages, ubuf->pagecount, -1); - kfree(pages); + pfn += ubuf->offsets[pg] >> PAGE_SHIFT; + pfns[pg] = pfn; + } + + vaddr = vmap_pfn(pfns, ubuf->pagecount, PAGE_KERNEL); + kvfree(pfns); if (!vaddr) return -EINVAL; @@ -159,34 +202,42 @@ static void unmap_udmabuf(struct dma_buf_attachment *at, return put_sg_table(at->dev, sg, direction); } -static void unpin_all_folios(struct list_head *unpin_list) +static void unpin_all_folios(struct udmabuf *ubuf) { - struct udmabuf_folio *ubuf_folio; + pgoff_t i; - while (!list_empty(unpin_list)) { - ubuf_folio = list_first_entry(unpin_list, - struct udmabuf_folio, list); - unpin_folio(ubuf_folio->folio); + for (i = 0; i < ubuf->nr_pinned; ++i) + unpin_folio(ubuf->pinned_folios[i]); - list_del(&ubuf_folio->list); - kfree(ubuf_folio); - } + kvfree(ubuf->pinned_folios); } -static int add_to_unpin_list(struct list_head *unpin_list, - struct folio *folio) +static __always_inline int init_udmabuf(struct udmabuf *ubuf, pgoff_t pgcnt) { - struct udmabuf_folio *ubuf_folio; - - ubuf_folio = kzalloc(sizeof(*ubuf_folio), GFP_KERNEL); - if (!ubuf_folio) + ubuf->folios = kvmalloc_array(pgcnt, sizeof(*ubuf->folios), GFP_KERNEL); + if (!ubuf->folios) + return -ENOMEM; + + ubuf->offsets = kvcalloc(pgcnt, sizeof(*ubuf->offsets), GFP_KERNEL); + if (!ubuf->offsets) + return -ENOMEM; + + ubuf->pinned_folios = kvmalloc_array(pgcnt, + sizeof(*ubuf->pinned_folios), + GFP_KERNEL); + if (!ubuf->pinned_folios) return -ENOMEM; - ubuf_folio->folio = folio; - list_add_tail(&ubuf_folio->list, unpin_list); return 0; } +static __always_inline void deinit_udmabuf(struct udmabuf *ubuf) +{ + unpin_all_folios(ubuf); + kvfree(ubuf->offsets); + kvfree(ubuf->folios); +} + static void release_udmabuf(struct dma_buf *buf) { struct udmabuf *ubuf = buf->priv; @@ -195,9 +246,7 @@ static void release_udmabuf(struct dma_buf *buf) if (ubuf->sg) put_sg_table(dev, ubuf->sg, DMA_BIDIRECTIONAL); - unpin_all_folios(&ubuf->unpin_list); - kfree(ubuf->offsets); - kfree(ubuf->folios); + deinit_udmabuf(ubuf); kfree(ubuf); } @@ -254,9 +303,6 @@ static int check_memfd_seals(struct file *memfd) { int seals; - if (!memfd) - return -EBADFD; - if (!shmem_file(memfd) && !is_file_hugepages(memfd)) return -EBADFD; @@ -291,100 +337,116 @@ static int export_udmabuf(struct udmabuf *ubuf, return dma_buf_fd(buf, flags); } +static long udmabuf_pin_folios(struct udmabuf *ubuf, struct file *memfd, + loff_t start, loff_t size, struct folio **folios) +{ + pgoff_t nr_pinned = ubuf->nr_pinned; + pgoff_t upgcnt = ubuf->pagecount; + u32 cur_folio, cur_pgcnt; + pgoff_t pgoff, pgcnt; + long nr_folios; + loff_t end; + + pgcnt = size >> PAGE_SHIFT; + end = start + (pgcnt << PAGE_SHIFT) - 1; + nr_folios = memfd_pin_folios(memfd, start, end, folios, pgcnt, &pgoff); + if (nr_folios <= 0) + return nr_folios ? nr_folios : -EINVAL; + + cur_pgcnt = 0; + for (cur_folio = 0; cur_folio < nr_folios; ++cur_folio) { + pgoff_t subpgoff = pgoff; + size_t fsize = folio_size(folios[cur_folio]); + + ubuf->pinned_folios[nr_pinned++] = folios[cur_folio]; + + for (; subpgoff < fsize; subpgoff += PAGE_SIZE) { + ubuf->folios[upgcnt] = folios[cur_folio]; + ubuf->offsets[upgcnt] = subpgoff; + ++upgcnt; + + if (++cur_pgcnt >= pgcnt) + goto end; + } + + /** + * In a given range, only the first subpage of the first folio + * has an offset, that is returned by memfd_pin_folios(). + * The first subpages of other folios (in the range) have an + * offset of 0. + */ + pgoff = 0; + } +end: + ubuf->pagecount = upgcnt; + ubuf->nr_pinned = nr_pinned; + return 0; +} + static long udmabuf_create(struct miscdevice *device, struct udmabuf_create_list *head, struct udmabuf_create_item *list) { - pgoff_t pgoff, pgcnt, pglimit, pgbuf = 0; - long nr_folios, ret = -EINVAL; - struct file *memfd = NULL; - struct folio **folios; + unsigned long max_nr_folios = 0; + struct folio **folios = NULL; + pgoff_t pgcnt = 0, pglimit; struct udmabuf *ubuf; - u32 i, j, k, flags; - loff_t end; + long ret = -EINVAL; + u32 i, flags; ubuf = kzalloc(sizeof(*ubuf), GFP_KERNEL); if (!ubuf) return -ENOMEM; - INIT_LIST_HEAD(&ubuf->unpin_list); pglimit = (size_limit_mb * 1024 * 1024) >> PAGE_SHIFT; for (i = 0; i < head->count; i++) { - if (!IS_ALIGNED(list[i].offset, PAGE_SIZE)) - goto err; - if (!IS_ALIGNED(list[i].size, PAGE_SIZE)) - goto err; - ubuf->pagecount += list[i].size >> PAGE_SHIFT; - if (ubuf->pagecount > pglimit) - goto err; + pgoff_t subpgcnt; + + if (!PAGE_ALIGNED(list[i].offset)) + goto err_noinit; + if (!PAGE_ALIGNED(list[i].size)) + goto err_noinit; + + subpgcnt = list[i].size >> PAGE_SHIFT; + pgcnt += subpgcnt; + if (pgcnt > pglimit) + goto err_noinit; + + max_nr_folios = max_t(unsigned long, subpgcnt, max_nr_folios); } - if (!ubuf->pagecount) + if (!pgcnt) + goto err_noinit; + + ret = init_udmabuf(ubuf, pgcnt); + if (ret) goto err; - ubuf->folios = kmalloc_array(ubuf->pagecount, sizeof(*ubuf->folios), - GFP_KERNEL); - if (!ubuf->folios) { - ret = -ENOMEM; - goto err; - } - ubuf->offsets = kcalloc(ubuf->pagecount, sizeof(*ubuf->offsets), - GFP_KERNEL); - if (!ubuf->offsets) { + folios = kvmalloc_array(max_nr_folios, sizeof(*folios), GFP_KERNEL); + if (!folios) { ret = -ENOMEM; goto err; } - pgbuf = 0; for (i = 0; i < head->count; i++) { - memfd = fget(list[i].memfd); + struct file *memfd = fget(list[i].memfd); + + if (!memfd) { + ret = -EBADFD; + goto err; + } + ret = check_memfd_seals(memfd); - if (ret < 0) - goto err; - - pgcnt = list[i].size >> PAGE_SHIFT; - folios = kmalloc_array(pgcnt, sizeof(*folios), GFP_KERNEL); - if (!folios) { - ret = -ENOMEM; + if (ret < 0) { + fput(memfd); goto err; } - end = list[i].offset + (pgcnt << PAGE_SHIFT) - 1; - ret = memfd_pin_folios(memfd, list[i].offset, end, - folios, pgcnt, &pgoff); - if (ret <= 0) { - kfree(folios); - if (!ret) - ret = -EINVAL; - goto err; - } - - nr_folios = ret; - pgoff >>= PAGE_SHIFT; - for (j = 0, k = 0; j < pgcnt; j++) { - ubuf->folios[pgbuf] = folios[k]; - ubuf->offsets[pgbuf] = pgoff << PAGE_SHIFT; - - if (j == 0 || ubuf->folios[pgbuf-1] != folios[k]) { - ret = add_to_unpin_list(&ubuf->unpin_list, - folios[k]); - if (ret < 0) { - kfree(folios); - goto err; - } - } - - pgbuf++; - if (++pgoff == folio_nr_pages(folios[k])) { - pgoff = 0; - if (++k == nr_folios) - break; - } - } - - kfree(folios); + ret = udmabuf_pin_folios(ubuf, memfd, list[i].offset, + list[i].size, folios); fput(memfd); - memfd = NULL; + if (ret) + goto err; } flags = head->flags & UDMABUF_FLAGS_CLOEXEC ? O_CLOEXEC : 0; @@ -392,15 +454,14 @@ static long udmabuf_create(struct miscdevice *device, if (ret < 0) goto err; + kvfree(folios); return ret; err: - if (memfd) - fput(memfd); - unpin_all_folios(&ubuf->unpin_list); - kfree(ubuf->offsets); - kfree(ubuf->folios); + deinit_udmabuf(ubuf); +err_noinit: kfree(ubuf); + kvfree(folios); return ret; } diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig index 1cb5a4f19293..5504721007cc 100644 --- a/drivers/gpu/drm/Kconfig +++ b/drivers/gpu/drm/Kconfig @@ -9,9 +9,6 @@ menuconfig DRM tristate "Direct Rendering Manager (XFree86 4.1.0 and higher DRI support)" depends on (AGP || AGP=n) && !EMULATED_CMPXCHG && HAS_DMA select DRM_PANEL_ORIENTATION_QUIRKS - select DRM_KMS_HELPER if DRM_FBDEV_EMULATION - select FB_CORE if DRM_FBDEV_EMULATION - select FB_SYSMEM_HELPERS_DEFERRED if DRM_FBDEV_EMULATION select HDMI select I2C select DMA_SHARED_BUFFER @@ -152,6 +149,7 @@ config DRM_PANIC_SCREEN config DRM_PANIC_SCREEN_QR_CODE bool "Add a panic screen with a QR code" depends on DRM_PANIC && RUST + select ZLIB_DEFLATE help This option adds a QR code generator, and a panic screen with a QR code. The QR code will contain the last lines of kmsg and other debug @@ -210,9 +208,47 @@ config DRM_DEBUG_MODESET_LOCK If in doubt, say "N". +config DRM_CLIENT + bool + depends on DRM + help + Enables support for DRM clients. DRM drivers that need + struct drm_client_dev and its interfaces should select this + option. Drivers that support the default clients should + select DRM_CLIENT_SELECTION instead. + +config DRM_CLIENT_LIB + tristate + depends on DRM + select DRM_KMS_HELPER if DRM_FBDEV_EMULATION + select FB_CORE if DRM_FBDEV_EMULATION + help + This option enables the DRM client library and selects all + modules and components according to the enabled clients. + +config DRM_CLIENT_SELECTION + tristate + depends on DRM + select DRM_CLIENT_LIB if DRM_FBDEV_EMULATION + help + Drivers that support in-kernel DRM clients have to select this + option. + +config DRM_CLIENT_SETUP + bool + depends on DRM_CLIENT_SELECTION + help + Enables the DRM client selection. DRM drivers that support the + default clients should select DRM_CLIENT_SELECTION instead. + +menu "Supported DRM clients" + depends on DRM_CLIENT_SELECTION + config DRM_FBDEV_EMULATION bool "Enable legacy fbdev support for your modesetting driver" - depends on DRM + depends on DRM_CLIENT_SELECTION + select DRM_CLIENT + select DRM_CLIENT_SETUP select FRAMEBUFFER_CONSOLE_DETECT_PRIMARY if FRAMEBUFFER_CONSOLE default FB help @@ -251,6 +287,8 @@ config DRM_FBDEV_LEAK_PHYS_SMEM If in doubt, say "N" or spread the word to your closed source library vendor. +endmenu + config DRM_LOAD_EDID_FIRMWARE bool "Allow to specify an EDID data set instead of probing for it" depends on DRM @@ -320,19 +358,21 @@ config DRM_TTM_HELPER tristate depends on DRM select DRM_TTM + select FB_SYSMEM_HELPERS_DEFERRED if DRM_FBDEV_EMULATION help Helpers for ttm-based gem objects config DRM_GEM_DMA_HELPER tristate depends on DRM - select FB_DMAMEM_HELPERS if DRM_FBDEV_EMULATION + select FB_DMAMEM_HELPERS_DEFERRED if DRM_FBDEV_EMULATION help Choose this if you need the GEM DMA helper functions config DRM_GEM_SHMEM_HELPER tristate depends on DRM && MMU + select FB_SYSMEM_HELPERS_DEFERRED if DRM_FBDEV_EMULATION help Choose this if you need the GEM shmem helper functions @@ -472,6 +512,7 @@ source "drivers/gpu/drm/imagination/Kconfig" config DRM_HYPERV tristate "DRM Support for Hyper-V synthetic video device" depends on DRM && PCI && MMU && HYPERV + select DRM_CLIENT_SELECTION select DRM_KMS_HELPER select DRM_GEM_SHMEM_HELPER help diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile index 784229d4504d..463afad1b5ca 100644 --- a/drivers/gpu/drm/Makefile +++ b/drivers/gpu/drm/Makefile @@ -34,15 +34,12 @@ endif subdir-ccflags-$(CONFIG_DRM_WERROR) += -Werror drm-y := \ - drm_aperture.o \ drm_atomic.o \ drm_atomic_uapi.o \ drm_auth.o \ drm_blend.o \ drm_bridge.o \ drm_cache.o \ - drm_client.o \ - drm_client_modeset.o \ drm_color_mgmt.o \ drm_connector.o \ drm_crtc.o \ @@ -68,6 +65,7 @@ drm-y := \ drm_prime.o \ drm_print.o \ drm_property.o \ + drm_rect.o \ drm_syncobj.o \ drm_sysfs.o \ drm_trace_points.o \ @@ -75,6 +73,10 @@ drm-y := \ drm_vblank_work.o \ drm_vma_manager.o \ drm_writeback.o +drm-$(CONFIG_DRM_CLIENT) += \ + drm_client.o \ + drm_client_event.o \ + drm_client_modeset.o drm-$(CONFIG_DRM_LIB_RANDOM) += lib/drm_random.o drm-$(CONFIG_COMPAT) += drm_ioc32.o drm-$(CONFIG_DRM_PANEL) += drm_panel.o @@ -140,13 +142,20 @@ drm_kms_helper-y := \ drm_modeset_helper.o \ drm_plane_helper.o \ drm_probe_helper.o \ - drm_rect.o \ drm_self_refresh_helper.o \ drm_simple_kms_helper.o drm_kms_helper-$(CONFIG_DRM_PANEL_BRIDGE) += bridge/panel.o drm_kms_helper-$(CONFIG_DRM_FBDEV_EMULATION) += drm_fb_helper.o obj-$(CONFIG_DRM_KMS_HELPER) += drm_kms_helper.o +# +# DRM clients +# + +drm_client_lib-y := drm_client_setup.o +drm_client_lib-$(CONFIG_DRM_FBDEV_EMULATION) += drm_fbdev_client.o +obj-$(CONFIG_DRM_CLIENT_LIB) += drm_client_lib.o + # # Drivers and the rest # diff --git a/drivers/gpu/drm/amd/amdgpu/Kconfig b/drivers/gpu/drm/amd/amdgpu/Kconfig index 0051fb1b437f..41fa3377d9cf 100644 --- a/drivers/gpu/drm/amd/amdgpu/Kconfig +++ b/drivers/gpu/drm/amd/amdgpu/Kconfig @@ -5,7 +5,10 @@ config DRM_AMDGPU depends on DRM && PCI && MMU depends on !UML select FW_LOADER + select DRM_CLIENT + select DRM_CLIENT_SELECTION select DRM_DISPLAY_DP_HELPER + select DRM_DISPLAY_DSC_HELPER select DRM_DISPLAY_HDMI_HELPER select DRM_DISPLAY_HDCP_HELPER select DRM_DISPLAY_HELPER diff --git a/drivers/gpu/drm/amd/amdgpu/aldebaran.c b/drivers/gpu/drm/amd/amdgpu/aldebaran.c index b0f95a7649bf..3a588fecb0c5 100644 --- a/drivers/gpu/drm/amd/amdgpu/aldebaran.c +++ b/drivers/gpu/drm/amd/amdgpu/aldebaran.c @@ -85,16 +85,9 @@ static int aldebaran_mode2_suspend_ip(struct amdgpu_device *adev) AMD_IP_BLOCK_TYPE_SDMA)) continue; - r = adev->ip_blocks[i].version->funcs->suspend(adev); - - if (r) { - dev_err(adev->dev, - "suspend of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); + r = amdgpu_ip_block_suspend(&adev->ip_blocks[i]); + if (r) return r; - } - - adev->ip_blocks[i].status.hw = false; } return 0; @@ -246,7 +239,7 @@ static int aldebaran_mode2_restore_ip(struct amdgpu_device *adev) dev_err(adev->dev, "Failed to get BIF handle\n"); return -EINVAL; } - r = cmn_block->version->funcs->resume(adev); + r = amdgpu_ip_block_resume(cmn_block); if (r) return r; @@ -282,15 +275,10 @@ static int aldebaran_mode2_restore_ip(struct amdgpu_device *adev) adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_SDMA)) continue; - r = adev->ip_blocks[i].version->funcs->resume(adev); - if (r) { - dev_err(adev->dev, - "resume of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); - return r; - } - adev->ip_blocks[i].status.hw = true; + r = amdgpu_ip_block_resume(&adev->ip_blocks[i]); + if (r) + return r; } for (i = 0; i < adev->num_ip_blocks; i++) { @@ -304,7 +292,7 @@ static int aldebaran_mode2_restore_ip(struct amdgpu_device *adev) if (adev->ip_blocks[i].version->funcs->late_init) { r = adev->ip_blocks[i].version->funcs->late_init( - (void *)adev); + &adev->ip_blocks[i]); if (r) { dev_err(adev->dev, "late_init of IP block <%s> failed %d after reset\n", @@ -417,6 +405,7 @@ static struct amdgpu_reset_handler aldebaran_mode2_handler = { static struct amdgpu_reset_handler *aldebaran_rst_handlers[AMDGPU_RESET_MAX_HANDLERS] = { &aldebaran_mode2_handler, + &xgmi_reset_on_init_handler, }; int aldebaran_reset_init(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 9b1e0ede05a4..d8bc6da50016 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -118,7 +118,7 @@ #define MAX_GPU_INSTANCE 64 -#define GFX_SLICE_PERIOD msecs_to_jiffies(250) +#define GFX_SLICE_PERIOD_MS 250 struct amdgpu_gpu_instance { struct amdgpu_device *adev; @@ -131,10 +131,6 @@ struct amdgpu_mgpu_info { uint32_t num_gpu; uint32_t num_dgpu; uint32_t num_apu; - - /* delayed reset_func for XGMI configuration if necessary */ - struct delayed_work delayed_reset_work; - bool pending_reset; }; enum amdgpu_ss { @@ -303,6 +299,12 @@ extern int amdgpu_wbrf; #define AMDGPU_RESET_VCE (1 << 13) #define AMDGPU_RESET_VCE1 (1 << 14) +/* reset mask */ +#define AMDGPU_RESET_TYPE_FULL (1 << 0) /* full adapter reset, mode1/mode2/BACO/etc. */ +#define AMDGPU_RESET_TYPE_SOFT_RESET (1 << 1) /* IP level soft reset */ +#define AMDGPU_RESET_TYPE_PER_QUEUE (1 << 2) /* per queue */ +#define AMDGPU_RESET_TYPE_PER_PIPE (1 << 3) /* per pipe */ + /* max cursor sizes (in pixels) */ #define CIK_CURSOR_WIDTH 128 #define CIK_CURSOR_HEIGHT 128 @@ -365,8 +367,11 @@ void amdgpu_device_ip_get_clockgating_state(struct amdgpu_device *adev, u64 *flags); int amdgpu_device_ip_wait_for_idle(struct amdgpu_device *adev, enum amd_ip_block_type block_type); -bool amdgpu_device_ip_is_idle(struct amdgpu_device *adev, +bool amdgpu_device_ip_is_valid(struct amdgpu_device *adev, enum amd_ip_block_type block_type); +int amdgpu_ip_block_suspend(struct amdgpu_ip_block *ip_block); + +int amdgpu_ip_block_resume(struct amdgpu_ip_block *ip_block); #define AMDGPU_MAX_IP_NUM 16 @@ -389,6 +394,7 @@ struct amdgpu_ip_block_version { struct amdgpu_ip_block { struct amdgpu_ip_block_status status; const struct amdgpu_ip_block_version *version; + struct amdgpu_device *adev; }; int amdgpu_device_ip_block_version_cmp(struct amdgpu_device *adev, @@ -563,6 +569,7 @@ enum amd_reset_method { AMD_RESET_METHOD_MODE2, AMD_RESET_METHOD_BACO, AMD_RESET_METHOD_PCI, + AMD_RESET_METHOD_ON_INIT, }; struct amdgpu_video_codec_info { @@ -821,6 +828,24 @@ struct amdgpu_mqd { struct amdgpu_mqd_prop *p); }; +/* + * Custom Init levels could be defined for different situations where a full + * initialization of all hardware blocks are not expected. Sample cases are + * custom init sequences after resume after S0i3/S3, reset on initialization, + * partial reset of blocks etc. Presently, this defines only two levels. Levels + * are described in corresponding struct definitions - amdgpu_init_default, + * amdgpu_init_minimal_xgmi. + */ +enum amdgpu_init_lvl_id { + AMDGPU_INIT_LEVEL_DEFAULT, + AMDGPU_INIT_LEVEL_MINIMAL_XGMI, +}; + +struct amdgpu_init_level { + enum amdgpu_init_lvl_id level; + uint32_t hwini_ip_block_mask; +}; + #define AMDGPU_RESET_MAGIC_NUM 64 #define AMDGPU_MAX_DF_PERFMONS 4 struct amdgpu_reset_domain; @@ -1092,8 +1117,6 @@ struct amdgpu_device { bool in_s3; bool in_s4; bool in_s0ix; - /* indicate amdgpu suspension status */ - bool suspend_complete; enum pp_mp1_state mp1_state; struct amdgpu_doorbell_index doorbell_index; @@ -1166,6 +1189,8 @@ struct amdgpu_device { bool enforce_isolation[MAX_XCP]; /* Added this mutex for cleaner shader isolation between GFX and compute processes */ struct mutex enforce_isolation_mutex; + + struct amdgpu_init_level *init_lvl; }; static inline uint32_t amdgpu_ip_version(const struct amdgpu_device *adev, @@ -1261,6 +1286,8 @@ int amdgpu_device_pre_asic_reset(struct amdgpu_device *adev, int amdgpu_do_asic_reset(struct list_head *device_list_handle, struct amdgpu_reset_context *reset_context); +int amdgpu_device_reinit_after_reset(struct amdgpu_reset_context *reset_context); + int emu_soc_asic_init(struct amdgpu_device *adev); /* @@ -1443,6 +1470,8 @@ struct dma_fence *amdgpu_device_get_gang(struct amdgpu_device *adev); struct dma_fence *amdgpu_device_switch_gang(struct amdgpu_device *adev, struct dma_fence *gang); bool amdgpu_device_has_display_hardware(struct amdgpu_device *adev); +ssize_t amdgpu_get_soft_full_reset_mask(struct amdgpu_ring *ring); +ssize_t amdgpu_show_reset_mask(char *buf, uint32_t supported_reset); /* atpx handler */ #if defined(CONFIG_VGA_SWITCHEROO) @@ -1450,23 +1479,15 @@ void amdgpu_register_atpx_handler(void); void amdgpu_unregister_atpx_handler(void); bool amdgpu_has_atpx_dgpu_power_cntl(void); bool amdgpu_is_atpx_hybrid(void); -bool amdgpu_atpx_dgpu_req_power_for_displays(void); bool amdgpu_has_atpx(void); #else static inline void amdgpu_register_atpx_handler(void) {} static inline void amdgpu_unregister_atpx_handler(void) {} static inline bool amdgpu_has_atpx_dgpu_power_cntl(void) { return false; } static inline bool amdgpu_is_atpx_hybrid(void) { return false; } -static inline bool amdgpu_atpx_dgpu_req_power_for_displays(void) { return false; } static inline bool amdgpu_has_atpx(void) { return false; } #endif -#if defined(CONFIG_VGA_SWITCHEROO) && defined(CONFIG_ACPI) -void *amdgpu_atpx_get_dhandle(void); -#else -static inline void *amdgpu_atpx_get_dhandle(void) { return NULL; } -#endif - /* * KMS */ @@ -1619,4 +1640,6 @@ extern const struct attribute_group amdgpu_vram_mgr_attr_group; extern const struct attribute_group amdgpu_gtt_mgr_attr_group; extern const struct attribute_group amdgpu_flash_attr_group; +void amdgpu_set_init_level(struct amdgpu_device *adev, + enum amdgpu_init_lvl_id lvl); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_aca.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_aca.c index 2ca127173135..9d6345146495 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_aca.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_aca.c @@ -158,7 +158,7 @@ static int aca_smu_get_valid_aca_banks(struct amdgpu_device *adev, enum aca_smu_ return -EINVAL; } - if (start + count >= max_count) + if (start + count > max_count) return -EINVAL; count = min_t(int, count, max_count); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c index bf6c4a0d0525..ec5e0dcf8613 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c @@ -98,9 +98,9 @@ enum { ACP_TILE_DSP2, }; -static int acp_sw_init(void *handle) +static int acp_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->acp.parent = adev->dev; @@ -112,9 +112,9 @@ static int acp_sw_init(void *handle) return 0; } -static int acp_sw_fini(void *handle) +static int acp_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->acp.cgs_device) amdgpu_cgs_destroy_device(adev->acp.cgs_device); @@ -219,10 +219,10 @@ static const struct dmi_system_id acp_quirk_table[] = { /** * acp_hw_init - start and test ACP block * - * @handle: handle used to pass amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * */ -static int acp_hw_init(void *handle) +static int acp_hw_init(struct amdgpu_ip_block *ip_block) { int r; u64 acp_base; @@ -230,13 +230,7 @@ static int acp_hw_init(void *handle) u32 count = 0; struct i2s_platform_data *i2s_pdata = NULL; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - const struct amdgpu_ip_block *ip_block = - amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_ACP); - - if (!ip_block) - return -EINVAL; + struct amdgpu_device *adev = ip_block->adev; r = amd_acp_hw_init(adev->acp.cgs_device, ip_block->version->major, ip_block->version->minor); @@ -503,14 +497,14 @@ failure: /** * acp_hw_fini - stop the hardware block * - * @handle: handle used to pass amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * */ -static int acp_hw_fini(void *handle) +static int acp_hw_fini(struct amdgpu_ip_block *ip_block) { u32 val = 0; u32 count = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* return early if no ACP */ if (!adev->acp.acp_genpd) { @@ -565,9 +559,9 @@ static int acp_hw_fini(void *handle) return 0; } -static int acp_suspend(void *handle) +static int acp_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* power up on suspend */ if (!adev->acp.acp_cell) @@ -575,9 +569,9 @@ static int acp_suspend(void *handle) return 0; } -static int acp_resume(void *handle) +static int acp_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* power down again on resume */ if (!adev->acp.acp_cell) @@ -585,26 +579,11 @@ static int acp_resume(void *handle) return 0; } -static int acp_early_init(void *handle) -{ - return 0; -} - static bool acp_is_idle(void *handle) { return true; } -static int acp_wait_for_idle(void *handle) -{ - return 0; -} - -static int acp_soft_reset(void *handle) -{ - return 0; -} - static int acp_set_clockgating_state(void *handle, enum amd_clockgating_state state) { @@ -624,8 +603,6 @@ static int acp_set_powergating_state(void *handle, static const struct amd_ip_funcs acp_ip_funcs = { .name = "acp_ip", - .early_init = acp_early_init, - .late_init = NULL, .sw_init = acp_sw_init, .sw_fini = acp_sw_fini, .hw_init = acp_hw_init, @@ -633,12 +610,8 @@ static const struct amd_ip_funcs acp_ip_funcs = { .suspend = acp_suspend, .resume = acp_resume, .is_idle = acp_is_idle, - .wait_for_idle = acp_wait_for_idle, - .soft_reset = acp_soft_reset, .set_clockgating_state = acp_set_clockgating_state, .set_powergating_state = acp_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; const struct amdgpu_ip_block_version acp_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c index 7dd55ed57c1d..b8d4e07d2043 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c @@ -800,6 +800,7 @@ int amdgpu_acpi_power_shift_control(struct amdgpu_device *adev, return -EIO; } + kfree(info); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index 4f08b153cb66..3afcd1e8aa54 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -834,6 +834,9 @@ int amdgpu_amdkfd_unmap_hiq(struct amdgpu_device *adev, u32 doorbell_off, if (!kiq->pmf || !kiq->pmf->kiq_unmap_queues) return -EINVAL; + if (!kiq_ring->sched.ready || adev->job_hang) + return 0; + ring_funcs = kzalloc(sizeof(*ring_funcs), GFP_KERNEL); if (!ring_funcs) return -ENOMEM; @@ -858,8 +861,14 @@ int amdgpu_amdkfd_unmap_hiq(struct amdgpu_device *adev, u32 doorbell_off, kiq->pmf->kiq_unmap_queues(kiq_ring, ring, RESET_QUEUES, 0, 0); - if (kiq_ring->sched.ready && !adev->job_hang) - r = amdgpu_ring_test_helper(kiq_ring); + /* Submit unmap queue packet */ + amdgpu_ring_commit(kiq_ring); + /* + * Ring test will do a basic scratch register change check. Just run + * this to ensure that unmap queues that is submitted before got + * processed successfully before returning. + */ + r = amdgpu_ring_test_helper(kiq_ring); spin_unlock(&kiq->ring_lock); @@ -889,3 +898,27 @@ int amdgpu_amdkfd_start_sched(struct amdgpu_device *adev, uint32_t node_id) return kgd2kfd_start_sched(adev->kfd.dev, node_id); } + +/* check if there are KFD queues active */ +bool amdgpu_amdkfd_compute_active(struct amdgpu_device *adev, uint32_t node_id) +{ + if (!adev->kfd.init_complete) + return false; + + return kgd2kfd_compute_active(adev->kfd.dev, node_id); +} + +/* Config CGTT_SQ_CLK_CTRL */ +int amdgpu_amdkfd_config_sq_perfmon(struct amdgpu_device *adev, uint32_t xcp_id, + bool core_override_enable, bool reg_override_enable, bool perfmon_override_enable) +{ + int r; + + if (!adev->kfd.init_complete) + return 0; + + r = psp_config_sq_perfmon(&adev->psp, xcp_id, core_override_enable, + reg_override_enable, perfmon_override_enable); + + return r; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index f9d119448442..4b80ad860639 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -266,6 +266,10 @@ int amdgpu_amdkfd_unmap_hiq(struct amdgpu_device *adev, u32 doorbell_off, u32 inst); int amdgpu_amdkfd_start_sched(struct amdgpu_device *adev, uint32_t node_id); int amdgpu_amdkfd_stop_sched(struct amdgpu_device *adev, uint32_t node_id); +int amdgpu_amdkfd_config_sq_perfmon(struct amdgpu_device *adev, uint32_t xcp_id, + bool core_override_enable, bool reg_override_enable, bool perfmon_override_enable); +bool amdgpu_amdkfd_compute_active(struct amdgpu_device *adev, uint32_t node_id); + /* Read user wptr from a specified user address space with page fault * disabled. The memory must be pinned and mapped to the hardware when @@ -428,6 +432,7 @@ int kgd2kfd_check_and_lock_kfd(void); void kgd2kfd_unlock_kfd(void); int kgd2kfd_start_sched(struct kfd_dev *kfd, uint32_t node_id); int kgd2kfd_stop_sched(struct kfd_dev *kfd, uint32_t node_id); +bool kgd2kfd_compute_active(struct kfd_dev *kfd, uint32_t node_id); #else static inline int kgd2kfd_init(void) { @@ -508,5 +513,10 @@ static inline int kgd2kfd_stop_sched(struct kfd_dev *kfd, uint32_t node_id) { return 0; } + +static inline bool kgd2kfd_compute_active(struct kfd_dev *kfd, uint32_t node_id) +{ + return false; +} #endif #endif /* AMDGPU_AMDKFD_H_INCLUDED */ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_arcturus.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_arcturus.c index 9435af2e6bdc..9abf29b58ac7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_arcturus.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_arcturus.c @@ -299,7 +299,7 @@ static int suspend_resume_compute_scheduler(struct amdgpu_device *adev, bool sus if (r) goto out; } else { - drm_sched_start(&ring->sched); + drm_sched_start(&ring->sched, 0); } } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c index 3bc0cbf45bc5..cc66ebb7bae1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c @@ -944,9 +944,7 @@ static void unlock_spi_csq_mutexes(struct amdgpu_device *adev) * * @adev: Handle of device whose registers are to be read * @queue_idx: Index of queue in the queue-map bit-field - * @wave_cnt: Output parameter updated with number of waves in flight - * @vmid: Output parameter updated with VMID of queue whose wave count - * is being collected + * @queue_cnt: Stores the wave count and doorbell offset for an active queue * @inst: xcc's instance number on a multi-XCC setup */ static void get_wave_count(struct amdgpu_device *adev, int queue_idx, @@ -1133,10 +1131,6 @@ uint64_t kgd_gfx_v9_hqd_get_pq_addr(struct amdgpu_device *adev, uint32_t low, high; uint64_t queue_addr = 0; - if (!adev->debug_exp_resets && - !adev->gfx.num_gfx_rings) - return 0; - kgd_gfx_v9_acquire_queue(adev, pipe_id, queue_id, inst); amdgpu_gfx_rlc_enter_safe_mode(adev, inst); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index fa572ba7f9fc..f30548f4c3b3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -2524,11 +2524,14 @@ int amdgpu_amdkfd_evict_userptr(struct mmu_interval_notifier *mni, /* First eviction, stop the queues */ r = kgd2kfd_quiesce_mm(mni->mm, KFD_QUEUE_EVICTION_TRIGGER_USERPTR); - if (r) + + if (r && r != -ESRCH) pr_err("Failed to quiesce KFD\n"); - queue_delayed_work(system_freezable_wq, - &process_info->restore_userptr_work, - msecs_to_jiffies(AMDGPU_USERPTR_RESTORE_DELAY_MS)); + + if (r != -ESRCH) + queue_delayed_work(system_freezable_wq, + &process_info->restore_userptr_work, + msecs_to_jiffies(AMDGPU_USERPTR_RESTORE_DELAY_MS)); } mutex_unlock(&process_info->notifier_lock); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c index 0c8975ac5af9..093141ad6ed0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c @@ -1145,8 +1145,8 @@ int amdgpu_atombios_get_memory_pll_dividers(struct amdgpu_device *adev, return 0; } -void amdgpu_atombios_set_engine_dram_timings(struct amdgpu_device *adev, - u32 eng_clock, u32 mem_clock) +int amdgpu_atombios_set_engine_dram_timings(struct amdgpu_device *adev, + u32 eng_clock, u32 mem_clock) { SET_ENGINE_CLOCK_PS_ALLOCATION args; int index = GetIndexIntoMasterTable(COMMAND, DynamicMemorySettings); @@ -1161,8 +1161,8 @@ void amdgpu_atombios_set_engine_dram_timings(struct amdgpu_device *adev, if (mem_clock) args.sReserved.ulClock = cpu_to_le32(mem_clock & SET_CLOCK_FREQ_MASK); - amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args, - sizeof(args)); + return amdgpu_atom_execute_table(adev->mode_info.atom_context, index, + (uint32_t *)&args, sizeof(args)); } void amdgpu_atombios_get_default_voltages(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h index 0811474e8fd3..0e16432d9a72 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h @@ -163,8 +163,8 @@ int amdgpu_atombios_get_memory_pll_dividers(struct amdgpu_device *adev, bool strobe_mode, struct atom_mpll_param *mpll_param); -void amdgpu_atombios_set_engine_dram_timings(struct amdgpu_device *adev, - u32 eng_clock, u32 mem_clock); +int amdgpu_atombios_set_engine_dram_timings(struct amdgpu_device *adev, + u32 eng_clock, u32 mem_clock); bool amdgpu_atombios_is_voltage_gpio(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c index 375f02002579..3893e6fc2f03 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c @@ -89,18 +89,6 @@ bool amdgpu_is_atpx_hybrid(void) return amdgpu_atpx_priv.atpx.is_hybrid; } -bool amdgpu_atpx_dgpu_req_power_for_displays(void) -{ - return amdgpu_atpx_priv.atpx.dgpu_req_power_for_displays; -} - -#if defined(CONFIG_ACPI) -void *amdgpu_atpx_get_dhandle(void) -{ - return amdgpu_atpx_priv.dhandle; -} -#endif - /** * amdgpu_atpx_call - call an ATPX method * diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c index 9da4414de617..a68338cb7b4a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c @@ -2095,6 +2095,11 @@ int amdgpu_debugfs_init(struct amdgpu_device *adev) if (amdgpu_umsch_mm & amdgpu_umsch_mm_fwlog) amdgpu_debugfs_umsch_fwlog_init(adev, &adev->umsch_mm); + amdgpu_debugfs_jpeg_sched_mask_init(adev); + amdgpu_debugfs_gfx_sched_mask_init(adev); + amdgpu_debugfs_compute_sched_mask_init(adev); + amdgpu_debugfs_sdma_sched_mask_init(adev); + amdgpu_ras_debugfs_create_all(adev); amdgpu_rap_debugfs_init(adev); amdgpu_securedisplay_debugfs_init(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c index 5ac59b62020c..946c48829f19 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c @@ -203,6 +203,7 @@ amdgpu_devcoredump_read(char *buffer, loff_t offset, size_t count, struct amdgpu_coredump_info *coredump = data; struct drm_print_iterator iter; struct amdgpu_vm_fault_info *fault_info; + struct amdgpu_ip_block *ip_block; int ver; iter.data = buffer; @@ -282,13 +283,10 @@ amdgpu_devcoredump_read(char *buffer, loff_t offset, size_t count, /* dump the ip state for each ip */ drm_printf(&p, "IP Dump\n"); for (int i = 0; i < coredump->adev->num_ip_blocks; i++) { - if (coredump->adev->ip_blocks[i].version->funcs->print_ip_state) { - drm_printf(&p, "IP: %s\n", - coredump->adev->ip_blocks[i] - .version->funcs->name); - coredump->adev->ip_blocks[i] - .version->funcs->print_ip_state( - (void *)coredump->adev, &p); + ip_block = &coredump->adev->ip_blocks[i]; + if (ip_block->version->funcs->print_ip_state) { + drm_printf(&p, "IP: %s\n", ip_block->version->funcs->name); + ip_block->version->funcs->print_ip_state(ip_block, &p); drm_printf(&p, "\n"); } } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index c2394c8b4d6b..0171d240fcb0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -25,6 +25,8 @@ * Alex Deucher * Jerome Glisse */ + +#include #include #include #include @@ -35,10 +37,9 @@ #include #include -#include #include +#include #include -#include #include #include #include @@ -144,6 +145,51 @@ const char *amdgpu_asic_name[] = { "LAST", }; +#define AMDGPU_IP_BLK_MASK_ALL GENMASK(AMDGPU_MAX_IP_NUM - 1, 0) +/* + * Default init level where all blocks are expected to be initialized. This is + * the level of initialization expected by default and also after a full reset + * of the device. + */ +struct amdgpu_init_level amdgpu_init_default = { + .level = AMDGPU_INIT_LEVEL_DEFAULT, + .hwini_ip_block_mask = AMDGPU_IP_BLK_MASK_ALL, +}; + +/* + * Minimal blocks needed to be initialized before a XGMI hive can be reset. This + * is used for cases like reset on initialization where the entire hive needs to + * be reset before first use. + */ +struct amdgpu_init_level amdgpu_init_minimal_xgmi = { + .level = AMDGPU_INIT_LEVEL_MINIMAL_XGMI, + .hwini_ip_block_mask = + BIT(AMD_IP_BLOCK_TYPE_GMC) | BIT(AMD_IP_BLOCK_TYPE_SMC) | + BIT(AMD_IP_BLOCK_TYPE_COMMON) | BIT(AMD_IP_BLOCK_TYPE_IH) | + BIT(AMD_IP_BLOCK_TYPE_PSP) +}; + +static inline bool amdgpu_ip_member_of_hwini(struct amdgpu_device *adev, + enum amd_ip_block_type block) +{ + return (adev->init_lvl->hwini_ip_block_mask & (1U << block)) != 0; +} + +void amdgpu_set_init_level(struct amdgpu_device *adev, + enum amdgpu_init_lvl_id lvl) +{ + switch (lvl) { + case AMDGPU_INIT_LEVEL_MINIMAL_XGMI: + adev->init_lvl = &amdgpu_init_minimal_xgmi; + break; + case AMDGPU_INIT_LEVEL_DEFAULT: + fallthrough; + default: + adev->init_lvl = &amdgpu_init_default; + break; + } +} + static inline void amdgpu_device_stop_pending_resets(struct amdgpu_device *adev); /** @@ -227,6 +273,42 @@ void amdgpu_reg_state_sysfs_fini(struct amdgpu_device *adev) sysfs_remove_bin_file(&adev->dev->kobj, &bin_attr_reg_state); } +int amdgpu_ip_block_suspend(struct amdgpu_ip_block *ip_block) +{ + int r; + + if (ip_block->version->funcs->suspend) { + r = ip_block->version->funcs->suspend(ip_block); + if (r) { + dev_err(ip_block->adev->dev, + "suspend of IP block <%s> failed %d\n", + ip_block->version->funcs->name, r); + return r; + } + } + + ip_block->status.hw = false; + return 0; +} + +int amdgpu_ip_block_resume(struct amdgpu_ip_block *ip_block) +{ + int r; + + if (ip_block->version->funcs->resume) { + r = ip_block->version->funcs->resume(ip_block); + if (r) { + dev_err(ip_block->adev->dev, + "resume of IP block <%s> failed %d\n", + ip_block->version->funcs->name, r); + return r; + } + } + + ip_block->status.hw = true; + return 0; +} + /** * DOC: board_info * @@ -1655,7 +1737,7 @@ bool amdgpu_device_need_post(struct amdgpu_device *adev) } /* Don't post if we need to reset whole hive on init */ - if (adev->gmc.xgmi.pending_reset) + if (adev->init_lvl->level == AMDGPU_INIT_LEVEL_MINIMAL_XGMI) return false; if (adev->has_hw_reset) { @@ -2159,9 +2241,12 @@ int amdgpu_device_ip_wait_for_idle(struct amdgpu_device *adev, if (!adev->ip_blocks[i].status.valid) continue; if (adev->ip_blocks[i].version->type == block_type) { - r = adev->ip_blocks[i].version->funcs->wait_for_idle((void *)adev); - if (r) - return r; + if (adev->ip_blocks[i].version->funcs->wait_for_idle) { + r = adev->ip_blocks[i].version->funcs->wait_for_idle( + &adev->ip_blocks[i]); + if (r) + return r; + } break; } } @@ -2170,26 +2255,24 @@ int amdgpu_device_ip_wait_for_idle(struct amdgpu_device *adev, } /** - * amdgpu_device_ip_is_idle - is the hardware IP idle + * amdgpu_device_ip_is_valid - is the hardware IP enabled * * @adev: amdgpu_device pointer * @block_type: Type of hardware IP (SMU, GFX, UVD, etc.) * - * Check if the hardware IP is idle or not. - * Returns true if it the IP is idle, false if not. + * Check if the hardware IP is enable or not. + * Returns true if it the IP is enable, false if not. */ -bool amdgpu_device_ip_is_idle(struct amdgpu_device *adev, - enum amd_ip_block_type block_type) +bool amdgpu_device_ip_is_valid(struct amdgpu_device *adev, + enum amd_ip_block_type block_type) { int i; for (i = 0; i < adev->num_ip_blocks; i++) { - if (!adev->ip_blocks[i].status.valid) - continue; if (adev->ip_blocks[i].version->type == block_type) - return adev->ip_blocks[i].version->funcs->is_idle((void *)adev); + return adev->ip_blocks[i].status.valid; } - return true; + return false; } @@ -2271,6 +2354,8 @@ int amdgpu_device_ip_block_add(struct amdgpu_device *adev, DRM_INFO("add ip block number %d <%s>\n", adev->num_ip_blocks, ip_block_version->funcs->name); + adev->ip_blocks[adev->num_ip_blocks].adev = adev; + adev->ip_blocks[adev->num_ip_blocks++].version = ip_block_version; return 0; @@ -2566,25 +2651,25 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) total = true; for (i = 0; i < adev->num_ip_blocks; i++) { + ip_block = &adev->ip_blocks[i]; + if ((amdgpu_ip_block_mask & (1 << i)) == 0) { DRM_WARN("disabled ip block: %d <%s>\n", i, adev->ip_blocks[i].version->funcs->name); adev->ip_blocks[i].status.valid = false; - } else { - if (adev->ip_blocks[i].version->funcs->early_init) { - r = adev->ip_blocks[i].version->funcs->early_init((void *)adev); - if (r == -ENOENT) { - adev->ip_blocks[i].status.valid = false; - } else if (r) { - DRM_ERROR("early_init of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); - total = false; - } else { - adev->ip_blocks[i].status.valid = true; - } + } else if (ip_block->version->funcs->early_init) { + r = ip_block->version->funcs->early_init(ip_block); + if (r == -ENOENT) { + adev->ip_blocks[i].status.valid = false; + } else if (r) { + DRM_ERROR("early_init of IP block <%s> failed %d\n", + adev->ip_blocks[i].version->funcs->name, r); + total = false; } else { adev->ip_blocks[i].status.valid = true; } + } else { + adev->ip_blocks[i].status.valid = true; } /* get the vbios after the asic_funcs are set up */ if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_COMMON) { @@ -2633,10 +2718,13 @@ static int amdgpu_device_ip_hw_init_phase1(struct amdgpu_device *adev) continue; if (adev->ip_blocks[i].status.hw) continue; + if (!amdgpu_ip_member_of_hwini( + adev, adev->ip_blocks[i].version->type)) + continue; if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_COMMON || (amdgpu_sriov_vf(adev) && (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_PSP)) || adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH) { - r = adev->ip_blocks[i].version->funcs->hw_init(adev); + r = adev->ip_blocks[i].version->funcs->hw_init(&adev->ip_blocks[i]); if (r) { DRM_ERROR("hw_init of IP block <%s> failed %d\n", adev->ip_blocks[i].version->funcs->name, r); @@ -2658,7 +2746,10 @@ static int amdgpu_device_ip_hw_init_phase2(struct amdgpu_device *adev) continue; if (adev->ip_blocks[i].status.hw) continue; - r = adev->ip_blocks[i].version->funcs->hw_init(adev); + if (!amdgpu_ip_member_of_hwini( + adev, adev->ip_blocks[i].version->type)) + continue; + r = adev->ip_blocks[i].version->funcs->hw_init(&adev->ip_blocks[i]); if (r) { DRM_ERROR("hw_init of IP block <%s> failed %d\n", adev->ip_blocks[i].version->funcs->name, r); @@ -2681,6 +2772,10 @@ static int amdgpu_device_fw_loading(struct amdgpu_device *adev) if (adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_PSP) continue; + if (!amdgpu_ip_member_of_hwini(adev, + AMD_IP_BLOCK_TYPE_PSP)) + break; + if (!adev->ip_blocks[i].status.sw) continue; @@ -2689,22 +2784,18 @@ static int amdgpu_device_fw_loading(struct amdgpu_device *adev) break; if (amdgpu_in_reset(adev) || adev->in_suspend) { - r = adev->ip_blocks[i].version->funcs->resume(adev); - if (r) { - DRM_ERROR("resume of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); + r = amdgpu_ip_block_resume(&adev->ip_blocks[i]); + if (r) return r; - } } else { - r = adev->ip_blocks[i].version->funcs->hw_init(adev); + r = adev->ip_blocks[i].version->funcs->hw_init(&adev->ip_blocks[i]); if (r) { DRM_ERROR("hw_init of IP block <%s> failed %d\n", adev->ip_blocks[i].version->funcs->name, r); return r; } + adev->ip_blocks[i].status.hw = true; } - - adev->ip_blocks[i].status.hw = true; break; } } @@ -2786,6 +2877,7 @@ static int amdgpu_device_init_schedulers(struct amdgpu_device *adev) */ static int amdgpu_device_ip_init(struct amdgpu_device *adev) { + bool init_badpage; int i, r; r = amdgpu_ras_init(adev); @@ -2795,17 +2887,23 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev) for (i = 0; i < adev->num_ip_blocks; i++) { if (!adev->ip_blocks[i].status.valid) continue; - r = adev->ip_blocks[i].version->funcs->sw_init((void *)adev); - if (r) { - DRM_ERROR("sw_init of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); - goto init_failed; + if (adev->ip_blocks[i].version->funcs->sw_init) { + r = adev->ip_blocks[i].version->funcs->sw_init(&adev->ip_blocks[i]); + if (r) { + DRM_ERROR("sw_init of IP block <%s> failed %d\n", + adev->ip_blocks[i].version->funcs->name, r); + goto init_failed; + } } adev->ip_blocks[i].status.sw = true; + if (!amdgpu_ip_member_of_hwini( + adev, adev->ip_blocks[i].version->type)) + continue; + if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_COMMON) { /* need to do common hw init early so everything is set up for gmc */ - r = adev->ip_blocks[i].version->funcs->hw_init((void *)adev); + r = adev->ip_blocks[i].version->funcs->hw_init(&adev->ip_blocks[i]); if (r) { DRM_ERROR("hw_init %d failed %d\n", i, r); goto init_failed; @@ -2822,7 +2920,7 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev) DRM_ERROR("amdgpu_mem_scratch_init failed %d\n", r); goto init_failed; } - r = adev->ip_blocks[i].version->funcs->hw_init((void *)adev); + r = adev->ip_blocks[i].version->funcs->hw_init(&adev->ip_blocks[i]); if (r) { DRM_ERROR("hw_init %d failed %d\n", i, r); goto init_failed; @@ -2895,7 +2993,8 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev) * Note: theoretically, this should be called before all vram allocations * to protect retired page from abusing */ - r = amdgpu_ras_recovery_init(adev); + init_badpage = (adev->init_lvl->level != AMDGPU_INIT_LEVEL_MINIMAL_XGMI); + r = amdgpu_ras_recovery_init(adev, init_badpage); if (r) goto init_failed; @@ -2935,7 +3034,7 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev) amdgpu_ttm_set_buffer_funcs_status(adev, true); /* Don't init kfd if whole hive need to be reset during init */ - if (!adev->gmc.xgmi.pending_reset) { + if (adev->init_lvl->level != AMDGPU_INIT_LEVEL_MINIMAL_XGMI) { kgd2kfd_init_zone_device(adev); amdgpu_amdkfd_device_init(adev); } @@ -3135,7 +3234,7 @@ static int amdgpu_device_ip_late_init(struct amdgpu_device *adev) if (!adev->ip_blocks[i].status.hw) continue; if (adev->ip_blocks[i].version->funcs->late_init) { - r = adev->ip_blocks[i].version->funcs->late_init((void *)adev); + r = adev->ip_blocks[i].version->funcs->late_init(&adev->ip_blocks[i]); if (r) { DRM_ERROR("late_init of IP block <%s> failed %d\n", adev->ip_blocks[i].version->funcs->name, r); @@ -3206,6 +3305,25 @@ static int amdgpu_device_ip_late_init(struct amdgpu_device *adev) return 0; } +static void amdgpu_ip_block_hw_fini(struct amdgpu_ip_block *ip_block) +{ + int r; + + if (!ip_block->version->funcs->hw_fini) { + DRM_ERROR("hw_fini of IP block <%s> not defined\n", + ip_block->version->funcs->name); + } else { + r = ip_block->version->funcs->hw_fini(ip_block); + /* XXX handle errors */ + if (r) { + DRM_DEBUG("hw_fini of IP block <%s> failed %d\n", + ip_block->version->funcs->name, r); + } + } + + ip_block->status.hw = false; +} + /** * amdgpu_device_smu_fini_early - smu hw_fini wrapper * @@ -3215,7 +3333,7 @@ static int amdgpu_device_ip_late_init(struct amdgpu_device *adev) */ static void amdgpu_device_smu_fini_early(struct amdgpu_device *adev) { - int i, r; + int i; if (amdgpu_ip_version(adev, GC_HWIP, 0) > IP_VERSION(9, 0, 0)) return; @@ -3224,13 +3342,7 @@ static void amdgpu_device_smu_fini_early(struct amdgpu_device *adev) if (!adev->ip_blocks[i].status.hw) continue; if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_SMC) { - r = adev->ip_blocks[i].version->funcs->hw_fini((void *)adev); - /* XXX handle errors */ - if (r) { - DRM_DEBUG("hw_fini of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); - } - adev->ip_blocks[i].status.hw = false; + amdgpu_ip_block_hw_fini(&adev->ip_blocks[i]); break; } } @@ -3244,7 +3356,7 @@ static int amdgpu_device_ip_fini_early(struct amdgpu_device *adev) if (!adev->ip_blocks[i].version->funcs->early_fini) continue; - r = adev->ip_blocks[i].version->funcs->early_fini((void *)adev); + r = adev->ip_blocks[i].version->funcs->early_fini(&adev->ip_blocks[i]); if (r) { DRM_DEBUG("early_fini of IP block <%s> failed %d\n", adev->ip_blocks[i].version->funcs->name, r); @@ -3263,14 +3375,7 @@ static int amdgpu_device_ip_fini_early(struct amdgpu_device *adev) if (!adev->ip_blocks[i].status.hw) continue; - r = adev->ip_blocks[i].version->funcs->hw_fini((void *)adev); - /* XXX handle errors */ - if (r) { - DRM_DEBUG("hw_fini of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); - } - - adev->ip_blocks[i].status.hw = false; + amdgpu_ip_block_hw_fini(&adev->ip_blocks[i]); } if (amdgpu_sriov_vf(adev)) { @@ -3316,12 +3421,13 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev) amdgpu_ib_pool_fini(adev); amdgpu_seq64_fini(adev); } - - r = adev->ip_blocks[i].version->funcs->sw_fini((void *)adev); - /* XXX handle errors */ - if (r) { - DRM_DEBUG("sw_fini of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); + if (adev->ip_blocks[i].version->funcs->sw_fini) { + r = adev->ip_blocks[i].version->funcs->sw_fini(&adev->ip_blocks[i]); + /* XXX handle errors */ + if (r) { + DRM_DEBUG("sw_fini of IP block <%s> failed %d\n", + adev->ip_blocks[i].version->funcs->name, r); + } } adev->ip_blocks[i].status.sw = false; adev->ip_blocks[i].status.valid = false; @@ -3331,7 +3437,7 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev) if (!adev->ip_blocks[i].status.late_initialized) continue; if (adev->ip_blocks[i].version->funcs->late_fini) - adev->ip_blocks[i].version->funcs->late_fini((void *)adev); + adev->ip_blocks[i].version->funcs->late_fini(&adev->ip_blocks[i]); adev->ip_blocks[i].status.late_initialized = false; } @@ -3403,15 +3509,9 @@ static int amdgpu_device_ip_suspend_phase1(struct amdgpu_device *adev) continue; /* XXX handle errors */ - r = adev->ip_blocks[i].version->funcs->suspend(adev); - /* XXX handle errors */ - if (r) { - DRM_ERROR("suspend of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); + r = amdgpu_ip_block_suspend(&adev->ip_blocks[i]); + if (r) return r; - } - - adev->ip_blocks[i].status.hw = false; } return 0; @@ -3449,14 +3549,9 @@ static int amdgpu_device_ip_suspend_phase2(struct amdgpu_device *adev) } /* skip unnecessary suspend if we do not initialize them yet */ - if (adev->gmc.xgmi.pending_reset && - !(adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC || - adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_SMC || - adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_COMMON || - adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH)) { - adev->ip_blocks[i].status.hw = false; + if (!amdgpu_ip_member_of_hwini( + adev, adev->ip_blocks[i].version->type)) continue; - } /* skip suspend of gfx/mes and psp for S0ix * gfx is in gfxoff state, so on resume it will exit gfxoff just @@ -3490,13 +3585,9 @@ static int amdgpu_device_ip_suspend_phase2(struct amdgpu_device *adev) continue; /* XXX handle errors */ - r = adev->ip_blocks[i].version->funcs->suspend(adev); - /* XXX handle errors */ - if (r) { - DRM_ERROR("suspend of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); - } + r = amdgpu_ip_block_suspend(&adev->ip_blocks[i]); adev->ip_blocks[i].status.hw = false; + /* handle putting the SMC in the appropriate state */ if (!amdgpu_sriov_vf(adev)) { if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_SMC) { @@ -3570,7 +3661,7 @@ static int amdgpu_device_ip_reinit_early_sriov(struct amdgpu_device *adev) !block->status.valid) continue; - r = block->version->funcs->hw_init(adev); + r = block->version->funcs->hw_init(&adev->ip_blocks[i]); DRM_INFO("RE-INIT-early: %s %s\n", block->version->funcs->name, r?"failed":"succeeded"); if (r) return r; @@ -3609,15 +3700,19 @@ static int amdgpu_device_ip_reinit_late_sriov(struct amdgpu_device *adev) block->status.hw) continue; - if (block->version->type == AMD_IP_BLOCK_TYPE_SMC) - r = block->version->funcs->resume(adev); - else - r = block->version->funcs->hw_init(adev); - - DRM_INFO("RE-INIT-late: %s %s\n", block->version->funcs->name, r?"failed":"succeeded"); - if (r) - return r; - block->status.hw = true; + if (block->version->type == AMD_IP_BLOCK_TYPE_SMC) { + r = amdgpu_ip_block_resume(&adev->ip_blocks[i]); + if (r) + return r; + } else { + r = block->version->funcs->hw_init(&adev->ip_blocks[i]); + if (r) { + DRM_ERROR("hw_init of IP block <%s> failed %d\n", + adev->ip_blocks[i].version->funcs->name, r); + return r; + } + block->status.hw = true; + } } } @@ -3648,13 +3743,9 @@ static int amdgpu_device_ip_resume_phase1(struct amdgpu_device *adev) adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH || (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_PSP && amdgpu_sriov_vf(adev))) { - r = adev->ip_blocks[i].version->funcs->resume(adev); - if (r) { - DRM_ERROR("resume of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); + r = amdgpu_ip_block_resume(&adev->ip_blocks[i]); + if (r) return r; - } - adev->ip_blocks[i].status.hw = true; } } @@ -3686,13 +3777,9 @@ static int amdgpu_device_ip_resume_phase2(struct amdgpu_device *adev) adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH || adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_PSP) continue; - r = adev->ip_blocks[i].version->funcs->resume(adev); - if (r) { - DRM_ERROR("resume of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); + r = amdgpu_ip_block_resume(&adev->ip_blocks[i]); + if (r) return r; - } - adev->ip_blocks[i].status.hw = true; } return 0; @@ -4149,7 +4236,10 @@ int amdgpu_device_init(struct amdgpu_device *adev, * for throttling interrupt) = 60 seconds. */ ratelimit_state_init(&adev->throttling_logging_rs, (60 - 1) * HZ, 1); + ratelimit_state_init(&adev->virt.ras_telemetry_rs, 5 * HZ, 1); + ratelimit_set_flags(&adev->throttling_logging_rs, RATELIMIT_MSG_ON_RELEASE); + ratelimit_set_flags(&adev->virt.ras_telemetry_rs, RATELIMIT_MSG_ON_RELEASE); /* Registers mapping */ /* TODO: block userspace mapping of io register */ @@ -4193,13 +4283,19 @@ int amdgpu_device_init(struct amdgpu_device *adev, amdgpu_device_set_mcbp(adev); + /* + * By default, use default mode where all blocks are expected to be + * initialized. At present a 'swinit' of blocks is required to be + * completed before the need for a different level is detected. + */ + amdgpu_set_init_level(adev, AMDGPU_INIT_LEVEL_DEFAULT); /* early init functions */ r = amdgpu_device_ip_early_init(adev); if (r) return r; /* Get rid of things like offb */ - r = drm_aperture_remove_conflicting_pci_framebuffers(adev->pdev, &amdgpu_kms_driver); + r = aperture_remove_conflicting_pci_devices(adev->pdev, amdgpu_kms_driver.name); if (r) return r; @@ -4265,20 +4361,8 @@ int amdgpu_device_init(struct amdgpu_device *adev, if (!amdgpu_sriov_vf(adev) && amdgpu_asic_need_reset_on_init(adev)) { if (adev->gmc.xgmi.num_physical_nodes) { dev_info(adev->dev, "Pending hive reset.\n"); - adev->gmc.xgmi.pending_reset = true; - /* Only need to init necessary block for SMU to handle the reset */ - for (i = 0; i < adev->num_ip_blocks; i++) { - if (!adev->ip_blocks[i].status.valid) - continue; - if (!(adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC || - adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_COMMON || - adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH || - adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_SMC)) { - DRM_DEBUG("IP %s disabled for hw_init.\n", - adev->ip_blocks[i].version->funcs->name); - adev->ip_blocks[i].status.hw = true; - } - } + amdgpu_set_init_level(adev, + AMDGPU_INIT_LEVEL_MINIMAL_XGMI); } else if (amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(13, 0, 10) && !amdgpu_device_has_display_hardware(adev)) { r = psp_gpu_reset(adev); @@ -4386,7 +4470,7 @@ fence_driver_init: /* enable clockgating, etc. after ib tests, etc. since some blocks require * explicit gating rather than handling it automatically. */ - if (!adev->gmc.xgmi.pending_reset) { + if (adev->init_lvl->level != AMDGPU_INIT_LEVEL_MINIMAL_XGMI) { r = amdgpu_device_ip_late_init(adev); if (r) { dev_err(adev->dev, "amdgpu_device_ip_late_init failed\n"); @@ -4436,6 +4520,7 @@ fence_driver_init: amdgpu_fru_sysfs_init(adev); amdgpu_reg_state_sysfs_init(adev); + amdgpu_xcp_cfg_sysfs_init(adev); if (IS_ENABLED(CONFIG_PERF_EVENTS)) r = amdgpu_pmu_init(adev); @@ -4463,9 +4548,8 @@ fence_driver_init: if (px) vga_switcheroo_init_domain_pm_ops(adev->dev, &adev->vga_pm_domain); - if (adev->gmc.xgmi.pending_reset) - queue_delayed_work(system_wq, &mgpu_info.delayed_reset_work, - msecs_to_jiffies(AMDGPU_RESUME_MS)); + if (adev->init_lvl->level == AMDGPU_INIT_LEVEL_MINIMAL_XGMI) + amdgpu_xgmi_reset_on_init(adev); amdgpu_device_check_iommu_direct_map(adev); @@ -4559,6 +4643,7 @@ void amdgpu_device_fini_hw(struct amdgpu_device *adev) amdgpu_fru_sysfs_fini(adev); amdgpu_reg_state_sysfs_fini(adev); + amdgpu_xcp_cfg_sysfs_fini(adev); /* disable ras feature must before hw fini */ amdgpu_ras_pre_fini(adev); @@ -4694,7 +4779,7 @@ int amdgpu_device_prepare(struct drm_device *dev) continue; if (!adev->ip_blocks[i].version->funcs->prepare_suspend) continue; - r = adev->ip_blocks[i].version->funcs->prepare_suspend((void *)adev); + r = adev->ip_blocks[i].version->funcs->prepare_suspend(&adev->ip_blocks[i]); if (r) goto unprepare; } @@ -4711,13 +4796,13 @@ unprepare: * amdgpu_device_suspend - initiate device suspend * * @dev: drm dev pointer - * @fbcon : notify the fbdev of suspend + * @notify_clients: notify in-kernel DRM clients * * Puts the hw in the suspend state (all asics). * Returns 0 for success or an error on failure. * Called at driver suspend. */ -int amdgpu_device_suspend(struct drm_device *dev, bool fbcon) +int amdgpu_device_suspend(struct drm_device *dev, bool notify_clients) { struct amdgpu_device *adev = drm_to_adev(dev); int r = 0; @@ -4737,8 +4822,8 @@ int amdgpu_device_suspend(struct drm_device *dev, bool fbcon) if (amdgpu_acpi_smart_shift_update(dev, AMDGPU_SS_DEV_D3)) DRM_WARN("smart shift update failed\n"); - if (fbcon) - drm_fb_helper_set_suspend_unlocked(adev_to_drm(adev)->fb_helper, true); + if (notify_clients) + drm_client_dev_suspend(adev_to_drm(adev), false); cancel_delayed_work_sync(&adev->delayed_init_work); @@ -4773,13 +4858,13 @@ int amdgpu_device_suspend(struct drm_device *dev, bool fbcon) * amdgpu_device_resume - initiate device resume * * @dev: drm dev pointer - * @fbcon : notify the fbdev of resume + * @notify_clients: notify in-kernel DRM clients * * Bring the hw back to operating state (all asics). * Returns 0 for success or an error on failure. * Called at driver resume. */ -int amdgpu_device_resume(struct drm_device *dev, bool fbcon) +int amdgpu_device_resume(struct drm_device *dev, bool notify_clients) { struct amdgpu_device *adev = drm_to_adev(dev); int r = 0; @@ -4835,8 +4920,8 @@ exit: /* Make sure IB tests flushed */ flush_delayed_work(&adev->delayed_init_work); - if (fbcon) - drm_fb_helper_set_suspend_unlocked(adev_to_drm(adev)->fb_helper, false); + if (notify_clients) + drm_client_dev_resume(adev_to_drm(adev), false); amdgpu_ras_resume(adev); @@ -4898,7 +4983,8 @@ static bool amdgpu_device_ip_check_soft_reset(struct amdgpu_device *adev) continue; if (adev->ip_blocks[i].version->funcs->check_soft_reset) adev->ip_blocks[i].status.hang = - adev->ip_blocks[i].version->funcs->check_soft_reset(adev); + adev->ip_blocks[i].version->funcs->check_soft_reset( + &adev->ip_blocks[i]); if (adev->ip_blocks[i].status.hang) { dev_info(adev->dev, "IP block:%s is hung!\n", adev->ip_blocks[i].version->funcs->name); asic_hang = true; @@ -4927,7 +5013,7 @@ static int amdgpu_device_ip_pre_soft_reset(struct amdgpu_device *adev) continue; if (adev->ip_blocks[i].status.hang && adev->ip_blocks[i].version->funcs->pre_soft_reset) { - r = adev->ip_blocks[i].version->funcs->pre_soft_reset(adev); + r = adev->ip_blocks[i].version->funcs->pre_soft_reset(&adev->ip_blocks[i]); if (r) return r; } @@ -4989,7 +5075,7 @@ static int amdgpu_device_ip_soft_reset(struct amdgpu_device *adev) continue; if (adev->ip_blocks[i].status.hang && adev->ip_blocks[i].version->funcs->soft_reset) { - r = adev->ip_blocks[i].version->funcs->soft_reset(adev); + r = adev->ip_blocks[i].version->funcs->soft_reset(&adev->ip_blocks[i]); if (r) return r; } @@ -5018,7 +5104,7 @@ static int amdgpu_device_ip_post_soft_reset(struct amdgpu_device *adev) continue; if (adev->ip_blocks[i].status.hang && adev->ip_blocks[i].version->funcs->post_soft_reset) - r = adev->ip_blocks[i].version->funcs->post_soft_reset(adev); + r = adev->ip_blocks[i].version->funcs->post_soft_reset(&adev->ip_blocks[i]); if (r) return r; } @@ -5103,6 +5189,9 @@ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev, amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(9, 4, 4) || amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(11, 0, 3)) amdgpu_ras_resume(adev); + + amdgpu_virt_ras_telemetry_post_reset(adev); + return 0; } @@ -5309,7 +5398,7 @@ int amdgpu_device_pre_asic_reset(struct amdgpu_device *adev, for (i = 0; i < tmp_adev->num_ip_blocks; i++) if (tmp_adev->ip_blocks[i].version->funcs->dump_ip_state) tmp_adev->ip_blocks[i].version->funcs - ->dump_ip_state((void *)tmp_adev); + ->dump_ip_state((void *)&tmp_adev->ip_blocks[i]); dev_info(tmp_adev->dev, "Dumping IP State Completed\n"); } @@ -5325,74 +5414,25 @@ int amdgpu_device_pre_asic_reset(struct amdgpu_device *adev, return r; } -int amdgpu_do_asic_reset(struct list_head *device_list_handle, - struct amdgpu_reset_context *reset_context) +int amdgpu_device_reinit_after_reset(struct amdgpu_reset_context *reset_context) { - struct amdgpu_device *tmp_adev = NULL; - bool need_full_reset, skip_hw_reset, vram_lost = false; - int r = 0; + struct list_head *device_list_handle; + bool full_reset, vram_lost = false; + struct amdgpu_device *tmp_adev; + int r; - /* Try reset handler method first */ - tmp_adev = list_first_entry(device_list_handle, struct amdgpu_device, - reset_list); + device_list_handle = reset_context->reset_device_list; - reset_context->reset_device_list = device_list_handle; - r = amdgpu_reset_perform_reset(tmp_adev, reset_context); - /* If reset handler not implemented, continue; otherwise return */ - if (r == -EOPNOTSUPP) - r = 0; - else - return r; + if (!device_list_handle) + return -EINVAL; - /* Reset handler not implemented, use the default method */ - need_full_reset = - test_bit(AMDGPU_NEED_FULL_RESET, &reset_context->flags); - skip_hw_reset = test_bit(AMDGPU_SKIP_HW_RESET, &reset_context->flags); - - /* - * ASIC reset has to be done on all XGMI hive nodes ASAP - * to allow proper links negotiation in FW (within 1 sec) - */ - if (!skip_hw_reset && need_full_reset) { - list_for_each_entry(tmp_adev, device_list_handle, reset_list) { - /* For XGMI run all resets in parallel to speed up the process */ - if (tmp_adev->gmc.xgmi.num_physical_nodes > 1) { - tmp_adev->gmc.xgmi.pending_reset = false; - if (!queue_work(system_unbound_wq, &tmp_adev->xgmi_reset_work)) - r = -EALREADY; - } else - r = amdgpu_asic_reset(tmp_adev); - - if (r) { - dev_err(tmp_adev->dev, "ASIC reset failed with error, %d for drm dev, %s", - r, adev_to_drm(tmp_adev)->unique); - goto out; - } - } - - /* For XGMI wait for all resets to complete before proceed */ - if (!r) { - list_for_each_entry(tmp_adev, device_list_handle, reset_list) { - if (tmp_adev->gmc.xgmi.num_physical_nodes > 1) { - flush_work(&tmp_adev->xgmi_reset_work); - r = tmp_adev->asic_reset_res; - if (r) - break; - } - } - } - } - - if (!r && amdgpu_ras_intr_triggered()) { - list_for_each_entry(tmp_adev, device_list_handle, reset_list) { - amdgpu_ras_reset_error_count(tmp_adev, AMDGPU_RAS_BLOCK__MMHUB); - } - - amdgpu_ras_intr_cleared(); - } + full_reset = test_bit(AMDGPU_NEED_FULL_RESET, &reset_context->flags); + r = 0; list_for_each_entry(tmp_adev, device_list_handle, reset_list) { - if (need_full_reset) { + /* After reset, it's default init level */ + amdgpu_set_init_level(tmp_adev, AMDGPU_INIT_LEVEL_DEFAULT); + if (full_reset) { /* post card */ amdgpu_ras_set_fed(tmp_adev, false); r = amdgpu_device_asic_init(tmp_adev); @@ -5448,7 +5488,7 @@ int amdgpu_do_asic_reset(struct list_head *device_list_handle, if (r) goto out; - drm_fb_helper_set_suspend_unlocked(adev_to_drm(tmp_adev)->fb_helper, false); + drm_client_dev_resume(adev_to_drm(tmp_adev), false); /* * The GPU enters bad state once faulty pages @@ -5482,7 +5522,6 @@ out: r = amdgpu_ib_ring_tests(tmp_adev); if (r) { dev_err(tmp_adev->dev, "ib ring test failed (%d).\n", r); - need_full_reset = true; r = -EAGAIN; goto end; } @@ -5493,10 +5532,85 @@ out: } end: - if (need_full_reset) + return r; +} + +int amdgpu_do_asic_reset(struct list_head *device_list_handle, + struct amdgpu_reset_context *reset_context) +{ + struct amdgpu_device *tmp_adev = NULL; + bool need_full_reset, skip_hw_reset; + int r = 0; + + /* Try reset handler method first */ + tmp_adev = list_first_entry(device_list_handle, struct amdgpu_device, + reset_list); + + reset_context->reset_device_list = device_list_handle; + r = amdgpu_reset_perform_reset(tmp_adev, reset_context); + /* If reset handler not implemented, continue; otherwise return */ + if (r == -EOPNOTSUPP) + r = 0; + else + return r; + + /* Reset handler not implemented, use the default method */ + need_full_reset = + test_bit(AMDGPU_NEED_FULL_RESET, &reset_context->flags); + skip_hw_reset = test_bit(AMDGPU_SKIP_HW_RESET, &reset_context->flags); + + /* + * ASIC reset has to be done on all XGMI hive nodes ASAP + * to allow proper links negotiation in FW (within 1 sec) + */ + if (!skip_hw_reset && need_full_reset) { + list_for_each_entry(tmp_adev, device_list_handle, reset_list) { + /* For XGMI run all resets in parallel to speed up the process */ + if (tmp_adev->gmc.xgmi.num_physical_nodes > 1) { + if (!queue_work(system_unbound_wq, + &tmp_adev->xgmi_reset_work)) + r = -EALREADY; + } else + r = amdgpu_asic_reset(tmp_adev); + + if (r) { + dev_err(tmp_adev->dev, + "ASIC reset failed with error, %d for drm dev, %s", + r, adev_to_drm(tmp_adev)->unique); + goto out; + } + } + + /* For XGMI wait for all resets to complete before proceed */ + if (!r) { + list_for_each_entry(tmp_adev, device_list_handle, + reset_list) { + if (tmp_adev->gmc.xgmi.num_physical_nodes > 1) { + flush_work(&tmp_adev->xgmi_reset_work); + r = tmp_adev->asic_reset_res; + if (r) + break; + } + } + } + } + + if (!r && amdgpu_ras_intr_triggered()) { + list_for_each_entry(tmp_adev, device_list_handle, reset_list) { + amdgpu_ras_reset_error_count(tmp_adev, + AMDGPU_RAS_BLOCK__MMHUB); + } + + amdgpu_ras_intr_cleared(); + } + + r = amdgpu_device_reinit_after_reset(reset_context); + if (r == -EAGAIN) set_bit(AMDGPU_NEED_FULL_RESET, &reset_context->flags); else clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context->flags); + +out: return r; } @@ -5734,7 +5848,7 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, */ amdgpu_unregister_gpu_instance(tmp_adev); - drm_fb_helper_set_suspend_unlocked(adev_to_drm(tmp_adev)->fb_helper, true); + drm_client_dev_suspend(adev_to_drm(tmp_adev), false); /* disable ras on ALL IPs */ if (!need_emergency_restart && @@ -5824,7 +5938,7 @@ skip_hw_reset: if (!amdgpu_ring_sched_ready(ring)) continue; - drm_sched_start(&ring->sched); + drm_sched_start(&ring->sched, 0); } if (!drm_drv_uses_atomic_modeset(adev_to_drm(tmp_adev)) && !job_signaled) @@ -6092,6 +6206,9 @@ bool amdgpu_device_is_peer_accessible(struct amdgpu_device *adev, bool p2p_access = !adev->gmc.xgmi.connected_to_cpu && !(pci_p2pdma_distance(adev->pdev, peer_adev->dev, false) < 0); + if (!p2p_access) + dev_info(adev->dev, "PCIe P2P access from peer device %s is not supported by the chipset\n", + pci_name(peer_adev->pdev)); bool is_large_bar = adev->gmc.visible_vram_size && adev->gmc.real_vram_size == adev->gmc.visible_vram_size; @@ -6331,7 +6448,7 @@ void amdgpu_pci_resume(struct pci_dev *pdev) if (!amdgpu_ring_sched_ready(ring)) continue; - drm_sched_start(&ring->sched); + drm_sched_start(&ring->sched, 0); } amdgpu_device_unset_mp1_state(adev); @@ -6344,6 +6461,9 @@ bool amdgpu_device_cache_pci_state(struct pci_dev *pdev) struct amdgpu_device *adev = drm_to_adev(dev); int r; + if (amdgpu_sriov_vf(adev)) + return false; + r = pci_save_state(pdev); if (!r) { kfree(adev->pci_state); @@ -6604,3 +6724,47 @@ uint32_t amdgpu_device_wait_on_rreg(struct amdgpu_device *adev, } return ret; } + +ssize_t amdgpu_get_soft_full_reset_mask(struct amdgpu_ring *ring) +{ + ssize_t size = 0; + + if (!ring || !ring->adev) + return size; + + if (amdgpu_device_should_recover_gpu(ring->adev)) + size |= AMDGPU_RESET_TYPE_FULL; + + if (unlikely(!ring->adev->debug_disable_soft_recovery) && + !amdgpu_sriov_vf(ring->adev) && ring->funcs->soft_recovery) + size |= AMDGPU_RESET_TYPE_SOFT_RESET; + + return size; +} + +ssize_t amdgpu_show_reset_mask(char *buf, uint32_t supported_reset) +{ + ssize_t size = 0; + + if (supported_reset == 0) { + size += sysfs_emit_at(buf, size, "unsupported"); + size += sysfs_emit_at(buf, size, "\n"); + return size; + + } + + if (supported_reset & AMDGPU_RESET_TYPE_SOFT_RESET) + size += sysfs_emit_at(buf, size, "soft "); + + if (supported_reset & AMDGPU_RESET_TYPE_PER_QUEUE) + size += sysfs_emit_at(buf, size, "queue "); + + if (supported_reset & AMDGPU_RESET_TYPE_PER_PIPE) + size += sysfs_emit_at(buf, size, "pipe "); + + if (supported_reset & AMDGPU_RESET_TYPE_FULL) + size += sysfs_emit_at(buf, size, "full "); + + size += sysfs_emit_at(buf, size, "\n"); + return size; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c index 4bd61c169ca8..1040204ac8b9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c @@ -1723,45 +1723,85 @@ union nps_info { struct nps_info_v1_0 v1; }; +static int amdgpu_discovery_refresh_nps_info(struct amdgpu_device *adev, + union nps_info *nps_data) +{ + uint64_t vram_size, pos, offset; + struct nps_info_header *nhdr; + struct binary_header bhdr; + uint16_t checksum; + + vram_size = (uint64_t)RREG32(mmRCC_CONFIG_MEMSIZE) << 20; + pos = vram_size - DISCOVERY_TMR_OFFSET; + amdgpu_device_vram_access(adev, pos, &bhdr, sizeof(bhdr), false); + + offset = le16_to_cpu(bhdr.table_list[NPS_INFO].offset); + checksum = le16_to_cpu(bhdr.table_list[NPS_INFO].checksum); + + amdgpu_device_vram_access(adev, (pos + offset), nps_data, + sizeof(*nps_data), false); + + nhdr = (struct nps_info_header *)(nps_data); + if (!amdgpu_discovery_verify_checksum((uint8_t *)nps_data, + le32_to_cpu(nhdr->size_bytes), + checksum)) { + dev_err(adev->dev, "nps data refresh, checksum mismatch\n"); + return -EINVAL; + } + + return 0; +} + int amdgpu_discovery_get_nps_info(struct amdgpu_device *adev, uint32_t *nps_type, struct amdgpu_gmc_memrange **ranges, - int *range_cnt) + int *range_cnt, bool refresh) { struct amdgpu_gmc_memrange *mem_ranges; struct binary_header *bhdr; union nps_info *nps_info; + union nps_info nps_data; u16 offset; - int i; + int i, r; if (!nps_type || !range_cnt || !ranges) return -EINVAL; - if (!adev->mman.discovery_bin) { - dev_err(adev->dev, - "fetch mem range failed, ip discovery uninitialized\n"); - return -EINVAL; + if (refresh) { + r = amdgpu_discovery_refresh_nps_info(adev, &nps_data); + if (r) + return r; + nps_info = &nps_data; + } else { + if (!adev->mman.discovery_bin) { + dev_err(adev->dev, + "fetch mem range failed, ip discovery uninitialized\n"); + return -EINVAL; + } + + bhdr = (struct binary_header *)adev->mman.discovery_bin; + offset = le16_to_cpu(bhdr->table_list[NPS_INFO].offset); + + if (!offset) + return -ENOENT; + + /* If verification fails, return as if NPS table doesn't exist */ + if (amdgpu_discovery_verify_npsinfo(adev, bhdr)) + return -ENOENT; + + nps_info = + (union nps_info *)(adev->mman.discovery_bin + offset); } - bhdr = (struct binary_header *)adev->mman.discovery_bin; - offset = le16_to_cpu(bhdr->table_list[NPS_INFO].offset); - - if (!offset) - return -ENOENT; - - /* If verification fails, return as if NPS table doesn't exist */ - if (amdgpu_discovery_verify_npsinfo(adev, bhdr)) - return -ENOENT; - - nps_info = (union nps_info *)(adev->mman.discovery_bin + offset); - switch (le16_to_cpu(nps_info->v1.header.version_major)) { case 1: + mem_ranges = kvcalloc(nps_info->v1.count, + sizeof(*mem_ranges), + GFP_KERNEL); + if (!mem_ranges) + return -ENOMEM; *nps_type = nps_info->v1.nps_type; *range_cnt = nps_info->v1.count; - mem_ranges = kvzalloc( - *range_cnt * sizeof(struct amdgpu_gmc_memrange), - GFP_KERNEL); for (i = 0; i < *range_cnt; i++) { mem_ranges[i].base_address = nps_info->v1.instance_info[i].base_address; @@ -2492,6 +2532,7 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev) adev->ip_versions[GC_HWIP][0] = IP_VERSION(9, 2, 2); adev->ip_versions[UVD_HWIP][0] = IP_VERSION(1, 0, 1); adev->ip_versions[DCE_HWIP][0] = IP_VERSION(1, 0, 1); + adev->ip_versions[ISP_HWIP][0] = IP_VERSION(2, 0, 0); } else { adev->ip_versions[MMHUB_HWIP][0] = IP_VERSION(9, 1, 0); adev->ip_versions[ATHUB_HWIP][0] = IP_VERSION(9, 1, 0); @@ -2508,6 +2549,7 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev) adev->ip_versions[GC_HWIP][0] = IP_VERSION(9, 1, 0); adev->ip_versions[UVD_HWIP][0] = IP_VERSION(1, 0, 0); adev->ip_versions[DCE_HWIP][0] = IP_VERSION(1, 0, 0); + adev->ip_versions[ISP_HWIP][0] = IP_VERSION(2, 0, 0); } break; case CHIP_VEGA20: diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.h index f5d36525ec3e..b44d56465c5b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.h @@ -33,6 +33,6 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev); int amdgpu_discovery_get_nps_info(struct amdgpu_device *adev, uint32_t *nps_type, struct amdgpu_gmc_memrange **ranges, - int *range_cnt); + int *range_cnt, bool refresh); #endif /* __AMDGPU_DISCOVERY__ */ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 81d9877c8735..38686203bea6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -23,6 +23,7 @@ */ #include +#include #include #include #include @@ -231,8 +232,6 @@ int amdgpu_wbrf = -1; int amdgpu_damage_clips = -1; /* auto */ int amdgpu_umsch_mm_fwlog; -static void amdgpu_drv_delayed_reset_work_handler(struct work_struct *work); - DECLARE_DYNDBG_CLASSMAP(drm_debug_classes, DD_CLASS_TYPE_DISJOINT_BITS, 0, "DRM_UT_CORE", "DRM_UT_DRIVER", @@ -247,9 +246,6 @@ DECLARE_DYNDBG_CLASSMAP(drm_debug_classes, DD_CLASS_TYPE_DISJOINT_BITS, 0, struct amdgpu_mgpu_info mgpu_info = { .mutex = __MUTEX_INITIALIZER(mgpu_info.mutex), - .delayed_reset_work = __DELAYED_WORK_INITIALIZER( - mgpu_info.delayed_reset_work, - amdgpu_drv_delayed_reset_work_handler, 0), }; int amdgpu_ras_enable = -1; uint amdgpu_ras_mask = 0xffffffff; @@ -892,7 +888,7 @@ module_param_named(visualconfirm, amdgpu_dc_visual_confirm, uint, 0444); * the ABM algorithm, with 1 being the least reduction and 4 being the most * reduction. * - * Defaults to -1, or disabled. Userspace can only override this level after + * Defaults to -1, or auto. Userspace can only override this level after * boot if it's set to auto. */ int amdgpu_dm_abm_level = -1; @@ -2365,11 +2361,15 @@ retry_init: */ if (adev->mode_info.mode_config_initialized && !list_empty(&adev_to_drm(adev)->mode_config.connector_list)) { + const struct drm_format_info *format; + /* select 8 bpp console on low vram cards */ if (adev->gmc.real_vram_size <= (32*1024*1024)) - drm_fbdev_ttm_setup(adev_to_drm(adev), 8); + format = drm_format_info(DRM_FORMAT_C8); else - drm_fbdev_ttm_setup(adev_to_drm(adev), 32); + format = NULL; + + drm_client_setup(adev_to_drm(adev), format); } ret = amdgpu_debugfs_init(adev); @@ -2434,6 +2434,7 @@ amdgpu_pci_remove(struct pci_dev *pdev) struct amdgpu_device *adev = drm_to_adev(dev); amdgpu_xcp_dev_unplug(adev); + amdgpu_gmc_prepare_nps_mode_change(adev); drm_dev_unplug(dev); if (adev->pm.rpm_mode != AMDGPU_RUNPM_NONE) { @@ -2472,82 +2473,6 @@ amdgpu_pci_shutdown(struct pci_dev *pdev) adev->mp1_state = PP_MP1_STATE_NONE; } -/** - * amdgpu_drv_delayed_reset_work_handler - work handler for reset - * - * @work: work_struct. - */ -static void amdgpu_drv_delayed_reset_work_handler(struct work_struct *work) -{ - struct list_head device_list; - struct amdgpu_device *adev; - int i, r; - struct amdgpu_reset_context reset_context; - - memset(&reset_context, 0, sizeof(reset_context)); - - mutex_lock(&mgpu_info.mutex); - if (mgpu_info.pending_reset == true) { - mutex_unlock(&mgpu_info.mutex); - return; - } - mgpu_info.pending_reset = true; - mutex_unlock(&mgpu_info.mutex); - - /* Use a common context, just need to make sure full reset is done */ - reset_context.method = AMD_RESET_METHOD_NONE; - set_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags); - - for (i = 0; i < mgpu_info.num_dgpu; i++) { - adev = mgpu_info.gpu_ins[i].adev; - reset_context.reset_req_dev = adev; - r = amdgpu_device_pre_asic_reset(adev, &reset_context); - if (r) { - dev_err(adev->dev, "GPU pre asic reset failed with err, %d for drm dev, %s ", - r, adev_to_drm(adev)->unique); - } - if (!queue_work(system_unbound_wq, &adev->xgmi_reset_work)) - r = -EALREADY; - } - for (i = 0; i < mgpu_info.num_dgpu; i++) { - adev = mgpu_info.gpu_ins[i].adev; - flush_work(&adev->xgmi_reset_work); - adev->gmc.xgmi.pending_reset = false; - } - - /* reset function will rebuild the xgmi hive info , clear it now */ - for (i = 0; i < mgpu_info.num_dgpu; i++) - amdgpu_xgmi_remove_device(mgpu_info.gpu_ins[i].adev); - - INIT_LIST_HEAD(&device_list); - - for (i = 0; i < mgpu_info.num_dgpu; i++) - list_add_tail(&mgpu_info.gpu_ins[i].adev->reset_list, &device_list); - - /* unregister the GPU first, reset function will add them back */ - list_for_each_entry(adev, &device_list, reset_list) - amdgpu_unregister_gpu_instance(adev); - - /* Use a common context, just need to make sure full reset is done */ - set_bit(AMDGPU_SKIP_HW_RESET, &reset_context.flags); - set_bit(AMDGPU_SKIP_COREDUMP, &reset_context.flags); - r = amdgpu_do_asic_reset(&device_list, &reset_context); - - if (r) { - DRM_ERROR("reinit gpus failure"); - return; - } - for (i = 0; i < mgpu_info.num_dgpu; i++) { - adev = mgpu_info.gpu_ins[i].adev; - if (!adev->kfd.init_complete) { - kgd2kfd_init_zone_device(adev); - amdgpu_amdkfd_device_init(adev); - amdgpu_amdkfd_drm_client_create(adev); - } - amdgpu_ttm_set_buffer_funcs_status(adev, true); - } -} - static int amdgpu_pmops_prepare(struct device *dev) { struct drm_device *drm_dev = dev_get_drvdata(dev); @@ -2580,7 +2505,6 @@ static int amdgpu_pmops_suspend(struct device *dev) struct drm_device *drm_dev = dev_get_drvdata(dev); struct amdgpu_device *adev = drm_to_adev(drm_dev); - adev->suspend_complete = false; if (amdgpu_acpi_is_s0ix_active(adev)) adev->in_s0ix = true; else if (amdgpu_acpi_is_s3_active(adev)) @@ -2595,7 +2519,6 @@ static int amdgpu_pmops_suspend_noirq(struct device *dev) struct drm_device *drm_dev = dev_get_drvdata(dev); struct amdgpu_device *adev = drm_to_adev(drm_dev); - adev->suspend_complete = true; if (amdgpu_acpi_should_gpu_reset(adev)) return amdgpu_asic_reset(adev); @@ -2982,6 +2905,7 @@ static const struct drm_driver amdgpu_kms_driver = { .num_ioctls = ARRAY_SIZE(amdgpu_ioctls_kms), .dumb_create = amdgpu_mode_dumb_create, .dumb_map_offset = amdgpu_mode_dumb_mmap, + DRM_FBDEV_TTM_DRIVER_OPS, .fops = &amdgpu_driver_kms_fops, .release = &amdgpu_driver_release_kms, #ifdef CONFIG_PROC_FS @@ -3008,6 +2932,7 @@ const struct drm_driver amdgpu_partition_driver = { .num_ioctls = ARRAY_SIZE(amdgpu_ioctls_kms), .dumb_create = amdgpu_mode_dumb_create, .dumb_map_offset = amdgpu_mode_dumb_mmap, + DRM_FBDEV_TTM_DRIVER_OPS, .fops = &amdgpu_driver_kms_fops, .release = &amdgpu_driver_release_kms, @@ -3068,6 +2993,12 @@ static int __init amdgpu_init(void) /* Ignore KFD init failures. Normal when CONFIG_HSA_AMD is not set. */ amdgpu_amdkfd_init(); + if (amdgpu_pp_feature_mask & PP_OVERDRIVE_MASK) { + add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_STILL_OK); + pr_crit("Overdrive is enabled, please disable it before " + "reporting any bugs unrelated to overdrive.\n"); + } + /* let modprobe override vga console setting */ return pci_register_driver(&amdgpu_kms_pci_driver); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.c index 35fee3e8cde2..8cd69836dd99 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.c @@ -200,7 +200,7 @@ static int amdgpu_eeprom_xfer(struct i2c_adapter *i2c_adap, u32 eeprom_addr, dev_err_ratelimited(&i2c_adap->dev, "maddr:0x%04X size:0x%02X:quirk max_%s_len must be > %d", eeprom_addr, buf_size, - read ? "read" : "write", EEPROM_OFFSET_SIZE); + str_read_write(read), EEPROM_OFFSET_SIZE); return -EINVAL; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fdinfo.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fdinfo.c index c7df7fa3459f..df2cf5c33925 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fdinfo.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fdinfo.c @@ -33,6 +33,7 @@ #include #include #include +#include #include "amdgpu.h" #include "amdgpu_vm.h" @@ -59,18 +60,25 @@ void amdgpu_show_fdinfo(struct drm_printer *p, struct drm_file *file) struct amdgpu_fpriv *fpriv = file->driver_priv; struct amdgpu_vm *vm = &fpriv->vm; - struct amdgpu_mem_stats stats; + struct amdgpu_mem_stats stats[__AMDGPU_PL_LAST + 1] = { }; ktime_t usage[AMDGPU_HW_IP_NUM]; - unsigned int hw_ip; + const char *pl_name[] = { + [TTM_PL_VRAM] = "vram", + [TTM_PL_TT] = "gtt", + [TTM_PL_SYSTEM] = "cpu", + [AMDGPU_PL_GDS] = "gds", + [AMDGPU_PL_GWS] = "gws", + [AMDGPU_PL_OA] = "oa", + [AMDGPU_PL_DOORBELL] = "doorbell", + }; + unsigned int hw_ip, i; int ret; - memset(&stats, 0, sizeof(stats)); - ret = amdgpu_bo_reserve(vm->root.bo, false); if (ret) return; - amdgpu_vm_get_memory(vm, &stats); + amdgpu_vm_get_memory(vm, stats, ARRAY_SIZE(stats)); amdgpu_bo_unreserve(vm->root.bo); amdgpu_ctx_mgr_usage(&fpriv->ctx_mgr, usage); @@ -82,24 +90,33 @@ void amdgpu_show_fdinfo(struct drm_printer *p, struct drm_file *file) */ drm_printf(p, "pasid:\t%u\n", fpriv->vm.pasid); - drm_printf(p, "drm-memory-vram:\t%llu KiB\n", stats.vram/1024UL); - drm_printf(p, "drm-memory-gtt: \t%llu KiB\n", stats.gtt/1024UL); - drm_printf(p, "drm-memory-cpu: \t%llu KiB\n", stats.cpu/1024UL); - drm_printf(p, "amd-memory-visible-vram:\t%llu KiB\n", - stats.visible_vram/1024UL); + + for (i = 0; i < ARRAY_SIZE(pl_name); i++) { + if (!pl_name[i]) + continue; + + drm_print_memory_stats(p, + &stats[i].drm, + DRM_GEM_OBJECT_RESIDENT | + DRM_GEM_OBJECT_PURGEABLE, + pl_name[i]); + } + + /* Legacy amdgpu keys, alias to drm-resident-memory-: */ + drm_printf(p, "drm-memory-vram:\t%llu KiB\n", + stats[TTM_PL_VRAM].drm.resident/1024UL); + drm_printf(p, "drm-memory-gtt: \t%llu KiB\n", + stats[TTM_PL_TT].drm.resident/1024UL); + drm_printf(p, "drm-memory-cpu: \t%llu KiB\n", + stats[TTM_PL_SYSTEM].drm.resident/1024UL); + + /* Amdgpu specific memory accounting keys: */ drm_printf(p, "amd-evicted-vram:\t%llu KiB\n", - stats.evicted_vram/1024UL); - drm_printf(p, "amd-evicted-visible-vram:\t%llu KiB\n", - stats.evicted_visible_vram/1024UL); + stats[TTM_PL_VRAM].evicted/1024UL); drm_printf(p, "amd-requested-vram:\t%llu KiB\n", - stats.requested_vram/1024UL); - drm_printf(p, "amd-requested-visible-vram:\t%llu KiB\n", - stats.requested_visible_vram/1024UL); + stats[TTM_PL_VRAM].requested/1024UL); drm_printf(p, "amd-requested-gtt:\t%llu KiB\n", - stats.requested_gtt/1024UL); - drm_printf(p, "drm-shared-vram:\t%llu KiB\n", stats.vram_shared/1024UL); - drm_printf(p, "drm-shared-gtt:\t%llu KiB\n", stats.gtt_shared/1024UL); - drm_printf(p, "drm-shared-cpu:\t%llu KiB\n", stats.cpu_shared/1024UL); + stats[TTM_PL_TT].requested/1024UL); for (hw_ip = 0; hw_ip < AMDGPU_HW_IP_NUM; ++hw_ip) { if (!usage[hw_ip]) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c index 256b95232de5..b2033f8352f5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c @@ -78,8 +78,9 @@ static int amdgpu_gart_dummy_page_init(struct amdgpu_device *adev) if (adev->dummy_page_addr) return 0; - adev->dummy_page_addr = dma_map_page(&adev->pdev->dev, dummy_page, 0, - PAGE_SIZE, DMA_BIDIRECTIONAL); + adev->dummy_page_addr = dma_map_page_attrs(&adev->pdev->dev, dummy_page, 0, + PAGE_SIZE, DMA_BIDIRECTIONAL, + DMA_ATTR_SKIP_CPU_SYNC); if (dma_mapping_error(&adev->pdev->dev, adev->dummy_page_addr)) { dev_err(&adev->pdev->dev, "Failed to DMA MAP the dummy page\n"); adev->dummy_page_addr = 0; @@ -99,8 +100,9 @@ void amdgpu_gart_dummy_page_fini(struct amdgpu_device *adev) { if (!adev->dummy_page_addr) return; - dma_unmap_page(&adev->pdev->dev, adev->dummy_page_addr, PAGE_SIZE, - DMA_BIDIRECTIONAL); + dma_unmap_page_attrs(&adev->pdev->dev, adev->dummy_page_addr, PAGE_SIZE, + DMA_BIDIRECTIONAL, + DMA_ATTR_SKIP_CPU_SYNC); adev->dummy_page_addr = 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c index f1ffab5a1eae..f57cc72c43cf 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c @@ -87,16 +87,6 @@ int amdgpu_gfx_me_queue_to_bit(struct amdgpu_device *adev, return bit; } -void amdgpu_gfx_bit_to_me_queue(struct amdgpu_device *adev, int bit, - int *me, int *pipe, int *queue) -{ - *queue = bit % adev->gfx.me.num_queue_per_pipe; - *pipe = (bit / adev->gfx.me.num_queue_per_pipe) - % adev->gfx.me.num_pipe_per_me; - *me = (bit / adev->gfx.me.num_queue_per_pipe) - / adev->gfx.me.num_pipe_per_me; -} - bool amdgpu_gfx_is_me_queue_enabled(struct amdgpu_device *adev, int me, int pipe, int queue) { @@ -415,7 +405,7 @@ int amdgpu_gfx_mqd_sw_init(struct amdgpu_device *adev, } /* prepare MQD backup */ - kiq->mqd_backup = kmalloc(mqd_size, GFP_KERNEL); + kiq->mqd_backup = kzalloc(mqd_size, GFP_KERNEL); if (!kiq->mqd_backup) { dev_warn(adev->dev, "no memory to create MQD backup for ring %s\n", ring->name); @@ -438,7 +428,7 @@ int amdgpu_gfx_mqd_sw_init(struct amdgpu_device *adev, ring->mqd_size = mqd_size; /* prepare MQD backup */ - adev->gfx.me.mqd_backup[i] = kmalloc(mqd_size, GFP_KERNEL); + adev->gfx.me.mqd_backup[i] = kzalloc(mqd_size, GFP_KERNEL); if (!adev->gfx.me.mqd_backup[i]) { dev_warn(adev->dev, "no memory to create MQD backup for ring %s\n", ring->name); return -ENOMEM; @@ -462,7 +452,7 @@ int amdgpu_gfx_mqd_sw_init(struct amdgpu_device *adev, ring->mqd_size = mqd_size; /* prepare MQD backup */ - adev->gfx.mec.mqd_backup[j] = kmalloc(mqd_size, GFP_KERNEL); + adev->gfx.mec.mqd_backup[j] = kzalloc(mqd_size, GFP_KERNEL); if (!adev->gfx.mec.mqd_backup[j]) { dev_warn(adev->dev, "no memory to create MQD backup for ring %s\n", ring->name); return -ENOMEM; @@ -525,6 +515,9 @@ int amdgpu_gfx_disable_kcq(struct amdgpu_device *adev, int xcc_id) if (!kiq->pmf || !kiq->pmf->kiq_unmap_queues) return -EINVAL; + if (!kiq_ring->sched.ready || adev->job_hang || amdgpu_in_reset(adev)) + return 0; + spin_lock(&kiq->ring_lock); if (amdgpu_ring_alloc(kiq_ring, kiq->pmf->unmap_queues_size * adev->gfx.num_compute_rings)) { @@ -538,20 +531,15 @@ int amdgpu_gfx_disable_kcq(struct amdgpu_device *adev, int xcc_id) &adev->gfx.compute_ring[j], RESET_QUEUES, 0, 0); } - - /** - * This is workaround: only skip kiq_ring test - * during ras recovery in suspend stage for gfx9.4.3 + /* Submit unmap queue packet */ + amdgpu_ring_commit(kiq_ring); + /* + * Ring test will do a basic scratch register change check. Just run + * this to ensure that unmap queues that is submitted before got + * processed successfully before returning. */ - if ((amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(9, 4, 3) || - amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(9, 4, 4)) && - amdgpu_ras_in_recovery(adev)) { - spin_unlock(&kiq->ring_lock); - return 0; - } + r = amdgpu_ring_test_helper(kiq_ring); - if (kiq_ring->sched.ready && !adev->job_hang) - r = amdgpu_ring_test_helper(kiq_ring); spin_unlock(&kiq->ring_lock); return r; @@ -579,8 +567,11 @@ int amdgpu_gfx_disable_kgq(struct amdgpu_device *adev, int xcc_id) if (!kiq->pmf || !kiq->pmf->kiq_unmap_queues) return -EINVAL; - spin_lock(&kiq->ring_lock); + if (!adev->gfx.kiq[0].ring.sched.ready || adev->job_hang) + return 0; + if (amdgpu_gfx_is_master_xcc(adev, xcc_id)) { + spin_lock(&kiq->ring_lock); if (amdgpu_ring_alloc(kiq_ring, kiq->pmf->unmap_queues_size * adev->gfx.num_gfx_rings)) { spin_unlock(&kiq->ring_lock); @@ -593,11 +584,17 @@ int amdgpu_gfx_disable_kgq(struct amdgpu_device *adev, int xcc_id) &adev->gfx.gfx_ring[j], PREEMPT_QUEUES, 0, 0); } - } + /* Submit unmap queue packet */ + amdgpu_ring_commit(kiq_ring); - if (adev->gfx.kiq[0].ring.sched.ready && !adev->job_hang) + /* + * Ring test will do a basic scratch register change check. + * Just run this to ensure that unmap queues that is submitted + * before got processed successfully before returning. + */ r = amdgpu_ring_test_helper(kiq_ring); - spin_unlock(&kiq->ring_lock); + spin_unlock(&kiq->ring_lock); + } return r; } @@ -702,7 +699,13 @@ int amdgpu_gfx_enable_kcq(struct amdgpu_device *adev, int xcc_id) kiq->pmf->kiq_map_queues(kiq_ring, &adev->gfx.compute_ring[j]); } - + /* Submit map queue packet */ + amdgpu_ring_commit(kiq_ring); + /* + * Ring test will do a basic scratch register change check. Just run + * this to ensure that map queues that is submitted before got + * processed successfully before returning. + */ r = amdgpu_ring_test_helper(kiq_ring); spin_unlock(&kiq->ring_lock); if (r) @@ -753,7 +756,13 @@ int amdgpu_gfx_enable_kgq(struct amdgpu_device *adev, int xcc_id) &adev->gfx.gfx_ring[j]); } } - + /* Submit map queue packet */ + amdgpu_ring_commit(kiq_ring); + /* + * Ring test will do a basic scratch register change check. Just run + * this to ensure that map queues that is submitted before got + * processed successfully before returning. + */ r = amdgpu_ring_test_helper(kiq_ring); spin_unlock(&kiq->ring_lock); if (r) @@ -895,6 +904,9 @@ int amdgpu_gfx_ras_late_init(struct amdgpu_device *adev, struct ras_common_if *r if (r) return r; + if (amdgpu_sriov_vf(adev)) + return r; + if (adev->gfx.cp_ecc_error_irq.funcs) { r = amdgpu_irq_get(adev, &adev->gfx.cp_ecc_error_irq, 0); if (r) @@ -1363,35 +1375,35 @@ static ssize_t amdgpu_gfx_set_compute_partition(struct device *dev, return count; } +static const char *xcp_desc[] = { + [AMDGPU_SPX_PARTITION_MODE] = "SPX", + [AMDGPU_DPX_PARTITION_MODE] = "DPX", + [AMDGPU_TPX_PARTITION_MODE] = "TPX", + [AMDGPU_QPX_PARTITION_MODE] = "QPX", + [AMDGPU_CPX_PARTITION_MODE] = "CPX", +}; + static ssize_t amdgpu_gfx_get_available_compute_partition(struct device *dev, struct device_attribute *addr, char *buf) { struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = drm_to_adev(ddev); - char *supported_partition; + struct amdgpu_xcp_mgr *xcp_mgr = adev->xcp_mgr; + int size = 0, mode; + char *sep = ""; - /* TBD */ - switch (NUM_XCC(adev->gfx.xcc_mask)) { - case 8: - supported_partition = "SPX, DPX, QPX, CPX"; - break; - case 6: - supported_partition = "SPX, TPX, CPX"; - break; - case 4: - supported_partition = "SPX, DPX, CPX"; - break; - /* this seems only existing in emulation phase */ - case 2: - supported_partition = "SPX, CPX"; - break; - default: - supported_partition = "Not supported"; - break; + if (!xcp_mgr || !xcp_mgr->avail_xcp_modes) + return sysfs_emit(buf, "Not supported\n"); + + for_each_inst(mode, xcp_mgr->avail_xcp_modes) { + size += sysfs_emit_at(buf, size, "%s%s", sep, xcp_desc[mode]); + sep = ", "; } - return sysfs_emit(buf, "%s\n", supported_partition); + size += sysfs_emit_at(buf, size, "\n"); + + return size; } static int amdgpu_gfx_run_cleaner_shader_job(struct amdgpu_ring *ring) @@ -1586,9 +1598,11 @@ static ssize_t amdgpu_gfx_set_enforce_isolation(struct device *dev, if (adev->enforce_isolation[i] && !partition_values[i]) { /* Going from enabled to disabled */ amdgpu_vmid_free_reserved(adev, AMDGPU_GFXHUB(i)); + amdgpu_mes_set_enforce_isolation(adev, i, false); } else if (!adev->enforce_isolation[i] && partition_values[i]) { /* Going from disabled to enabled */ amdgpu_vmid_alloc_reserved(adev, AMDGPU_GFXHUB(i)); + amdgpu_mes_set_enforce_isolation(adev, i, true); } adev->enforce_isolation[i] = partition_values[i]; } @@ -1598,6 +1612,32 @@ static ssize_t amdgpu_gfx_set_enforce_isolation(struct device *dev, return count; } +static ssize_t amdgpu_gfx_get_gfx_reset_mask(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = drm_to_adev(ddev); + + if (!adev) + return -ENODEV; + + return amdgpu_show_reset_mask(buf, adev->gfx.gfx_supported_reset); +} + +static ssize_t amdgpu_gfx_get_compute_reset_mask(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = drm_to_adev(ddev); + + if (!adev) + return -ENODEV; + + return amdgpu_show_reset_mask(buf, adev->gfx.compute_supported_reset); +} + static DEVICE_ATTR(run_cleaner_shader, 0200, NULL, amdgpu_gfx_set_run_cleaner_shader); @@ -1611,45 +1651,136 @@ static DEVICE_ATTR(current_compute_partition, 0644, static DEVICE_ATTR(available_compute_partition, 0444, amdgpu_gfx_get_available_compute_partition, NULL); +static DEVICE_ATTR(gfx_reset_mask, 0444, + amdgpu_gfx_get_gfx_reset_mask, NULL); -int amdgpu_gfx_sysfs_init(struct amdgpu_device *adev) +static DEVICE_ATTR(compute_reset_mask, 0444, + amdgpu_gfx_get_compute_reset_mask, NULL); + +static int amdgpu_gfx_sysfs_xcp_init(struct amdgpu_device *adev) { + struct amdgpu_xcp_mgr *xcp_mgr = adev->xcp_mgr; + bool xcp_switch_supported; int r; + if (!xcp_mgr) + return 0; + + xcp_switch_supported = + (xcp_mgr->funcs && xcp_mgr->funcs->switch_partition_mode); + + if (!xcp_switch_supported) + dev_attr_current_compute_partition.attr.mode &= + ~(S_IWUSR | S_IWGRP | S_IWOTH); + r = device_create_file(adev->dev, &dev_attr_current_compute_partition); if (r) return r; - r = device_create_file(adev->dev, &dev_attr_available_compute_partition); + if (xcp_switch_supported) + r = device_create_file(adev->dev, + &dev_attr_available_compute_partition); return r; } -void amdgpu_gfx_sysfs_fini(struct amdgpu_device *adev) +static void amdgpu_gfx_sysfs_xcp_fini(struct amdgpu_device *adev) { + struct amdgpu_xcp_mgr *xcp_mgr = adev->xcp_mgr; + bool xcp_switch_supported; + + if (!xcp_mgr) + return; + + xcp_switch_supported = + (xcp_mgr->funcs && xcp_mgr->funcs->switch_partition_mode); device_remove_file(adev->dev, &dev_attr_current_compute_partition); - device_remove_file(adev->dev, &dev_attr_available_compute_partition); + + if (xcp_switch_supported) + device_remove_file(adev->dev, + &dev_attr_available_compute_partition); } -int amdgpu_gfx_sysfs_isolation_shader_init(struct amdgpu_device *adev) +static int amdgpu_gfx_sysfs_isolation_shader_init(struct amdgpu_device *adev) { int r; r = device_create_file(adev->dev, &dev_attr_enforce_isolation); if (r) return r; + if (adev->gfx.enable_cleaner_shader) + r = device_create_file(adev->dev, &dev_attr_run_cleaner_shader); - r = device_create_file(adev->dev, &dev_attr_run_cleaner_shader); - if (r) - return r; - - return 0; + return r; } -void amdgpu_gfx_sysfs_isolation_shader_fini(struct amdgpu_device *adev) +static void amdgpu_gfx_sysfs_isolation_shader_fini(struct amdgpu_device *adev) { device_remove_file(adev->dev, &dev_attr_enforce_isolation); - device_remove_file(adev->dev, &dev_attr_run_cleaner_shader); + if (adev->gfx.enable_cleaner_shader) + device_remove_file(adev->dev, &dev_attr_run_cleaner_shader); +} + +static int amdgpu_gfx_sysfs_reset_mask_init(struct amdgpu_device *adev) +{ + int r = 0; + + if (!amdgpu_gpu_recovery) + return r; + + if (adev->gfx.num_gfx_rings) { + r = device_create_file(adev->dev, &dev_attr_gfx_reset_mask); + if (r) + return r; + } + + if (adev->gfx.num_compute_rings) { + r = device_create_file(adev->dev, &dev_attr_compute_reset_mask); + if (r) + return r; + } + + return r; +} + +static void amdgpu_gfx_sysfs_reset_mask_fini(struct amdgpu_device *adev) +{ + if (!amdgpu_gpu_recovery) + return; + + if (adev->gfx.num_gfx_rings) + device_remove_file(adev->dev, &dev_attr_gfx_reset_mask); + + if (adev->gfx.num_compute_rings) + device_remove_file(adev->dev, &dev_attr_compute_reset_mask); +} + +int amdgpu_gfx_sysfs_init(struct amdgpu_device *adev) +{ + int r; + + r = amdgpu_gfx_sysfs_xcp_init(adev); + if (r) { + dev_err(adev->dev, "failed to create xcp sysfs files"); + return r; + } + + r = amdgpu_gfx_sysfs_isolation_shader_init(adev); + if (r) + dev_err(adev->dev, "failed to create isolation sysfs files"); + + r = amdgpu_gfx_sysfs_reset_mask_init(adev); + if (r) + dev_err(adev->dev, "failed to create reset mask sysfs files"); + + return r; +} + +void amdgpu_gfx_sysfs_fini(struct amdgpu_device *adev) +{ + amdgpu_gfx_sysfs_xcp_fini(adev); + amdgpu_gfx_sysfs_isolation_shader_fini(adev); + amdgpu_gfx_sysfs_reset_mask_fini(adev); } int amdgpu_gfx_cleaner_shader_sw_init(struct amdgpu_device *adev, @@ -1737,7 +1868,7 @@ static void amdgpu_gfx_kfd_sch_ctrl(struct amdgpu_device *adev, u32 idx, if (adev->gfx.kfd_sch_req_count[idx] == 0 && adev->gfx.kfd_sch_inactive[idx]) { schedule_delayed_work(&adev->gfx.enforce_isolation[idx].work, - GFX_SLICE_PERIOD); + msecs_to_jiffies(adev->gfx.enforce_isolation_time[idx])); } } else { if (adev->gfx.kfd_sch_req_count[idx] == 0) { @@ -1792,8 +1923,9 @@ void amdgpu_gfx_enforce_isolation_handler(struct work_struct *work) fences += amdgpu_fence_count_emitted(&adev->gfx.compute_ring[i]); } if (fences) { + /* we've already had our timeslice, so let's wrap this up */ schedule_delayed_work(&adev->gfx.enforce_isolation[idx].work, - GFX_SLICE_PERIOD); + msecs_to_jiffies(1)); } else { /* Tell KFD to resume the runqueue */ if (adev->kfd.init_complete) { @@ -1806,6 +1938,51 @@ void amdgpu_gfx_enforce_isolation_handler(struct work_struct *work) mutex_unlock(&adev->enforce_isolation_mutex); } +static void +amdgpu_gfx_enforce_isolation_wait_for_kfd(struct amdgpu_device *adev, + u32 idx) +{ + unsigned long cjiffies; + bool wait = false; + + mutex_lock(&adev->enforce_isolation_mutex); + if (adev->enforce_isolation[idx]) { + /* set the initial values if nothing is set */ + if (!adev->gfx.enforce_isolation_jiffies[idx]) { + adev->gfx.enforce_isolation_jiffies[idx] = jiffies; + adev->gfx.enforce_isolation_time[idx] = GFX_SLICE_PERIOD_MS; + } + /* Make sure KFD gets a chance to run */ + if (amdgpu_amdkfd_compute_active(adev, idx)) { + cjiffies = jiffies; + if (time_after(cjiffies, adev->gfx.enforce_isolation_jiffies[idx])) { + cjiffies -= adev->gfx.enforce_isolation_jiffies[idx]; + if ((jiffies_to_msecs(cjiffies) >= GFX_SLICE_PERIOD_MS)) { + /* if our time is up, let KGD work drain before scheduling more */ + wait = true; + /* reset the timer period */ + adev->gfx.enforce_isolation_time[idx] = GFX_SLICE_PERIOD_MS; + } else { + /* set the timer period to what's left in our time slice */ + adev->gfx.enforce_isolation_time[idx] = + GFX_SLICE_PERIOD_MS - jiffies_to_msecs(cjiffies); + } + } else { + /* if jiffies wrap around we will just wait a little longer */ + adev->gfx.enforce_isolation_jiffies[idx] = jiffies; + } + } else { + /* if there is no KFD work, then set the full slice period */ + adev->gfx.enforce_isolation_jiffies[idx] = jiffies; + adev->gfx.enforce_isolation_time[idx] = GFX_SLICE_PERIOD_MS; + } + } + mutex_unlock(&adev->enforce_isolation_mutex); + + if (wait) + msleep(GFX_SLICE_PERIOD_MS); +} + void amdgpu_gfx_enforce_isolation_ring_begin_use(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; @@ -1822,6 +1999,9 @@ void amdgpu_gfx_enforce_isolation_ring_begin_use(struct amdgpu_ring *ring) if (idx >= MAX_XCP) return; + /* Don't submit more work until KFD has had some time */ + amdgpu_gfx_enforce_isolation_wait_for_kfd(adev, idx); + mutex_lock(&adev->enforce_isolation_mutex); if (adev->enforce_isolation[idx]) { if (adev->kfd.init_complete) @@ -1853,3 +2033,144 @@ void amdgpu_gfx_enforce_isolation_ring_end_use(struct amdgpu_ring *ring) } mutex_unlock(&adev->enforce_isolation_mutex); } + +/* + * debugfs for to enable/disable gfx job submission to specific core. + */ +#if defined(CONFIG_DEBUG_FS) +static int amdgpu_debugfs_gfx_sched_mask_set(void *data, u64 val) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)data; + u32 i; + u64 mask = 0; + struct amdgpu_ring *ring; + + if (!adev) + return -ENODEV; + + mask = (1 << adev->gfx.num_gfx_rings) - 1; + if ((val & mask) == 0) + return -EINVAL; + + for (i = 0; i < adev->gfx.num_gfx_rings; ++i) { + ring = &adev->gfx.gfx_ring[i]; + if (val & (1 << i)) + ring->sched.ready = true; + else + ring->sched.ready = false; + } + /* publish sched.ready flag update effective immediately across smp */ + smp_rmb(); + return 0; +} + +static int amdgpu_debugfs_gfx_sched_mask_get(void *data, u64 *val) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)data; + u32 i; + u64 mask = 0; + struct amdgpu_ring *ring; + + if (!adev) + return -ENODEV; + for (i = 0; i < adev->gfx.num_gfx_rings; ++i) { + ring = &adev->gfx.gfx_ring[i]; + if (ring->sched.ready) + mask |= 1 << i; + } + + *val = mask; + return 0; +} + +DEFINE_DEBUGFS_ATTRIBUTE(amdgpu_debugfs_gfx_sched_mask_fops, + amdgpu_debugfs_gfx_sched_mask_get, + amdgpu_debugfs_gfx_sched_mask_set, "%llx\n"); + +#endif + +void amdgpu_debugfs_gfx_sched_mask_init(struct amdgpu_device *adev) +{ +#if defined(CONFIG_DEBUG_FS) + struct drm_minor *minor = adev_to_drm(adev)->primary; + struct dentry *root = minor->debugfs_root; + char name[32]; + + if (!(adev->gfx.num_gfx_rings > 1)) + return; + sprintf(name, "amdgpu_gfx_sched_mask"); + debugfs_create_file(name, 0600, root, adev, + &amdgpu_debugfs_gfx_sched_mask_fops); +#endif +} + +/* + * debugfs for to enable/disable compute job submission to specific core. + */ +#if defined(CONFIG_DEBUG_FS) +static int amdgpu_debugfs_compute_sched_mask_set(void *data, u64 val) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)data; + u32 i; + u64 mask = 0; + struct amdgpu_ring *ring; + + if (!adev) + return -ENODEV; + + mask = (1 << adev->gfx.num_compute_rings) - 1; + if ((val & mask) == 0) + return -EINVAL; + + for (i = 0; i < adev->gfx.num_compute_rings; ++i) { + ring = &adev->gfx.compute_ring[i]; + if (val & (1 << i)) + ring->sched.ready = true; + else + ring->sched.ready = false; + } + + /* publish sched.ready flag update effective immediately across smp */ + smp_rmb(); + return 0; +} + +static int amdgpu_debugfs_compute_sched_mask_get(void *data, u64 *val) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)data; + u32 i; + u64 mask = 0; + struct amdgpu_ring *ring; + + if (!adev) + return -ENODEV; + for (i = 0; i < adev->gfx.num_compute_rings; ++i) { + ring = &adev->gfx.compute_ring[i]; + if (ring->sched.ready) + mask |= 1 << i; + } + + *val = mask; + return 0; +} + +DEFINE_DEBUGFS_ATTRIBUTE(amdgpu_debugfs_compute_sched_mask_fops, + amdgpu_debugfs_compute_sched_mask_get, + amdgpu_debugfs_compute_sched_mask_set, "%llx\n"); + +#endif + +void amdgpu_debugfs_compute_sched_mask_init(struct amdgpu_device *adev) +{ +#if defined(CONFIG_DEBUG_FS) + struct drm_minor *minor = adev_to_drm(adev)->primary; + struct dentry *root = minor->debugfs_root; + char name[32]; + + if (!(adev->gfx.num_compute_rings > 1)) + return; + sprintf(name, "amdgpu_compute_sched_mask"); + debugfs_create_file(name, 0600, root, adev, + &amdgpu_debugfs_compute_sched_mask_fops); +#endif +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h index 5644e10a86a9..8b5bd63b5773 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h @@ -424,6 +424,8 @@ struct amdgpu_gfx { /* reset mask */ uint32_t grbm_soft_reset; uint32_t srbm_soft_reset; + uint32_t gfx_supported_reset; + uint32_t compute_supported_reset; /* gfx off */ bool gfx_off_state; /* true: enabled, false: disabled */ @@ -472,6 +474,8 @@ struct amdgpu_gfx { struct mutex kfd_sch_mutex; u64 kfd_sch_req_count[MAX_XCP]; bool kfd_sch_inactive[MAX_XCP]; + unsigned long enforce_isolation_jiffies[MAX_XCP]; + unsigned long enforce_isolation_time[MAX_XCP]; }; struct amdgpu_gfx_ras_reg_entry { @@ -540,8 +544,6 @@ bool amdgpu_gfx_is_high_priority_graphics_queue(struct amdgpu_device *adev, struct amdgpu_ring *ring); int amdgpu_gfx_me_queue_to_bit(struct amdgpu_device *adev, int me, int pipe, int queue); -void amdgpu_gfx_bit_to_me_queue(struct amdgpu_device *adev, int bit, - int *me, int *pipe, int *queue); bool amdgpu_gfx_is_me_queue_enabled(struct amdgpu_device *adev, int me, int pipe, int queue); void amdgpu_gfx_off_ctrl(struct amdgpu_device *adev, bool enable); @@ -579,11 +581,11 @@ void amdgpu_gfx_cleaner_shader_sw_fini(struct amdgpu_device *adev); void amdgpu_gfx_cleaner_shader_init(struct amdgpu_device *adev, unsigned int cleaner_shader_size, const void *cleaner_shader_ptr); -int amdgpu_gfx_sysfs_isolation_shader_init(struct amdgpu_device *adev); -void amdgpu_gfx_sysfs_isolation_shader_fini(struct amdgpu_device *adev); void amdgpu_gfx_enforce_isolation_handler(struct work_struct *work); void amdgpu_gfx_enforce_isolation_ring_begin_use(struct amdgpu_ring *ring); void amdgpu_gfx_enforce_isolation_ring_end_use(struct amdgpu_ring *ring); +void amdgpu_debugfs_gfx_sched_mask_init(struct amdgpu_device *adev); +void amdgpu_debugfs_compute_sched_mask_init(struct amdgpu_device *adev); static inline const char *amdgpu_gfx_compute_mode_desc(int mode) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c index 17a19d49d30a..1c19a65e6553 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c @@ -1065,18 +1065,6 @@ uint64_t amdgpu_gmc_vram_pa(struct amdgpu_device *adev, struct amdgpu_bo *bo) return amdgpu_gmc_vram_mc2pa(adev, amdgpu_bo_gpu_offset(bo)); } -/** - * amdgpu_gmc_vram_cpu_pa - calculate vram buffer object's physical address - * from CPU's view - * - * @adev: amdgpu_device pointer - * @bo: amdgpu buffer object - */ -uint64_t amdgpu_gmc_vram_cpu_pa(struct amdgpu_device *adev, struct amdgpu_bo *bo) -{ - return amdgpu_bo_gpu_offset(bo) - adev->gmc.vram_start + adev->gmc.aper_base; -} - int amdgpu_gmc_vram_checking(struct amdgpu_device *adev) { struct amdgpu_bo *vram_bo = NULL; @@ -1130,6 +1118,79 @@ release_buffer: return ret; } +static const char *nps_desc[] = { + [AMDGPU_NPS1_PARTITION_MODE] = "NPS1", + [AMDGPU_NPS2_PARTITION_MODE] = "NPS2", + [AMDGPU_NPS3_PARTITION_MODE] = "NPS3", + [AMDGPU_NPS4_PARTITION_MODE] = "NPS4", + [AMDGPU_NPS6_PARTITION_MODE] = "NPS6", + [AMDGPU_NPS8_PARTITION_MODE] = "NPS8", +}; + +static ssize_t available_memory_partition_show(struct device *dev, + struct device_attribute *addr, + char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = drm_to_adev(ddev); + int size = 0, mode; + char *sep = ""; + + for_each_inst(mode, adev->gmc.supported_nps_modes) { + size += sysfs_emit_at(buf, size, "%s%s", sep, nps_desc[mode]); + sep = ", "; + } + size += sysfs_emit_at(buf, size, "\n"); + + return size; +} + +static ssize_t current_memory_partition_store(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t count) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = drm_to_adev(ddev); + enum amdgpu_memory_partition mode; + struct amdgpu_hive_info *hive; + int i; + + mode = UNKNOWN_MEMORY_PARTITION_MODE; + for_each_inst(i, adev->gmc.supported_nps_modes) { + if (!strncasecmp(nps_desc[i], buf, strlen(nps_desc[i]))) { + mode = i; + break; + } + } + + if (mode == UNKNOWN_MEMORY_PARTITION_MODE) + return -EINVAL; + + if (mode == adev->gmc.gmc_funcs->query_mem_partition_mode(adev)) { + dev_info( + adev->dev, + "requested NPS mode is same as current NPS mode, skipping\n"); + return count; + } + + /* If device is part of hive, all devices in the hive should request the + * same mode. Hence store the requested mode in hive. + */ + hive = amdgpu_get_xgmi_hive(adev); + if (hive) { + atomic_set(&hive->requested_nps_mode, mode); + amdgpu_put_xgmi_hive(hive); + } else { + adev->gmc.requested_nps_mode = mode; + } + + dev_info( + adev->dev, + "NPS mode change requested, please remove and reload the driver\n"); + + return count; +} + static ssize_t current_memory_partition_show( struct device *dev, struct device_attribute *addr, char *buf) { @@ -1138,53 +1199,65 @@ static ssize_t current_memory_partition_show( enum amdgpu_memory_partition mode; mode = adev->gmc.gmc_funcs->query_mem_partition_mode(adev); - switch (mode) { - case AMDGPU_NPS1_PARTITION_MODE: - return sysfs_emit(buf, "NPS1\n"); - case AMDGPU_NPS2_PARTITION_MODE: - return sysfs_emit(buf, "NPS2\n"); - case AMDGPU_NPS3_PARTITION_MODE: - return sysfs_emit(buf, "NPS3\n"); - case AMDGPU_NPS4_PARTITION_MODE: - return sysfs_emit(buf, "NPS4\n"); - case AMDGPU_NPS6_PARTITION_MODE: - return sysfs_emit(buf, "NPS6\n"); - case AMDGPU_NPS8_PARTITION_MODE: - return sysfs_emit(buf, "NPS8\n"); - default: + if ((mode >= ARRAY_SIZE(nps_desc)) || + (BIT(mode) & AMDGPU_ALL_NPS_MASK) != BIT(mode)) return sysfs_emit(buf, "UNKNOWN\n"); - } + + return sysfs_emit(buf, "%s\n", nps_desc[mode]); } -static DEVICE_ATTR_RO(current_memory_partition); +static DEVICE_ATTR_RW(current_memory_partition); +static DEVICE_ATTR_RO(available_memory_partition); int amdgpu_gmc_sysfs_init(struct amdgpu_device *adev) { + bool nps_switch_support; + int r = 0; + if (!adev->gmc.gmc_funcs->query_mem_partition_mode) return 0; + nps_switch_support = (hweight32(adev->gmc.supported_nps_modes & + AMDGPU_ALL_NPS_MASK) > 1); + if (!nps_switch_support) + dev_attr_current_memory_partition.attr.mode &= + ~(S_IWUSR | S_IWGRP | S_IWOTH); + else + r = device_create_file(adev->dev, + &dev_attr_available_memory_partition); + + if (r) + return r; + return device_create_file(adev->dev, &dev_attr_current_memory_partition); } void amdgpu_gmc_sysfs_fini(struct amdgpu_device *adev) { + if (!adev->gmc.gmc_funcs->query_mem_partition_mode) + return; + device_remove_file(adev->dev, &dev_attr_current_memory_partition); + device_remove_file(adev->dev, &dev_attr_available_memory_partition); } int amdgpu_gmc_get_nps_memranges(struct amdgpu_device *adev, struct amdgpu_mem_partition_info *mem_ranges, - int exp_ranges) + uint8_t *exp_ranges) { struct amdgpu_gmc_memrange *ranges; int range_cnt, ret, i, j; uint32_t nps_type; + bool refresh; - if (!mem_ranges) + if (!mem_ranges || !exp_ranges) return -EINVAL; + refresh = (adev->init_lvl->level != AMDGPU_INIT_LEVEL_MINIMAL_XGMI) && + (adev->gmc.reset_flags & AMDGPU_GMC_INIT_RESET_NPS); ret = amdgpu_discovery_get_nps_info(adev, &nps_type, &ranges, - &range_cnt); + &range_cnt, refresh); if (ret) return ret; @@ -1192,16 +1265,16 @@ int amdgpu_gmc_get_nps_memranges(struct amdgpu_device *adev, /* TODO: For now, expect ranges and partition count to be the same. * Adjust if there are holes expected in any NPS domain. */ - if (range_cnt != exp_ranges) { + if (*exp_ranges && (range_cnt != *exp_ranges)) { dev_warn( adev->dev, "NPS config mismatch - expected ranges: %d discovery - nps mode: %d, nps ranges: %d", - exp_ranges, nps_type, range_cnt); + *exp_ranges, nps_type, range_cnt); ret = -EINVAL; goto err; } - for (i = 0; i < exp_ranges; ++i) { + for (i = 0; i < range_cnt; ++i) { if (ranges[i].base_address >= ranges[i].limit_address) { dev_warn( adev->dev, @@ -1242,8 +1315,81 @@ int amdgpu_gmc_get_nps_memranges(struct amdgpu_device *adev, ranges[i].limit_address - ranges[i].base_address + 1; } + if (!*exp_ranges) + *exp_ranges = range_cnt; err: kfree(ranges); return ret; } + +int amdgpu_gmc_request_memory_partition(struct amdgpu_device *adev, + int nps_mode) +{ + /* Not supported on VF devices and APUs */ + if (amdgpu_sriov_vf(adev) || (adev->flags & AMD_IS_APU)) + return -EOPNOTSUPP; + + if (!adev->psp.funcs) { + dev_err(adev->dev, + "PSP interface not available for nps mode change request"); + return -EINVAL; + } + + return psp_memory_partition(&adev->psp, nps_mode); +} + +static inline bool amdgpu_gmc_need_nps_switch_req(struct amdgpu_device *adev, + int req_nps_mode, + int cur_nps_mode) +{ + return (((BIT(req_nps_mode) & adev->gmc.supported_nps_modes) == + BIT(req_nps_mode)) && + req_nps_mode != cur_nps_mode); +} + +void amdgpu_gmc_prepare_nps_mode_change(struct amdgpu_device *adev) +{ + int req_nps_mode, cur_nps_mode, r; + struct amdgpu_hive_info *hive; + + if (amdgpu_sriov_vf(adev) || !adev->gmc.supported_nps_modes || + !adev->gmc.gmc_funcs->request_mem_partition_mode) + return; + + cur_nps_mode = adev->gmc.gmc_funcs->query_mem_partition_mode(adev); + hive = amdgpu_get_xgmi_hive(adev); + if (hive) { + req_nps_mode = atomic_read(&hive->requested_nps_mode); + if (!amdgpu_gmc_need_nps_switch_req(adev, req_nps_mode, + cur_nps_mode)) { + amdgpu_put_xgmi_hive(hive); + return; + } + r = amdgpu_xgmi_request_nps_change(adev, hive, req_nps_mode); + amdgpu_put_xgmi_hive(hive); + goto out; + } + + req_nps_mode = adev->gmc.requested_nps_mode; + if (!amdgpu_gmc_need_nps_switch_req(adev, req_nps_mode, cur_nps_mode)) + return; + + /* even if this fails, we should let driver unload w/o blocking */ + r = adev->gmc.gmc_funcs->request_mem_partition_mode(adev, req_nps_mode); +out: + if (r) + dev_err(adev->dev, "NPS mode change request failed\n"); + else + dev_info( + adev->dev, + "NPS mode change request done, reload driver to complete the change\n"); +} + +bool amdgpu_gmc_need_reset_on_init(struct amdgpu_device *adev) +{ + if (adev->gmc.gmc_funcs->need_reset_on_init) + return adev->gmc.gmc_funcs->need_reset_on_init(adev); + + return false; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h index 4d951a1baefa..459a30fe239f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h @@ -73,6 +73,13 @@ enum amdgpu_memory_partition { AMDGPU_NPS8_PARTITION_MODE = 8, }; +#define AMDGPU_ALL_NPS_MASK \ + (BIT(AMDGPU_NPS1_PARTITION_MODE) | BIT(AMDGPU_NPS2_PARTITION_MODE) | \ + BIT(AMDGPU_NPS3_PARTITION_MODE) | BIT(AMDGPU_NPS4_PARTITION_MODE) | \ + BIT(AMDGPU_NPS6_PARTITION_MODE) | BIT(AMDGPU_NPS8_PARTITION_MODE)) + +#define AMDGPU_GMC_INIT_RESET_NPS BIT(0) + /* * GMC page fault information */ @@ -161,6 +168,10 @@ struct amdgpu_gmc_funcs { enum amdgpu_memory_partition (*query_mem_partition_mode)( struct amdgpu_device *adev); + /* Request NPS mode */ + int (*request_mem_partition_mode)(struct amdgpu_device *adev, + int nps_mode); + bool (*need_reset_on_init)(struct amdgpu_device *adev); }; struct amdgpu_xgmi_ras { @@ -182,7 +193,6 @@ struct amdgpu_xgmi { bool supported; struct ras_common_if *ras_if; bool connected_to_cpu; - bool pending_reset; struct amdgpu_xgmi_ras *ras; }; @@ -305,6 +315,9 @@ struct amdgpu_gmc { struct amdgpu_mem_partition_info *mem_partitions; uint8_t num_mem_partitions; const struct amdgpu_gmc_funcs *gmc_funcs; + enum amdgpu_memory_partition requested_nps_mode; + uint32_t supported_nps_modes; + uint32_t reset_flags; struct amdgpu_xgmi xgmi; struct amdgpu_irq_src ecc_irq; @@ -447,13 +460,17 @@ void amdgpu_gmc_get_vbios_allocations(struct amdgpu_device *adev); void amdgpu_gmc_init_pdb0(struct amdgpu_device *adev); uint64_t amdgpu_gmc_vram_mc2pa(struct amdgpu_device *adev, uint64_t mc_addr); uint64_t amdgpu_gmc_vram_pa(struct amdgpu_device *adev, struct amdgpu_bo *bo); -uint64_t amdgpu_gmc_vram_cpu_pa(struct amdgpu_device *adev, struct amdgpu_bo *bo); int amdgpu_gmc_vram_checking(struct amdgpu_device *adev); int amdgpu_gmc_sysfs_init(struct amdgpu_device *adev); void amdgpu_gmc_sysfs_fini(struct amdgpu_device *adev); int amdgpu_gmc_get_nps_memranges(struct amdgpu_device *adev, struct amdgpu_mem_partition_info *mem_ranges, - int exp_ranges); + uint8_t *exp_ranges); + +int amdgpu_gmc_request_memory_partition(struct amdgpu_device *adev, + int nps_mode); +void amdgpu_gmc_prepare_nps_mode_change(struct amdgpu_device *adev); +bool amdgpu_gmc_need_reset_on_init(struct amdgpu_device *adev); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.c index 00d6211e0fbf..f0765ccde668 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.c @@ -225,15 +225,6 @@ void amdgpu_i2c_destroy(struct amdgpu_i2c_chan *i2c) kfree(i2c); } -/* Add the default buses */ -void amdgpu_i2c_init(struct amdgpu_device *adev) -{ - if (amdgpu_hw_i2c) - DRM_INFO("hw_i2c forced on, you may experience display detection problems!\n"); - - amdgpu_atombios_i2c_init(adev); -} - /* remove all the buses */ void amdgpu_i2c_fini(struct amdgpu_device *adev) { @@ -247,22 +238,6 @@ void amdgpu_i2c_fini(struct amdgpu_device *adev) } } -/* Add additional buses */ -void amdgpu_i2c_add(struct amdgpu_device *adev, - const struct amdgpu_i2c_bus_rec *rec, - const char *name) -{ - struct drm_device *dev = adev_to_drm(adev); - int i; - - for (i = 0; i < AMDGPU_MAX_I2C_BUS; i++) { - if (!adev->i2c_bus[i]) { - adev->i2c_bus[i] = amdgpu_i2c_create(dev, rec, name); - return; - } - } -} - /* looks up bus based on id */ struct amdgpu_i2c_chan * amdgpu_i2c_lookup(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.h index 63c2ff7499e1..21e3d1dad0a1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.h @@ -28,11 +28,7 @@ struct amdgpu_i2c_chan *amdgpu_i2c_create(struct drm_device *dev, const struct amdgpu_i2c_bus_rec *rec, const char *name); void amdgpu_i2c_destroy(struct amdgpu_i2c_chan *i2c); -void amdgpu_i2c_init(struct amdgpu_device *adev); void amdgpu_i2c_fini(struct amdgpu_device *adev); -void amdgpu_i2c_add(struct amdgpu_device *adev, - const struct amdgpu_i2c_bus_rec *rec, - const char *name); struct amdgpu_i2c_chan * amdgpu_i2c_lookup(struct amdgpu_device *adev, const struct amdgpu_i2c_bus_rec *i2c_bus); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c index 92d27d32de41..8e712a11aba5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c @@ -342,15 +342,13 @@ static int amdgpu_vmid_grab_reserved(struct amdgpu_vm *vm, * @ring: ring we want to submit job to * @job: job who wants to use the VMID * @id: resulting VMID - * @fence: fence to wait for if no id could be grabbed * * Try to reuse a VMID for this submission. */ static int amdgpu_vmid_grab_used(struct amdgpu_vm *vm, struct amdgpu_ring *ring, struct amdgpu_job *job, - struct amdgpu_vmid **id, - struct dma_fence **fence) + struct amdgpu_vmid **id) { struct amdgpu_device *adev = ring->adev; unsigned vmhub = ring->vm_hub; @@ -429,7 +427,7 @@ int amdgpu_vmid_grab(struct amdgpu_vm *vm, struct amdgpu_ring *ring, if (r || !id) goto error; } else { - r = amdgpu_vmid_grab_used(vm, ring, job, &id, fence); + r = amdgpu_vmid_grab_used(vm, ring, job, &id); if (r) goto error; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_isp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_isp.c index 4766e99dd98f..263ce1811cc8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_isp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_isp.c @@ -33,33 +33,17 @@ #include "isp_v4_1_0.h" #include "isp_v4_1_1.h" -static int isp_sw_init(void *handle) -{ - return 0; -} - -static int isp_sw_fini(void *handle) -{ - return 0; -} - /** * isp_hw_init - start and test isp block * - * @handle: handle for amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * */ -static int isp_hw_init(void *handle) +static int isp_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_isp *isp = &adev->isp; - const struct amdgpu_ip_block *ip_block = - amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_ISP); - - if (!ip_block) - return -EINVAL; - if (isp->funcs->hw_init != NULL) return isp->funcs->hw_init(isp); @@ -69,13 +53,12 @@ static int isp_hw_init(void *handle) /** * isp_hw_fini - stop the hardware block * - * @handle: handle for amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * */ -static int isp_hw_fini(void *handle) +static int isp_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - struct amdgpu_isp *isp = &adev->isp; + struct amdgpu_isp *isp = &ip_block->adev->isp; if (isp->funcs->hw_fini != NULL) return isp->funcs->hw_fini(isp); @@ -83,16 +66,6 @@ static int isp_hw_fini(void *handle) return -ENODEV; } -static int isp_suspend(void *handle) -{ - return 0; -} - -static int isp_resume(void *handle) -{ - return 0; -} - static int isp_load_fw_by_psp(struct amdgpu_device *adev) { const struct common_firmware_header *hdr; @@ -122,9 +95,10 @@ static int isp_load_fw_by_psp(struct amdgpu_device *adev) return r; } -static int isp_early_init(void *handle) +static int isp_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_isp *isp = &adev->isp; switch (amdgpu_ip_version(adev, ISP_HWIP, 0)) { @@ -154,16 +128,6 @@ static bool isp_is_idle(void *handle) return true; } -static int isp_wait_for_idle(void *handle) -{ - return 0; -} - -static int isp_soft_reset(void *handle) -{ - return 0; -} - static int isp_set_clockgating_state(void *handle, enum amd_clockgating_state state) { @@ -179,16 +143,9 @@ static int isp_set_powergating_state(void *handle, static const struct amd_ip_funcs isp_ip_funcs = { .name = "isp_ip", .early_init = isp_early_init, - .late_init = NULL, - .sw_init = isp_sw_init, - .sw_fini = isp_sw_fini, .hw_init = isp_hw_init, .hw_fini = isp_hw_fini, - .suspend = isp_suspend, - .resume = isp_resume, .is_idle = isp_is_idle, - .wait_for_idle = isp_wait_for_idle, - .soft_reset = isp_soft_reset, .set_clockgating_state = isp_set_clockgating_state, .set_powergating_state = isp_set_powergating_state, }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c index 16f2605ac50b..b9d08bc96581 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c @@ -42,7 +42,7 @@ static void amdgpu_job_do_core_dump(struct amdgpu_device *adev, for (i = 0; i < adev->num_ip_blocks; i++) if (adev->ip_blocks[i].version->funcs->dump_ip_state) adev->ip_blocks[i].version->funcs - ->dump_ip_state((void *)adev); + ->dump_ip_state((void *)&adev->ip_blocks[i]); dev_info(adev->dev, "Dumping IP State Completed\n"); amdgpu_coredump(adev, true, false, job); @@ -137,6 +137,7 @@ static enum drm_gpu_sched_stat amdgpu_job_timedout(struct drm_sched_job *s_job) /* attempt a per ring reset */ if (amdgpu_gpu_recovery && ring->funcs->reset) { + dev_err(adev->dev, "Starting %s ring reset\n", s_job->sched->name); /* stop the scheduler, but don't mess with the * bad job yet because if ring reset fails * we'll fall back to full GPU reset. @@ -149,9 +150,10 @@ static enum drm_gpu_sched_stat amdgpu_job_timedout(struct drm_sched_job *s_job) atomic_inc(&ring->adev->gpu_reset_counter); amdgpu_fence_driver_force_completion(ring); if (amdgpu_ring_sched_ready(ring)) - drm_sched_start(&ring->sched); + drm_sched_start(&ring->sched, 0); goto exit; } + dev_err(adev->dev, "Ring %s reset failure\n", ring->sched.name); } if (amdgpu_device_should_recover_gpu(ring->adev)) { @@ -356,10 +358,10 @@ amdgpu_job_prepare_job(struct drm_sched_job *sched_job, if (r) goto error; - if (!fence && job->gang_submit) + if (job->gang_submit) fence = amdgpu_device_switch_gang(ring->adev, job->gang_submit); - while (!fence && job->vm && !job->vmid) { + if (!fence && job->vm && !job->vmid) { r = amdgpu_vmid_grab(job->vm, ring, job, &fence); if (r) { dev_err(ring->adev->dev, "Error getting VM ID (%d)\n", r); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.c index 6df99cb00d9a..04eb51674596 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.c @@ -47,7 +47,7 @@ int amdgpu_jpeg_sw_init(struct amdgpu_device *adev) adev->jpeg.indirect_sram = true; for (i = 0; i < adev->jpeg.num_jpeg_inst; i++) { - if (adev->jpeg.harvest_config & (1 << i)) + if (adev->jpeg.harvest_config & (1U << i)) continue; if (adev->jpeg.indirect_sram) { @@ -73,7 +73,7 @@ int amdgpu_jpeg_sw_fini(struct amdgpu_device *adev) int i, j; for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) { - if (adev->jpeg.harvest_config & (1 << i)) + if (adev->jpeg.harvest_config & (1U << i)) continue; amdgpu_bo_free_kernel( @@ -110,7 +110,7 @@ static void amdgpu_jpeg_idle_work_handler(struct work_struct *work) unsigned int i, j; for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) { - if (adev->jpeg.harvest_config & (1 << i)) + if (adev->jpeg.harvest_config & (1U << i)) continue; for (j = 0; j < adev->jpeg.num_jpeg_rings; ++j) @@ -342,3 +342,111 @@ int amdgpu_jpeg_psp_update_sram(struct amdgpu_device *adev, int inst_idx, return psp_execute_ip_fw_load(&adev->psp, &ucode); } + +/* + * debugfs for to enable/disable jpeg job submission to specific core. + */ +#if defined(CONFIG_DEBUG_FS) +static int amdgpu_debugfs_jpeg_sched_mask_set(void *data, u64 val) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)data; + u32 i, j; + u64 mask = 0; + struct amdgpu_ring *ring; + + if (!adev) + return -ENODEV; + + mask = (1ULL << (adev->jpeg.num_jpeg_inst * adev->jpeg.num_jpeg_rings)) - 1; + if ((val & mask) == 0) + return -EINVAL; + + for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) { + for (j = 0; j < adev->jpeg.num_jpeg_rings; ++j) { + ring = &adev->jpeg.inst[i].ring_dec[j]; + if (val & (1 << ((i * adev->jpeg.num_jpeg_rings) + j))) + ring->sched.ready = true; + else + ring->sched.ready = false; + } + } + /* publish sched.ready flag update effective immediately across smp */ + smp_rmb(); + return 0; +} + +static int amdgpu_debugfs_jpeg_sched_mask_get(void *data, u64 *val) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)data; + u32 i, j; + u64 mask = 0; + struct amdgpu_ring *ring; + + if (!adev) + return -ENODEV; + for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) { + for (j = 0; j < adev->jpeg.num_jpeg_rings; ++j) { + ring = &adev->jpeg.inst[i].ring_dec[j]; + if (ring->sched.ready) + mask |= 1ULL << ((i * adev->jpeg.num_jpeg_rings) + j); + } + } + *val = mask; + return 0; +} + +DEFINE_DEBUGFS_ATTRIBUTE(amdgpu_debugfs_jpeg_sched_mask_fops, + amdgpu_debugfs_jpeg_sched_mask_get, + amdgpu_debugfs_jpeg_sched_mask_set, "%llx\n"); + +#endif + +void amdgpu_debugfs_jpeg_sched_mask_init(struct amdgpu_device *adev) +{ +#if defined(CONFIG_DEBUG_FS) + struct drm_minor *minor = adev_to_drm(adev)->primary; + struct dentry *root = minor->debugfs_root; + char name[32]; + + if (!(adev->jpeg.num_jpeg_inst > 1) && !(adev->jpeg.num_jpeg_rings > 1)) + return; + sprintf(name, "amdgpu_jpeg_sched_mask"); + debugfs_create_file(name, 0600, root, adev, + &amdgpu_debugfs_jpeg_sched_mask_fops); +#endif +} + +static ssize_t amdgpu_get_jpeg_reset_mask(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = drm_to_adev(ddev); + + if (!adev) + return -ENODEV; + + return amdgpu_show_reset_mask(buf, adev->jpeg.supported_reset); +} + +static DEVICE_ATTR(jpeg_reset_mask, 0444, + amdgpu_get_jpeg_reset_mask, NULL); + +int amdgpu_jpeg_sysfs_reset_mask_init(struct amdgpu_device *adev) +{ + int r = 0; + + if (adev->jpeg.num_jpeg_inst) { + r = device_create_file(adev->dev, &dev_attr_jpeg_reset_mask); + if (r) + return r; + } + + return r; +} + +void amdgpu_jpeg_sysfs_reset_mask_fini(struct amdgpu_device *adev) +{ + if (adev->jpeg.num_jpeg_inst) + device_remove_file(adev->dev, &dev_attr_jpeg_reset_mask); +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.h index f9cdd873ac9b..3eb4a4653fce 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.h @@ -128,6 +128,7 @@ struct amdgpu_jpeg { uint16_t inst_mask; uint8_t num_inst_per_aid; bool indirect_sram; + uint32_t supported_reset; }; int amdgpu_jpeg_sw_init(struct amdgpu_device *adev); @@ -149,5 +150,8 @@ int amdgpu_jpeg_ras_late_init(struct amdgpu_device *adev, int amdgpu_jpeg_ras_sw_init(struct amdgpu_device *adev); int amdgpu_jpeg_psp_update_sram(struct amdgpu_device *adev, int inst_idx, enum AMDGPU_UCODE_ID ucode_id); +void amdgpu_debugfs_jpeg_sched_mask_init(struct amdgpu_device *adev); +int amdgpu_jpeg_sysfs_reset_mask_init(struct amdgpu_device *adev); +void amdgpu_jpeg_sysfs_reset_mask_fini(struct amdgpu_device *adev); #endif /*__AMDGPU_JPEG_H__*/ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c index 18ee60378727..3ca03b5e0f91 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c @@ -348,6 +348,24 @@ static bool amdgpu_mca_bank_should_update(struct amdgpu_device *adev, enum amdgp return ret; } +static bool amdgpu_mca_bank_should_dump(struct amdgpu_device *adev, enum amdgpu_mca_error_type type, + struct mca_bank_entry *entry) +{ + bool ret; + + switch (type) { + case AMDGPU_MCA_ERROR_TYPE_CE: + ret = amdgpu_mca_is_deferred_error(adev, entry->regs[MCA_REG_IDX_STATUS]); + break; + case AMDGPU_MCA_ERROR_TYPE_UE: + default: + ret = true; + break; + } + + return ret; +} + static int amdgpu_mca_smu_get_mca_set(struct amdgpu_device *adev, enum amdgpu_mca_error_type type, struct mca_bank_set *mca_set, struct ras_query_context *qctx) { @@ -373,7 +391,8 @@ static int amdgpu_mca_smu_get_mca_set(struct amdgpu_device *adev, enum amdgpu_mc amdgpu_mca_bank_set_add_entry(mca_set, &entry); - amdgpu_mca_smu_mca_bank_dump(adev, i, &entry, qctx); + if (amdgpu_mca_bank_should_dump(adev, type, &entry)) + amdgpu_mca_smu_mca_bank_dump(adev, i, &entry, qctx); } return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c index 7d4b540340e0..59ec20b07a6a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c @@ -104,7 +104,7 @@ static int amdgpu_mes_event_log_init(struct amdgpu_device *adev) return 0; r = amdgpu_bo_create_kernel(adev, adev->mes.event_log_size, PAGE_SIZE, - AMDGPU_GEM_DOMAIN_GTT, + AMDGPU_GEM_DOMAIN_VRAM, &adev->mes.event_log_gpu_obj, &adev->mes.event_log_gpu_addr, &adev->mes.event_log_cpu_addr); @@ -192,17 +192,6 @@ int amdgpu_mes_init(struct amdgpu_device *adev) (uint64_t *)&adev->wb.wb[adev->mes.query_status_fence_offs[i]]; } - r = amdgpu_device_wb_get(adev, &adev->mes.read_val_offs); - if (r) { - dev_err(adev->dev, - "(%d) read_val_offs alloc failed\n", r); - goto error; - } - adev->mes.read_val_gpu_addr = - adev->wb.gpu_addr + (adev->mes.read_val_offs * 4); - adev->mes.read_val_ptr = - (uint32_t *)&adev->wb.wb[adev->mes.read_val_offs]; - r = amdgpu_mes_doorbell_init(adev); if (r) goto error; @@ -223,8 +212,6 @@ error: amdgpu_device_wb_free(adev, adev->mes.query_status_fence_offs[i]); } - if (adev->mes.read_val_ptr) - amdgpu_device_wb_free(adev, adev->mes.read_val_offs); idr_destroy(&adev->mes.pasid_idr); idr_destroy(&adev->mes.gang_id_idr); @@ -249,8 +236,6 @@ void amdgpu_mes_fini(struct amdgpu_device *adev) amdgpu_device_wb_free(adev, adev->mes.query_status_fence_offs[i]); } - if (adev->mes.read_val_ptr) - amdgpu_device_wb_free(adev, adev->mes.read_val_offs); amdgpu_mes_doorbell_free(adev); @@ -905,7 +890,7 @@ int amdgpu_mes_reset_legacy_queue(struct amdgpu_device *adev, queue_input.me_id = ring->me; queue_input.pipe_id = ring->pipe; queue_input.queue_id = ring->queue; - queue_input.mqd_addr = amdgpu_bo_gpu_offset(ring->mqd_obj); + queue_input.mqd_addr = ring->mqd_obj ? amdgpu_bo_gpu_offset(ring->mqd_obj) : 0; queue_input.wptr_addr = ring->wptr_gpu_addr; queue_input.vmid = vmid; queue_input.use_mmio = use_mmio; @@ -921,10 +906,19 @@ uint32_t amdgpu_mes_rreg(struct amdgpu_device *adev, uint32_t reg) { struct mes_misc_op_input op_input; int r, val = 0; + uint32_t addr_offset = 0; + uint64_t read_val_gpu_addr; + uint32_t *read_val_ptr; + if (amdgpu_device_wb_get(adev, &addr_offset)) { + DRM_ERROR("critical bug! too many mes readers\n"); + goto error; + } + read_val_gpu_addr = adev->wb.gpu_addr + (addr_offset * 4); + read_val_ptr = (uint32_t *)&adev->wb.wb[addr_offset]; op_input.op = MES_MISC_OP_READ_REG; op_input.read_reg.reg_offset = reg; - op_input.read_reg.buffer_addr = adev->mes.read_val_gpu_addr; + op_input.read_reg.buffer_addr = read_val_gpu_addr; if (!adev->mes.funcs->misc_op) { DRM_ERROR("mes rreg is not supported!\n"); @@ -935,9 +929,11 @@ uint32_t amdgpu_mes_rreg(struct amdgpu_device *adev, uint32_t reg) if (r) DRM_ERROR("failed to read reg (0x%x)\n", reg); else - val = *(adev->mes.read_val_ptr); + val = *(read_val_ptr); error: + if (addr_offset) + amdgpu_device_wb_free(adev, addr_offset); return val; } @@ -1594,6 +1590,7 @@ int amdgpu_mes_init_microcode(struct amdgpu_device *adev, int pipe) char ucode_prefix[30]; char fw_name[50]; bool need_retry = false; + u32 *ucode_ptr; int r; amdgpu_ucode_ip_version_decode(adev, GC_HWIP, ucode_prefix, @@ -1631,6 +1628,10 @@ int amdgpu_mes_init_microcode(struct amdgpu_device *adev, int pipe) adev->mes.data_start_addr[pipe] = le32_to_cpu(mes_hdr->mes_data_start_addr_lo) | ((uint64_t)(le32_to_cpu(mes_hdr->mes_data_start_addr_hi)) << 32); + ucode_ptr = (u32 *)(adev->mes.fw[pipe]->data + + sizeof(union amdgpu_firmware_header)); + adev->mes.fw_version[pipe] = + le32_to_cpu(ucode_ptr[24]) & AMDGPU_MES_VERSION_MASK; if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { int ucode, ucode_data; @@ -1677,6 +1678,29 @@ bool amdgpu_mes_suspend_resume_all_supported(struct amdgpu_device *adev) return is_supported; } +/* Fix me -- node_id is used to identify the correct MES instances in the future */ +int amdgpu_mes_set_enforce_isolation(struct amdgpu_device *adev, uint32_t node_id, bool enable) +{ + struct mes_misc_op_input op_input = {0}; + int r; + + op_input.op = MES_MISC_OP_CHANGE_CONFIG; + op_input.change_config.option.limit_single_process = enable ? 1 : 0; + + if (!adev->mes.funcs->misc_op) { + dev_err(adev->dev, "mes change config is not supported!\n"); + r = -EINVAL; + goto error; + } + + r = adev->mes.funcs->misc_op(&adev->mes, &op_input); + if (r) + dev_err(adev->dev, "failed to change_config.\n"); + +error: + return r; +} + #if defined(CONFIG_DEBUG_FS) static int amdgpu_debugfs_mes_event_log_show(struct seq_file *m, void *unused) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h index 96788c0f42f1..c6f93cbd6739 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h @@ -40,6 +40,7 @@ #define AMDGPU_MES_VERSION_MASK 0x00000fff #define AMDGPU_MES_API_VERSION_MASK 0x00fff000 #define AMDGPU_MES_FEAT_VERSION_MASK 0xff000000 +#define AMDGPU_MES_MSCRATCH_SIZE 0x8000 enum amdgpu_mes_priority_level { AMDGPU_MES_PRIORITY_LEVEL_LOW = 0, @@ -75,6 +76,7 @@ struct amdgpu_mes { uint32_t sched_version; uint32_t kiq_version; + uint32_t fw_version[AMDGPU_MAX_MES_PIPES]; bool enable_legacy_queue_map; uint32_t total_max_queue; @@ -119,9 +121,6 @@ struct amdgpu_mes { uint32_t query_status_fence_offs[AMDGPU_MAX_MES_PIPES]; uint64_t query_status_fence_gpu_addr[AMDGPU_MAX_MES_PIPES]; uint64_t *query_status_fence_ptr[AMDGPU_MAX_MES_PIPES]; - uint32_t read_val_offs; - uint64_t read_val_gpu_addr; - uint32_t *read_val_ptr; uint32_t saved_flags; @@ -310,6 +309,7 @@ enum mes_misc_opcode { MES_MISC_OP_WRM_REG_WAIT, MES_MISC_OP_WRM_REG_WR_WAIT, MES_MISC_OP_SET_SHADER_DEBUGGER, + MES_MISC_OP_CHANGE_CONFIG, }; struct mes_misc_op_input { @@ -348,6 +348,21 @@ struct mes_misc_op_input { uint32_t tcp_watch_cntl[4]; uint32_t trap_en; } set_shader_debugger; + + struct { + union { + struct { + uint32_t limit_single_process : 1; + uint32_t enable_hws_logging_buffer : 1; + uint32_t reserved : 30; + }; + uint32_t all; + } option; + struct { + uint32_t tdr_level; + uint32_t tdr_delay; + } tdr_config; + } change_config; }; }; @@ -518,4 +533,7 @@ static inline void amdgpu_mes_unlock(struct amdgpu_mes *mes) } bool amdgpu_mes_suspend_resume_all_supported(struct amdgpu_device *adev); + +int amdgpu_mes_set_enforce_isolation(struct amdgpu_device *adev, uint32_t node_id, bool enable); + #endif /* __AMDGPU_MES_H__ */ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h index f61d117b0caf..79c2f807b9fe 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h @@ -101,6 +101,7 @@ struct amdgpu_nbio_funcs { int (*get_compute_partition_mode)(struct amdgpu_device *adev); u32 (*get_memory_partition_mode)(struct amdgpu_device *adev, u32 *supp_modes); + bool (*is_nps_switch_requested)(struct amdgpu_device *adev); u64 (*get_pcie_replay_count)(struct amdgpu_device *adev); void (*set_reg_remap)(struct amdgpu_device *adev); }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index 971419e3a9bb..6852d50caa89 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -40,6 +40,7 @@ #include "amdgpu_trace.h" #include "amdgpu_amdkfd.h" #include "amdgpu_vram_mgr.h" +#include "amdgpu_vm.h" /** * DOC: amdgpu_object @@ -1171,54 +1172,71 @@ void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, } void amdgpu_bo_get_memory(struct amdgpu_bo *bo, - struct amdgpu_mem_stats *stats) + struct amdgpu_mem_stats *stats, + unsigned int sz) { + const unsigned int domain_to_pl[] = { + [ilog2(AMDGPU_GEM_DOMAIN_CPU)] = TTM_PL_SYSTEM, + [ilog2(AMDGPU_GEM_DOMAIN_GTT)] = TTM_PL_TT, + [ilog2(AMDGPU_GEM_DOMAIN_VRAM)] = TTM_PL_VRAM, + [ilog2(AMDGPU_GEM_DOMAIN_GDS)] = AMDGPU_PL_GDS, + [ilog2(AMDGPU_GEM_DOMAIN_GWS)] = AMDGPU_PL_GWS, + [ilog2(AMDGPU_GEM_DOMAIN_OA)] = AMDGPU_PL_OA, + [ilog2(AMDGPU_GEM_DOMAIN_DOORBELL)] = AMDGPU_PL_DOORBELL, + }; struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); struct ttm_resource *res = bo->tbo.resource; + struct drm_gem_object *obj = &bo->tbo.base; uint64_t size = amdgpu_bo_size(bo); - struct drm_gem_object *obj; - bool shared; + unsigned int type; - /* Abort if the BO doesn't currently have a backing store */ - if (!res) - return; - - obj = &bo->tbo.base; - shared = drm_gem_object_is_shared_for_memory_stats(obj); - - switch (res->mem_type) { - case TTM_PL_VRAM: - stats->vram += size; - if (amdgpu_res_cpu_visible(adev, res)) - stats->visible_vram += size; - if (shared) - stats->vram_shared += size; - break; - case TTM_PL_TT: - stats->gtt += size; - if (shared) - stats->gtt_shared += size; - break; - case TTM_PL_SYSTEM: - default: - stats->cpu += size; - if (shared) - stats->cpu_shared += size; - break; + if (!res) { + /* + * If no backing store use one of the preferred domain for basic + * stats. We take the MSB since that should give a reasonable + * view. + */ + BUILD_BUG_ON(TTM_PL_VRAM < TTM_PL_TT || + TTM_PL_VRAM < TTM_PL_SYSTEM); + type = fls(bo->preferred_domains & AMDGPU_GEM_DOMAIN_MASK); + if (!type) + return; + type--; + if (drm_WARN_ON_ONCE(&adev->ddev, + type >= ARRAY_SIZE(domain_to_pl))) + return; + type = domain_to_pl[type]; + } else { + type = res->mem_type; } - if (bo->preferred_domains & AMDGPU_GEM_DOMAIN_VRAM) { - stats->requested_vram += size; - if (bo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) - stats->requested_visible_vram += size; + if (drm_WARN_ON_ONCE(&adev->ddev, type >= sz)) + return; - if (res->mem_type != TTM_PL_VRAM) { - stats->evicted_vram += size; - if (bo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) - stats->evicted_visible_vram += size; - } + /* DRM stats common fields: */ + + if (drm_gem_object_is_shared_for_memory_stats(obj)) + stats[type].drm.shared += size; + else + stats[type].drm.private += size; + + if (res) { + stats[type].drm.resident += size; + + if (!dma_resv_test_signaled(obj->resv, DMA_RESV_USAGE_BOOKKEEP)) + stats[type].drm.active += size; + else if (bo->flags & AMDGPU_GEM_CREATE_DISCARDABLE) + stats[type].drm.purgeable += size; + } + + /* amdgpu specific stats: */ + + if (bo->preferred_domains & AMDGPU_GEM_DOMAIN_VRAM) { + stats[TTM_PL_VRAM].requested += size; + if (type != TTM_PL_VRAM) + stats[TTM_PL_VRAM].evicted += size; } else if (bo->preferred_domains & AMDGPU_GEM_DOMAIN_GTT) { - stats->requested_gtt += size; + stats[TTM_PL_TT].requested += size; } } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h index 717e47b46167..be6769852ece 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h @@ -139,33 +139,6 @@ struct amdgpu_bo_vm { struct amdgpu_vm_bo_base entries[]; }; -struct amdgpu_mem_stats { - /* current VRAM usage, includes visible VRAM */ - uint64_t vram; - /* current shared VRAM usage, includes visible VRAM */ - uint64_t vram_shared; - /* current visible VRAM usage */ - uint64_t visible_vram; - /* current GTT usage */ - uint64_t gtt; - /* current shared GTT usage */ - uint64_t gtt_shared; - /* current system memory usage */ - uint64_t cpu; - /* current shared system memory usage */ - uint64_t cpu_shared; - /* sum of evicted buffers, includes visible VRAM */ - uint64_t evicted_vram; - /* sum of evicted buffers due to CPU access */ - uint64_t evicted_visible_vram; - /* how much userspace asked for, includes vis.VRAM */ - uint64_t requested_vram; - /* how much userspace asked for */ - uint64_t requested_visible_vram; - /* how much userspace asked for */ - uint64_t requested_gtt; -}; - static inline struct amdgpu_bo *ttm_to_amdgpu_bo(struct ttm_buffer_object *tbo) { return container_of(tbo, struct amdgpu_bo, tbo); @@ -328,7 +301,8 @@ int amdgpu_bo_sync_wait(struct amdgpu_bo *bo, void *owner, bool intr); u64 amdgpu_bo_gpu_offset(struct amdgpu_bo *bo); u64 amdgpu_bo_gpu_offset_no_check(struct amdgpu_bo *bo); void amdgpu_bo_get_memory(struct amdgpu_bo *bo, - struct amdgpu_mem_stats *stats); + struct amdgpu_mem_stats *stats, + unsigned int size); uint32_t amdgpu_bo_get_preferred_domain(struct amdgpu_device *adev, uint32_t domain); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index 0b28b2cf1517..17cf10c0b72b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -159,9 +159,9 @@ static int psp_init_sriov_microcode(struct psp_context *psp) return ret; } -static int psp_early_init(void *handle) +static int psp_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct psp_context *psp = &adev->psp; psp->autoload_supported = true; @@ -421,9 +421,9 @@ static bool psp_get_runtime_db_entry(struct amdgpu_device *adev, return ret; } -static int psp_sw_init(void *handle) +static int psp_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct psp_context *psp = &adev->psp; int ret; struct psp_runtime_boot_cfg_entry boot_cfg_entry; @@ -527,9 +527,9 @@ failed1: return ret; } -static int psp_sw_fini(void *handle) +static int psp_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct psp_context *psp = &adev->psp; struct psp_gfx_cmd_resp *cmd = psp->cmd; @@ -639,6 +639,8 @@ static const char *psp_gfx_cmd_name(enum psp_gfx_cmd_id cmd_id) return "AUTOLOAD_RLC"; case GFX_CMD_ID_BOOT_CFG: return "BOOT_CFG"; + case GFX_CMD_ID_CONFIG_SQ_PERFMON: + return "CONFIG_SQ_PERFMON"; default: return "UNKNOWN CMD"; } @@ -1043,6 +1045,31 @@ static int psp_rl_load(struct amdgpu_device *adev) return ret; } +int psp_memory_partition(struct psp_context *psp, int mode) +{ + struct psp_gfx_cmd_resp *cmd; + int ret; + + if (amdgpu_sriov_vf(psp->adev)) + return 0; + + cmd = acquire_psp_cmd_buf(psp); + + cmd->cmd_id = GFX_CMD_ID_FB_NPS_MODE; + cmd->cmd.cmd_memory_part.mode = mode; + + dev_info(psp->adev->dev, + "Requesting %d memory partition change through PSP", mode); + ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr); + if (ret) + dev_err(psp->adev->dev, + "PSP request failed to change to NPS%d mode\n", mode); + + release_psp_cmd_buf(psp); + + return ret; +} + int psp_spatial_partition(struct psp_context *psp, int mode) { struct psp_gfx_cmd_resp *cmd; @@ -1807,6 +1834,9 @@ int psp_ras_initialize(struct psp_context *psp) ras_cmd->ras_in_message.init_flags.xcc_mask = adev->gfx.xcc_mask; ras_cmd->ras_in_message.init_flags.channel_dis_num = hweight32(adev->gmc.m_half_use) * 2; + if (adev->gmc.gmc_funcs->query_mem_partition_mode) + ras_cmd->ras_in_message.init_flags.nps_mode = + adev->gmc.gmc_funcs->query_mem_partition_mode(adev); ret = psp_ta_load(psp, &psp->ras_context.context); @@ -2264,6 +2294,19 @@ bool amdgpu_psp_get_ras_capability(struct psp_context *psp) } } +bool amdgpu_psp_tos_reload_needed(struct amdgpu_device *adev) +{ + struct psp_context *psp = &adev->psp; + + if (amdgpu_sriov_vf(adev) || (adev->flags & AMD_IS_APU)) + return false; + + if (psp->funcs && psp->funcs->is_reload_needed) + return psp->funcs->is_reload_needed(psp); + + return false; +} + static int psp_hw_start(struct psp_context *psp) { struct amdgpu_device *adev = psp->adev; @@ -2958,10 +3001,10 @@ failed: return ret; } -static int psp_hw_init(void *handle) +static int psp_hw_init(struct amdgpu_ip_block *ip_block) { int ret; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; mutex_lock(&adev->firmware.mutex); /* @@ -2987,9 +3030,9 @@ failed: return -EINVAL; } -static int psp_hw_fini(void *handle) +static int psp_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct psp_context *psp = &adev->psp; if (psp->ta_fw) { @@ -3011,10 +3054,10 @@ static int psp_hw_fini(void *handle) return 0; } -static int psp_suspend(void *handle) +static int psp_suspend(struct amdgpu_ip_block *ip_block) { int ret = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct psp_context *psp = &adev->psp; if (adev->gmc.xgmi.num_physical_nodes > 1 && @@ -3074,10 +3117,10 @@ out: return ret; } -static int psp_resume(void *handle) +static int psp_resume(struct amdgpu_ip_block *ip_block) { int ret; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct psp_context *psp = &adev->psp; dev_info(adev->dev, "PSP is resuming...\n"); @@ -3523,6 +3566,36 @@ out: return err; } +static bool is_ta_fw_applicable(struct psp_context *psp, + const struct psp_fw_bin_desc *desc) +{ + struct amdgpu_device *adev = psp->adev; + uint32_t fw_version; + + switch (desc->fw_type) { + case TA_FW_TYPE_PSP_XGMI: + case TA_FW_TYPE_PSP_XGMI_AUX: + /* for now, AUX TA only exists on 13.0.6 ta bin, + * from v20.00.0x.14 + */ + if (amdgpu_ip_version(adev, MP0_HWIP, 0) == + IP_VERSION(13, 0, 6)) { + fw_version = le32_to_cpu(desc->fw_version); + + if (adev->flags & AMD_IS_APU && + (fw_version & 0xff) >= 0x14) + return desc->fw_type == TA_FW_TYPE_PSP_XGMI_AUX; + else + return desc->fw_type == TA_FW_TYPE_PSP_XGMI; + } + break; + default: + break; + } + + return true; +} + static int parse_ta_bin_descriptor(struct psp_context *psp, const struct psp_fw_bin_desc *desc, const struct ta_firmware_header_v2_0 *ta_hdr) @@ -3532,6 +3605,9 @@ static int parse_ta_bin_descriptor(struct psp_context *psp, if (!psp || !desc || !ta_hdr) return -EINVAL; + if (!is_ta_fw_applicable(psp, desc)) + return 0; + ucode_start_addr = (uint8_t *)ta_hdr + le32_to_cpu(desc->offset_bytes) + le32_to_cpu(ta_hdr->header.ucode_array_offset_bytes); @@ -3544,6 +3620,7 @@ static int parse_ta_bin_descriptor(struct psp_context *psp, psp->asd_context.bin_desc.start_addr = ucode_start_addr; break; case TA_FW_TYPE_PSP_XGMI: + case TA_FW_TYPE_PSP_XGMI_AUX: psp->xgmi_context.context.bin_desc.fw_version = le32_to_cpu(desc->fw_version); psp->xgmi_context.context.bin_desc.size_bytes = le32_to_cpu(desc->size_bytes); psp->xgmi_context.context.bin_desc.start_addr = ucode_start_addr; @@ -3736,8 +3813,44 @@ out: return err; } +int psp_config_sq_perfmon(struct psp_context *psp, + uint32_t xcp_id, bool core_override_enable, + bool reg_override_enable, bool perfmon_override_enable) +{ + int ret; + + if (amdgpu_sriov_vf(psp->adev)) + return 0; + + if (xcp_id > MAX_XCP) { + dev_err(psp->adev->dev, "invalid xcp_id %d\n", xcp_id); + return -EINVAL; + } + + if (amdgpu_ip_version(psp->adev, MP0_HWIP, 0) != IP_VERSION(13, 0, 6)) { + dev_err(psp->adev->dev, "Unsupported MP0 version 0x%x for CONFIG_SQ_PERFMON command\n", + amdgpu_ip_version(psp->adev, MP0_HWIP, 0)); + return -EINVAL; + } + struct psp_gfx_cmd_resp *cmd = acquire_psp_cmd_buf(psp); + + cmd->cmd_id = GFX_CMD_ID_CONFIG_SQ_PERFMON; + cmd->cmd.config_sq_perfmon.gfx_xcp_mask = BIT_MASK(xcp_id); + cmd->cmd.config_sq_perfmon.core_override = core_override_enable; + cmd->cmd.config_sq_perfmon.reg_override = reg_override_enable; + cmd->cmd.config_sq_perfmon.perfmon_override = perfmon_override_enable; + + ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr); + if (ret) + dev_warn(psp->adev->dev, "PSP failed to config sq: xcp%d core%d reg%d perfmon%d\n", + xcp_id, core_override_enable, reg_override_enable, perfmon_override_enable); + + release_psp_cmd_buf(psp); + return ret; +} + static int psp_set_clockgating_state(void *handle, - enum amd_clockgating_state state) + enum amd_clockgating_state state) { return 0; } @@ -4019,17 +4132,12 @@ const struct attribute_group amdgpu_flash_attr_group = { const struct amd_ip_funcs psp_ip_funcs = { .name = "psp", .early_init = psp_early_init, - .late_init = NULL, .sw_init = psp_sw_init, .sw_fini = psp_sw_fini, .hw_init = psp_hw_init, .hw_fini = psp_hw_fini, .suspend = psp_suspend, .resume = psp_resume, - .is_idle = NULL, - .check_soft_reset = NULL, - .wait_for_idle = NULL, - .soft_reset = NULL, .set_clockgating_state = psp_set_clockgating_state, .set_powergating_state = psp_set_powergating_state, }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h index e8abbbcb4326..567cb1f924ca 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h @@ -139,6 +139,7 @@ struct psp_funcs { int (*fatal_error_recovery_quirk)(struct psp_context *psp); bool (*get_ras_capability)(struct psp_context *psp); bool (*is_aux_sos_load_required)(struct psp_context *psp); + bool (*is_reload_needed)(struct psp_context *psp); }; struct ta_funcs { @@ -552,9 +553,15 @@ int psp_load_fw_list(struct psp_context *psp, void psp_copy_fw(struct psp_context *psp, uint8_t *start_addr, uint32_t bin_size); int psp_spatial_partition(struct psp_context *psp, int mode); +int psp_memory_partition(struct psp_context *psp, int mode); int is_psp_fw_valid(struct psp_bin_desc bin); int amdgpu_psp_wait_for_bootloader(struct amdgpu_device *adev); bool amdgpu_psp_get_ras_capability(struct psp_context *psp); + +int psp_config_sq_perfmon(struct psp_context *psp, uint32_t xcp_id, + bool core_override_enable, bool reg_override_enable, bool perfmon_override_enable); +bool amdgpu_psp_tos_reload_needed(struct amdgpu_device *adev); + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index 1a1395c5fff1..1bc95b0cdbb8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -1214,6 +1214,42 @@ static void amdgpu_ras_error_generate_report(struct amdgpu_device *adev, } } +static void amdgpu_ras_virt_error_generate_report(struct amdgpu_device *adev, + struct ras_query_if *query_if, + struct ras_err_data *err_data, + struct ras_query_context *qctx) +{ + unsigned long new_ue, new_ce, new_de; + struct ras_manager *obj = amdgpu_ras_find_obj(adev, &query_if->head); + const char *blk_name = get_ras_block_str(&query_if->head); + u64 event_id = qctx->evid.event_id; + + new_ce = err_data->ce_count - obj->err_data.ce_count; + new_ue = err_data->ue_count - obj->err_data.ue_count; + new_de = err_data->de_count - obj->err_data.de_count; + + if (new_ce) { + RAS_EVENT_LOG(adev, event_id, "%lu correctable hardware errors " + "detected in %s block\n", + new_ce, + blk_name); + } + + if (new_ue) { + RAS_EVENT_LOG(adev, event_id, "%lu uncorrectable hardware errors " + "detected in %s block\n", + new_ue, + blk_name); + } + + if (new_de) { + RAS_EVENT_LOG(adev, event_id, "%lu deferred hardware errors " + "detected in %s block\n", + new_de, + blk_name); + } +} + static void amdgpu_rasmgr_error_data_statistic_update(struct ras_manager *obj, struct ras_err_data *err_data) { struct ras_err_node *err_node; @@ -1237,6 +1273,15 @@ static void amdgpu_rasmgr_error_data_statistic_update(struct ras_manager *obj, s } } +static void amdgpu_ras_mgr_virt_error_data_statistics_update(struct ras_manager *obj, + struct ras_err_data *err_data) +{ + /* Host reports absolute counts */ + obj->err_data.ue_count = err_data->ue_count; + obj->err_data.ce_count = err_data->ce_count; + obj->err_data.de_count = err_data->de_count; +} + static struct ras_manager *get_ras_manager(struct amdgpu_device *adev, enum amdgpu_ras_block blk) { struct ras_common_if head; @@ -1323,7 +1368,9 @@ static int amdgpu_ras_query_error_status_helper(struct amdgpu_device *adev, if (error_query_mode == AMDGPU_RAS_INVALID_ERROR_QUERY) return -EINVAL; - if (error_query_mode == AMDGPU_RAS_DIRECT_ERROR_QUERY) { + if (error_query_mode == AMDGPU_RAS_VIRT_ERROR_COUNT_QUERY) { + return amdgpu_virt_req_ras_err_count(adev, blk, err_data); + } else if (error_query_mode == AMDGPU_RAS_DIRECT_ERROR_QUERY) { if (info->head.block == AMDGPU_RAS_BLOCK__UMC) { amdgpu_ras_get_ecc_info(adev, err_data); } else { @@ -1405,14 +1452,22 @@ static int amdgpu_ras_query_error_status_with_event(struct amdgpu_device *adev, if (ret) goto out_fini_err_data; - amdgpu_rasmgr_error_data_statistic_update(obj, &err_data); + if (error_query_mode != AMDGPU_RAS_VIRT_ERROR_COUNT_QUERY) { + amdgpu_rasmgr_error_data_statistic_update(obj, &err_data); + amdgpu_ras_error_generate_report(adev, info, &err_data, &qctx); + } else { + /* Host provides absolute error counts. First generate the report + * using the previous VF internal count against new host count. + * Then Update VF internal count. + */ + amdgpu_ras_virt_error_generate_report(adev, info, &err_data, &qctx); + amdgpu_ras_mgr_virt_error_data_statistics_update(obj, &err_data); + } info->ue_count = obj->err_data.ue_count; info->ce_count = obj->err_data.ce_count; info->de_count = obj->err_data.de_count; - amdgpu_ras_error_generate_report(adev, info, &err_data, &qctx); - out_fini_err_data: amdgpu_ras_error_data_fini(&err_data); @@ -2605,6 +2660,7 @@ static void amdgpu_ras_do_recovery(struct work_struct *work) reset_context.method = AMD_RESET_METHOD_NONE; reset_context.reset_req_dev = adev; reset_context.src = AMDGPU_RESET_SRC_RAS; + set_bit(AMDGPU_SKIP_COREDUMP, &reset_context.flags); /* Perform full reset in fatal error mode */ if (!amdgpu_ras_is_poison_mode_supported(ras->adev)) @@ -3146,7 +3202,42 @@ static int amdgpu_ras_page_retirement_thread(void *param) return 0; } -int amdgpu_ras_recovery_init(struct amdgpu_device *adev) +int amdgpu_ras_init_badpage_info(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + int ret; + + if (!con || amdgpu_sriov_vf(adev)) + return 0; + + ret = amdgpu_ras_eeprom_init(&con->eeprom_control); + + if (ret) + return ret; + + /* HW not usable */ + if (amdgpu_ras_is_rma(adev)) + return -EHWPOISON; + + if (con->eeprom_control.ras_num_recs) { + ret = amdgpu_ras_load_bad_pages(adev); + if (ret) + return ret; + + amdgpu_dpm_send_hbm_bad_pages_num( + adev, con->eeprom_control.ras_num_recs); + + if (con->update_channel_flag == true) { + amdgpu_dpm_send_hbm_bad_channel_flag( + adev, con->eeprom_control.bad_channel_bitmap); + con->update_channel_flag = false; + } + } + + return ret; +} + +int amdgpu_ras_recovery_init(struct amdgpu_device *adev, bool init_bp_info) { struct amdgpu_ras *con = amdgpu_ras_get_context(adev); struct ras_err_handler_data **data; @@ -3181,31 +3272,10 @@ int amdgpu_ras_recovery_init(struct amdgpu_device *adev) max_eeprom_records_count = amdgpu_ras_eeprom_max_record_count(&con->eeprom_control); amdgpu_ras_validate_threshold(adev, max_eeprom_records_count); - /* Todo: During test the SMU might fail to read the eeprom through I2C - * when the GPU is pending on XGMI reset during probe time - * (Mostly after second bus reset), skip it now - */ - if (adev->gmc.xgmi.pending_reset) - return 0; - ret = amdgpu_ras_eeprom_init(&con->eeprom_control); - /* - * This calling fails when is_rma is true or - * ret != 0. - */ - if (amdgpu_ras_is_rma(adev) || ret) - goto free; - - if (con->eeprom_control.ras_num_recs) { - ret = amdgpu_ras_load_bad_pages(adev); + if (init_bp_info) { + ret = amdgpu_ras_init_badpage_info(adev); if (ret) goto free; - - amdgpu_dpm_send_hbm_bad_pages_num(adev, con->eeprom_control.ras_num_recs); - - if (con->update_channel_flag == true) { - amdgpu_dpm_send_hbm_bad_channel_flag(adev, con->eeprom_control.bad_channel_bitmap); - con->update_channel_flag = false; - } } mutex_init(&con->page_rsv_lock); @@ -3438,6 +3508,11 @@ static void amdgpu_ras_check_supported(struct amdgpu_device *adev) if (!amdgpu_ras_asic_supported(adev)) return; + if (amdgpu_sriov_vf(adev)) { + if (amdgpu_virt_get_ras_capability(adev)) + goto init_ras_enabled_flag; + } + /* query ras capability from psp */ if (amdgpu_psp_get_ras_capability(&adev->psp)) goto init_ras_enabled_flag; @@ -3910,7 +3985,7 @@ int amdgpu_ras_late_init(struct amdgpu_device *adev) } /* Guest side doesn't need init ras feature */ - if (amdgpu_sriov_vf(adev)) + if (amdgpu_sriov_vf(adev) && !amdgpu_sriov_ras_telemetry_en(adev)) return 0; list_for_each_entry_safe(node, tmp, &adev->ras_list, node) { @@ -4294,8 +4369,27 @@ int amdgpu_ras_reset_gpu(struct amdgpu_device *adev) ras->gpu_reset_flags |= AMDGPU_RAS_GPU_RESET_MODE1_RESET; } - if (atomic_cmpxchg(&ras->in_recovery, 0, 1) == 0) + if (atomic_cmpxchg(&ras->in_recovery, 0, 1) == 0) { + struct amdgpu_hive_info *hive = amdgpu_get_xgmi_hive(adev); + int hive_ras_recovery = 0; + + if (hive) { + hive_ras_recovery = atomic_read(&hive->ras_recovery); + amdgpu_put_xgmi_hive(hive); + } + /* In the case of multiple GPUs, after a GPU has started + * resetting all GPUs on hive, other GPUs do not need to + * trigger GPU reset again. + */ + if (!hive_ras_recovery) + amdgpu_reset_domain_schedule(ras->adev->reset_domain, &ras->recovery_work); + else + atomic_set(&ras->in_recovery, 0); + } else { + flush_work(&ras->recovery_work); amdgpu_reset_domain_schedule(ras->adev->reset_domain, &ras->recovery_work); + } + return 0; } @@ -4358,11 +4452,14 @@ bool amdgpu_ras_get_error_query_mode(struct amdgpu_device *adev, return false; } - if ((smu_funcs && smu_funcs->set_debug_mode) || (mca_funcs && mca_funcs->mca_set_debug_mode)) + if (amdgpu_sriov_vf(adev)) { + *error_query_mode = AMDGPU_RAS_VIRT_ERROR_COUNT_QUERY; + } else if ((smu_funcs && smu_funcs->set_debug_mode) || (mca_funcs && mca_funcs->mca_set_debug_mode)) { *error_query_mode = (con->is_aca_debug_mode) ? AMDGPU_RAS_DIRECT_ERROR_QUERY : AMDGPU_RAS_FIRMWARE_ERROR_QUERY; - else + } else { *error_query_mode = AMDGPU_RAS_DIRECT_ERROR_QUERY; + } return true; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h index 669720a9c60a..6db772ecfee4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h @@ -365,6 +365,7 @@ enum amdgpu_ras_error_query_mode { AMDGPU_RAS_INVALID_ERROR_QUERY = 0, AMDGPU_RAS_DIRECT_ERROR_QUERY = 1, AMDGPU_RAS_FIRMWARE_ERROR_QUERY = 2, + AMDGPU_RAS_VIRT_ERROR_COUNT_QUERY = 3, }; /* ras error status reisger fields */ @@ -736,8 +737,8 @@ struct amdgpu_ras_block_hw_ops { * 8: feature disable */ - -int amdgpu_ras_recovery_init(struct amdgpu_device *adev); +int amdgpu_ras_init_badpage_info(struct amdgpu_device *adev); +int amdgpu_ras_recovery_init(struct amdgpu_device *adev, bool init_bp_info); void amdgpu_ras_resume(struct amdgpu_device *adev); void amdgpu_ras_suspend(struct amdgpu_device *adev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c index 66c1a868c0e1..24dae7cdbe95 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c @@ -26,6 +26,156 @@ #include "sienna_cichlid.h" #include "smu_v13_0_10.h" +static int amdgpu_reset_xgmi_reset_on_init_suspend(struct amdgpu_device *adev) +{ + int i; + + for (i = adev->num_ip_blocks - 1; i >= 0; i--) { + if (!adev->ip_blocks[i].status.valid) + continue; + if (!adev->ip_blocks[i].status.hw) + continue; + /* displays are handled in phase1 */ + if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_DCE) + continue; + + /* XXX handle errors */ + amdgpu_ip_block_suspend(&adev->ip_blocks[i]); + adev->ip_blocks[i].status.hw = false; + } + + /* VCN FW shared region is in frambuffer, there are some flags + * initialized in that region during sw_init. Make sure the region is + * backed up. + */ + amdgpu_vcn_save_vcpu_bo(adev); + + return 0; +} + +static int amdgpu_reset_xgmi_reset_on_init_prep_hwctxt( + struct amdgpu_reset_control *reset_ctl, + struct amdgpu_reset_context *reset_context) +{ + struct list_head *reset_device_list = reset_context->reset_device_list; + struct amdgpu_device *tmp_adev; + int r; + + list_for_each_entry(tmp_adev, reset_device_list, reset_list) { + amdgpu_unregister_gpu_instance(tmp_adev); + r = amdgpu_reset_xgmi_reset_on_init_suspend(tmp_adev); + if (r) { + dev_err(tmp_adev->dev, + "xgmi reset on init: prepare for reset failed"); + return r; + } + } + + return r; +} + +static int amdgpu_reset_xgmi_reset_on_init_restore_hwctxt( + struct amdgpu_reset_control *reset_ctl, + struct amdgpu_reset_context *reset_context) +{ + struct list_head *reset_device_list = reset_context->reset_device_list; + struct amdgpu_device *tmp_adev = NULL; + int r; + + r = amdgpu_device_reinit_after_reset(reset_context); + if (r) + return r; + list_for_each_entry(tmp_adev, reset_device_list, reset_list) { + if (!tmp_adev->kfd.init_complete) { + kgd2kfd_init_zone_device(tmp_adev); + amdgpu_amdkfd_device_init(tmp_adev); + amdgpu_amdkfd_drm_client_create(tmp_adev); + } + } + + return r; +} + +static int amdgpu_reset_xgmi_reset_on_init_perform_reset( + struct amdgpu_reset_control *reset_ctl, + struct amdgpu_reset_context *reset_context) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)reset_ctl->handle; + struct list_head *reset_device_list = reset_context->reset_device_list; + struct amdgpu_device *tmp_adev = NULL; + int r; + + dev_dbg(adev->dev, "xgmi roi - hw reset\n"); + + list_for_each_entry(tmp_adev, reset_device_list, reset_list) { + mutex_lock(&tmp_adev->reset_cntl->reset_lock); + tmp_adev->reset_cntl->active_reset = + amdgpu_asic_reset_method(adev); + } + r = 0; + /* Mode1 reset needs to be triggered on all devices together */ + list_for_each_entry(tmp_adev, reset_device_list, reset_list) { + /* For XGMI run all resets in parallel to speed up the process */ + if (!queue_work(system_unbound_wq, &tmp_adev->xgmi_reset_work)) + r = -EALREADY; + if (r) { + dev_err(tmp_adev->dev, + "xgmi reset on init: reset failed with error, %d", + r); + break; + } + } + + /* For XGMI wait for all resets to complete before proceed */ + if (!r) { + list_for_each_entry(tmp_adev, reset_device_list, reset_list) { + flush_work(&tmp_adev->xgmi_reset_work); + r = tmp_adev->asic_reset_res; + if (r) + break; + } + } + + list_for_each_entry(tmp_adev, reset_device_list, reset_list) { + mutex_unlock(&tmp_adev->reset_cntl->reset_lock); + tmp_adev->reset_cntl->active_reset = AMD_RESET_METHOD_NONE; + } + + return r; +} + +int amdgpu_reset_do_xgmi_reset_on_init( + struct amdgpu_reset_context *reset_context) +{ + struct list_head *reset_device_list = reset_context->reset_device_list; + struct amdgpu_device *adev; + int r; + + if (!reset_device_list || list_empty(reset_device_list) || + list_is_singular(reset_device_list)) + return -EINVAL; + + adev = list_first_entry(reset_device_list, struct amdgpu_device, + reset_list); + r = amdgpu_reset_prepare_hwcontext(adev, reset_context); + if (r) + return r; + + r = amdgpu_reset_perform_reset(adev, reset_context); + + return r; +} + +struct amdgpu_reset_handler xgmi_reset_on_init_handler = { + .reset_method = AMD_RESET_METHOD_ON_INIT, + .prepare_env = NULL, + .prepare_hwcontext = amdgpu_reset_xgmi_reset_on_init_prep_hwctxt, + .perform_reset = amdgpu_reset_xgmi_reset_on_init_perform_reset, + .restore_hwcontext = amdgpu_reset_xgmi_reset_on_init_restore_hwctxt, + .restore_env = NULL, + .do_reset = NULL, +}; + int amdgpu_reset_init(struct amdgpu_device *adev) { int ret = 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h index 1cb920abc2fe..f8628bc898df 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h @@ -153,4 +153,9 @@ void amdgpu_reset_get_desc(struct amdgpu_reset_context *rst_ctxt, char *buf, for (i = 0; (i < AMDGPU_RESET_MAX_HANDLERS) && \ (handler = (*reset_ctl->reset_handlers)[i]); \ ++i) + +extern struct amdgpu_reset_handler xgmi_reset_on_init_handler; +int amdgpu_reset_do_xgmi_reset_on_init( + struct amdgpu_reset_context *reset_context); + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c index 690976665cf6..a6e28fe3f8d6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c @@ -108,10 +108,22 @@ int amdgpu_ring_alloc(struct amdgpu_ring *ring, unsigned int ndw) */ void amdgpu_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count) { - int i; + uint32_t occupied, chunk1, chunk2; - for (i = 0; i < count; i++) - amdgpu_ring_write(ring, ring->funcs->nop); + occupied = ring->wptr & ring->buf_mask; + chunk1 = ring->buf_mask + 1 - occupied; + chunk1 = (chunk1 >= count) ? count : chunk1; + chunk2 = count - chunk1; + + if (chunk1) + memset32(&ring->ring[occupied], ring->funcs->nop, chunk1); + + if (chunk2) + memset32(ring->ring, ring->funcs->nop, chunk2); + + ring->wptr += count; + ring->wptr &= ring->ptr_mask; + ring->count_dw -= count; } /** @@ -141,6 +153,9 @@ void amdgpu_ring_commit(struct amdgpu_ring *ring) { uint32_t count; + if (ring->count_dw < 0) + DRM_ERROR("amdgpu: writing more dwords to the ring than expected!\n"); + /* We pad to match fetch size */ count = ring->funcs->align_mask + 1 - (ring->wptr & ring->funcs->align_mask); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h index f93f51002201..36fc9578c53c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h @@ -246,7 +246,7 @@ struct amdgpu_ring { struct drm_gpu_scheduler sched; struct amdgpu_bo *ring_obj; - volatile uint32_t *ring; + uint32_t *ring; unsigned rptr_offs; u64 rptr_gpu_addr; volatile u32 *rptr_cpu_addr; @@ -288,7 +288,7 @@ struct amdgpu_ring { u64 cond_exe_gpu_addr; volatile u32 *cond_exe_cpu_addr; unsigned int set_q_mode_offs; - volatile u32 *set_q_mode_ptr; + u32 *set_q_mode_ptr; u64 set_q_mode_token; unsigned vm_hub; unsigned vm_inv_eng; @@ -377,8 +377,6 @@ static inline void amdgpu_ring_clear_ring(struct amdgpu_ring *ring) static inline void amdgpu_ring_write(struct amdgpu_ring *ring, uint32_t v) { - if (ring->count_dw <= 0) - DRM_ERROR("amdgpu: writing more dwords to the ring than expected!\n"); ring->ring[ring->wptr++ & ring->buf_mask] = v; ring->wptr &= ring->ptr_mask; ring->count_dw--; @@ -388,13 +386,8 @@ static inline void amdgpu_ring_write_multiple(struct amdgpu_ring *ring, void *src, int count_dw) { unsigned occupied, chunk1, chunk2; - void *dst; - - if (unlikely(ring->count_dw < count_dw)) - DRM_ERROR("amdgpu: writing more dwords to the ring than expected!\n"); occupied = ring->wptr & ring->buf_mask; - dst = (void *)&ring->ring[occupied]; chunk1 = ring->buf_mask + 1 - occupied; chunk1 = (chunk1 >= count_dw) ? count_dw : chunk1; chunk2 = count_dw - chunk1; @@ -402,12 +395,11 @@ static inline void amdgpu_ring_write_multiple(struct amdgpu_ring *ring, chunk2 <<= 2; if (chunk1) - memcpy(dst, src, chunk1); + memcpy(&ring->ring[occupied], src, chunk1); if (chunk2) { src += chunk1; - dst = (void *)ring->ring; - memcpy(dst, src, chunk2); + memcpy(ring->ring, src, chunk2); } ring->wptr += count_dw; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.c index 183a976ba29d..8c89b69edc20 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.c @@ -343,3 +343,114 @@ int amdgpu_sdma_ras_sw_init(struct amdgpu_device *adev) return 0; } + +/* + * debugfs for to enable/disable sdma job submission to specific core. + */ +#if defined(CONFIG_DEBUG_FS) +static int amdgpu_debugfs_sdma_sched_mask_set(void *data, u64 val) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)data; + u32 i; + u64 mask = 0; + struct amdgpu_ring *ring; + + if (!adev) + return -ENODEV; + + mask = (1 << adev->sdma.num_instances) - 1; + if ((val & mask) == 0) + return -EINVAL; + + for (i = 0; i < adev->sdma.num_instances; ++i) { + ring = &adev->sdma.instance[i].ring; + if (val & (1 << i)) + ring->sched.ready = true; + else + ring->sched.ready = false; + } + /* publish sched.ready flag update effective immediately across smp */ + smp_rmb(); + return 0; +} + +static int amdgpu_debugfs_sdma_sched_mask_get(void *data, u64 *val) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)data; + u32 i; + u64 mask = 0; + struct amdgpu_ring *ring; + + if (!adev) + return -ENODEV; + for (i = 0; i < adev->sdma.num_instances; ++i) { + ring = &adev->sdma.instance[i].ring; + if (ring->sched.ready) + mask |= 1 << i; + } + + *val = mask; + return 0; +} + +DEFINE_DEBUGFS_ATTRIBUTE(amdgpu_debugfs_sdma_sched_mask_fops, + amdgpu_debugfs_sdma_sched_mask_get, + amdgpu_debugfs_sdma_sched_mask_set, "%llx\n"); + +#endif + +void amdgpu_debugfs_sdma_sched_mask_init(struct amdgpu_device *adev) +{ +#if defined(CONFIG_DEBUG_FS) + struct drm_minor *minor = adev_to_drm(adev)->primary; + struct dentry *root = minor->debugfs_root; + char name[32]; + + if (!(adev->sdma.num_instances > 1)) + return; + sprintf(name, "amdgpu_sdma_sched_mask"); + debugfs_create_file(name, 0600, root, adev, + &amdgpu_debugfs_sdma_sched_mask_fops); +#endif +} + +static ssize_t amdgpu_get_sdma_reset_mask(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = drm_to_adev(ddev); + + if (!adev) + return -ENODEV; + + return amdgpu_show_reset_mask(buf, adev->sdma.supported_reset); +} + +static DEVICE_ATTR(sdma_reset_mask, 0444, + amdgpu_get_sdma_reset_mask, NULL); + +int amdgpu_sdma_sysfs_reset_mask_init(struct amdgpu_device *adev) +{ + int r = 0; + + if (!amdgpu_gpu_recovery) + return r; + + if (adev->sdma.num_instances) { + r = device_create_file(adev->dev, &dev_attr_sdma_reset_mask); + if (r) + return r; + } + + return r; +} + +void amdgpu_sdma_sysfs_reset_mask_fini(struct amdgpu_device *adev) +{ + if (!amdgpu_gpu_recovery) + return; + + if (adev->sdma.num_instances) + device_remove_file(adev->dev, &dev_attr_sdma_reset_mask); +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h index 087ce0f6fa07..2db58b5812a8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h @@ -116,6 +116,7 @@ struct amdgpu_sdma { struct ras_common_if *ras_if; struct amdgpu_sdma_ras *ras; uint32_t *ip_dump; + uint32_t supported_reset; }; /* @@ -175,5 +176,7 @@ int amdgpu_sdma_init_microcode(struct amdgpu_device *adev, u32 instance, void amdgpu_sdma_destroy_inst_ctx(struct amdgpu_device *adev, bool duplicate); int amdgpu_sdma_ras_sw_init(struct amdgpu_device *adev); - +void amdgpu_debugfs_sdma_sched_mask_init(struct amdgpu_device *adev); +int amdgpu_sdma_sysfs_reset_mask_init(struct amdgpu_device *adev); +void amdgpu_sdma_sysfs_reset_mask_fini(struct amdgpu_device *adev); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index 74adb983ab03..9f922ec50ea2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -812,7 +812,7 @@ static int amdgpu_ttm_tt_pin_userptr(struct ttm_device *bdev, /* Map SG to device */ r = dma_map_sgtable(adev->dev, ttm->sg, direction, 0); if (r) - goto release_sg; + goto release_sg_table; /* convert SG to linear array of pages and dma addresses */ drm_prime_sg_to_dma_addr_array(ttm->sg, gtt->ttm.dma_address, @@ -820,6 +820,8 @@ static int amdgpu_ttm_tt_pin_userptr(struct ttm_device *bdev, return 0; +release_sg_table: + sg_free_table(ttm->sg); release_sg: kfree(ttm->sg); ttm->sg = NULL; @@ -1849,6 +1851,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev) mutex_init(&adev->mman.gtt_window_lock); + dma_set_max_seg_size(adev->dev, UINT_MAX); /* No others user of address space so set it to 0 */ r = ttm_device_init(&adev->mman.bdev, &amdgpu_bo_driver, adev->dev, adev_to_drm(adev)->anon_inode->i_mapping, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h index 138d80017f35..2852a6064c9a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h @@ -34,6 +34,7 @@ #define AMDGPU_PL_OA (TTM_PL_PRIV + 2) #define AMDGPU_PL_PREEMPT (TTM_PL_PRIV + 3) #define AMDGPU_PL_DOORBELL (TTM_PL_PRIV + 4) +#define __AMDGPU_PL_LAST (TTM_PL_PRIV + 4) #define AMDGPU_GTT_MAX_TRANSFER_SIZE 512 #define AMDGPU_GTT_NUM_TRANSFER_WINDOWS 2 diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h index 4e23419b92d4..4150ec0aa10d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h @@ -163,6 +163,7 @@ enum ta_fw_type { TA_FW_TYPE_PSP_DTM, TA_FW_TYPE_PSP_RAP, TA_FW_TYPE_PSP_SECUREDISPLAY, + TA_FW_TYPE_PSP_XGMI_AUX, TA_FW_TYPE_MAX_INDEX, }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c index bb7b9b2eaac1..896f3609b0ee 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c @@ -318,6 +318,9 @@ int amdgpu_umc_ras_late_init(struct amdgpu_device *adev, struct ras_common_if *r if (r) return r; + if (amdgpu_sriov_vf(adev)) + return r; + if (amdgpu_ras_is_supported(adev, ras_block->block)) { r = amdgpu_irq_get(adev, &adev->gmc.ecc_irq, 0); if (r) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umsch_mm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_umsch_mm.c index 6162582d0aa2..bd2d3863c3ed 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umsch_mm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umsch_mm.c @@ -765,9 +765,9 @@ static int umsch_mm_init(struct amdgpu_device *adev) } -static int umsch_mm_early_init(void *handle) +static int umsch_mm_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; switch (amdgpu_ip_version(adev, VCN_HWIP, 0)) { case IP_VERSION(4, 0, 5): @@ -784,9 +784,9 @@ static int umsch_mm_early_init(void *handle) return 0; } -static int umsch_mm_late_init(void *handle) +static int umsch_mm_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_in_reset(adev) || adev->in_s0ix || adev->in_suspend) return 0; @@ -794,9 +794,9 @@ static int umsch_mm_late_init(void *handle) return umsch_mm_test(adev); } -static int umsch_mm_sw_init(void *handle) +static int umsch_mm_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = umsch_mm_init(adev); @@ -815,9 +815,9 @@ static int umsch_mm_sw_init(void *handle) return 0; } -static int umsch_mm_sw_fini(void *handle) +static int umsch_mm_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; release_firmware(adev->umsch_mm.fw); adev->umsch_mm.fw = NULL; @@ -839,9 +839,9 @@ static int umsch_mm_sw_fini(void *handle) return 0; } -static int umsch_mm_hw_init(void *handle) +static int umsch_mm_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = umsch_mm_load_microcode(&adev->umsch_mm); @@ -857,9 +857,9 @@ static int umsch_mm_hw_init(void *handle) return 0; } -static int umsch_mm_hw_fini(void *handle) +static int umsch_mm_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; umsch_mm_ring_stop(&adev->umsch_mm); @@ -873,18 +873,14 @@ static int umsch_mm_hw_fini(void *handle) return 0; } -static int umsch_mm_suspend(void *handle) +static int umsch_mm_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return umsch_mm_hw_fini(adev); + return umsch_mm_hw_fini(ip_block); } -static int umsch_mm_resume(void *handle) +static int umsch_mm_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return umsch_mm_hw_init(adev); + return umsch_mm_hw_init(ip_block); } void amdgpu_umsch_fwlog_init(struct amdgpu_umsch_mm *umsch_mm) @@ -997,8 +993,6 @@ static const struct amd_ip_funcs umsch_mm_v4_0_ip_funcs = { .hw_fini = umsch_mm_hw_fini, .suspend = umsch_mm_suspend, .resume = umsch_mm_resume, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; const struct amdgpu_ip_block_version umsch_mm_v4_0_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c index 43f44cc201cb..aecb78e0519f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c @@ -294,21 +294,12 @@ bool amdgpu_vcn_is_disabled_vcn(struct amdgpu_device *adev, enum vcn_ring_type t return ret; } -int amdgpu_vcn_suspend(struct amdgpu_device *adev) +int amdgpu_vcn_save_vcpu_bo(struct amdgpu_device *adev) { unsigned int size; void *ptr; int i, idx; - bool in_ras_intr = amdgpu_ras_intr_triggered(); - - cancel_delayed_work_sync(&adev->vcn.idle_work); - - /* err_event_athub will corrupt VCPU buffer, so we need to - * restore fw data and clear buffer in amdgpu_vcn_resume() */ - if (in_ras_intr) - return 0; - for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { if (adev->vcn.harvest_config & (1 << i)) continue; @@ -327,9 +318,24 @@ int amdgpu_vcn_suspend(struct amdgpu_device *adev) drm_dev_exit(idx); } } + return 0; } +int amdgpu_vcn_suspend(struct amdgpu_device *adev) +{ + bool in_ras_intr = amdgpu_ras_intr_triggered(); + + cancel_delayed_work_sync(&adev->vcn.idle_work); + + /* err_event_athub will corrupt VCPU buffer, so we need to + * restore fw data and clear buffer in amdgpu_vcn_resume() */ + if (in_ras_intr) + return 0; + + return amdgpu_vcn_save_vcpu_bo(adev); +} + int amdgpu_vcn_resume(struct amdgpu_device *adev) { unsigned int size; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h index 2a1f3dbb14d3..765b809d48a2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h @@ -518,5 +518,6 @@ int amdgpu_vcn_ras_sw_init(struct amdgpu_device *adev); int amdgpu_vcn_psp_update_sram(struct amdgpu_device *adev, int inst_idx, enum AMDGPU_UCODE_ID ucode_id); +int amdgpu_vcn_save_vcpu_bo(struct amdgpu_device *adev); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index b6397d3229e1..c704e9803e11 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -523,6 +523,9 @@ static int amdgpu_virt_read_pf2vf_data(struct amdgpu_device *adev) adev->unique_id = ((struct amd_sriov_msg_pf2vf_info *)pf2vf_info)->uuid; + adev->virt.ras_en_caps.all = ((struct amd_sriov_msg_pf2vf_info *)pf2vf_info)->ras_en_caps.all; + adev->virt.ras_telemetry_en_caps.all = + ((struct amd_sriov_msg_pf2vf_info *)pf2vf_info)->ras_telemetry_en_caps.all; break; default: dev_err(adev->dev, "invalid pf2vf version: 0x%x\n", pf2vf_info->version); @@ -703,6 +706,8 @@ void amdgpu_virt_exchange_data(struct amdgpu_device *adev) adev->virt.fw_reserve.p_vf2pf = (struct amd_sriov_msg_vf2pf_info_header *) (adev->mman.fw_vram_usage_va + (AMD_SRIOV_MSG_VF2PF_OFFSET_KB << 10)); + adev->virt.fw_reserve.ras_telemetry = + (adev->mman.fw_vram_usage_va + (AMD_SRIOV_MSG_RAS_TELEMETRY_OFFSET_KB << 10)); } else if (adev->mman.drv_vram_usage_va) { adev->virt.fw_reserve.p_pf2vf = (struct amd_sriov_msg_pf2vf_info_header *) @@ -710,6 +715,8 @@ void amdgpu_virt_exchange_data(struct amdgpu_device *adev) adev->virt.fw_reserve.p_vf2pf = (struct amd_sriov_msg_vf2pf_info_header *) (adev->mman.drv_vram_usage_va + (AMD_SRIOV_MSG_VF2PF_OFFSET_KB << 10)); + adev->virt.fw_reserve.ras_telemetry = + (adev->mman.drv_vram_usage_va + (AMD_SRIOV_MSG_RAS_TELEMETRY_OFFSET_KB << 10)); } amdgpu_virt_read_pf2vf_data(adev); @@ -1144,3 +1151,185 @@ bool amdgpu_sriov_xnack_support(struct amdgpu_device *adev) return xnack_mode; } + +bool amdgpu_virt_get_ras_capability(struct amdgpu_device *adev) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + + if (!amdgpu_sriov_ras_caps_en(adev)) + return false; + + if (adev->virt.ras_en_caps.bits.block_umc) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__UMC); + if (adev->virt.ras_en_caps.bits.block_sdma) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__SDMA); + if (adev->virt.ras_en_caps.bits.block_gfx) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__GFX); + if (adev->virt.ras_en_caps.bits.block_mmhub) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__MMHUB); + if (adev->virt.ras_en_caps.bits.block_athub) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__ATHUB); + if (adev->virt.ras_en_caps.bits.block_pcie_bif) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__PCIE_BIF); + if (adev->virt.ras_en_caps.bits.block_hdp) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__HDP); + if (adev->virt.ras_en_caps.bits.block_xgmi_wafl) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__XGMI_WAFL); + if (adev->virt.ras_en_caps.bits.block_df) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__DF); + if (adev->virt.ras_en_caps.bits.block_smn) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__SMN); + if (adev->virt.ras_en_caps.bits.block_sem) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__SEM); + if (adev->virt.ras_en_caps.bits.block_mp0) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__MP0); + if (adev->virt.ras_en_caps.bits.block_mp1) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__MP1); + if (adev->virt.ras_en_caps.bits.block_fuse) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__FUSE); + if (adev->virt.ras_en_caps.bits.block_mca) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__MCA); + if (adev->virt.ras_en_caps.bits.block_vcn) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__VCN); + if (adev->virt.ras_en_caps.bits.block_jpeg) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__JPEG); + if (adev->virt.ras_en_caps.bits.block_ih) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__IH); + if (adev->virt.ras_en_caps.bits.block_mpio) + adev->ras_hw_enabled |= BIT(AMDGPU_RAS_BLOCK__MPIO); + + if (adev->virt.ras_en_caps.bits.poison_propogation_mode) + con->poison_supported = true; /* Poison is handled by host */ + + return true; +} + +static inline enum amd_sriov_ras_telemetry_gpu_block +amdgpu_ras_block_to_sriov(struct amdgpu_device *adev, enum amdgpu_ras_block block) { + switch (block) { + case AMDGPU_RAS_BLOCK__UMC: + return RAS_TELEMETRY_GPU_BLOCK_UMC; + case AMDGPU_RAS_BLOCK__SDMA: + return RAS_TELEMETRY_GPU_BLOCK_SDMA; + case AMDGPU_RAS_BLOCK__GFX: + return RAS_TELEMETRY_GPU_BLOCK_GFX; + case AMDGPU_RAS_BLOCK__MMHUB: + return RAS_TELEMETRY_GPU_BLOCK_MMHUB; + case AMDGPU_RAS_BLOCK__ATHUB: + return RAS_TELEMETRY_GPU_BLOCK_ATHUB; + case AMDGPU_RAS_BLOCK__PCIE_BIF: + return RAS_TELEMETRY_GPU_BLOCK_PCIE_BIF; + case AMDGPU_RAS_BLOCK__HDP: + return RAS_TELEMETRY_GPU_BLOCK_HDP; + case AMDGPU_RAS_BLOCK__XGMI_WAFL: + return RAS_TELEMETRY_GPU_BLOCK_XGMI_WAFL; + case AMDGPU_RAS_BLOCK__DF: + return RAS_TELEMETRY_GPU_BLOCK_DF; + case AMDGPU_RAS_BLOCK__SMN: + return RAS_TELEMETRY_GPU_BLOCK_SMN; + case AMDGPU_RAS_BLOCK__SEM: + return RAS_TELEMETRY_GPU_BLOCK_SEM; + case AMDGPU_RAS_BLOCK__MP0: + return RAS_TELEMETRY_GPU_BLOCK_MP0; + case AMDGPU_RAS_BLOCK__MP1: + return RAS_TELEMETRY_GPU_BLOCK_MP1; + case AMDGPU_RAS_BLOCK__FUSE: + return RAS_TELEMETRY_GPU_BLOCK_FUSE; + case AMDGPU_RAS_BLOCK__MCA: + return RAS_TELEMETRY_GPU_BLOCK_MCA; + case AMDGPU_RAS_BLOCK__VCN: + return RAS_TELEMETRY_GPU_BLOCK_VCN; + case AMDGPU_RAS_BLOCK__JPEG: + return RAS_TELEMETRY_GPU_BLOCK_JPEG; + case AMDGPU_RAS_BLOCK__IH: + return RAS_TELEMETRY_GPU_BLOCK_IH; + case AMDGPU_RAS_BLOCK__MPIO: + return RAS_TELEMETRY_GPU_BLOCK_MPIO; + default: + dev_err(adev->dev, "Unsupported SRIOV RAS telemetry block 0x%x\n", block); + return RAS_TELEMETRY_GPU_BLOCK_COUNT; + } +} + +static int amdgpu_virt_cache_host_error_counts(struct amdgpu_device *adev, + struct amdsriov_ras_telemetry *host_telemetry) +{ + struct amd_sriov_ras_telemetry_error_count *tmp = NULL; + uint32_t checksum, used_size; + + checksum = host_telemetry->header.checksum; + used_size = host_telemetry->header.used_size; + + if (used_size > (AMD_SRIOV_RAS_TELEMETRY_SIZE_KB << 10)) + return 0; + + tmp = kmalloc(used_size, GFP_KERNEL); + if (!tmp) + return -ENOMEM; + + memcpy(tmp, &host_telemetry->body.error_count, used_size); + + if (checksum != amd_sriov_msg_checksum(tmp, used_size, 0, 0)) + goto out; + + memcpy(&adev->virt.count_cache, tmp, + min(used_size, sizeof(adev->virt.count_cache))); +out: + kfree(tmp); + + return 0; +} + +static int amdgpu_virt_req_ras_err_count_internal(struct amdgpu_device *adev, bool force_update) +{ + struct amdgpu_virt *virt = &adev->virt; + + /* Host allows 15 ras telemetry requests per 60 seconds. Afterwhich, the Host + * will ignore incoming guest messages. Ratelimit the guest messages to + * prevent guest self DOS. + */ + if (__ratelimit(&adev->virt.ras_telemetry_rs) || force_update) { + if (!virt->ops->req_ras_err_count(adev)) + amdgpu_virt_cache_host_error_counts(adev, + adev->virt.fw_reserve.ras_telemetry); + } + + return 0; +} + +/* Bypass ACA interface and query ECC counts directly from host */ +int amdgpu_virt_req_ras_err_count(struct amdgpu_device *adev, enum amdgpu_ras_block block, + struct ras_err_data *err_data) +{ + enum amd_sriov_ras_telemetry_gpu_block sriov_block; + + sriov_block = amdgpu_ras_block_to_sriov(adev, block); + + if (sriov_block >= RAS_TELEMETRY_GPU_BLOCK_COUNT || + !amdgpu_sriov_ras_telemetry_block_en(adev, sriov_block)) + return -EOPNOTSUPP; + + /* Host Access may be lost during reset, just return last cached data. */ + if (down_read_trylock(&adev->reset_domain->sem)) { + amdgpu_virt_req_ras_err_count_internal(adev, false); + up_read(&adev->reset_domain->sem); + } + + err_data->ue_count = adev->virt.count_cache.block[sriov_block].ue_count; + err_data->ce_count = adev->virt.count_cache.block[sriov_block].ce_count; + err_data->de_count = adev->virt.count_cache.block[sriov_block].de_count; + + return 0; +} + +int amdgpu_virt_ras_telemetry_post_reset(struct amdgpu_device *adev) +{ + unsigned long ue_count, ce_count; + + if (amdgpu_sriov_ras_telemetry_en(adev)) { + amdgpu_virt_req_ras_err_count_internal(adev, true); + amdgpu_ras_query_error_count(adev, &ce_count, &ue_count, NULL); + } + + return 0; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h index b650a2032c42..5381b8d596e6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h @@ -95,6 +95,7 @@ struct amdgpu_virt_ops { void (*ras_poison_handler)(struct amdgpu_device *adev, enum amdgpu_ras_block block); bool (*rcvd_ras_intr)(struct amdgpu_device *adev); + int (*req_ras_err_count)(struct amdgpu_device *adev); }; /* @@ -103,6 +104,7 @@ struct amdgpu_virt_ops { struct amdgpu_virt_fw_reserve { struct amd_sriov_msg_pf2vf_info_header *p_pf2vf; struct amd_sriov_msg_vf2pf_info_header *p_vf2pf; + void *ras_telemetry; unsigned int checksum_key; }; @@ -136,6 +138,8 @@ enum AMDGIM_FEATURE_FLAG { AMDGIM_FEATURE_VCN_RB_DECOUPLE = (1 << 7), /* MES info */ AMDGIM_FEATURE_MES_INFO_ENABLE = (1 << 8), + AMDGIM_FEATURE_RAS_CAPS = (1 << 9), + AMDGIM_FEATURE_RAS_TELEMETRY = (1 << 10), }; enum AMDGIM_REG_ACCESS_FLAG { @@ -276,6 +280,12 @@ struct amdgpu_virt { uint32_t autoload_ucode_id; struct mutex rlcg_reg_lock; + + union amd_sriov_ras_caps ras_en_caps; + union amd_sriov_ras_caps ras_telemetry_en_caps; + + struct ratelimit_state ras_telemetry_rs; + struct amd_sriov_ras_telemetry_error_count count_cache; }; struct amdgpu_video_codec_info; @@ -320,6 +330,15 @@ struct amdgpu_video_codec_info; #define amdgpu_sriov_vf_mmio_access_protection(adev) \ ((adev)->virt.caps & AMDGPU_VF_MMIO_ACCESS_PROTECT) +#define amdgpu_sriov_ras_caps_en(adev) \ +((adev)->virt.gim_feature & AMDGIM_FEATURE_RAS_CAPS) + +#define amdgpu_sriov_ras_telemetry_en(adev) \ +(((adev)->virt.gim_feature & AMDGIM_FEATURE_RAS_TELEMETRY) && (adev)->virt.fw_reserve.ras_telemetry) + +#define amdgpu_sriov_ras_telemetry_block_en(adev, sriov_blk) \ +(amdgpu_sriov_ras_telemetry_en((adev)) && (adev)->virt.ras_telemetry_en_caps.all & BIT(sriov_blk)) + static inline bool is_virtual_machine(void) { #if defined(CONFIG_X86) @@ -383,4 +402,8 @@ bool amdgpu_virt_get_rlcg_reg_access_flag(struct amdgpu_device *adev, u32 acc_flags, u32 hwip, bool write, u32 *rlcg_flag); u32 amdgpu_virt_rlcg_reg_rw(struct amdgpu_device *adev, u32 offset, u32 v, u32 flag, u32 xcc_id); +bool amdgpu_virt_get_ras_capability(struct amdgpu_device *adev); +int amdgpu_virt_req_ras_err_count(struct amdgpu_device *adev, enum amdgpu_ras_block block, + struct ras_err_data *err_data); +int amdgpu_virt_ras_telemetry_post_reset(struct amdgpu_device *adev); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c index d4c2afafbb73..8bf28d336807 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c @@ -493,10 +493,10 @@ const struct drm_mode_config_funcs amdgpu_vkms_mode_funcs = { .atomic_commit = drm_atomic_helper_commit, }; -static int amdgpu_vkms_sw_init(void *handle) +static int amdgpu_vkms_sw_init(struct amdgpu_ip_block *ip_block) { int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->amdgpu_vkms_output = kcalloc(adev->mode_info.num_crtc, sizeof(struct amdgpu_vkms_output), GFP_KERNEL); @@ -536,9 +536,9 @@ static int amdgpu_vkms_sw_init(void *handle) return 0; } -static int amdgpu_vkms_sw_fini(void *handle) +static int amdgpu_vkms_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i = 0; for (i = 0; i < adev->mode_info.num_crtc; i++) @@ -555,9 +555,9 @@ static int amdgpu_vkms_sw_fini(void *handle) return 0; } -static int amdgpu_vkms_hw_init(void *handle) +static int amdgpu_vkms_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; switch (adev->asic_type) { #ifdef CONFIG_DRM_AMDGPU_SI @@ -600,31 +600,31 @@ static int amdgpu_vkms_hw_init(void *handle) return 0; } -static int amdgpu_vkms_hw_fini(void *handle) +static int amdgpu_vkms_hw_fini(struct amdgpu_ip_block *ip_block) { return 0; } -static int amdgpu_vkms_suspend(void *handle) +static int amdgpu_vkms_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = drm_mode_config_helper_suspend(adev_to_drm(adev)); if (r) return r; - return amdgpu_vkms_hw_fini(handle); + + return 0; } -static int amdgpu_vkms_resume(void *handle) +static int amdgpu_vkms_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = amdgpu_vkms_hw_init(handle); + r = amdgpu_vkms_hw_init(ip_block); if (r) return r; - return drm_mode_config_helper_resume(adev_to_drm(adev)); + return drm_mode_config_helper_resume(adev_to_drm(ip_block->adev)); } static bool amdgpu_vkms_is_idle(void *handle) @@ -632,16 +632,6 @@ static bool amdgpu_vkms_is_idle(void *handle) return true; } -static int amdgpu_vkms_wait_for_idle(void *handle) -{ - return 0; -} - -static int amdgpu_vkms_soft_reset(void *handle) -{ - return 0; -} - static int amdgpu_vkms_set_clockgating_state(void *handle, enum amd_clockgating_state state) { @@ -656,8 +646,6 @@ static int amdgpu_vkms_set_powergating_state(void *handle, static const struct amd_ip_funcs amdgpu_vkms_ip_funcs = { .name = "amdgpu_vkms", - .early_init = NULL, - .late_init = NULL, .sw_init = amdgpu_vkms_sw_init, .sw_fini = amdgpu_vkms_sw_fini, .hw_init = amdgpu_vkms_hw_init, @@ -665,12 +653,8 @@ static const struct amd_ip_funcs amdgpu_vkms_ip_funcs = { .suspend = amdgpu_vkms_suspend, .resume = amdgpu_vkms_resume, .is_idle = amdgpu_vkms_is_idle, - .wait_for_idle = amdgpu_vkms_wait_for_idle, - .soft_reset = amdgpu_vkms_soft_reset, .set_clockgating_state = amdgpu_vkms_set_clockgating_state, .set_powergating_state = amdgpu_vkms_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; const struct amdgpu_ip_block_version amdgpu_vkms_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 6005280f5f38..8d9bf7a0857f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -1083,7 +1083,8 @@ error_free: } static void amdgpu_vm_bo_get_memory(struct amdgpu_bo_va *bo_va, - struct amdgpu_mem_stats *stats) + struct amdgpu_mem_stats *stats, + unsigned int size) { struct amdgpu_vm *vm = bo_va->base.vm; struct amdgpu_bo *bo = bo_va->base.bo; @@ -1099,34 +1100,35 @@ static void amdgpu_vm_bo_get_memory(struct amdgpu_bo_va *bo_va, !dma_resv_trylock(bo->tbo.base.resv)) return; - amdgpu_bo_get_memory(bo, stats); + amdgpu_bo_get_memory(bo, stats, size); if (!amdgpu_vm_is_bo_always_valid(vm, bo)) dma_resv_unlock(bo->tbo.base.resv); } void amdgpu_vm_get_memory(struct amdgpu_vm *vm, - struct amdgpu_mem_stats *stats) + struct amdgpu_mem_stats *stats, + unsigned int size) { struct amdgpu_bo_va *bo_va, *tmp; spin_lock(&vm->status_lock); list_for_each_entry_safe(bo_va, tmp, &vm->idle, base.vm_status) - amdgpu_vm_bo_get_memory(bo_va, stats); + amdgpu_vm_bo_get_memory(bo_va, stats, size); list_for_each_entry_safe(bo_va, tmp, &vm->evicted, base.vm_status) - amdgpu_vm_bo_get_memory(bo_va, stats); + amdgpu_vm_bo_get_memory(bo_va, stats, size); list_for_each_entry_safe(bo_va, tmp, &vm->relocated, base.vm_status) - amdgpu_vm_bo_get_memory(bo_va, stats); + amdgpu_vm_bo_get_memory(bo_va, stats, size); list_for_each_entry_safe(bo_va, tmp, &vm->moved, base.vm_status) - amdgpu_vm_bo_get_memory(bo_va, stats); + amdgpu_vm_bo_get_memory(bo_va, stats, size); list_for_each_entry_safe(bo_va, tmp, &vm->invalidated, base.vm_status) - amdgpu_vm_bo_get_memory(bo_va, stats); + amdgpu_vm_bo_get_memory(bo_va, stats, size); list_for_each_entry_safe(bo_va, tmp, &vm->done, base.vm_status) - amdgpu_vm_bo_get_memory(bo_va, stats); + amdgpu_vm_bo_get_memory(bo_va, stats, size); spin_unlock(&vm->status_lock); } @@ -1159,7 +1161,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, int r; amdgpu_sync_create(&sync); - if (clear || !bo) { + if (clear) { mem = NULL; /* Implicitly sync to command submissions in the same VM before @@ -1174,6 +1176,10 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, if (r) goto error_free; } + } else if (!bo) { + mem = NULL; + + /* PRT map operations don't need to sync to anything. */ } else { struct drm_gem_object *obj = &bo->tbo.base; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index 52dd7cdfdc81..5d119ac26c4f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -42,7 +42,6 @@ struct amdgpu_bo_va; struct amdgpu_job; struct amdgpu_bo_list_entry; struct amdgpu_bo_vm; -struct amdgpu_mem_stats; /* * GPUVM handling @@ -322,6 +321,16 @@ struct amdgpu_vm_fault_info { unsigned int vmhub; }; +struct amdgpu_mem_stats { + struct drm_memory_stats drm; + + /* buffers that requested this placement */ + uint64_t requested; + /* buffers that requested this placement + * but are currently evicted */ + uint64_t evicted; +}; + struct amdgpu_vm { /* tree of virtual addresses mapped */ struct rb_root_cached va; @@ -567,7 +576,8 @@ void amdgpu_vm_set_task_info(struct amdgpu_vm *vm); void amdgpu_vm_move_to_lru_tail(struct amdgpu_device *adev, struct amdgpu_vm *vm); void amdgpu_vm_get_memory(struct amdgpu_vm *vm, - struct amdgpu_mem_stats *stats); + struct amdgpu_mem_stats *stats, + unsigned int size); int amdgpu_vm_pt_clear(struct amdgpu_device *adev, struct amdgpu_vm *vm, struct amdgpu_bo_vm *vmbo, bool immediate); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.c index 5acd20ff5979..3e6f9dfb61bb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.c @@ -295,9 +295,9 @@ int amdgpu_vpe_ring_fini(struct amdgpu_vpe *vpe) return 0; } -static int vpe_early_init(void *handle) +static int vpe_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_vpe *vpe = &adev->vpe; switch (amdgpu_ip_version(adev, VPE_HWIP, 0)) { @@ -356,9 +356,9 @@ static int vpe_common_init(struct amdgpu_vpe *vpe) return 0; } -static int vpe_sw_init(void *handle) +static int vpe_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_vpe *vpe = &adev->vpe; int ret; @@ -377,18 +377,26 @@ static int vpe_sw_init(void *handle) ret = vpe_init_microcode(vpe); if (ret) goto out; + + /* TODO: Add queue reset mask when FW fully supports it */ + adev->vpe.supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->vpe.ring); + ret = amdgpu_vpe_sysfs_reset_mask_init(adev); + if (ret) + goto out; out: return ret; } -static int vpe_sw_fini(void *handle) +static int vpe_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_vpe *vpe = &adev->vpe; release_firmware(vpe->fw); vpe->fw = NULL; + amdgpu_vpe_sysfs_reset_mask_fini(adev); vpe_ring_fini(vpe); amdgpu_bo_free_kernel(&adev->vpe.cmdbuf_obj, @@ -398,9 +406,9 @@ static int vpe_sw_fini(void *handle) return 0; } -static int vpe_hw_init(void *handle) +static int vpe_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_vpe *vpe = &adev->vpe; int ret; @@ -421,9 +429,9 @@ static int vpe_hw_init(void *handle) return 0; } -static int vpe_hw_fini(void *handle) +static int vpe_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_vpe *vpe = &adev->vpe; vpe_ring_stop(vpe); @@ -434,20 +442,18 @@ static int vpe_hw_fini(void *handle) return 0; } -static int vpe_suspend(void *handle) +static int vpe_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->vpe.idle_work); - return vpe_hw_fini(adev); + return vpe_hw_fini(ip_block); } -static int vpe_resume(void *handle) +static int vpe_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return vpe_hw_init(adev); + return vpe_hw_init(ip_block); } static void vpe_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count) @@ -867,6 +873,41 @@ static void vpe_ring_end_use(struct amdgpu_ring *ring) schedule_delayed_work(&adev->vpe.idle_work, VPE_IDLE_TIMEOUT); } +static ssize_t amdgpu_get_vpe_reset_mask(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = drm_to_adev(ddev); + + if (!adev) + return -ENODEV; + + return amdgpu_show_reset_mask(buf, adev->vpe.supported_reset); +} + +static DEVICE_ATTR(vpe_reset_mask, 0444, + amdgpu_get_vpe_reset_mask, NULL); + +int amdgpu_vpe_sysfs_reset_mask_init(struct amdgpu_device *adev) +{ + int r = 0; + + if (adev->vpe.num_instances) { + r = device_create_file(adev->dev, &dev_attr_vpe_reset_mask); + if (r) + return r; + } + + return r; +} + +void amdgpu_vpe_sysfs_reset_mask_fini(struct amdgpu_device *adev) +{ + if (adev->vpe.num_instances) + device_remove_file(adev->dev, &dev_attr_vpe_reset_mask); +} + static const struct amdgpu_ring_funcs vpe_ring_funcs = { .type = AMDGPU_RING_TYPE_VPE, .align_mask = 0xf, @@ -908,14 +949,12 @@ static void vpe_set_ring_funcs(struct amdgpu_device *adev) const struct amd_ip_funcs vpe_ip_funcs = { .name = "vpe_v6_1", .early_init = vpe_early_init, - .late_init = NULL, .sw_init = vpe_sw_init, .sw_fini = vpe_sw_fini, .hw_init = vpe_hw_init, .hw_fini = vpe_hw_fini, .suspend = vpe_suspend, .resume = vpe_resume, - .soft_reset = NULL, .set_clockgating_state = vpe_set_clockgating_state, .set_powergating_state = vpe_set_powergating_state, }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.h index 231d86d0953e..695da740a97e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.h @@ -79,6 +79,7 @@ struct amdgpu_vpe { uint32_t num_instances; bool collaborate_mode; + uint32_t supported_reset; }; int amdgpu_vpe_psp_update_sram(struct amdgpu_device *adev); @@ -86,6 +87,8 @@ int amdgpu_vpe_init_microcode(struct amdgpu_vpe *vpe); int amdgpu_vpe_ring_init(struct amdgpu_vpe *vpe); int amdgpu_vpe_ring_fini(struct amdgpu_vpe *vpe); int amdgpu_vpe_configure_dpm(struct amdgpu_vpe *vpe); +void amdgpu_vpe_sysfs_reset_mask_fini(struct amdgpu_device *adev); +int amdgpu_vpe_sysfs_reset_mask_init(struct amdgpu_device *adev); #define vpe_ring_init(vpe) ((vpe)->funcs->ring_init ? (vpe)->funcs->ring_init((vpe)) : 0) #define vpe_ring_start(vpe) ((vpe)->funcs->ring_start ? (vpe)->funcs->ring_start((vpe)) : 0) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xcp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_xcp.c index a6d456ec6aeb..e209b5e101df 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xcp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xcp.c @@ -433,3 +433,292 @@ void amdgpu_xcp_release_sched(struct amdgpu_device *adev, } } +#define XCP_CFG_SYSFS_RES_ATTR_SHOW(_name) \ + static ssize_t amdgpu_xcp_res_sysfs_##_name##_show( \ + struct amdgpu_xcp_res_details *xcp_res, char *buf) \ + { \ + return sysfs_emit(buf, "%d\n", xcp_res->_name); \ + } + +struct amdgpu_xcp_res_sysfs_attribute { + struct attribute attr; + ssize_t (*show)(struct amdgpu_xcp_res_details *xcp_res, char *buf); +}; + +#define XCP_CFG_SYSFS_RES_ATTR(_name) \ + struct amdgpu_xcp_res_sysfs_attribute xcp_res_sysfs_attr_##_name = { \ + .attr = { .name = __stringify(_name), .mode = 0400 }, \ + .show = amdgpu_xcp_res_sysfs_##_name##_show, \ + } + +XCP_CFG_SYSFS_RES_ATTR_SHOW(num_inst) +XCP_CFG_SYSFS_RES_ATTR(num_inst); +XCP_CFG_SYSFS_RES_ATTR_SHOW(num_shared) +XCP_CFG_SYSFS_RES_ATTR(num_shared); + +#define XCP_CFG_SYSFS_RES_ATTR_PTR(_name) xcp_res_sysfs_attr_##_name.attr + +static struct attribute *xcp_cfg_res_sysfs_attrs[] = { + &XCP_CFG_SYSFS_RES_ATTR_PTR(num_inst), + &XCP_CFG_SYSFS_RES_ATTR_PTR(num_shared), NULL +}; + +static const char *xcp_desc[] = { + [AMDGPU_SPX_PARTITION_MODE] = "SPX", + [AMDGPU_DPX_PARTITION_MODE] = "DPX", + [AMDGPU_TPX_PARTITION_MODE] = "TPX", + [AMDGPU_QPX_PARTITION_MODE] = "QPX", + [AMDGPU_CPX_PARTITION_MODE] = "CPX", +}; + +static const char *nps_desc[] = { + [UNKNOWN_MEMORY_PARTITION_MODE] = "UNKNOWN", + [AMDGPU_NPS1_PARTITION_MODE] = "NPS1", + [AMDGPU_NPS2_PARTITION_MODE] = "NPS2", + [AMDGPU_NPS3_PARTITION_MODE] = "NPS3", + [AMDGPU_NPS4_PARTITION_MODE] = "NPS4", + [AMDGPU_NPS6_PARTITION_MODE] = "NPS6", + [AMDGPU_NPS8_PARTITION_MODE] = "NPS8", +}; + +ATTRIBUTE_GROUPS(xcp_cfg_res_sysfs); + +#define to_xcp_attr(x) \ + container_of(x, struct amdgpu_xcp_res_sysfs_attribute, attr) +#define to_xcp_res(x) container_of(x, struct amdgpu_xcp_res_details, kobj) + +static ssize_t xcp_cfg_res_sysfs_attr_show(struct kobject *kobj, + struct attribute *attr, char *buf) +{ + struct amdgpu_xcp_res_sysfs_attribute *attribute; + struct amdgpu_xcp_res_details *xcp_res; + + attribute = to_xcp_attr(attr); + xcp_res = to_xcp_res(kobj); + + if (!attribute->show) + return -EIO; + + return attribute->show(xcp_res, buf); +} + +static const struct sysfs_ops xcp_cfg_res_sysfs_ops = { + .show = xcp_cfg_res_sysfs_attr_show, +}; + +static const struct kobj_type xcp_cfg_res_sysfs_ktype = { + .sysfs_ops = &xcp_cfg_res_sysfs_ops, + .default_groups = xcp_cfg_res_sysfs_groups, +}; + +const char *xcp_res_names[] = { + [AMDGPU_XCP_RES_XCC] = "xcc", + [AMDGPU_XCP_RES_DMA] = "dma", + [AMDGPU_XCP_RES_DEC] = "dec", + [AMDGPU_XCP_RES_JPEG] = "jpeg", +}; + +static int amdgpu_xcp_get_res_info(struct amdgpu_xcp_mgr *xcp_mgr, + int mode, + struct amdgpu_xcp_cfg *xcp_cfg) +{ + if (xcp_mgr->funcs && xcp_mgr->funcs->get_xcp_res_info) + return xcp_mgr->funcs->get_xcp_res_info(xcp_mgr, mode, xcp_cfg); + + return -EOPNOTSUPP; +} + +#define to_xcp_cfg(x) container_of(x, struct amdgpu_xcp_cfg, kobj) +static ssize_t supported_xcp_configs_show(struct kobject *kobj, + struct kobj_attribute *attr, char *buf) +{ + struct amdgpu_xcp_cfg *xcp_cfg = to_xcp_cfg(kobj); + struct amdgpu_xcp_mgr *xcp_mgr = xcp_cfg->xcp_mgr; + int size = 0, mode; + char *sep = ""; + + if (!xcp_mgr || !xcp_mgr->supp_xcp_modes) + return sysfs_emit(buf, "Not supported\n"); + + for_each_inst(mode, xcp_mgr->supp_xcp_modes) { + size += sysfs_emit_at(buf, size, "%s%s", sep, xcp_desc[mode]); + sep = ", "; + } + + size += sysfs_emit_at(buf, size, "\n"); + + return size; +} + +static ssize_t supported_nps_configs_show(struct kobject *kobj, + struct kobj_attribute *attr, char *buf) +{ + struct amdgpu_xcp_cfg *xcp_cfg = to_xcp_cfg(kobj); + int size = 0, mode; + char *sep = ""; + + if (!xcp_cfg || !xcp_cfg->compatible_nps_modes) + return sysfs_emit(buf, "Not supported\n"); + + for_each_inst(mode, xcp_cfg->compatible_nps_modes) { + size += sysfs_emit_at(buf, size, "%s%s", sep, nps_desc[mode]); + sep = ", "; + } + + size += sysfs_emit_at(buf, size, "\n"); + + return size; +} + +static ssize_t xcp_config_show(struct kobject *kobj, + struct kobj_attribute *attr, char *buf) +{ + struct amdgpu_xcp_cfg *xcp_cfg = to_xcp_cfg(kobj); + + return sysfs_emit(buf, "%s\n", + amdgpu_gfx_compute_mode_desc(xcp_cfg->mode)); +} + +static ssize_t xcp_config_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t size) +{ + struct amdgpu_xcp_cfg *xcp_cfg = to_xcp_cfg(kobj); + int mode, r; + + if (!strncasecmp("SPX", buf, strlen("SPX"))) + mode = AMDGPU_SPX_PARTITION_MODE; + else if (!strncasecmp("DPX", buf, strlen("DPX"))) + mode = AMDGPU_DPX_PARTITION_MODE; + else if (!strncasecmp("TPX", buf, strlen("TPX"))) + mode = AMDGPU_TPX_PARTITION_MODE; + else if (!strncasecmp("QPX", buf, strlen("QPX"))) + mode = AMDGPU_QPX_PARTITION_MODE; + else if (!strncasecmp("CPX", buf, strlen("CPX"))) + mode = AMDGPU_CPX_PARTITION_MODE; + else + return -EINVAL; + + r = amdgpu_xcp_get_res_info(xcp_cfg->xcp_mgr, mode, xcp_cfg); + + if (r) + return r; + + xcp_cfg->mode = mode; + return size; +} + +static struct kobj_attribute xcp_cfg_sysfs_mode = + __ATTR_RW_MODE(xcp_config, 0644); + +static void xcp_cfg_sysfs_release(struct kobject *kobj) +{ + struct amdgpu_xcp_cfg *xcp_cfg = to_xcp_cfg(kobj); + + kfree(xcp_cfg); +} + +static const struct kobj_type xcp_cfg_sysfs_ktype = { + .release = xcp_cfg_sysfs_release, + .sysfs_ops = &kobj_sysfs_ops, +}; + +static struct kobj_attribute supp_part_sysfs_mode = + __ATTR_RO(supported_xcp_configs); + +static struct kobj_attribute supp_nps_sysfs_mode = + __ATTR_RO(supported_nps_configs); + +static const struct attribute *xcp_attrs[] = { + &supp_part_sysfs_mode.attr, + &xcp_cfg_sysfs_mode.attr, + NULL, +}; + +void amdgpu_xcp_cfg_sysfs_init(struct amdgpu_device *adev) +{ + struct amdgpu_xcp_res_details *xcp_res; + struct amdgpu_xcp_cfg *xcp_cfg; + int i, r, j, rid, mode; + + if (!adev->xcp_mgr) + return; + + xcp_cfg = kzalloc(sizeof(*xcp_cfg), GFP_KERNEL); + if (!xcp_cfg) + return; + xcp_cfg->xcp_mgr = adev->xcp_mgr; + + r = kobject_init_and_add(&xcp_cfg->kobj, &xcp_cfg_sysfs_ktype, + &adev->dev->kobj, "compute_partition_config"); + if (r) + goto err1; + + r = sysfs_create_files(&xcp_cfg->kobj, xcp_attrs); + if (r) + goto err1; + + if (adev->gmc.supported_nps_modes != 0) { + r = sysfs_create_file(&xcp_cfg->kobj, &supp_nps_sysfs_mode.attr); + if (r) { + sysfs_remove_files(&xcp_cfg->kobj, xcp_attrs); + goto err1; + } + } + + mode = (xcp_cfg->xcp_mgr->mode == + AMDGPU_UNKNOWN_COMPUTE_PARTITION_MODE) ? + AMDGPU_SPX_PARTITION_MODE : + xcp_cfg->xcp_mgr->mode; + r = amdgpu_xcp_get_res_info(xcp_cfg->xcp_mgr, mode, xcp_cfg); + if (r) { + sysfs_remove_file(&xcp_cfg->kobj, &supp_nps_sysfs_mode.attr); + sysfs_remove_files(&xcp_cfg->kobj, xcp_attrs); + goto err1; + } + + xcp_cfg->mode = mode; + for (i = 0; i < xcp_cfg->num_res; i++) { + xcp_res = &xcp_cfg->xcp_res[i]; + rid = xcp_res->id; + r = kobject_init_and_add(&xcp_res->kobj, + &xcp_cfg_res_sysfs_ktype, + &xcp_cfg->kobj, "%s", + xcp_res_names[rid]); + if (r) + goto err; + } + + adev->xcp_mgr->xcp_cfg = xcp_cfg; + return; +err: + for (j = 0; j < i; j++) { + xcp_res = &xcp_cfg->xcp_res[i]; + kobject_put(&xcp_res->kobj); + } + + sysfs_remove_file(&xcp_cfg->kobj, &supp_nps_sysfs_mode.attr); + sysfs_remove_files(&xcp_cfg->kobj, xcp_attrs); +err1: + kobject_put(&xcp_cfg->kobj); +} + +void amdgpu_xcp_cfg_sysfs_fini(struct amdgpu_device *adev) +{ + struct amdgpu_xcp_res_details *xcp_res; + struct amdgpu_xcp_cfg *xcp_cfg; + int i; + + if (!adev->xcp_mgr) + return; + + xcp_cfg = adev->xcp_mgr->xcp_cfg; + for (i = 0; i < xcp_cfg->num_res; i++) { + xcp_res = &xcp_cfg->xcp_res[i]; + kobject_put(&xcp_res->kobj); + } + + sysfs_remove_file(&xcp_cfg->kobj, &supp_nps_sysfs_mode.attr); + sysfs_remove_files(&xcp_cfg->kobj, xcp_attrs); + kobject_put(&xcp_cfg->kobj); +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xcp.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_xcp.h index 32775260556f..b63f53242c57 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xcp.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xcp.h @@ -56,6 +56,30 @@ enum AMDGPU_XCP_STATE { AMDGPU_XCP_RESUME, }; +enum amdgpu_xcp_res_id { + AMDGPU_XCP_RES_XCC, + AMDGPU_XCP_RES_DMA, + AMDGPU_XCP_RES_DEC, + AMDGPU_XCP_RES_JPEG, + AMDGPU_XCP_RES_MAX, +}; + +struct amdgpu_xcp_res_details { + enum amdgpu_xcp_res_id id; + u8 num_inst; + u8 num_shared; + struct kobject kobj; +}; + +struct amdgpu_xcp_cfg { + u8 mode; + struct amdgpu_xcp_res_details xcp_res[AMDGPU_XCP_RES_MAX]; + u8 num_res; + struct amdgpu_xcp_mgr *xcp_mgr; + struct kobject kobj; + u16 compatible_nps_modes; +}; + struct amdgpu_xcp_ip_funcs { int (*prepare_suspend)(void *handle, uint32_t inst_mask); int (*suspend)(void *handle, uint32_t inst_mask); @@ -97,6 +121,9 @@ struct amdgpu_xcp_mgr { /* Used to determine KFD memory size limits per XCP */ unsigned int num_xcp_per_mem_partition; + struct amdgpu_xcp_cfg *xcp_cfg; + uint32_t supp_xcp_modes; + uint32_t avail_xcp_modes; }; struct amdgpu_xcp_mgr_funcs { @@ -108,7 +135,9 @@ struct amdgpu_xcp_mgr_funcs { struct amdgpu_xcp_ip *ip); int (*get_xcp_mem_id)(struct amdgpu_xcp_mgr *xcp_mgr, struct amdgpu_xcp *xcp, uint8_t *mem_id); - + int (*get_xcp_res_info)(struct amdgpu_xcp_mgr *xcp_mgr, + int mode, + struct amdgpu_xcp_cfg *xcp_cfg); int (*prepare_suspend)(struct amdgpu_xcp_mgr *xcp_mgr, int xcp_id); int (*suspend)(struct amdgpu_xcp_mgr *xcp_mgr, int xcp_id); int (*prepare_resume)(struct amdgpu_xcp_mgr *xcp_mgr, int xcp_id); @@ -146,6 +175,9 @@ int amdgpu_xcp_open_device(struct amdgpu_device *adev, void amdgpu_xcp_release_sched(struct amdgpu_device *adev, struct amdgpu_ctx_entity *entity); +void amdgpu_xcp_cfg_sysfs_init(struct amdgpu_device *adev); +void amdgpu_xcp_cfg_sysfs_fini(struct amdgpu_device *adev); + #define amdgpu_xcp_select_scheds(adev, e, c, d, x, y) \ ((adev)->xcp_mgr && (adev)->xcp_mgr->funcs && \ (adev)->xcp_mgr->funcs->select_scheds ? \ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c index 7de449fae1e3..b47422b0b5b1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c @@ -667,6 +667,7 @@ struct amdgpu_hive_info *amdgpu_get_xgmi_hive(struct amdgpu_device *adev) task_barrier_init(&hive->tb); hive->pstate = AMDGPU_XGMI_PSTATE_UNKNOWN; hive->hi_req_gpu = NULL; + atomic_set(&hive->requested_nps_mode, UNKNOWN_MEMORY_PARTITION_MODE); /* * hive pstate on boot is high in vega20 so we have to go to low @@ -800,6 +801,23 @@ int amdgpu_xgmi_get_num_links(struct amdgpu_device *adev, return -EINVAL; } +bool amdgpu_xgmi_get_is_sharing_enabled(struct amdgpu_device *adev, + struct amdgpu_device *peer_adev) +{ + struct psp_xgmi_topology_info *top = &adev->psp.xgmi_context.top_info; + int i; + + /* Sharing should always be enabled for non-SRIOV. */ + if (!amdgpu_sriov_vf(adev)) + return true; + + for (i = 0 ; i < top->num_nodes; ++i) + if (top->nodes[i].node_id == peer_adev->gmc.xgmi.node_id) + return !!top->nodes[i].is_sharing_enabled; + + return false; +} + /* * Devices that support extended data require the entire hive to initialize with * the shared memory buffer flag set. @@ -860,8 +878,7 @@ int amdgpu_xgmi_add_device(struct amdgpu_device *adev) if (!adev->gmc.xgmi.supported) return 0; - if (!adev->gmc.xgmi.pending_reset && - amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_PSP)) { + if (amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_PSP)) { ret = psp_xgmi_initialize(&adev->psp, false, true); if (ret) { dev_err(adev->dev, @@ -907,8 +924,7 @@ int amdgpu_xgmi_add_device(struct amdgpu_device *adev) task_barrier_add_task(&hive->tb); - if (!adev->gmc.xgmi.pending_reset && - amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_PSP)) { + if (amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_PSP)) { list_for_each_entry(tmp_adev, &hive->device_list, gmc.xgmi.head) { /* update node list for other device in the hive */ if (tmp_adev != adev) { @@ -985,7 +1001,7 @@ int amdgpu_xgmi_add_device(struct amdgpu_device *adev) } } - if (!ret && !adev->gmc.xgmi.pending_reset) + if (!ret) ret = amdgpu_xgmi_sysfs_add_dev_info(adev, hive); exit_unlock: @@ -1500,3 +1516,117 @@ int amdgpu_xgmi_ras_sw_init(struct amdgpu_device *adev) return 0; } + +static void amdgpu_xgmi_reset_on_init_work(struct work_struct *work) +{ + struct amdgpu_hive_info *hive = + container_of(work, struct amdgpu_hive_info, reset_on_init_work); + struct amdgpu_reset_context reset_context; + struct amdgpu_device *tmp_adev; + struct list_head device_list; + int r; + + mutex_lock(&hive->hive_lock); + + INIT_LIST_HEAD(&device_list); + list_for_each_entry(tmp_adev, &hive->device_list, gmc.xgmi.head) + list_add_tail(&tmp_adev->reset_list, &device_list); + + tmp_adev = list_first_entry(&device_list, struct amdgpu_device, + reset_list); + amdgpu_device_lock_reset_domain(tmp_adev->reset_domain); + + reset_context.method = AMD_RESET_METHOD_ON_INIT; + reset_context.reset_req_dev = tmp_adev; + reset_context.hive = hive; + reset_context.reset_device_list = &device_list; + set_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags); + set_bit(AMDGPU_SKIP_COREDUMP, &reset_context.flags); + + amdgpu_reset_do_xgmi_reset_on_init(&reset_context); + mutex_unlock(&hive->hive_lock); + amdgpu_device_unlock_reset_domain(tmp_adev->reset_domain); + + list_for_each_entry(tmp_adev, &hive->device_list, gmc.xgmi.head) { + r = amdgpu_ras_init_badpage_info(tmp_adev); + if (r && r != -EHWPOISON) + dev_err(tmp_adev->dev, + "error during bad page data initialization"); + } +} + +static void amdgpu_xgmi_schedule_reset_on_init(struct amdgpu_hive_info *hive) +{ + INIT_WORK(&hive->reset_on_init_work, amdgpu_xgmi_reset_on_init_work); + amdgpu_reset_domain_schedule(hive->reset_domain, + &hive->reset_on_init_work); +} + +int amdgpu_xgmi_reset_on_init(struct amdgpu_device *adev) +{ + struct amdgpu_hive_info *hive; + bool reset_scheduled; + int num_devs; + + hive = amdgpu_get_xgmi_hive(adev); + if (!hive) + return -EINVAL; + + mutex_lock(&hive->hive_lock); + num_devs = atomic_read(&hive->number_devices); + reset_scheduled = false; + if (num_devs == adev->gmc.xgmi.num_physical_nodes) { + amdgpu_xgmi_schedule_reset_on_init(hive); + reset_scheduled = true; + } + + mutex_unlock(&hive->hive_lock); + amdgpu_put_xgmi_hive(hive); + + if (reset_scheduled) + flush_work(&hive->reset_on_init_work); + + return 0; +} + +int amdgpu_xgmi_request_nps_change(struct amdgpu_device *adev, + struct amdgpu_hive_info *hive, + int req_nps_mode) +{ + struct amdgpu_device *tmp_adev; + int cur_nps_mode, r; + + /* This is expected to be called only during unload of driver. The + * request needs to be placed only once for all devices in the hive. If + * one of them fail, revert the request for previous successful devices. + * After placing the request, make hive mode as UNKNOWN so that other + * devices don't request anymore. + */ + mutex_lock(&hive->hive_lock); + if (atomic_read(&hive->requested_nps_mode) == + UNKNOWN_MEMORY_PARTITION_MODE) { + dev_dbg(adev->dev, "Unexpected entry for hive NPS change"); + mutex_unlock(&hive->hive_lock); + return 0; + } + list_for_each_entry(tmp_adev, &hive->device_list, gmc.xgmi.head) { + r = adev->gmc.gmc_funcs->request_mem_partition_mode( + tmp_adev, req_nps_mode); + if (r) + break; + } + if (r) { + /* Request back current mode if one of the requests failed */ + cur_nps_mode = + adev->gmc.gmc_funcs->query_mem_partition_mode(tmp_adev); + list_for_each_entry_continue_reverse( + tmp_adev, &hive->device_list, gmc.xgmi.head) + adev->gmc.gmc_funcs->request_mem_partition_mode( + tmp_adev, cur_nps_mode); + } + /* Set to UNKNOWN so that other devices don't request anymore */ + atomic_set(&hive->requested_nps_mode, UNKNOWN_MEMORY_PARTITION_MODE); + mutex_unlock(&hive->hive_lock); + + return r; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h index a3bfc16de6d4..8cc7ab38db7c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h @@ -45,6 +45,8 @@ struct amdgpu_hive_info { struct amdgpu_reset_domain *reset_domain; atomic_t ras_recovery; struct ras_event_manager event_mgr; + struct work_struct reset_on_init_work; + atomic_t requested_nps_mode; }; struct amdgpu_pcs_ras_field { @@ -64,6 +66,8 @@ int amdgpu_xgmi_get_hops_count(struct amdgpu_device *adev, struct amdgpu_device *peer_adev); int amdgpu_xgmi_get_num_links(struct amdgpu_device *adev, struct amdgpu_device *peer_adev); +bool amdgpu_xgmi_get_is_sharing_enabled(struct amdgpu_device *adev, + struct amdgpu_device *peer_adev); uint64_t amdgpu_xgmi_get_relative_phy_addr(struct amdgpu_device *adev, uint64_t addr); static inline bool amdgpu_xgmi_same_hive(struct amdgpu_device *adev, @@ -75,5 +79,10 @@ static inline bool amdgpu_xgmi_same_hive(struct amdgpu_device *adev, adev->gmc.xgmi.hive_id == bo_adev->gmc.xgmi.hive_id); } int amdgpu_xgmi_ras_sw_init(struct amdgpu_device *adev); +int amdgpu_xgmi_reset_on_init(struct amdgpu_device *adev); + +int amdgpu_xgmi_request_nps_change(struct amdgpu_device *adev, + struct amdgpu_hive_info *hive, + int req_nps_mode); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h b/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h index 6e9eeaeb3de1..b4f9c2f4e92c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h @@ -28,17 +28,21 @@ #define AMD_SRIOV_MSG_VBIOS_SIZE_KB 64 #define AMD_SRIOV_MSG_DATAEXCHANGE_OFFSET_KB AMD_SRIOV_MSG_VBIOS_SIZE_KB #define AMD_SRIOV_MSG_DATAEXCHANGE_SIZE_KB 4 - +#define AMD_SRIOV_MSG_TMR_OFFSET_KB 2048 +#define AMD_SRIOV_MSG_BAD_PAGE_SIZE_KB 2 +#define AMD_SRIOV_RAS_TELEMETRY_SIZE_KB 64 /* * layout - * 0 64KB 65KB 66KB - * | VBIOS | PF2VF | VF2PF | Bad Page | ... - * | 64KB | 1KB | 1KB | + * 0 64KB 65KB 66KB 68KB 132KB + * | VBIOS | PF2VF | VF2PF | Bad Page | RAS Telemetry Region | ... + * | 64KB | 1KB | 1KB | 2KB | 64KB | ... */ + #define AMD_SRIOV_MSG_SIZE_KB 1 #define AMD_SRIOV_MSG_PF2VF_OFFSET_KB AMD_SRIOV_MSG_DATAEXCHANGE_OFFSET_KB #define AMD_SRIOV_MSG_VF2PF_OFFSET_KB (AMD_SRIOV_MSG_PF2VF_OFFSET_KB + AMD_SRIOV_MSG_SIZE_KB) #define AMD_SRIOV_MSG_BAD_PAGE_OFFSET_KB (AMD_SRIOV_MSG_VF2PF_OFFSET_KB + AMD_SRIOV_MSG_SIZE_KB) +#define AMD_SRIOV_MSG_RAS_TELEMETRY_OFFSET_KB (AMD_SRIOV_MSG_BAD_PAGE_OFFSET_KB + AMD_SRIOV_MSG_BAD_PAGE_SIZE_KB) /* * PF2VF history log: @@ -86,30 +90,59 @@ enum amd_sriov_ucode_engine_id { union amd_sriov_msg_feature_flags { struct { - uint32_t error_log_collect : 1; - uint32_t host_load_ucodes : 1; - uint32_t host_flr_vramlost : 1; - uint32_t mm_bw_management : 1; - uint32_t pp_one_vf_mode : 1; - uint32_t reg_indirect_acc : 1; - uint32_t av1_support : 1; - uint32_t vcn_rb_decouple : 1; - uint32_t mes_info_enable : 1; - uint32_t reserved : 23; + uint32_t error_log_collect : 1; + uint32_t host_load_ucodes : 1; + uint32_t host_flr_vramlost : 1; + uint32_t mm_bw_management : 1; + uint32_t pp_one_vf_mode : 1; + uint32_t reg_indirect_acc : 1; + uint32_t av1_support : 1; + uint32_t vcn_rb_decouple : 1; + uint32_t mes_info_dump_enable : 1; + uint32_t ras_caps : 1; + uint32_t ras_telemetry : 1; + uint32_t reserved : 21; } flags; uint32_t all; }; union amd_sriov_reg_access_flags { struct { - uint32_t vf_reg_access_ih : 1; - uint32_t vf_reg_access_mmhub : 1; - uint32_t vf_reg_access_gc : 1; - uint32_t reserved : 29; + uint32_t vf_reg_access_ih : 1; + uint32_t vf_reg_access_mmhub : 1; + uint32_t vf_reg_access_gc : 1; + uint32_t reserved : 29; } flags; uint32_t all; }; +union amd_sriov_ras_caps { + struct { + uint64_t block_umc : 1; + uint64_t block_sdma : 1; + uint64_t block_gfx : 1; + uint64_t block_mmhub : 1; + uint64_t block_athub : 1; + uint64_t block_pcie_bif : 1; + uint64_t block_hdp : 1; + uint64_t block_xgmi_wafl : 1; + uint64_t block_df : 1; + uint64_t block_smn : 1; + uint64_t block_sem : 1; + uint64_t block_mp0 : 1; + uint64_t block_mp1 : 1; + uint64_t block_fuse : 1; + uint64_t block_mca : 1; + uint64_t block_vcn : 1; + uint64_t block_jpeg : 1; + uint64_t block_ih : 1; + uint64_t block_mpio : 1; + uint64_t poison_propogation_mode : 1; + uint64_t reserved : 44; + } bits; + uint64_t all; +}; + union amd_sriov_msg_os_info { struct { uint32_t windows : 1; @@ -158,7 +191,7 @@ struct amd_sriov_msg_pf2vf_info_header { uint32_t reserved[2]; }; -#define AMD_SRIOV_MSG_PF2VF_INFO_FILLED_SIZE (49) +#define AMD_SRIOV_MSG_PF2VF_INFO_FILLED_SIZE (55) struct amd_sriov_msg_pf2vf_info { /* header contains size and version */ struct amd_sriov_msg_pf2vf_info_header header; @@ -211,6 +244,12 @@ struct amd_sriov_msg_pf2vf_info { uint32_t pcie_atomic_ops_support_flags; /* Portion of GPU memory occupied by VF. MAX value is 65535, but set to uint32_t to maintain alignment with reserved size */ uint32_t gpu_capacity; + /* vf bdf on host pci tree for debug only */ + uint32_t bdf_on_host; + uint32_t more_bp; //Reserved for future use. + union amd_sriov_ras_caps ras_en_caps; + union amd_sriov_ras_caps ras_telemetry_en_caps; + /* reserved */ uint32_t reserved[256 - AMD_SRIOV_MSG_PF2VF_INFO_FILLED_SIZE]; } __packed; @@ -283,8 +322,12 @@ enum amd_sriov_mailbox_request_message { MB_REQ_MSG_REL_GPU_FINI_ACCESS, MB_REQ_MSG_REQ_GPU_RESET_ACCESS, MB_REQ_MSG_REQ_GPU_INIT_DATA, + MB_REQ_MSG_PSP_VF_CMD_RELAY, MB_REQ_MSG_LOG_VF_ERROR = 200, + MB_REQ_MSG_READY_TO_RESET = 201, + MB_REQ_MSG_RAS_POISON = 202, + MB_REQ_RAS_ERROR_COUNT = 203, }; /* mailbox message send from host to guest */ @@ -297,10 +340,60 @@ enum amd_sriov_mailbox_response_message { MB_RES_MSG_FAIL, MB_RES_MSG_QUERY_ALIVE, MB_RES_MSG_GPU_INIT_DATA_READY, + MB_RES_MSG_RAS_ERROR_COUNT_READY = 11, MB_RES_MSG_TEXT_MESSAGE = 255 }; +enum amd_sriov_ras_telemetry_gpu_block { + RAS_TELEMETRY_GPU_BLOCK_UMC = 0, + RAS_TELEMETRY_GPU_BLOCK_SDMA = 1, + RAS_TELEMETRY_GPU_BLOCK_GFX = 2, + RAS_TELEMETRY_GPU_BLOCK_MMHUB = 3, + RAS_TELEMETRY_GPU_BLOCK_ATHUB = 4, + RAS_TELEMETRY_GPU_BLOCK_PCIE_BIF = 5, + RAS_TELEMETRY_GPU_BLOCK_HDP = 6, + RAS_TELEMETRY_GPU_BLOCK_XGMI_WAFL = 7, + RAS_TELEMETRY_GPU_BLOCK_DF = 8, + RAS_TELEMETRY_GPU_BLOCK_SMN = 9, + RAS_TELEMETRY_GPU_BLOCK_SEM = 10, + RAS_TELEMETRY_GPU_BLOCK_MP0 = 11, + RAS_TELEMETRY_GPU_BLOCK_MP1 = 12, + RAS_TELEMETRY_GPU_BLOCK_FUSE = 13, + RAS_TELEMETRY_GPU_BLOCK_MCA = 14, + RAS_TELEMETRY_GPU_BLOCK_VCN = 15, + RAS_TELEMETRY_GPU_BLOCK_JPEG = 16, + RAS_TELEMETRY_GPU_BLOCK_IH = 17, + RAS_TELEMETRY_GPU_BLOCK_MPIO = 18, + RAS_TELEMETRY_GPU_BLOCK_COUNT = 19, +}; + +struct amd_sriov_ras_telemetry_header { + uint32_t checksum; + uint32_t used_size; + uint32_t reserved[2]; +}; + +struct amd_sriov_ras_telemetry_error_count { + struct { + uint32_t ce_count; + uint32_t ue_count; + uint32_t de_count; + uint32_t ce_overflow_count; + uint32_t ue_overflow_count; + uint32_t de_overflow_count; + uint32_t reserved[6]; + } block[RAS_TELEMETRY_GPU_BLOCK_COUNT]; +}; + +struct amdsriov_ras_telemetry { + struct amd_sriov_ras_telemetry_header header; + + union { + struct amd_sriov_ras_telemetry_error_count error_count; + } body; +}; + /* version data stored in MAILBOX_MSGBUF_RCV_DW1 for future expansion */ enum amd_sriov_gpu_init_data_version { GPU_INIT_DATA_READY_V1 = 1, diff --git a/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c b/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c index ccfd2a4b4acc..e157d6d857b6 100644 --- a/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c +++ b/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c @@ -447,6 +447,72 @@ static int __aqua_vanjaram_get_xcp_ip_info(struct amdgpu_xcp_mgr *xcp_mgr, int x return 0; } +static int aqua_vanjaram_get_xcp_res_info(struct amdgpu_xcp_mgr *xcp_mgr, + int mode, + struct amdgpu_xcp_cfg *xcp_cfg) +{ + struct amdgpu_device *adev = xcp_mgr->adev; + int max_res[AMDGPU_XCP_RES_MAX] = {}; + bool res_lt_xcp; + int num_xcp, i; + u16 nps_modes; + + if (!(xcp_mgr->supp_xcp_modes & BIT(mode))) + return -EINVAL; + + max_res[AMDGPU_XCP_RES_XCC] = NUM_XCC(adev->gfx.xcc_mask); + max_res[AMDGPU_XCP_RES_DMA] = adev->sdma.num_instances; + max_res[AMDGPU_XCP_RES_DEC] = adev->vcn.num_vcn_inst; + max_res[AMDGPU_XCP_RES_JPEG] = adev->jpeg.num_jpeg_inst; + + switch (mode) { + case AMDGPU_SPX_PARTITION_MODE: + num_xcp = 1; + nps_modes = BIT(AMDGPU_NPS1_PARTITION_MODE); + break; + case AMDGPU_DPX_PARTITION_MODE: + num_xcp = 2; + nps_modes = BIT(AMDGPU_NPS1_PARTITION_MODE); + break; + case AMDGPU_TPX_PARTITION_MODE: + num_xcp = 3; + nps_modes = BIT(AMDGPU_NPS1_PARTITION_MODE) | + BIT(AMDGPU_NPS4_PARTITION_MODE); + break; + case AMDGPU_QPX_PARTITION_MODE: + num_xcp = 4; + nps_modes = BIT(AMDGPU_NPS1_PARTITION_MODE) | + BIT(AMDGPU_NPS4_PARTITION_MODE); + break; + case AMDGPU_CPX_PARTITION_MODE: + num_xcp = NUM_XCC(adev->gfx.xcc_mask); + nps_modes = BIT(AMDGPU_NPS1_PARTITION_MODE) | + BIT(AMDGPU_NPS4_PARTITION_MODE); + break; + default: + return -EINVAL; + } + + xcp_cfg->compatible_nps_modes = + (adev->gmc.supported_nps_modes & nps_modes); + xcp_cfg->num_res = ARRAY_SIZE(max_res); + + for (i = 0; i < xcp_cfg->num_res; i++) { + res_lt_xcp = max_res[i] < num_xcp; + xcp_cfg->xcp_res[i].id = i; + xcp_cfg->xcp_res[i].num_inst = + res_lt_xcp ? 1 : max_res[i] / num_xcp; + xcp_cfg->xcp_res[i].num_inst = + i == AMDGPU_XCP_RES_JPEG ? + xcp_cfg->xcp_res[i].num_inst * + adev->jpeg.num_jpeg_rings : xcp_cfg->xcp_res[i].num_inst; + xcp_cfg->xcp_res[i].num_shared = + res_lt_xcp ? num_xcp / max_res[i] : 1; + } + + return 0; +} + static enum amdgpu_gfx_partition __aqua_vanjaram_get_auto_mode(struct amdgpu_xcp_mgr *xcp_mgr) { @@ -530,6 +596,57 @@ static int __aqua_vanjaram_post_partition_switch(struct amdgpu_xcp_mgr *xcp_mgr, return ret; } +static void +__aqua_vanjaram_update_supported_modes(struct amdgpu_xcp_mgr *xcp_mgr) +{ + struct amdgpu_device *adev = xcp_mgr->adev; + + xcp_mgr->supp_xcp_modes = 0; + + switch (NUM_XCC(adev->gfx.xcc_mask)) { + case 8: + xcp_mgr->supp_xcp_modes = BIT(AMDGPU_SPX_PARTITION_MODE) | + BIT(AMDGPU_DPX_PARTITION_MODE) | + BIT(AMDGPU_QPX_PARTITION_MODE) | + BIT(AMDGPU_CPX_PARTITION_MODE); + break; + case 6: + xcp_mgr->supp_xcp_modes = BIT(AMDGPU_SPX_PARTITION_MODE) | + BIT(AMDGPU_TPX_PARTITION_MODE) | + BIT(AMDGPU_CPX_PARTITION_MODE); + break; + case 4: + xcp_mgr->supp_xcp_modes = BIT(AMDGPU_SPX_PARTITION_MODE) | + BIT(AMDGPU_DPX_PARTITION_MODE) | + BIT(AMDGPU_CPX_PARTITION_MODE); + break; + /* this seems only existing in emulation phase */ + case 2: + xcp_mgr->supp_xcp_modes = BIT(AMDGPU_SPX_PARTITION_MODE) | + BIT(AMDGPU_CPX_PARTITION_MODE); + break; + case 1: + xcp_mgr->supp_xcp_modes = BIT(AMDGPU_SPX_PARTITION_MODE) | + BIT(AMDGPU_CPX_PARTITION_MODE); + break; + + default: + break; + } +} + +static void __aqua_vanjaram_update_available_partition_mode(struct amdgpu_xcp_mgr *xcp_mgr) +{ + int mode; + + xcp_mgr->avail_xcp_modes = 0; + + for_each_inst(mode, xcp_mgr->supp_xcp_modes) { + if (__aqua_vanjaram_is_valid_mode(xcp_mgr, mode)) + xcp_mgr->avail_xcp_modes |= BIT(mode); + } +} + static int aqua_vanjaram_switch_partition_mode(struct amdgpu_xcp_mgr *xcp_mgr, int mode, int *num_xcps) { @@ -578,6 +695,8 @@ static int aqua_vanjaram_switch_partition_mode(struct amdgpu_xcp_mgr *xcp_mgr, amdgpu_xcp_init(xcp_mgr, *num_xcps, mode); ret = __aqua_vanjaram_post_partition_switch(xcp_mgr, flags); + if (!ret) + __aqua_vanjaram_update_available_partition_mode(xcp_mgr); unlock: if (flags & AMDGPU_XCP_OPS_KFD) amdgpu_amdkfd_unlock_kfd(adev); @@ -656,9 +775,11 @@ struct amdgpu_xcp_mgr_funcs aqua_vanjaram_xcp_funcs = { .switch_partition_mode = &aqua_vanjaram_switch_partition_mode, .query_partition_mode = &aqua_vanjaram_query_partition_mode, .get_ip_details = &aqua_vanjaram_get_xcp_ip_details, + .get_xcp_res_info = &aqua_vanjaram_get_xcp_res_info, .get_xcp_mem_id = &aqua_vanjaram_get_xcp_mem_id, .select_scheds = &aqua_vanjaram_select_scheds, - .update_partition_sched_list = &aqua_vanjaram_update_partition_sched_list + .update_partition_sched_list = + &aqua_vanjaram_update_partition_sched_list }; static int aqua_vanjaram_xcp_mgr_init(struct amdgpu_device *adev) @@ -673,6 +794,7 @@ static int aqua_vanjaram_xcp_mgr_init(struct amdgpu_device *adev) if (ret) return ret; + __aqua_vanjaram_update_supported_modes(adev->xcp_mgr); /* TODO: Default memory node affinity init */ return ret; diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c index cf1d5d462b67..e2cb1f080e88 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik.c +++ b/drivers/gpu/drm/amd/amdgpu/cik.c @@ -1985,9 +1985,9 @@ static const struct amdgpu_asic_funcs cik_asic_funcs = .query_video_codecs = &cik_query_video_codecs, }; -static int cik_common_early_init(void *handle) +static int cik_common_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->smc_rreg = &cik_smc_rreg; adev->smc_wreg = &cik_smc_wreg; @@ -2124,19 +2124,9 @@ static int cik_common_early_init(void *handle) return 0; } -static int cik_common_sw_init(void *handle) +static int cik_common_hw_init(struct amdgpu_ip_block *ip_block) { - return 0; -} - -static int cik_common_sw_fini(void *handle) -{ - return 0; -} - -static int cik_common_hw_init(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* move the golden regs per IP block */ cik_init_golden_registers(adev); @@ -2148,23 +2138,14 @@ static int cik_common_hw_init(void *handle) return 0; } -static int cik_common_hw_fini(void *handle) +static int cik_common_hw_fini(struct amdgpu_ip_block *ip_block) { return 0; } -static int cik_common_suspend(void *handle) +static int cik_common_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return cik_common_hw_fini(adev); -} - -static int cik_common_resume(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return cik_common_hw_init(adev); + return cik_common_hw_init(ip_block); } static bool cik_common_is_idle(void *handle) @@ -2172,12 +2153,9 @@ static bool cik_common_is_idle(void *handle) return true; } -static int cik_common_wait_for_idle(void *handle) -{ - return 0; -} -static int cik_common_soft_reset(void *handle) + +static int cik_common_soft_reset(struct amdgpu_ip_block *ip_block) { /* XXX hard reset?? */ return 0; @@ -2198,20 +2176,13 @@ static int cik_common_set_powergating_state(void *handle, static const struct amd_ip_funcs cik_common_ip_funcs = { .name = "cik_common", .early_init = cik_common_early_init, - .late_init = NULL, - .sw_init = cik_common_sw_init, - .sw_fini = cik_common_sw_fini, .hw_init = cik_common_hw_init, .hw_fini = cik_common_hw_fini, - .suspend = cik_common_suspend, .resume = cik_common_resume, .is_idle = cik_common_is_idle, - .wait_for_idle = cik_common_wait_for_idle, .soft_reset = cik_common_soft_reset, .set_clockgating_state = cik_common_set_clockgating_state, .set_powergating_state = cik_common_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ip_block_version cik_common_ip_block = diff --git a/drivers/gpu/drm/amd/amdgpu/cik_ih.c b/drivers/gpu/drm/amd/amdgpu/cik_ih.c index 576baa9dbb0e..1da17755ad53 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/cik_ih.c @@ -283,9 +283,9 @@ static void cik_ih_set_rptr(struct amdgpu_device *adev, WREG32(mmIH_RB_RPTR, ih->rptr); } -static int cik_ih_early_init(void *handle) +static int cik_ih_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int ret; ret = amdgpu_irq_add_domain(adev); @@ -297,10 +297,10 @@ static int cik_ih_early_init(void *handle) return 0; } -static int cik_ih_sw_init(void *handle) +static int cik_ih_sw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_ih_ring_init(adev, &adev->irq.ih, 64 * 1024, false); if (r) @@ -311,9 +311,9 @@ static int cik_ih_sw_init(void *handle) return r; } -static int cik_ih_sw_fini(void *handle) +static int cik_ih_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_fini_sw(adev); amdgpu_irq_remove_domain(adev); @@ -321,34 +321,28 @@ static int cik_ih_sw_fini(void *handle) return 0; } -static int cik_ih_hw_init(void *handle) +static int cik_ih_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return cik_ih_irq_init(adev); } -static int cik_ih_hw_fini(void *handle) +static int cik_ih_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - cik_ih_irq_disable(adev); + cik_ih_irq_disable(ip_block->adev); return 0; } -static int cik_ih_suspend(void *handle) +static int cik_ih_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return cik_ih_hw_fini(adev); + return cik_ih_hw_fini(ip_block); } -static int cik_ih_resume(void *handle) +static int cik_ih_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return cik_ih_hw_init(adev); + return cik_ih_hw_init(ip_block); } static bool cik_ih_is_idle(void *handle) @@ -362,11 +356,11 @@ static bool cik_ih_is_idle(void *handle) return true; } -static int cik_ih_wait_for_idle(void *handle) +static int cik_ih_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { /* read MC_STATUS */ @@ -378,9 +372,9 @@ static int cik_ih_wait_for_idle(void *handle) return -ETIMEDOUT; } -static int cik_ih_soft_reset(void *handle) +static int cik_ih_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 srbm_soft_reset = 0; u32 tmp = RREG32(mmSRBM_STATUS); @@ -423,7 +417,6 @@ static int cik_ih_set_powergating_state(void *handle, static const struct amd_ip_funcs cik_ih_ip_funcs = { .name = "cik_ih", .early_init = cik_ih_early_init, - .late_init = NULL, .sw_init = cik_ih_sw_init, .sw_fini = cik_ih_sw_fini, .hw_init = cik_ih_hw_init, @@ -435,8 +428,6 @@ static const struct amd_ip_funcs cik_ih_ip_funcs = { .soft_reset = cik_ih_soft_reset, .set_clockgating_state = cik_ih_set_clockgating_state, .set_powergating_state = cik_ih_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs cik_ih_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c index 952737de9411..ede1a028d48d 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c +++ b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c @@ -54,7 +54,7 @@ static void cik_sdma_set_ring_funcs(struct amdgpu_device *adev); static void cik_sdma_set_irq_funcs(struct amdgpu_device *adev); static void cik_sdma_set_buffer_funcs(struct amdgpu_device *adev); static void cik_sdma_set_vm_pte_funcs(struct amdgpu_device *adev); -static int cik_sdma_soft_reset(void *handle); +static int cik_sdma_soft_reset(struct amdgpu_ip_block *ip_block); MODULE_FIRMWARE("amdgpu/bonaire_sdma.bin"); MODULE_FIRMWARE("amdgpu/bonaire_sdma1.bin"); @@ -918,9 +918,9 @@ static void cik_enable_sdma_mgls(struct amdgpu_device *adev, } } -static int cik_sdma_early_init(void *handle) +static int cik_sdma_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; adev->sdma.num_instances = SDMA_MAX_INSTANCE; @@ -937,10 +937,10 @@ static int cik_sdma_early_init(void *handle) return 0; } -static int cik_sdma_sw_init(void *handle) +static int cik_sdma_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r, i; /* SDMA trap event */ @@ -977,9 +977,9 @@ static int cik_sdma_sw_init(void *handle) return r; } -static int cik_sdma_sw_fini(void *handle) +static int cik_sdma_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; for (i = 0; i < adev->sdma.num_instances; i++) @@ -989,10 +989,10 @@ static int cik_sdma_sw_fini(void *handle) return 0; } -static int cik_sdma_hw_init(void *handle) +static int cik_sdma_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = cik_sdma_start(adev); if (r) @@ -1001,9 +1001,9 @@ static int cik_sdma_hw_init(void *handle) return r; } -static int cik_sdma_hw_fini(void *handle) +static int cik_sdma_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cik_ctx_switch_enable(adev, false); cik_sdma_enable(adev, false); @@ -1011,20 +1011,16 @@ static int cik_sdma_hw_fini(void *handle) return 0; } -static int cik_sdma_suspend(void *handle) +static int cik_sdma_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return cik_sdma_hw_fini(adev); + return cik_sdma_hw_fini(ip_block); } -static int cik_sdma_resume(void *handle) +static int cik_sdma_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + cik_sdma_soft_reset(ip_block); - cik_sdma_soft_reset(handle); - - return cik_sdma_hw_init(adev); + return cik_sdma_hw_init(ip_block); } static bool cik_sdma_is_idle(void *handle) @@ -1039,11 +1035,11 @@ static bool cik_sdma_is_idle(void *handle) return true; } -static int cik_sdma_wait_for_idle(void *handle) +static int cik_sdma_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { tmp = RREG32(mmSRBM_STATUS2) & (SRBM_STATUS2__SDMA_BUSY_MASK | @@ -1056,10 +1052,10 @@ static int cik_sdma_wait_for_idle(void *handle) return -ETIMEDOUT; } -static int cik_sdma_soft_reset(void *handle) +static int cik_sdma_soft_reset(struct amdgpu_ip_block *ip_block) { u32 srbm_soft_reset = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 tmp; /* sdma0 */ @@ -1217,7 +1213,6 @@ static int cik_sdma_set_powergating_state(void *handle, static const struct amd_ip_funcs cik_sdma_ip_funcs = { .name = "cik_sdma", .early_init = cik_sdma_early_init, - .late_init = NULL, .sw_init = cik_sdma_sw_init, .sw_fini = cik_sdma_sw_fini, .hw_init = cik_sdma_hw_init, @@ -1229,8 +1224,6 @@ static const struct amd_ip_funcs cik_sdma_ip_funcs = { .soft_reset = cik_sdma_soft_reset, .set_clockgating_state = cik_sdma_set_clockgating_state, .set_powergating_state = cik_sdma_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs cik_sdma_ring_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/cz_ih.c b/drivers/gpu/drm/amd/amdgpu/cz_ih.c index 072643787384..d72973bd570d 100644 --- a/drivers/gpu/drm/amd/amdgpu/cz_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/cz_ih.c @@ -274,9 +274,9 @@ static void cz_ih_set_rptr(struct amdgpu_device *adev, WREG32(mmIH_RB_RPTR, ih->rptr); } -static int cz_ih_early_init(void *handle) +static int cz_ih_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int ret; ret = amdgpu_irq_add_domain(adev); @@ -288,10 +288,10 @@ static int cz_ih_early_init(void *handle) return 0; } -static int cz_ih_sw_init(void *handle) +static int cz_ih_sw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_ih_ring_init(adev, &adev->irq.ih, 64 * 1024, false); if (r) @@ -302,9 +302,9 @@ static int cz_ih_sw_init(void *handle) return r; } -static int cz_ih_sw_fini(void *handle) +static int cz_ih_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_fini_sw(adev); amdgpu_irq_remove_domain(adev); @@ -312,10 +312,10 @@ static int cz_ih_sw_fini(void *handle) return 0; } -static int cz_ih_hw_init(void *handle) +static int cz_ih_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = cz_ih_irq_init(adev); if (r) @@ -324,27 +324,21 @@ static int cz_ih_hw_init(void *handle) return 0; } -static int cz_ih_hw_fini(void *handle) +static int cz_ih_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - cz_ih_irq_disable(adev); + cz_ih_irq_disable(ip_block->adev); return 0; } -static int cz_ih_suspend(void *handle) +static int cz_ih_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return cz_ih_hw_fini(adev); + return cz_ih_hw_fini(ip_block); } -static int cz_ih_resume(void *handle) +static int cz_ih_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return cz_ih_hw_init(adev); + return cz_ih_hw_init(ip_block); } static bool cz_ih_is_idle(void *handle) @@ -358,11 +352,11 @@ static bool cz_ih_is_idle(void *handle) return true; } -static int cz_ih_wait_for_idle(void *handle) +static int cz_ih_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { /* read MC_STATUS */ @@ -374,10 +368,10 @@ static int cz_ih_wait_for_idle(void *handle) return -ETIMEDOUT; } -static int cz_ih_soft_reset(void *handle) +static int cz_ih_soft_reset(struct amdgpu_ip_block *ip_block) { u32 srbm_soft_reset = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 tmp = RREG32(mmSRBM_STATUS); if (tmp & SRBM_STATUS__IH_BUSY_MASK) @@ -421,7 +415,6 @@ static int cz_ih_set_powergating_state(void *handle, static const struct amd_ip_funcs cz_ih_ip_funcs = { .name = "cz_ih", .early_init = cz_ih_early_init, - .late_init = NULL, .sw_init = cz_ih_sw_init, .sw_fini = cz_ih_sw_fini, .hw_init = cz_ih_hw_init, @@ -433,8 +426,6 @@ static const struct amd_ip_funcs cz_ih_ip_funcs = { .soft_reset = cz_ih_soft_reset, .set_clockgating_state = cz_ih_set_clockgating_state, .set_powergating_state = cz_ih_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs cz_ih_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c index 70c1399f738d..5098c50d54c8 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c @@ -2738,9 +2738,9 @@ static int dce_v10_0_crtc_init(struct amdgpu_device *adev, int index) return 0; } -static int dce_v10_0_early_init(void *handle) +static int dce_v10_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->audio_endpt_rreg = &dce_v10_0_audio_endpt_rreg; adev->audio_endpt_wreg = &dce_v10_0_audio_endpt_wreg; @@ -2765,10 +2765,10 @@ static int dce_v10_0_early_init(void *handle) return 0; } -static int dce_v10_0_sw_init(void *handle) +static int dce_v10_0_sw_init(struct amdgpu_ip_block *ip_block) { int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->mode_info.num_crtc; i++) { r = amdgpu_irq_add_id(adev, AMDGPU_IRQ_CLIENTID_LEGACY, i + 1, &adev->crtc_irq); @@ -2844,9 +2844,9 @@ static int dce_v10_0_sw_init(void *handle) return 0; } -static int dce_v10_0_sw_fini(void *handle) +static int dce_v10_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; drm_edid_free(adev->mode_info.bios_hardcoded_edid); @@ -2862,10 +2862,10 @@ static int dce_v10_0_sw_fini(void *handle) return 0; } -static int dce_v10_0_hw_init(void *handle) +static int dce_v10_0_hw_init(struct amdgpu_ip_block *ip_block) { int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; dce_v10_0_init_golden_registers(adev); @@ -2887,10 +2887,10 @@ static int dce_v10_0_hw_init(void *handle) return 0; } -static int dce_v10_0_hw_fini(void *handle) +static int dce_v10_0_hw_fini(struct amdgpu_ip_block *ip_block) { int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; dce_v10_0_hpd_fini(adev); @@ -2905,9 +2905,9 @@ static int dce_v10_0_hw_fini(void *handle) return 0; } -static int dce_v10_0_suspend(void *handle) +static int dce_v10_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_display_suspend_helper(adev); @@ -2917,18 +2917,18 @@ static int dce_v10_0_suspend(void *handle) adev->mode_info.bl_level = amdgpu_atombios_encoder_get_backlight_level_from_reg(adev); - return dce_v10_0_hw_fini(handle); + return dce_v10_0_hw_fini(ip_block); } -static int dce_v10_0_resume(void *handle) +static int dce_v10_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int ret; amdgpu_atombios_encoder_set_backlight_level_to_reg(adev, adev->mode_info.bl_level); - ret = dce_v10_0_hw_init(handle); + ret = dce_v10_0_hw_init(ip_block); /* turn on the BL */ if (adev->mode_info.bl_encoder) { @@ -2948,22 +2948,17 @@ static bool dce_v10_0_is_idle(void *handle) return true; } -static int dce_v10_0_wait_for_idle(void *handle) +static bool dce_v10_0_check_soft_reset(struct amdgpu_ip_block *ip_block) { - return 0; -} - -static bool dce_v10_0_check_soft_reset(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return dce_v10_0_is_display_hung(adev); } -static int dce_v10_0_soft_reset(void *handle) +static int dce_v10_0_soft_reset(struct amdgpu_ip_block *ip_block) { u32 srbm_soft_reset = 0, tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (dce_v10_0_is_display_hung(adev)) srbm_soft_reset |= SRBM_SOFT_RESET__SOFT_RESET_DC_MASK; @@ -3322,7 +3317,6 @@ static int dce_v10_0_set_powergating_state(void *handle, static const struct amd_ip_funcs dce_v10_0_ip_funcs = { .name = "dce_v10_0", .early_init = dce_v10_0_early_init, - .late_init = NULL, .sw_init = dce_v10_0_sw_init, .sw_fini = dce_v10_0_sw_fini, .hw_init = dce_v10_0_hw_init, @@ -3330,13 +3324,10 @@ static const struct amd_ip_funcs dce_v10_0_ip_funcs = { .suspend = dce_v10_0_suspend, .resume = dce_v10_0_resume, .is_idle = dce_v10_0_is_idle, - .wait_for_idle = dce_v10_0_wait_for_idle, .check_soft_reset = dce_v10_0_check_soft_reset, .soft_reset = dce_v10_0_soft_reset, .set_clockgating_state = dce_v10_0_set_clockgating_state, .set_powergating_state = dce_v10_0_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static void diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c index f154c24499c8..c5680ff4ab9f 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c @@ -2851,9 +2851,9 @@ static int dce_v11_0_crtc_init(struct amdgpu_device *adev, int index) return 0; } -static int dce_v11_0_early_init(void *handle) +static int dce_v11_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->audio_endpt_rreg = &dce_v11_0_audio_endpt_rreg; adev->audio_endpt_wreg = &dce_v11_0_audio_endpt_wreg; @@ -2891,10 +2891,10 @@ static int dce_v11_0_early_init(void *handle) return 0; } -static int dce_v11_0_sw_init(void *handle) +static int dce_v11_0_sw_init(struct amdgpu_ip_block *ip_block) { int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->mode_info.num_crtc; i++) { r = amdgpu_irq_add_id(adev, AMDGPU_IRQ_CLIENTID_LEGACY, i + 1, &adev->crtc_irq); @@ -2971,9 +2971,9 @@ static int dce_v11_0_sw_init(void *handle) return 0; } -static int dce_v11_0_sw_fini(void *handle) +static int dce_v11_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; drm_edid_free(adev->mode_info.bios_hardcoded_edid); @@ -2989,10 +2989,10 @@ static int dce_v11_0_sw_fini(void *handle) return 0; } -static int dce_v11_0_hw_init(void *handle) +static int dce_v11_0_hw_init(struct amdgpu_ip_block *ip_block) { int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; dce_v11_0_init_golden_registers(adev); @@ -3025,10 +3025,10 @@ static int dce_v11_0_hw_init(void *handle) return 0; } -static int dce_v11_0_hw_fini(void *handle) +static int dce_v11_0_hw_fini(struct amdgpu_ip_block *ip_block) { int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; dce_v11_0_hpd_fini(adev); @@ -3043,9 +3043,9 @@ static int dce_v11_0_hw_fini(void *handle) return 0; } -static int dce_v11_0_suspend(void *handle) +static int dce_v11_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_display_suspend_helper(adev); @@ -3055,18 +3055,18 @@ static int dce_v11_0_suspend(void *handle) adev->mode_info.bl_level = amdgpu_atombios_encoder_get_backlight_level_from_reg(adev); - return dce_v11_0_hw_fini(handle); + return dce_v11_0_hw_fini(ip_block); } -static int dce_v11_0_resume(void *handle) +static int dce_v11_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int ret; amdgpu_atombios_encoder_set_backlight_level_to_reg(adev, adev->mode_info.bl_level); - ret = dce_v11_0_hw_init(handle); + ret = dce_v11_0_hw_init(ip_block); /* turn on the BL */ if (adev->mode_info.bl_encoder) { @@ -3086,15 +3086,10 @@ static bool dce_v11_0_is_idle(void *handle) return true; } -static int dce_v11_0_wait_for_idle(void *handle) -{ - return 0; -} - -static int dce_v11_0_soft_reset(void *handle) +static int dce_v11_0_soft_reset(struct amdgpu_ip_block *ip_block) { u32 srbm_soft_reset = 0, tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (dce_v11_0_is_display_hung(adev)) srbm_soft_reset |= SRBM_SOFT_RESET__SOFT_RESET_DC_MASK; @@ -3454,7 +3449,6 @@ static int dce_v11_0_set_powergating_state(void *handle, static const struct amd_ip_funcs dce_v11_0_ip_funcs = { .name = "dce_v11_0", .early_init = dce_v11_0_early_init, - .late_init = NULL, .sw_init = dce_v11_0_sw_init, .sw_fini = dce_v11_0_sw_fini, .hw_init = dce_v11_0_hw_init, @@ -3462,12 +3456,9 @@ static const struct amd_ip_funcs dce_v11_0_ip_funcs = { .suspend = dce_v11_0_suspend, .resume = dce_v11_0_resume, .is_idle = dce_v11_0_is_idle, - .wait_for_idle = dce_v11_0_wait_for_idle, .soft_reset = dce_v11_0_soft_reset, .set_clockgating_state = dce_v11_0_set_clockgating_state, .set_powergating_state = dce_v11_0_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static void diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c index a7fcb135827f..eb7de9122d99 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c @@ -2633,9 +2633,9 @@ static int dce_v6_0_crtc_init(struct amdgpu_device *adev, int index) return 0; } -static int dce_v6_0_early_init(void *handle) +static int dce_v6_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->audio_endpt_rreg = &dce_v6_0_audio_endpt_rreg; adev->audio_endpt_wreg = &dce_v6_0_audio_endpt_wreg; @@ -2664,11 +2664,11 @@ static int dce_v6_0_early_init(void *handle) return 0; } -static int dce_v6_0_sw_init(void *handle) +static int dce_v6_0_sw_init(struct amdgpu_ip_block *ip_block) { int r, i; bool ret; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->mode_info.num_crtc; i++) { r = amdgpu_irq_add_id(adev, AMDGPU_IRQ_CLIENTID_LEGACY, i + 1, &adev->crtc_irq); @@ -2743,9 +2743,9 @@ static int dce_v6_0_sw_init(void *handle) return r; } -static int dce_v6_0_sw_fini(void *handle) +static int dce_v6_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; drm_edid_free(adev->mode_info.bios_hardcoded_edid); @@ -2760,10 +2760,10 @@ static int dce_v6_0_sw_fini(void *handle) return 0; } -static int dce_v6_0_hw_init(void *handle) +static int dce_v6_0_hw_init(struct amdgpu_ip_block *ip_block) { int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* disable vga render */ dce_v6_0_set_vga_render_state(adev, false); @@ -2783,10 +2783,10 @@ static int dce_v6_0_hw_init(void *handle) return 0; } -static int dce_v6_0_hw_fini(void *handle) +static int dce_v6_0_hw_fini(struct amdgpu_ip_block *ip_block) { int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; dce_v6_0_hpd_fini(adev); @@ -2801,9 +2801,9 @@ static int dce_v6_0_hw_fini(void *handle) return 0; } -static int dce_v6_0_suspend(void *handle) +static int dce_v6_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_display_suspend_helper(adev); @@ -2812,18 +2812,18 @@ static int dce_v6_0_suspend(void *handle) adev->mode_info.bl_level = amdgpu_atombios_encoder_get_backlight_level_from_reg(adev); - return dce_v6_0_hw_fini(handle); + return dce_v6_0_hw_fini(ip_block); } -static int dce_v6_0_resume(void *handle) +static int dce_v6_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int ret; amdgpu_atombios_encoder_set_backlight_level_to_reg(adev, adev->mode_info.bl_level); - ret = dce_v6_0_hw_init(handle); + ret = dce_v6_0_hw_init(ip_block); /* turn on the BL */ if (adev->mode_info.bl_encoder) { @@ -2843,12 +2843,7 @@ static bool dce_v6_0_is_idle(void *handle) return true; } -static int dce_v6_0_wait_for_idle(void *handle) -{ - return 0; -} - -static int dce_v6_0_soft_reset(void *handle) +static int dce_v6_0_soft_reset(struct amdgpu_ip_block *ip_block) { DRM_INFO("xxxx: dce_v6_0_soft_reset --- no impl!!\n"); return 0; @@ -3144,7 +3139,6 @@ static int dce_v6_0_set_powergating_state(void *handle, static const struct amd_ip_funcs dce_v6_0_ip_funcs = { .name = "dce_v6_0", .early_init = dce_v6_0_early_init, - .late_init = NULL, .sw_init = dce_v6_0_sw_init, .sw_fini = dce_v6_0_sw_fini, .hw_init = dce_v6_0_hw_init, @@ -3152,12 +3146,9 @@ static const struct amd_ip_funcs dce_v6_0_ip_funcs = { .suspend = dce_v6_0_suspend, .resume = dce_v6_0_resume, .is_idle = dce_v6_0_is_idle, - .wait_for_idle = dce_v6_0_wait_for_idle, .soft_reset = dce_v6_0_soft_reset, .set_clockgating_state = dce_v6_0_set_clockgating_state, .set_powergating_state = dce_v6_0_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static void diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c index 77ac3f114d24..04b79ff87f75 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c @@ -2644,9 +2644,9 @@ static int dce_v8_0_crtc_init(struct amdgpu_device *adev, int index) return 0; } -static int dce_v8_0_early_init(void *handle) +static int dce_v8_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->audio_endpt_rreg = &dce_v8_0_audio_endpt_rreg; adev->audio_endpt_wreg = &dce_v8_0_audio_endpt_wreg; @@ -2680,10 +2680,10 @@ static int dce_v8_0_early_init(void *handle) return 0; } -static int dce_v8_0_sw_init(void *handle) +static int dce_v8_0_sw_init(struct amdgpu_ip_block *ip_block) { int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->mode_info.num_crtc; i++) { r = amdgpu_irq_add_id(adev, AMDGPU_IRQ_CLIENTID_LEGACY, i + 1, &adev->crtc_irq); @@ -2764,9 +2764,9 @@ static int dce_v8_0_sw_init(void *handle) return 0; } -static int dce_v8_0_sw_fini(void *handle) +static int dce_v8_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; drm_edid_free(adev->mode_info.bios_hardcoded_edid); @@ -2782,10 +2782,10 @@ static int dce_v8_0_sw_fini(void *handle) return 0; } -static int dce_v8_0_hw_init(void *handle) +static int dce_v8_0_hw_init(struct amdgpu_ip_block *ip_block) { int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* disable vga render */ dce_v8_0_set_vga_render_state(adev, false); @@ -2805,10 +2805,10 @@ static int dce_v8_0_hw_init(void *handle) return 0; } -static int dce_v8_0_hw_fini(void *handle) +static int dce_v8_0_hw_fini(struct amdgpu_ip_block *ip_block) { int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; dce_v8_0_hpd_fini(adev); @@ -2823,9 +2823,9 @@ static int dce_v8_0_hw_fini(void *handle) return 0; } -static int dce_v8_0_suspend(void *handle) +static int dce_v8_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_display_suspend_helper(adev); @@ -2835,18 +2835,18 @@ static int dce_v8_0_suspend(void *handle) adev->mode_info.bl_level = amdgpu_atombios_encoder_get_backlight_level_from_reg(adev); - return dce_v8_0_hw_fini(handle); + return dce_v8_0_hw_fini(ip_block); } -static int dce_v8_0_resume(void *handle) +static int dce_v8_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int ret; amdgpu_atombios_encoder_set_backlight_level_to_reg(adev, adev->mode_info.bl_level); - ret = dce_v8_0_hw_init(handle); + ret = dce_v8_0_hw_init(ip_block); /* turn on the BL */ if (adev->mode_info.bl_encoder) { @@ -2866,15 +2866,10 @@ static bool dce_v8_0_is_idle(void *handle) return true; } -static int dce_v8_0_wait_for_idle(void *handle) -{ - return 0; -} - -static int dce_v8_0_soft_reset(void *handle) +static int dce_v8_0_soft_reset(struct amdgpu_ip_block *ip_block) { u32 srbm_soft_reset = 0, tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (dce_v8_0_is_display_hung(adev)) srbm_soft_reset |= SRBM_SOFT_RESET__SOFT_RESET_DC_MASK; @@ -3232,7 +3227,6 @@ static int dce_v8_0_set_powergating_state(void *handle, static const struct amd_ip_funcs dce_v8_0_ip_funcs = { .name = "dce_v8_0", .early_init = dce_v8_0_early_init, - .late_init = NULL, .sw_init = dce_v8_0_sw_init, .sw_fini = dce_v8_0_sw_fini, .hw_init = dce_v8_0_hw_init, @@ -3240,12 +3234,9 @@ static const struct amd_ip_funcs dce_v8_0_ip_funcs = { .suspend = dce_v8_0_suspend, .resume = dce_v8_0_resume, .is_idle = dce_v8_0_is_idle, - .wait_for_idle = dce_v8_0_wait_for_idle, .soft_reset = dce_v8_0_soft_reset, .set_clockgating_state = dce_v8_0_set_clockgating_state, .set_powergating_state = dce_v8_0_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static void diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c index 45ed97038df0..24dce803a829 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c @@ -3677,13 +3677,19 @@ static int gfx_v10_0_set_powergating_state(void *handle, enum amd_powergating_state state); static void gfx10_kiq_set_resources(struct amdgpu_ring *kiq_ring, uint64_t queue_mask) { + struct amdgpu_device *adev = kiq_ring->adev; + u64 shader_mc_addr; + + /* Cleaner shader MC address */ + shader_mc_addr = adev->gfx.cleaner_shader_gpu_addr >> 8; + amdgpu_ring_write(kiq_ring, PACKET3(PACKET3_SET_RESOURCES, 6)); amdgpu_ring_write(kiq_ring, PACKET3_SET_RESOURCES_VMID_MASK(0) | PACKET3_SET_RESOURCES_QUEUE_TYPE(0)); /* vmid_mask:0 queue_type:0 (KIQ) */ amdgpu_ring_write(kiq_ring, lower_32_bits(queue_mask)); /* queue mask lo */ amdgpu_ring_write(kiq_ring, upper_32_bits(queue_mask)); /* queue mask hi */ - amdgpu_ring_write(kiq_ring, 0); /* gws mask lo */ - amdgpu_ring_write(kiq_ring, 0); /* gws mask hi */ + amdgpu_ring_write(kiq_ring, lower_32_bits(shader_mc_addr)); /* cleaner shader addr lo */ + amdgpu_ring_write(kiq_ring, upper_32_bits(shader_mc_addr)); /* cleaner shader addr hi */ amdgpu_ring_write(kiq_ring, 0); /* oac mask */ amdgpu_ring_write(kiq_ring, 0); /* gds heap base:0, gds heap size:0 */ } @@ -4683,11 +4689,11 @@ static void gfx_v10_0_alloc_ip_dump(struct amdgpu_device *adev) } } -static int gfx_v10_0_sw_init(void *handle) +static int gfx_v10_0_sw_init(struct amdgpu_ip_block *ip_block) { int i, j, k, r, ring_id = 0; int xcc_id = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { case IP_VERSION(10, 1, 10): @@ -4726,6 +4732,11 @@ static int gfx_v10_0_sw_init(void *handle) adev->gfx.mec.num_queue_per_pipe = 8; break; } + switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { + default: + adev->gfx.enable_cleaner_shader = false; + break; + } /* KIQ event */ r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, @@ -4814,6 +4825,11 @@ static int gfx_v10_0_sw_init(void *handle) } } } + /* TODO: Add queue reset mask when FW fully supports it */ + adev->gfx.gfx_supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->gfx.gfx_ring[0]); + adev->gfx.compute_supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->gfx.compute_ring[0]); r = amdgpu_gfx_kiq_init(adev, GFX10_MEC_HPD_SIZE, 0); if (r) { @@ -4842,6 +4858,10 @@ static int gfx_v10_0_sw_init(void *handle) gfx_v10_0_alloc_ip_dump(adev); + r = amdgpu_gfx_sysfs_init(adev); + if (r) + return r; + return 0; } @@ -4866,10 +4886,10 @@ static void gfx_v10_0_me_fini(struct amdgpu_device *adev) (void **)&adev->gfx.me.me_fw_ptr); } -static int gfx_v10_0_sw_fini(void *handle) +static int gfx_v10_0_sw_fini(struct amdgpu_ip_block *ip_block) { int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->gfx.num_gfx_rings; i++) amdgpu_ring_fini(&adev->gfx.gfx_ring[i]); @@ -4881,6 +4901,8 @@ static int gfx_v10_0_sw_fini(void *handle) amdgpu_gfx_kiq_free_ring(&adev->gfx.kiq[0].ring); amdgpu_gfx_kiq_fini(adev, 0); + amdgpu_gfx_cleaner_shader_sw_fini(adev); + gfx_v10_0_pfp_fini(adev); gfx_v10_0_ce_fini(adev); gfx_v10_0_me_fini(adev); @@ -4891,6 +4913,7 @@ static int gfx_v10_0_sw_fini(void *handle) gfx_v10_0_rlc_backdoor_autoload_buffer_fini(adev); gfx_v10_0_free_microcode(adev); + amdgpu_gfx_sysfs_fini(adev); kfree(adev->gfx.ip_dump_core); kfree(adev->gfx.ip_dump_compute_queues); @@ -6374,7 +6397,7 @@ static int gfx_v10_0_cp_gfx_resume(struct amdgpu_device *adev) WREG32_SOC15(GC, 0, mmCP_RB0_WPTR, lower_32_bits(ring->wptr)); WREG32_SOC15(GC, 0, mmCP_RB0_WPTR_HI, upper_32_bits(ring->wptr)); - /* set the wb address wether it's enabled or not */ + /* set the wb address whether it's enabled or not */ rptr_addr = ring->rptr_gpu_addr; WREG32_SOC15(GC, 0, mmCP_RB0_RPTR_ADDR, lower_32_bits(rptr_addr)); WREG32_SOC15(GC, 0, mmCP_RB0_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & @@ -6412,7 +6435,7 @@ static int gfx_v10_0_cp_gfx_resume(struct amdgpu_device *adev) ring->wptr = 0; WREG32_SOC15(GC, 0, mmCP_RB1_WPTR, lower_32_bits(ring->wptr)); WREG32_SOC15(GC, 0, mmCP_RB1_WPTR_HI, upper_32_bits(ring->wptr)); - /* Set the wb address wether it's enabled or not */ + /* Set the wb address whether it's enabled or not */ rptr_addr = ring->rptr_gpu_addr; WREG32_SOC15(GC, 0, mmCP_RB1_RPTR_ADDR, lower_32_bits(rptr_addr)); WREG32_SOC15(GC, 0, mmCP_RB1_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & @@ -7366,14 +7389,17 @@ static void gfx_v10_0_disable_gpa_mode(struct amdgpu_device *adev) WREG32_SOC15(GC, 0, mmCPG_PSP_DEBUG, data); } -static int gfx_v10_0_hw_init(void *handle) +static int gfx_v10_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (!amdgpu_emu_mode) gfx_v10_0_init_golden_registers(adev); + amdgpu_gfx_cleaner_shader_init(adev, adev->gfx.cleaner_shader_size, + adev->gfx.cleaner_shader_ptr); + if (adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT) { /** * For gfx 10, rlc firmware loading relies on smu firmware is @@ -7418,9 +7444,9 @@ static int gfx_v10_0_hw_init(void *handle) return r; } -static int gfx_v10_0_hw_fini(void *handle) +static int gfx_v10_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0); amdgpu_irq_put(adev, &adev->gfx.priv_inst_irq, 0); @@ -7431,7 +7457,7 @@ static int gfx_v10_0_hw_fini(void *handle) * otherwise the gfxoff disallowing will be failed to set. */ if (amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(10, 3, 1)) - gfx_v10_0_set_powergating_state(handle, AMD_PG_STATE_UNGATE); + gfx_v10_0_set_powergating_state(ip_block->adev, AMD_PG_STATE_UNGATE); if (!adev->no_hw_access) { if (amdgpu_async_gfx_ring) { @@ -7456,14 +7482,14 @@ static int gfx_v10_0_hw_fini(void *handle) return 0; } -static int gfx_v10_0_suspend(void *handle) +static int gfx_v10_0_suspend(struct amdgpu_ip_block *ip_block) { - return gfx_v10_0_hw_fini(handle); + return gfx_v10_0_hw_fini(ip_block); } -static int gfx_v10_0_resume(void *handle) +static int gfx_v10_0_resume(struct amdgpu_ip_block *ip_block) { - return gfx_v10_0_hw_init(handle); + return gfx_v10_0_hw_init(ip_block); } static bool gfx_v10_0_is_idle(void *handle) @@ -7477,11 +7503,11 @@ static bool gfx_v10_0_is_idle(void *handle) return true; } -static int gfx_v10_0_wait_for_idle(void *handle) +static int gfx_v10_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned int i; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { /* read MC_STATUS */ @@ -7495,11 +7521,11 @@ static int gfx_v10_0_wait_for_idle(void *handle) return -ETIMEDOUT; } -static int gfx_v10_0_soft_reset(void *handle) +static int gfx_v10_0_soft_reset(struct amdgpu_ip_block *ip_block) { u32 grbm_soft_reset = 0; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* GRBM_STATUS */ tmp = RREG32_SOC15(GC, 0, mmGRBM_STATUS); @@ -7678,9 +7704,9 @@ static void gfx_v10_0_ring_emit_gds_switch(struct amdgpu_ring *ring, (1 << (oa_size + oa_base)) - (1 << oa_base)); } -static int gfx_v10_0_early_init(void *handle) +static int gfx_v10_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->gfx.funcs = &gfx_v10_0_gfx_funcs; @@ -7722,9 +7748,9 @@ static int gfx_v10_0_early_init(void *handle) return gfx_v10_0_init_microcode(adev); } -static int gfx_v10_0_late_init(void *handle) +static int gfx_v10_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_irq_get(adev, &adev->gfx.priv_reg_irq, 0); @@ -9402,8 +9428,6 @@ static void gfx_v10_0_emit_mem_sync(struct amdgpu_ring *ring) static void gfx_v10_ring_insert_nop(struct amdgpu_ring *ring, uint32_t num_nop) { - int i; - /* Header itself is a NOP packet */ if (num_nop == 1) { amdgpu_ring_write(ring, ring->funcs->nop); @@ -9414,8 +9438,7 @@ static void gfx_v10_ring_insert_nop(struct amdgpu_ring *ring, uint32_t num_nop) amdgpu_ring_write(ring, PACKET3(PACKET3_NOP, min(num_nop - 2, 0x3ffe))); /* Header is at index 0, followed by num_nops - 1 NOP packet's */ - for (i = 1; i < num_nop; i++) - amdgpu_ring_write(ring, ring->funcs->nop); + amdgpu_ring_insert_nop(ring, num_nop - 1); } static int gfx_v10_0_reset_kgq(struct amdgpu_ring *ring, unsigned int vmid) @@ -9568,9 +9591,9 @@ static int gfx_v10_0_reset_kcq(struct amdgpu_ring *ring, return amdgpu_ring_test_ring(ring); } -static void gfx_v10_ip_print(void *handle, struct drm_printer *p) +static void gfx_v10_ip_print(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t i, j, k, reg, index = 0; uint32_t reg_count = ARRAY_SIZE(gc_reg_list_10_1); @@ -9632,9 +9655,9 @@ static void gfx_v10_ip_print(void *handle, struct drm_printer *p) } } -static void gfx_v10_ip_dump(void *handle) +static void gfx_v10_ip_dump(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t i, j, k, reg, index = 0; uint32_t reg_count = ARRAY_SIZE(gc_reg_list_10_1); @@ -9699,6 +9722,13 @@ static void gfx_v10_ip_dump(void *handle) amdgpu_gfx_off_ctrl(adev, true); } +static void gfx_v10_0_ring_emit_cleaner_shader(struct amdgpu_ring *ring) +{ + /* Emit the cleaner shader */ + amdgpu_ring_write(ring, PACKET3(PACKET3_RUN_CLEANER_SHADER, 0)); + amdgpu_ring_write(ring, 0); /* RESERVED field, programmed to zero */ +} + static const struct amd_ip_funcs gfx_v10_0_ip_funcs = { .name = "gfx_v10_0", .early_init = gfx_v10_0_early_init, @@ -9749,7 +9779,8 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_gfx = { 5 + /* HDP_INVL */ 8 + 8 + /* FENCE x2 */ 2 + /* SWITCH_BUFFER */ - 8, /* gfx_v10_0_emit_mem_sync */ + 8 + /* gfx_v10_0_emit_mem_sync */ + 2, /* gfx_v10_0_ring_emit_cleaner_shader */ .emit_ib_size = 4, /* gfx_v10_0_ring_emit_ib_gfx */ .emit_ib = gfx_v10_0_ring_emit_ib_gfx, .emit_fence = gfx_v10_0_ring_emit_fence, @@ -9772,6 +9803,9 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_gfx = { .soft_recovery = gfx_v10_0_ring_soft_recovery, .emit_mem_sync = gfx_v10_0_emit_mem_sync, .reset = gfx_v10_0_reset_kgq, + .emit_cleaner_shader = gfx_v10_0_ring_emit_cleaner_shader, + .begin_use = amdgpu_gfx_enforce_isolation_ring_begin_use, + .end_use = amdgpu_gfx_enforce_isolation_ring_end_use, }; static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_compute = { @@ -9791,7 +9825,8 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_compute = { SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 + 2 + /* gfx_v10_0_ring_emit_vm_flush */ 8 + 8 + 8 + /* gfx_v10_0_ring_emit_fence x3 for user fence, vm fence */ - 8, /* gfx_v10_0_emit_mem_sync */ + 8 + /* gfx_v10_0_emit_mem_sync */ + 2, /* gfx_v10_0_ring_emit_cleaner_shader */ .emit_ib_size = 7, /* gfx_v10_0_ring_emit_ib_compute */ .emit_ib = gfx_v10_0_ring_emit_ib_compute, .emit_fence = gfx_v10_0_ring_emit_fence, @@ -9809,6 +9844,9 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_compute = { .soft_recovery = gfx_v10_0_ring_soft_recovery, .emit_mem_sync = gfx_v10_0_emit_mem_sync, .reset = gfx_v10_0_reset_kcq, + .emit_cleaner_shader = gfx_v10_0_ring_emit_cleaner_shader, + .begin_use = amdgpu_gfx_enforce_isolation_ring_begin_use, + .end_use = amdgpu_gfx_enforce_isolation_ring_end_use, }; static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_kiq = { diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c index d3e8be82a172..2ae058a224f4 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c @@ -46,6 +46,7 @@ #include "clearstate_gfx11.h" #include "v11_structs.h" #include "gfx_v11_0.h" +#include "gfx_v11_0_cleaner_shader.h" #include "gfx_v11_0_3.h" #include "nbio_v4_3.h" #include "mes_v11_0.h" @@ -293,14 +294,20 @@ static void gfx_v11_0_update_perf_clk(struct amdgpu_device *adev, static void gfx11_kiq_set_resources(struct amdgpu_ring *kiq_ring, uint64_t queue_mask) { + struct amdgpu_device *adev = kiq_ring->adev; + u64 shader_mc_addr; + + /* Cleaner shader MC address */ + shader_mc_addr = adev->gfx.cleaner_shader_gpu_addr >> 8; + amdgpu_ring_write(kiq_ring, PACKET3(PACKET3_SET_RESOURCES, 6)); amdgpu_ring_write(kiq_ring, PACKET3_SET_RESOURCES_VMID_MASK(0) | PACKET3_SET_RESOURCES_UNMAP_LATENTY(0xa) | /* unmap_latency: 0xa (~ 1s) */ PACKET3_SET_RESOURCES_QUEUE_TYPE(0)); /* vmid_mask:0 queue_type:0 (KIQ) */ amdgpu_ring_write(kiq_ring, lower_32_bits(queue_mask)); /* queue mask lo */ amdgpu_ring_write(kiq_ring, upper_32_bits(queue_mask)); /* queue mask hi */ - amdgpu_ring_write(kiq_ring, 0); /* gws mask lo */ - amdgpu_ring_write(kiq_ring, 0); /* gws mask hi */ + amdgpu_ring_write(kiq_ring, lower_32_bits(shader_mc_addr)); /* cleaner shader addr lo */ + amdgpu_ring_write(kiq_ring, upper_32_bits(shader_mc_addr)); /* cleaner shader addr hi */ amdgpu_ring_write(kiq_ring, 0); /* oac mask */ amdgpu_ring_write(kiq_ring, 0); /* gds heap base:0, gds heap size:0 */ } @@ -483,8 +490,6 @@ static void gfx_v11_0_wait_reg_mem(struct amdgpu_ring *ring, int eng_sel, static void gfx_v11_ring_insert_nop(struct amdgpu_ring *ring, uint32_t num_nop) { - int i; - /* Header itself is a NOP packet */ if (num_nop == 1) { amdgpu_ring_write(ring, ring->funcs->nop); @@ -495,8 +500,7 @@ static void gfx_v11_ring_insert_nop(struct amdgpu_ring *ring, uint32_t num_nop) amdgpu_ring_write(ring, PACKET3(PACKET3_NOP, min(num_nop - 2, 0x3ffe))); /* Header is at index 0, followed by num_nops - 1 NOP packet's */ - for (i = 1; i < num_nop; i++) - amdgpu_ring_write(ring, ring->funcs->nop); + amdgpu_ring_insert_nop(ring, num_nop - 1); } static int gfx_v11_0_ring_test_ring(struct amdgpu_ring *ring) @@ -1536,11 +1540,11 @@ static void gfx_v11_0_alloc_ip_dump(struct amdgpu_device *adev) } } -static int gfx_v11_0_sw_init(void *handle) +static int gfx_v11_0_sw_init(struct amdgpu_ip_block *ip_block) { int i, j, k, r, ring_id = 0; int xcc_id = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { case IP_VERSION(11, 0, 0): @@ -1575,6 +1579,29 @@ static int gfx_v11_0_sw_init(void *handle) break; } + switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { + case IP_VERSION(11, 0, 0): + case IP_VERSION(11, 0, 2): + case IP_VERSION(11, 0, 3): + adev->gfx.cleaner_shader_ptr = gfx_11_0_3_cleaner_shader_hex; + adev->gfx.cleaner_shader_size = sizeof(gfx_11_0_3_cleaner_shader_hex); + if (adev->gfx.me_fw_version >= 2280 && + adev->gfx.pfp_fw_version >= 2370 && + adev->gfx.mec_fw_version >= 2450 && + adev->mes.fw_version[0] >= 99) { + adev->gfx.enable_cleaner_shader = true; + r = amdgpu_gfx_cleaner_shader_sw_init(adev, adev->gfx.cleaner_shader_size); + if (r) { + adev->gfx.enable_cleaner_shader = false; + dev_err(adev->dev, "Failed to initialize cleaner shader\n"); + } + } + break; + default: + adev->gfx.enable_cleaner_shader = false; + break; + } + /* Enable CG flag in one VF mode for enabling RLC safe mode enter/exit */ if (amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(11, 0, 3) && amdgpu_sriov_is_pp_one_vf(adev)) @@ -1666,6 +1693,24 @@ static int gfx_v11_0_sw_init(void *handle) } } + adev->gfx.gfx_supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->gfx.gfx_ring[0]); + adev->gfx.compute_supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->gfx.compute_ring[0]); + switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { + case IP_VERSION(11, 0, 0): + case IP_VERSION(11, 0, 2): + case IP_VERSION(11, 0, 3): + if ((adev->gfx.me_fw_version >= 2280) && + (adev->gfx.mec_fw_version >= 2410)) { + adev->gfx.compute_supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE; + adev->gfx.gfx_supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE; + } + break; + default: + break; + } + if (!adev->enable_mes_kiq) { r = amdgpu_gfx_kiq_init(adev, GFX11_MEC_HPD_SIZE, 0); if (r) { @@ -1700,6 +1745,10 @@ static int gfx_v11_0_sw_init(void *handle) gfx_v11_0_alloc_ip_dump(adev); + r = amdgpu_gfx_sysfs_init(adev); + if (r) + return r; + return 0; } @@ -1732,10 +1781,10 @@ static void gfx_v11_0_rlc_autoload_buffer_fini(struct amdgpu_device *adev) (void **)&adev->gfx.rlc.rlc_autoload_ptr); } -static int gfx_v11_0_sw_fini(void *handle) +static int gfx_v11_0_sw_fini(struct amdgpu_ip_block *ip_block) { int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->gfx.num_gfx_rings; i++) amdgpu_ring_fini(&adev->gfx.gfx_ring[i]); @@ -1749,6 +1798,8 @@ static int gfx_v11_0_sw_fini(void *handle) amdgpu_gfx_kiq_fini(adev, 0); } + amdgpu_gfx_cleaner_shader_sw_fini(adev); + gfx_v11_0_pfp_fini(adev); gfx_v11_0_me_fini(adev); gfx_v11_0_rlc_fini(adev); @@ -1759,6 +1810,8 @@ static int gfx_v11_0_sw_fini(void *handle) gfx_v11_0_free_microcode(adev); + amdgpu_gfx_sysfs_fini(adev); + kfree(adev->gfx.ip_dump_core); kfree(adev->gfx.ip_dump_compute_queues); kfree(adev->gfx.ip_dump_gfx_queues); @@ -1893,8 +1946,10 @@ static void gfx_v11_0_init_compute_vmid(struct amdgpu_device *adev) soc21_grbm_select(adev, 0, 0, 0, 0); mutex_unlock(&adev->srbm_mutex); - /* Initialize all compute VMIDs to have no GDS, GWS, or OA - acccess. These should be enabled by FW for target VMIDs. */ + /* + * Initialize all compute VMIDs to have no GDS, GWS, or OA + * access. These should be enabled by FW for target VMIDs. + */ for (i = adev->vm_manager.first_kfd_vmid; i < AMDGPU_NUM_VMID; i++) { WREG32_SOC15_OFFSET(GC, 0, regGDS_VMID0_BASE, 2 * i, 0); WREG32_SOC15_OFFSET(GC, 0, regGDS_VMID0_SIZE, 2 * i, 0); @@ -3555,7 +3610,7 @@ static int gfx_v11_0_cp_gfx_resume(struct amdgpu_device *adev) WREG32_SOC15(GC, 0, regCP_RB0_WPTR, lower_32_bits(ring->wptr)); WREG32_SOC15(GC, 0, regCP_RB0_WPTR_HI, upper_32_bits(ring->wptr)); - /* set the wb address wether it's enabled or not */ + /* set the wb address whether it's enabled or not */ rptr_addr = ring->rptr_gpu_addr; WREG32_SOC15(GC, 0, regCP_RB0_RPTR_ADDR, lower_32_bits(rptr_addr)); WREG32_SOC15(GC, 0, regCP_RB0_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & @@ -3593,7 +3648,7 @@ static int gfx_v11_0_cp_gfx_resume(struct amdgpu_device *adev) ring->wptr = 0; WREG32_SOC15(GC, 0, regCP_RB1_WPTR, lower_32_bits(ring->wptr)); WREG32_SOC15(GC, 0, regCP_RB1_WPTR_HI, upper_32_bits(ring->wptr)); - /* Set the wb address wether it's enabled or not */ + /* Set the wb address whether it's enabled or not */ rptr_addr = ring->rptr_gpu_addr; WREG32_SOC15(GC, 0, regCP_RB1_RPTR_ADDR, lower_32_bits(rptr_addr)); WREG32_SOC15(GC, 0, regCP_RB1_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & @@ -4568,10 +4623,13 @@ static void gfx_v11_0_disable_gpa_mode(struct amdgpu_device *adev) WREG32_SOC15(GC, 0, regCPG_PSP_DEBUG, data); } -static int gfx_v11_0_hw_init(void *handle) +static int gfx_v11_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; + + amdgpu_gfx_cleaner_shader_init(adev, adev->gfx.cleaner_shader_size, + adev->gfx.cleaner_shader_ptr); if (adev->firmware.load_type == AMDGPU_FW_LOAD_RLC_BACKDOOR_AUTO) { if (adev->gfx.imu.funcs) { @@ -4665,9 +4723,9 @@ static int gfx_v11_0_hw_init(void *handle) return r; } -static int gfx_v11_0_hw_fini(void *handle) +static int gfx_v11_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0); amdgpu_irq_put(adev, &adev->gfx.priv_inst_irq, 0); @@ -4703,14 +4761,14 @@ static int gfx_v11_0_hw_fini(void *handle) return 0; } -static int gfx_v11_0_suspend(void *handle) +static int gfx_v11_0_suspend(struct amdgpu_ip_block *ip_block) { - return gfx_v11_0_hw_fini(handle); + return gfx_v11_0_hw_fini(ip_block); } -static int gfx_v11_0_resume(void *handle) +static int gfx_v11_0_resume(struct amdgpu_ip_block *ip_block) { - return gfx_v11_0_hw_init(handle); + return gfx_v11_0_hw_init(ip_block); } static bool gfx_v11_0_is_idle(void *handle) @@ -4724,11 +4782,11 @@ static bool gfx_v11_0_is_idle(void *handle) return true; } -static int gfx_v11_0_wait_for_idle(void *handle) +static int gfx_v11_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { /* read MC_STATUS */ @@ -4774,12 +4832,12 @@ int gfx_v11_0_request_gfx_index_mutex(struct amdgpu_device *adev, return 0; } -static int gfx_v11_0_soft_reset(void *handle) +static int gfx_v11_0_soft_reset(struct amdgpu_ip_block *ip_block) { u32 grbm_soft_reset = 0; u32 tmp; int r, i, j, k; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_gfx_rlc_enter_safe_mode(adev, 0); @@ -4905,10 +4963,10 @@ static int gfx_v11_0_soft_reset(void *handle) return gfx_v11_0_cp_resume(adev); } -static bool gfx_v11_0_check_soft_reset(void *handle) +static bool gfx_v11_0_check_soft_reset(struct amdgpu_ip_block *ip_block) { int i, r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; long tmo = msecs_to_jiffies(1000); @@ -4929,12 +4987,13 @@ static bool gfx_v11_0_check_soft_reset(void *handle) return false; } -static int gfx_v11_0_post_soft_reset(void *handle) +static int gfx_v11_0_post_soft_reset(struct amdgpu_ip_block *ip_block) { + struct amdgpu_device *adev = ip_block->adev; /** * GFX soft reset will impact MES, need resume MES when do GFX soft reset */ - return amdgpu_mes_resume((struct amdgpu_device *)handle); + return amdgpu_mes_resume(adev); } static uint64_t gfx_v11_0_get_gpu_clock_counter(struct amdgpu_device *adev) @@ -4995,9 +5054,9 @@ static void gfx_v11_0_ring_emit_gds_switch(struct amdgpu_ring *ring, (1 << (oa_size + oa_base)) - (1 << oa_base)); } -static int gfx_v11_0_early_init(void *handle) +static int gfx_v11_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->gfx.funcs = &gfx_v11_0_gfx_funcs; @@ -5018,9 +5077,9 @@ static int gfx_v11_0_early_init(void *handle) return gfx_v11_0_init_microcode(adev); } -static int gfx_v11_0_late_init(void *handle) +static int gfx_v11_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_irq_get(adev, &adev->gfx.priv_reg_irq, 0); @@ -6639,9 +6698,9 @@ static int gfx_v11_0_reset_kcq(struct amdgpu_ring *ring, unsigned int vmid) return amdgpu_ring_test_ring(ring); } -static void gfx_v11_ip_print(void *handle, struct drm_printer *p) +static void gfx_v11_ip_print(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t i, j, k, reg, index = 0; uint32_t reg_count = ARRAY_SIZE(gc_reg_list_11_0); @@ -6703,9 +6762,9 @@ static void gfx_v11_ip_print(void *handle, struct drm_printer *p) } } -static void gfx_v11_ip_dump(void *handle) +static void gfx_v11_ip_dump(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t i, j, k, reg, index = 0; uint32_t reg_count = ARRAY_SIZE(gc_reg_list_11_0); @@ -6769,6 +6828,13 @@ static void gfx_v11_ip_dump(void *handle) amdgpu_gfx_off_ctrl(adev, true); } +static void gfx_v11_0_ring_emit_cleaner_shader(struct amdgpu_ring *ring) +{ + /* Emit the cleaner shader */ + amdgpu_ring_write(ring, PACKET3(PACKET3_RUN_CLEANER_SHADER, 0)); + amdgpu_ring_write(ring, 0); /* RESERVED field, programmed to zero */ +} + static const struct amd_ip_funcs gfx_v11_0_ip_funcs = { .name = "gfx_v11_0", .early_init = gfx_v11_0_early_init, @@ -6818,7 +6884,8 @@ static const struct amdgpu_ring_funcs gfx_v11_0_ring_funcs_gfx = { 5 + /* HDP_INVL */ 22 + /* SET_Q_PREEMPTION_MODE */ 8 + 8 + /* FENCE x2 */ - 8, /* gfx_v11_0_emit_mem_sync */ + 8 + /* gfx_v11_0_emit_mem_sync */ + 2, /* gfx_v11_0_ring_emit_cleaner_shader */ .emit_ib_size = 4, /* gfx_v11_0_ring_emit_ib_gfx */ .emit_ib = gfx_v11_0_ring_emit_ib_gfx, .emit_fence = gfx_v11_0_ring_emit_fence, @@ -6841,6 +6908,9 @@ static const struct amdgpu_ring_funcs gfx_v11_0_ring_funcs_gfx = { .soft_recovery = gfx_v11_0_ring_soft_recovery, .emit_mem_sync = gfx_v11_0_emit_mem_sync, .reset = gfx_v11_0_reset_kgq, + .emit_cleaner_shader = gfx_v11_0_ring_emit_cleaner_shader, + .begin_use = amdgpu_gfx_enforce_isolation_ring_begin_use, + .end_use = amdgpu_gfx_enforce_isolation_ring_end_use, }; static const struct amdgpu_ring_funcs gfx_v11_0_ring_funcs_compute = { @@ -6861,7 +6931,8 @@ static const struct amdgpu_ring_funcs gfx_v11_0_ring_funcs_compute = { SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 + 2 + /* gfx_v11_0_ring_emit_vm_flush */ 8 + 8 + 8 + /* gfx_v11_0_ring_emit_fence x3 for user fence, vm fence */ - 8, /* gfx_v11_0_emit_mem_sync */ + 8 + /* gfx_v11_0_emit_mem_sync */ + 2, /* gfx_v11_0_ring_emit_cleaner_shader */ .emit_ib_size = 7, /* gfx_v11_0_ring_emit_ib_compute */ .emit_ib = gfx_v11_0_ring_emit_ib_compute, .emit_fence = gfx_v11_0_ring_emit_fence, @@ -6879,6 +6950,9 @@ static const struct amdgpu_ring_funcs gfx_v11_0_ring_funcs_compute = { .soft_recovery = gfx_v11_0_ring_soft_recovery, .emit_mem_sync = gfx_v11_0_emit_mem_sync, .reset = gfx_v11_0_reset_kcq, + .emit_cleaner_shader = gfx_v11_0_ring_emit_cleaner_shader, + .begin_use = amdgpu_gfx_enforce_isolation_ring_begin_use, + .end_use = amdgpu_gfx_enforce_isolation_ring_end_use, }; static const struct amdgpu_ring_funcs gfx_v11_0_ring_funcs_kiq = { diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0_3_cleaner_shader.asm b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0_3_cleaner_shader.asm new file mode 100644 index 000000000000..9b90b66368c7 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0_3_cleaner_shader.asm @@ -0,0 +1,118 @@ +/* SPDX-License-Identifier: MIT */ +/* + * Copyright 2024 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +// This shader is to clean LDS, SGPRs and VGPRs. It is first 64 Dwords or 256 bytes of 192 Dwords cleaner shader. +//To turn this shader program on for complitaion change this to main and lower shader main to main_1 + +// Navi3 : Clear SGPRs, VGPRs and LDS +// Launch 32 waves per CU (16 per SIMD) as a workgroup (threadgroup) to fill every wave slot +// Waves are "wave32" and have 64 VGPRs each, which uses all 1024 VGPRs per SIMD +// Waves are launched in "CU" mode, and the workgroup shares 64KB of LDS (half of the WGP's LDS) +// It takes 2 workgroups to use all of LDS: one on each CU of the WGP +// Each wave clears SGPRs 0 - 107 +// Each wave clears VGPRs 0 - 63 +// The first wave of the workgroup clears its 64KB of LDS +// The shader starts with "S_BARRIER" to ensure SPI has launched all waves of the workgroup +// before any wave in the workgroup could end. Without this, it is possible not all SGPRs get cleared. + +shader main + asic(GFX11) + type(CS) + wave_size(32) +// Note: original source code from SQ team + +// Takes about 2500 clocks to run. +// (theorhetical fastest = 1024clks vgpr + 640lds = 1660 clks) +// + S_BARRIER + + // + // CLEAR VGPRs + // + s_mov_b32 m0, 0x00000058 // Loop 96/8=12 times (loop unrolled for performance) + +label_0005: + v_movreld_b32 v0, 0 + v_movreld_b32 v1, 0 + v_movreld_b32 v2, 0 + v_movreld_b32 v3, 0 + v_movreld_b32 v4, 0 + v_movreld_b32 v5, 0 + v_movreld_b32 v6, 0 + v_movreld_b32 v7, 0 + s_sub_u32 m0, m0, 8 + s_cbranch_scc0 label_0005 + // + // + + s_mov_b32 s2, 0x80000000 // Bit31 is first_wave + s_and_b32 s2, s2, s0 // sgpr0 has tg_size (first_wave) term as in ucode only COMPUTE_PGM_RSRC2.tg_size_en is set + s_cbranch_scc0 label_0023 // Clean LDS if its first wave of ThreadGroup/WorkGroup + // CLEAR LDS + // + s_mov_b32 exec_lo, 0xffffffff + s_mov_b32 exec_hi, 0xffffffff + v_mbcnt_lo_u32_b32 v1, exec_hi, 0 // Set V1 to thread-ID (0..63) + v_mbcnt_hi_u32_b32 v1, exec_lo, v1 // Set V1 to thread-ID (0..63) + v_mul_u32_u24 v1, 0x00000008, v1 // * 8, so each thread is a double-dword address (8byte) + s_mov_b32 s2, 0x00000003f // 64 loop iterations + s_mov_b32 m0, 0xffffffff + // Clear all of LDS space + // Each FirstWave of WorkGroup clears 64kbyte block + +label_001F: + ds_write2_b64 v1, v[2:3], v[2:3] offset1:32 + ds_write2_b64 v1, v[4:5], v[4:5] offset0:64 offset1:96 + v_add_co_u32 v1, vcc, 0x00000400, v1 + s_sub_u32 s2, s2, 1 + s_cbranch_scc0 label_001F + // + // CLEAR SGPRs + // +label_0023: + s_mov_b32 m0, 0x00000068 // Loop 108/4=27 times (loop unrolled for performance) +label_sgpr_loop: + s_movreld_b32 s0, 0 + s_movreld_b32 s1, 0 + s_movreld_b32 s2, 0 + s_movreld_b32 s3, 0 + s_sub_u32 m0, m0, 4 + s_cbranch_scc0 label_sgpr_loop + + //clear vcc + s_mov_b64 vcc, 0 //clear vcc + s_mov_b32 flat_scratch_lo, 0 //clear flat scratch lo SGPR + s_mov_b32 flat_scratch_hi, 0 //clear flat scratch hi SGPR + s_mov_b64 ttmp0, 0 //Clear ttmp0 and ttmp1 + s_mov_b64 ttmp2, 0 //Clear ttmp2 and ttmp3 + s_mov_b64 ttmp4, 0 //Clear ttmp4 and ttmp5 + s_mov_b64 ttmp6, 0 //Clear ttmp6 and ttmp7 + s_mov_b64 ttmp8, 0 //Clear ttmp8 and ttmp9 + s_mov_b64 ttmp10, 0 //Clear ttmp10 and ttmp11 + s_mov_b64 ttmp12, 0 //Clear ttmp12 and ttmp13 + s_mov_b64 ttmp14, 0 //Clear ttmp14 and ttmp15 + + s_endpgm + +end + diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0_cleaner_shader.h b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0_cleaner_shader.h new file mode 100644 index 000000000000..3218cc04f543 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0_cleaner_shader.h @@ -0,0 +1,56 @@ +/* SPDX-License-Identifier: MIT */ +/* + * Copyright 2024 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +/* Define the cleaner shader gfx_11_0_3 */ +static const u32 gfx_11_0_3_cleaner_shader_hex[] = { + 0xb0804006, 0xbe8200ff, + 0x00000058, 0xbefd0080, + 0x7e008480, 0x7e028480, + 0x7e048480, 0x7e068480, + 0x7e088480, 0x7e0a8480, + 0x7e0c8480, 0x7e0e8480, + 0xbefd0002, 0x80828802, + 0xbfa1fff5, 0xbe8200ff, + 0x80000000, 0x8b020002, + 0xbfa10012, 0xbefe00c1, + 0xbeff00c1, 0xd71f0001, + 0x0001007f, 0xd7200001, + 0x0002027e, 0x16020288, + 0xbe8200bf, 0xbefd00c1, + 0xd9382000, 0x00020201, + 0xd9386040, 0x00040401, + 0xd7006a01, 0x000202ff, + 0x00000400, 0x80828102, + 0xbfa1fff7, 0xbefd00ff, + 0x00000068, 0xbe804280, + 0xbe814280, 0xbe824280, + 0xbe834280, 0x80fd847d, + 0xbfa1fffa, 0xbeea0180, + 0xbeec0180, 0xbeee0180, + 0xbef00180, 0xbef20180, + 0xbef40180, 0xbef60180, + 0xbef80180, 0xbefa0180, + 0xbfb00000, 0xbf9f0000, + 0xbf9f0000, 0xbf9f0000, + 0xbf9f0000, 0xbf9f0000, +}; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c index 47b47d21f464..fe7c48f2fb2a 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c @@ -1319,12 +1319,12 @@ static void gfx_v12_0_alloc_ip_dump(struct amdgpu_device *adev) } } -static int gfx_v12_0_sw_init(void *handle) +static int gfx_v12_0_sw_init(struct amdgpu_ip_block *ip_block) { int i, j, k, r, ring_id = 0; unsigned num_compute_rings; int xcc_id = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { case IP_VERSION(12, 0, 0): @@ -1346,6 +1346,12 @@ static int gfx_v12_0_sw_init(void *handle) break; } + switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { + default: + adev->gfx.enable_cleaner_shader = false; + break; + } + /* recalculate compute rings to use based on hardware configuration */ num_compute_rings = (adev->gfx.mec.num_pipe_per_mec * adev->gfx.mec.num_queue_per_pipe) / 2; @@ -1431,6 +1437,12 @@ static int gfx_v12_0_sw_init(void *handle) } } + /* TODO: Add queue reset mask when FW fully supports it */ + adev->gfx.gfx_supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->gfx.gfx_ring[0]); + adev->gfx.compute_supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->gfx.compute_ring[0]); + if (!adev->enable_mes_kiq) { r = amdgpu_gfx_kiq_init(adev, GFX12_MEC_HPD_SIZE, 0); if (r) { @@ -1460,6 +1472,10 @@ static int gfx_v12_0_sw_init(void *handle) gfx_v12_0_alloc_ip_dump(adev); + r = amdgpu_gfx_sysfs_init(adev); + if (r) + return r; + return 0; } @@ -1492,10 +1508,10 @@ static void gfx_v12_0_rlc_autoload_buffer_fini(struct amdgpu_device *adev) (void **)&adev->gfx.rlc.rlc_autoload_ptr); } -static int gfx_v12_0_sw_fini(void *handle) +static int gfx_v12_0_sw_fini(struct amdgpu_ip_block *ip_block) { int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->gfx.num_gfx_rings; i++) amdgpu_ring_fini(&adev->gfx.gfx_ring[i]); @@ -1519,6 +1535,8 @@ static int gfx_v12_0_sw_fini(void *handle) gfx_v12_0_free_microcode(adev); + amdgpu_gfx_sysfs_fini(adev); + kfree(adev->gfx.ip_dump_core); kfree(adev->gfx.ip_dump_compute_queues); kfree(adev->gfx.ip_dump_gfx_queues); @@ -2601,7 +2619,7 @@ static int gfx_v12_0_cp_gfx_resume(struct amdgpu_device *adev) WREG32_SOC15(GC, 0, regCP_RB0_WPTR, lower_32_bits(ring->wptr)); WREG32_SOC15(GC, 0, regCP_RB0_WPTR_HI, upper_32_bits(ring->wptr)); - /* set the wb address wether it's enabled or not */ + /* set the wb address whether it's enabled or not */ rptr_addr = ring->rptr_gpu_addr; WREG32_SOC15(GC, 0, regCP_RB0_RPTR_ADDR, lower_32_bits(rptr_addr)); WREG32_SOC15(GC, 0, regCP_RB0_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & @@ -3513,10 +3531,10 @@ static void gfx_v12_0_init_golden_registers(struct amdgpu_device *adev) } } -static int gfx_v12_0_hw_init(void *handle) +static int gfx_v12_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->firmware.load_type == AMDGPU_FW_LOAD_RLC_BACKDOOR_AUTO) { if (adev->gfx.imu.funcs && (amdgpu_dpm > 0)) { @@ -3603,9 +3621,9 @@ static int gfx_v12_0_hw_init(void *handle) return r; } -static int gfx_v12_0_hw_fini(void *handle) +static int gfx_v12_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t tmp; amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0); @@ -3643,14 +3661,14 @@ static int gfx_v12_0_hw_fini(void *handle) return 0; } -static int gfx_v12_0_suspend(void *handle) +static int gfx_v12_0_suspend(struct amdgpu_ip_block *ip_block) { - return gfx_v12_0_hw_fini(handle); + return gfx_v12_0_hw_fini(ip_block); } -static int gfx_v12_0_resume(void *handle) +static int gfx_v12_0_resume(struct amdgpu_ip_block *ip_block) { - return gfx_v12_0_hw_init(handle); + return gfx_v12_0_hw_init(ip_block); } static bool gfx_v12_0_is_idle(void *handle) @@ -3664,11 +3682,11 @@ static bool gfx_v12_0_is_idle(void *handle) return true; } -static int gfx_v12_0_wait_for_idle(void *handle) +static int gfx_v12_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { /* read MC_STATUS */ @@ -3695,9 +3713,9 @@ static uint64_t gfx_v12_0_get_gpu_clock_counter(struct amdgpu_device *adev) return clock; } -static int gfx_v12_0_early_init(void *handle) +static int gfx_v12_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->gfx.funcs = &gfx_v12_0_gfx_funcs; @@ -3717,9 +3735,9 @@ static int gfx_v12_0_early_init(void *handle) return gfx_v12_0_init_microcode(adev); } -static int gfx_v12_0_late_init(void *handle) +static int gfx_v12_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_irq_get(adev, &adev->gfx.priv_reg_irq, 0); @@ -5022,8 +5040,6 @@ static void gfx_v12_0_emit_mem_sync(struct amdgpu_ring *ring) static void gfx_v12_ring_insert_nop(struct amdgpu_ring *ring, uint32_t num_nop) { - int i; - /* Header itself is a NOP packet */ if (num_nop == 1) { amdgpu_ring_write(ring, ring->funcs->nop); @@ -5034,13 +5050,19 @@ static void gfx_v12_ring_insert_nop(struct amdgpu_ring *ring, uint32_t num_nop) amdgpu_ring_write(ring, PACKET3(PACKET3_NOP, min(num_nop - 2, 0x3ffe))); /* Header is at index 0, followed by num_nops - 1 NOP packet's */ - for (i = 1; i < num_nop; i++) - amdgpu_ring_write(ring, ring->funcs->nop); + amdgpu_ring_insert_nop(ring, num_nop - 1); } -static void gfx_v12_ip_print(void *handle, struct drm_printer *p) +static void gfx_v12_0_ring_emit_cleaner_shader(struct amdgpu_ring *ring) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + /* Emit the cleaner shader */ + amdgpu_ring_write(ring, PACKET3(PACKET3_RUN_CLEANER_SHADER, 0)); + amdgpu_ring_write(ring, 0); /* RESERVED field, programmed to zero */ +} + +static void gfx_v12_ip_print(struct amdgpu_ip_block *ip_block, struct drm_printer *p) +{ + struct amdgpu_device *adev = ip_block->adev; uint32_t i, j, k, reg, index = 0; uint32_t reg_count = ARRAY_SIZE(gc_reg_list_12_0); @@ -5102,9 +5124,9 @@ static void gfx_v12_ip_print(void *handle, struct drm_printer *p) } } -static void gfx_v12_ip_dump(void *handle) +static void gfx_v12_ip_dump(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t i, j, k, reg, index = 0; uint32_t reg_count = ARRAY_SIZE(gc_reg_list_12_0); @@ -5297,7 +5319,8 @@ static const struct amdgpu_ring_funcs gfx_v12_0_ring_funcs_gfx = { 3 + /* CNTX_CTRL */ 5 + /* HDP_INVL */ 8 + 8 + /* FENCE x2 */ - 8, /* gfx_v12_0_emit_mem_sync */ + 8 + /* gfx_v12_0_emit_mem_sync */ + 2, /* gfx_v12_0_ring_emit_cleaner_shader */ .emit_ib_size = 4, /* gfx_v12_0_ring_emit_ib_gfx */ .emit_ib = gfx_v12_0_ring_emit_ib_gfx, .emit_fence = gfx_v12_0_ring_emit_fence, @@ -5318,6 +5341,9 @@ static const struct amdgpu_ring_funcs gfx_v12_0_ring_funcs_gfx = { .soft_recovery = gfx_v12_0_ring_soft_recovery, .emit_mem_sync = gfx_v12_0_emit_mem_sync, .reset = gfx_v12_0_reset_kgq, + .emit_cleaner_shader = gfx_v12_0_ring_emit_cleaner_shader, + .begin_use = amdgpu_gfx_enforce_isolation_ring_begin_use, + .end_use = amdgpu_gfx_enforce_isolation_ring_end_use, }; static const struct amdgpu_ring_funcs gfx_v12_0_ring_funcs_compute = { @@ -5336,7 +5362,8 @@ static const struct amdgpu_ring_funcs gfx_v12_0_ring_funcs_compute = { SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 + 2 + /* gfx_v12_0_ring_emit_vm_flush */ 8 + 8 + 8 + /* gfx_v12_0_ring_emit_fence x3 for user fence, vm fence */ - 8, /* gfx_v12_0_emit_mem_sync */ + 8 + /* gfx_v12_0_emit_mem_sync */ + 2, /* gfx_v12_0_ring_emit_cleaner_shader */ .emit_ib_size = 7, /* gfx_v12_0_ring_emit_ib_compute */ .emit_ib = gfx_v12_0_ring_emit_ib_compute, .emit_fence = gfx_v12_0_ring_emit_fence, @@ -5353,6 +5380,9 @@ static const struct amdgpu_ring_funcs gfx_v12_0_ring_funcs_compute = { .soft_recovery = gfx_v12_0_ring_soft_recovery, .emit_mem_sync = gfx_v12_0_emit_mem_sync, .reset = gfx_v12_0_reset_kcq, + .emit_cleaner_shader = gfx_v12_0_ring_emit_cleaner_shader, + .begin_use = amdgpu_gfx_enforce_isolation_ring_begin_use, + .end_use = amdgpu_gfx_enforce_isolation_ring_end_use, }; static const struct amdgpu_ring_funcs gfx_v12_0_ring_funcs_kiq = { diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c index 564f0b9336b6..41f50bf380c4 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c @@ -3023,9 +3023,9 @@ static const struct amdgpu_rlc_funcs gfx_v6_0_rlc_funcs = { .start = gfx_v6_0_rlc_start }; -static int gfx_v6_0_early_init(void *handle) +static int gfx_v6_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->gfx.xcc_mask = 1; adev->gfx.num_gfx_rings = GFX6_NUM_GFX_RINGS; @@ -3039,10 +3039,10 @@ static int gfx_v6_0_early_init(void *handle) return 0; } -static int gfx_v6_0_sw_init(void *handle) +static int gfx_v6_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, r; r = amdgpu_irq_add_id(adev, AMDGPU_IRQ_CLIENTID_LEGACY, 181, &adev->gfx.eop_irq); @@ -3107,10 +3107,10 @@ static int gfx_v6_0_sw_init(void *handle) return r; } -static int gfx_v6_0_sw_fini(void *handle) +static int gfx_v6_0_sw_fini(struct amdgpu_ip_block *ip_block) { int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->gfx.num_gfx_rings; i++) amdgpu_ring_fini(&adev->gfx.gfx_ring[i]); @@ -3122,10 +3122,10 @@ static int gfx_v6_0_sw_fini(void *handle) return 0; } -static int gfx_v6_0_hw_init(void *handle) +static int gfx_v6_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gfx_v6_0_constants_init(adev); @@ -3142,9 +3142,9 @@ static int gfx_v6_0_hw_init(void *handle) return r; } -static int gfx_v6_0_hw_fini(void *handle) +static int gfx_v6_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gfx_v6_0_cp_enable(adev, false); adev->gfx.rlc.funcs->stop(adev); @@ -3153,18 +3153,14 @@ static int gfx_v6_0_hw_fini(void *handle) return 0; } -static int gfx_v6_0_suspend(void *handle) +static int gfx_v6_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return gfx_v6_0_hw_fini(adev); + return gfx_v6_0_hw_fini(ip_block); } -static int gfx_v6_0_resume(void *handle) +static int gfx_v6_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return gfx_v6_0_hw_init(adev); + return gfx_v6_0_hw_init(ip_block); } static bool gfx_v6_0_is_idle(void *handle) @@ -3177,24 +3173,19 @@ static bool gfx_v6_0_is_idle(void *handle) return true; } -static int gfx_v6_0_wait_for_idle(void *handle) +static int gfx_v6_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { - if (gfx_v6_0_is_idle(handle)) + if (gfx_v6_0_is_idle(adev)) return 0; udelay(1); } return -ETIMEDOUT; } -static int gfx_v6_0_soft_reset(void *handle) -{ - return 0; -} - static void gfx_v6_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev, enum amdgpu_interrupt_state state) { @@ -3444,7 +3435,6 @@ static void gfx_v6_0_emit_mem_sync(struct amdgpu_ring *ring) static const struct amd_ip_funcs gfx_v6_0_ip_funcs = { .name = "gfx_v6_0", .early_init = gfx_v6_0_early_init, - .late_init = NULL, .sw_init = gfx_v6_0_sw_init, .sw_fini = gfx_v6_0_sw_fini, .hw_init = gfx_v6_0_hw_init, @@ -3453,11 +3443,8 @@ static const struct amd_ip_funcs gfx_v6_0_ip_funcs = { .resume = gfx_v6_0_resume, .is_idle = gfx_v6_0_is_idle, .wait_for_idle = gfx_v6_0_wait_for_idle, - .soft_reset = gfx_v6_0_soft_reset, .set_clockgating_state = gfx_v6_0_set_clockgating_state, .set_powergating_state = gfx_v6_0_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs gfx_v6_0_ring_funcs_gfx = { diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c index f146806c4633..824d5913103b 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c @@ -2559,7 +2559,7 @@ static int gfx_v7_0_cp_gfx_resume(struct amdgpu_device *adev) ring->wptr = 0; WREG32(mmCP_RB0_WPTR, lower_32_bits(ring->wptr)); - /* set the wb address wether it's enabled or not */ + /* set the wb address whether it's enabled or not */ rptr_addr = ring->rptr_gpu_addr; WREG32(mmCP_RB0_RPTR_ADDR, lower_32_bits(rptr_addr)); WREG32(mmCP_RB0_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & 0xFF); @@ -2876,7 +2876,7 @@ static void gfx_v7_0_mqd_init(struct amdgpu_device *adev, mqd->cp_hqd_pq_wptr_poll_addr_lo = wb_gpu_addr & 0xfffffffc; mqd->cp_hqd_pq_wptr_poll_addr_hi = upper_32_bits(wb_gpu_addr) & 0xffff; - /* set the wb address wether it's enabled or not */ + /* set the wb address whether it's enabled or not */ wb_gpu_addr = ring->rptr_gpu_addr; mqd->cp_hqd_pq_rptr_report_addr_lo = wb_gpu_addr & 0xfffffffc; mqd->cp_hqd_pq_rptr_report_addr_hi = @@ -4134,9 +4134,9 @@ static const struct amdgpu_rlc_funcs gfx_v7_0_rlc_funcs = { .update_spm_vmid = gfx_v7_0_update_spm_vmid }; -static int gfx_v7_0_early_init(void *handle) +static int gfx_v7_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->gfx.xcc_mask = 1; adev->gfx.num_gfx_rings = GFX7_NUM_GFX_RINGS; @@ -4151,9 +4151,9 @@ static int gfx_v7_0_early_init(void *handle) return 0; } -static int gfx_v7_0_late_init(void *handle) +static int gfx_v7_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_irq_get(adev, &adev->gfx.priv_reg_irq, 0); @@ -4343,10 +4343,10 @@ static int gfx_v7_0_compute_ring_init(struct amdgpu_device *adev, int ring_id, return 0; } -static int gfx_v7_0_sw_init(void *handle) +static int gfx_v7_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j, k, r, ring_id; switch (adev->asic_type) { @@ -4439,9 +4439,9 @@ static int gfx_v7_0_sw_init(void *handle) return r; } -static int gfx_v7_0_sw_fini(void *handle) +static int gfx_v7_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; for (i = 0; i < adev->gfx.num_gfx_rings; i++) @@ -4465,10 +4465,10 @@ static int gfx_v7_0_sw_fini(void *handle) return 0; } -static int gfx_v7_0_hw_init(void *handle) +static int gfx_v7_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gfx_v7_0_constants_init(adev); @@ -4486,9 +4486,9 @@ static int gfx_v7_0_hw_init(void *handle) return r; } -static int gfx_v7_0_hw_fini(void *handle) +static int gfx_v7_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0); amdgpu_irq_put(adev, &adev->gfx.priv_inst_irq, 0); @@ -4499,18 +4499,14 @@ static int gfx_v7_0_hw_fini(void *handle) return 0; } -static int gfx_v7_0_suspend(void *handle) +static int gfx_v7_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return gfx_v7_0_hw_fini(adev); + return gfx_v7_0_hw_fini(ip_block); } -static int gfx_v7_0_resume(void *handle) +static int gfx_v7_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return gfx_v7_0_hw_init(adev); + return gfx_v7_0_hw_init(ip_block); } static bool gfx_v7_0_is_idle(void *handle) @@ -4523,11 +4519,11 @@ static bool gfx_v7_0_is_idle(void *handle) return true; } -static int gfx_v7_0_wait_for_idle(void *handle) +static int gfx_v7_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { /* read MC_STATUS */ @@ -4540,11 +4536,11 @@ static int gfx_v7_0_wait_for_idle(void *handle) return -ETIMEDOUT; } -static int gfx_v7_0_soft_reset(void *handle) +static int gfx_v7_0_soft_reset(struct amdgpu_ip_block *ip_block) { u32 grbm_soft_reset = 0, srbm_soft_reset = 0; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* GRBM_STATUS */ tmp = RREG32(mmGRBM_STATUS); @@ -5009,8 +5005,6 @@ static const struct amd_ip_funcs gfx_v7_0_ip_funcs = { .soft_reset = gfx_v7_0_soft_reset, .set_clockgating_state = gfx_v7_0_set_clockgating_state, .set_powergating_state = gfx_v7_0_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_gfx = { diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index bc8295812cc8..b7006c41e270 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -1894,12 +1894,12 @@ static int gfx_v8_0_compute_ring_init(struct amdgpu_device *adev, int ring_id, static void gfx_v8_0_sq_irq_work_func(struct work_struct *work); -static int gfx_v8_0_sw_init(void *handle) +static int gfx_v8_0_sw_init(struct amdgpu_ip_block *ip_block) { int i, j, k, r, ring_id; int xcc_id = 0; struct amdgpu_ring *ring; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; switch (adev->asic_type) { case CHIP_TONGA: @@ -2037,9 +2037,9 @@ static int gfx_v8_0_sw_init(void *handle) return 0; } -static int gfx_v8_0_sw_fini(void *handle) +static int gfx_v8_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; for (i = 0; i < adev->gfx.num_gfx_rings; i++) @@ -4260,7 +4260,7 @@ static int gfx_v8_0_cp_gfx_resume(struct amdgpu_device *adev) ring->wptr = 0; WREG32(mmCP_RB0_WPTR, lower_32_bits(ring->wptr)); - /* set the wb address wether it's enabled or not */ + /* set the wb address whether it's enabled or not */ rptr_addr = ring->rptr_gpu_addr; WREG32(mmCP_RB0_RPTR_ADDR, lower_32_bits(rptr_addr)); WREG32(mmCP_RB0_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & 0xFF); @@ -4783,10 +4783,10 @@ static void gfx_v8_0_cp_enable(struct amdgpu_device *adev, bool enable) gfx_v8_0_cp_compute_enable(adev, enable); } -static int gfx_v8_0_hw_init(void *handle) +static int gfx_v8_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gfx_v8_0_init_golden_registers(adev); gfx_v8_0_constants_init(adev); @@ -4823,6 +4823,13 @@ static int gfx_v8_0_kcq_disable(struct amdgpu_device *adev) amdgpu_ring_write(kiq_ring, 0); amdgpu_ring_write(kiq_ring, 0); } + /* Submit unmap queue packet */ + amdgpu_ring_commit(kiq_ring); + /* + * Ring test will do a basic scratch register change check. Just run + * this to ensure that unmap queues that is submitted before got + * processed successfully before returning. + */ r = amdgpu_ring_test_helper(kiq_ring); if (r) DRM_ERROR("KCQ disable failed\n"); @@ -4865,13 +4872,13 @@ static int gfx_v8_0_wait_for_rlc_idle(void *handle) return -ETIMEDOUT; } -static int gfx_v8_0_wait_for_idle(void *handle) +static int gfx_v8_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { - if (gfx_v8_0_is_idle(handle)) + if (gfx_v8_0_is_idle(adev)) return 0; udelay(1); @@ -4879,9 +4886,9 @@ static int gfx_v8_0_wait_for_idle(void *handle) return -ETIMEDOUT; } -static int gfx_v8_0_hw_fini(void *handle) +static int gfx_v8_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0); amdgpu_irq_put(adev, &adev->gfx.priv_inst_irq, 0); @@ -4897,8 +4904,9 @@ static int gfx_v8_0_hw_fini(void *handle) pr_debug("For SRIOV client, shouldn't do anything.\n"); return 0; } + amdgpu_gfx_rlc_enter_safe_mode(adev, 0); - if (!gfx_v8_0_wait_for_idle(adev)) + if (!gfx_v8_0_wait_for_idle(ip_block)) gfx_v8_0_cp_enable(adev, false); else pr_err("cp is busy, skip halt cp\n"); @@ -4911,19 +4919,19 @@ static int gfx_v8_0_hw_fini(void *handle) return 0; } -static int gfx_v8_0_suspend(void *handle) +static int gfx_v8_0_suspend(struct amdgpu_ip_block *ip_block) { - return gfx_v8_0_hw_fini(handle); + return gfx_v8_0_hw_fini(ip_block); } -static int gfx_v8_0_resume(void *handle) +static int gfx_v8_0_resume(struct amdgpu_ip_block *ip_block) { - return gfx_v8_0_hw_init(handle); + return gfx_v8_0_hw_init(ip_block); } -static bool gfx_v8_0_check_soft_reset(void *handle) +static bool gfx_v8_0_check_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 grbm_soft_reset = 0, srbm_soft_reset = 0; u32 tmp; @@ -4983,9 +4991,9 @@ static bool gfx_v8_0_check_soft_reset(void *handle) } } -static int gfx_v8_0_pre_soft_reset(void *handle) +static int gfx_v8_0_pre_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 grbm_soft_reset = 0; if ((!adev->gfx.grbm_soft_reset) && @@ -5024,9 +5032,9 @@ static int gfx_v8_0_pre_soft_reset(void *handle) return 0; } -static int gfx_v8_0_soft_reset(void *handle) +static int gfx_v8_0_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 grbm_soft_reset = 0, srbm_soft_reset = 0; u32 tmp; @@ -5086,9 +5094,9 @@ static int gfx_v8_0_soft_reset(void *handle) return 0; } -static int gfx_v8_0_post_soft_reset(void *handle) +static int gfx_v8_0_post_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 grbm_soft_reset = 0; if ((!adev->gfx.grbm_soft_reset) && @@ -5254,9 +5262,9 @@ static const struct amdgpu_gfx_funcs gfx_v8_0_gfx_funcs = { .select_me_pipe_q = &gfx_v8_0_select_me_pipe_q }; -static int gfx_v8_0_early_init(void *handle) +static int gfx_v8_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->gfx.xcc_mask = 1; adev->gfx.num_gfx_rings = GFX8_NUM_GFX_RINGS; @@ -5271,9 +5279,9 @@ static int gfx_v8_0_early_init(void *handle) return 0; } -static int gfx_v8_0_late_init(void *handle) +static int gfx_v8_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_irq_get(adev, &adev->gfx.priv_reg_irq, 0); @@ -6947,8 +6955,6 @@ static const struct amd_ip_funcs gfx_v8_0_ip_funcs = { .set_clockgating_state = gfx_v8_0_set_clockgating_state, .set_powergating_state = gfx_v8_0_set_powergating_state, .get_clockgating_state = gfx_v8_0_get_clockgating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_gfx = { diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 23f0573ae47b..0b6f09f2cc9b 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -2198,12 +2198,12 @@ static void gfx_v9_0_alloc_ip_dump(struct amdgpu_device *adev) } } -static int gfx_v9_0_sw_init(void *handle) +static int gfx_v9_0_sw_init(struct amdgpu_ip_block *ip_block) { int i, j, k, r, ring_id; int xcc_id = 0; struct amdgpu_ring *ring; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; unsigned int hw_prio; switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { @@ -2223,6 +2223,18 @@ static int gfx_v9_0_sw_init(void *handle) } switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { + case IP_VERSION(9, 4, 2): + adev->gfx.cleaner_shader_ptr = gfx_9_4_2_cleaner_shader_hex; + adev->gfx.cleaner_shader_size = sizeof(gfx_9_4_2_cleaner_shader_hex); + if (adev->gfx.mec_fw_version >= 88) { + adev->gfx.enable_cleaner_shader = true; + r = amdgpu_gfx_cleaner_shader_sw_init(adev, adev->gfx.cleaner_shader_size); + if (r) { + adev->gfx.enable_cleaner_shader = false; + dev_err(adev->dev, "Failed to initialize cleaner shader\n"); + } + } + break; default: adev->gfx.enable_cleaner_shader = false; break; @@ -2362,6 +2374,12 @@ static int gfx_v9_0_sw_init(void *handle) } } + /* TODO: Add queue reset mask when FW fully supports it */ + adev->gfx.gfx_supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->gfx.gfx_ring[0]); + adev->gfx.compute_supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->gfx.compute_ring[0]); + r = amdgpu_gfx_kiq_init(adev, GFX9_MEC_HPD_SIZE, 0); if (r) { DRM_ERROR("Failed to init KIQ BOs!\n"); @@ -2390,7 +2408,7 @@ static int gfx_v9_0_sw_init(void *handle) gfx_v9_0_alloc_ip_dump(adev); - r = amdgpu_gfx_sysfs_isolation_shader_init(adev); + r = amdgpu_gfx_sysfs_init(adev); if (r) return r; @@ -2398,10 +2416,10 @@ static int gfx_v9_0_sw_init(void *handle) } -static int gfx_v9_0_sw_fini(void *handle) +static int gfx_v9_0_sw_fini(struct amdgpu_ip_block *ip_block) { int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->gfx.mcbp && adev->gfx.num_gfx_rings) { for (i = 0; i < GFX9_NUM_SW_GFX_RINGS; i++) @@ -2418,6 +2436,8 @@ static int gfx_v9_0_sw_fini(void *handle) amdgpu_gfx_kiq_free_ring(&adev->gfx.kiq[0].ring); amdgpu_gfx_kiq_fini(adev, 0); + amdgpu_gfx_cleaner_shader_sw_fini(adev); + gfx_v9_0_mec_fini(adev); amdgpu_bo_free_kernel(&adev->gfx.rlc.clear_state_obj, &adev->gfx.rlc.clear_state_gpu_addr, @@ -2429,7 +2449,7 @@ static int gfx_v9_0_sw_fini(void *handle) } gfx_v9_0_free_microcode(adev); - amdgpu_gfx_sysfs_isolation_shader_fini(adev); + amdgpu_gfx_sysfs_fini(adev); kfree(adev->gfx.ip_dump_core); kfree(adev->gfx.ip_dump_compute_queues); @@ -3184,6 +3204,15 @@ static void gfx_v9_0_cp_gfx_enable(struct amdgpu_device *adev, bool enable) { u32 tmp = RREG32_SOC15(GC, 0, mmCP_ME_CNTL); + tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, CE_INVALIDATE_ICACHE, enable ? 0 : 1); + tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, PFP_INVALIDATE_ICACHE, enable ? 0 : 1); + tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, ME_INVALIDATE_ICACHE, enable ? 0 : 1); + tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, CE_PIPE0_RESET, enable ? 0 : 1); + tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, CE_PIPE1_RESET, enable ? 0 : 1); + tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, PFP_PIPE0_RESET, enable ? 0 : 1); + tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, PFP_PIPE1_RESET, enable ? 0 : 1); + tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, ME_PIPE0_RESET, enable ? 0 : 1); + tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, ME_PIPE1_RESET, enable ? 0 : 1); tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, ME_HALT, enable ? 0 : 1); tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, PFP_HALT, enable ? 0 : 1); tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, CE_HALT, enable ? 0 : 1); @@ -3265,8 +3294,8 @@ static int gfx_v9_0_cp_gfx_start(struct amdgpu_device *adev) * confirmed that the APU gfx10/gfx11 needn't such update. */ if (adev->flags & AMD_IS_APU && - adev->in_s3 && !adev->suspend_complete) { - DRM_INFO(" Will skip the CSB packet resubmit\n"); + adev->in_s3 && !pm_resume_via_firmware()) { + DRM_INFO("Will skip the CSB packet resubmit\n"); return 0; } r = amdgpu_ring_alloc(ring, gfx_v9_0_get_csb_size(adev) + 4 + 3); @@ -3346,7 +3375,7 @@ static int gfx_v9_0_cp_gfx_resume(struct amdgpu_device *adev) WREG32_SOC15(GC, 0, mmCP_RB0_WPTR, lower_32_bits(ring->wptr)); WREG32_SOC15(GC, 0, mmCP_RB0_WPTR_HI, upper_32_bits(ring->wptr)); - /* set the wb address wether it's enabled or not */ + /* set the wb address whether it's enabled or not */ rptr_addr = ring->rptr_gpu_addr; WREG32_SOC15(GC, 0, mmCP_RB0_RPTR_ADDR, lower_32_bits(rptr_addr)); WREG32_SOC15(GC, 0, mmCP_RB0_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & CP_RB_RPTR_ADDR_HI__RB_RPTR_ADDR_HI_MASK); @@ -3393,7 +3422,15 @@ static void gfx_v9_0_cp_compute_enable(struct amdgpu_device *adev, bool enable) WREG32_SOC15_RLC(GC, 0, mmCP_MEC_CNTL, 0); } else { WREG32_SOC15_RLC(GC, 0, mmCP_MEC_CNTL, - (CP_MEC_CNTL__MEC_ME1_HALT_MASK | CP_MEC_CNTL__MEC_ME2_HALT_MASK)); + (CP_MEC_CNTL__MEC_INVALIDATE_ICACHE_MASK | + CP_MEC_CNTL__MEC_ME1_PIPE0_RESET_MASK | + CP_MEC_CNTL__MEC_ME1_PIPE1_RESET_MASK | + CP_MEC_CNTL__MEC_ME1_PIPE2_RESET_MASK | + CP_MEC_CNTL__MEC_ME1_PIPE3_RESET_MASK | + CP_MEC_CNTL__MEC_ME2_PIPE0_RESET_MASK | + CP_MEC_CNTL__MEC_ME2_PIPE1_RESET_MASK | + CP_MEC_CNTL__MEC_ME1_HALT_MASK | + CP_MEC_CNTL__MEC_ME2_HALT_MASK)); adev->gfx.kiq[0].ring.sched.ready = false; } udelay(50); @@ -3914,6 +3951,10 @@ static int gfx_v9_0_cp_resume(struct amdgpu_device *adev) return r; } + if (adev->gfx.num_gfx_rings) + gfx_v9_0_cp_gfx_enable(adev, false); + gfx_v9_0_cp_compute_enable(adev, false); + r = gfx_v9_0_kiq_resume(adev); if (r) return r; @@ -3970,10 +4011,10 @@ static void gfx_v9_0_cp_enable(struct amdgpu_device *adev, bool enable) gfx_v9_0_cp_compute_enable(adev, enable); } -static int gfx_v9_0_hw_init(void *handle) +static int gfx_v9_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_gfx_cleaner_shader_init(adev, adev->gfx.cleaner_shader_size, adev->gfx.cleaner_shader_ptr); @@ -3999,9 +4040,9 @@ static int gfx_v9_0_hw_init(void *handle) return r; } -static int gfx_v9_0_hw_fini(void *handle) +static int gfx_v9_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__GFX)) amdgpu_irq_put(adev, &adev->gfx.cp_ecc_error_irq, 0); @@ -4051,14 +4092,14 @@ static int gfx_v9_0_hw_fini(void *handle) return 0; } -static int gfx_v9_0_suspend(void *handle) +static int gfx_v9_0_suspend(struct amdgpu_ip_block *ip_block) { - return gfx_v9_0_hw_fini(handle); + return gfx_v9_0_hw_fini(ip_block); } -static int gfx_v9_0_resume(void *handle) +static int gfx_v9_0_resume(struct amdgpu_ip_block *ip_block) { - return gfx_v9_0_hw_init(handle); + return gfx_v9_0_hw_init(ip_block); } static bool gfx_v9_0_is_idle(void *handle) @@ -4072,24 +4113,24 @@ static bool gfx_v9_0_is_idle(void *handle) return true; } -static int gfx_v9_0_wait_for_idle(void *handle) +static int gfx_v9_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { - if (gfx_v9_0_is_idle(handle)) + if (gfx_v9_0_is_idle(adev)) return 0; udelay(1); } return -ETIMEDOUT; } -static int gfx_v9_0_soft_reset(void *handle) +static int gfx_v9_0_soft_reset(struct amdgpu_ip_block *ip_block) { u32 grbm_soft_reset = 0; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* GRBM_STATUS */ tmp = RREG32_SOC15(GC, 0, mmGRBM_STATUS); @@ -4745,9 +4786,9 @@ fail: return r; } -static int gfx_v9_0_early_init(void *handle) +static int gfx_v9_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->gfx.funcs = &gfx_v9_0_gfx_funcs; @@ -4771,9 +4812,9 @@ static int gfx_v9_0_early_init(void *handle) return gfx_v9_0_init_microcode(adev); } -static int gfx_v9_0_ecc_late_init(void *handle) +static int gfx_v9_0_ecc_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; /* @@ -4805,9 +4846,9 @@ static int gfx_v9_0_ecc_late_init(void *handle) return 0; } -static int gfx_v9_0_late_init(void *handle) +static int gfx_v9_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_irq_get(adev, &adev->gfx.priv_reg_irq, 0); @@ -4822,7 +4863,7 @@ static int gfx_v9_0_late_init(void *handle) if (r) return r; - r = gfx_v9_0_ecc_late_init(handle); + r = gfx_v9_0_ecc_late_init(ip_block); if (r) return r; @@ -7167,8 +7208,6 @@ static void gfx_v9_0_emit_wave_limit(struct amdgpu_ring *ring, bool enable) static void gfx_v9_ring_insert_nop(struct amdgpu_ring *ring, uint32_t num_nop) { - int i; - /* Header itself is a NOP packet */ if (num_nop == 1) { amdgpu_ring_write(ring, ring->funcs->nop); @@ -7179,8 +7218,7 @@ static void gfx_v9_ring_insert_nop(struct amdgpu_ring *ring, uint32_t num_nop) amdgpu_ring_write(ring, PACKET3(PACKET3_NOP, min(num_nop - 2, 0x3ffe))); /* Header is at index 0, followed by num_nops - 1 NOP packet's */ - for (i = 1; i < num_nop; i++) - amdgpu_ring_write(ring, ring->funcs->nop); + amdgpu_ring_insert_nop(ring, num_nop - 1); } static int gfx_v9_0_reset_kgq(struct amdgpu_ring *ring, unsigned int vmid) @@ -7237,10 +7275,6 @@ static int gfx_v9_0_reset_kcq(struct amdgpu_ring *ring, unsigned long flags; int i, r; - if (!adev->debug_exp_resets && - !adev->gfx.num_gfx_rings) - return -EINVAL; - if (amdgpu_sriov_vf(adev)) return -EINVAL; @@ -7316,9 +7350,9 @@ static int gfx_v9_0_reset_kcq(struct amdgpu_ring *ring, return amdgpu_ring_test_ring(ring); } -static void gfx_v9_ip_print(void *handle, struct drm_printer *p) +static void gfx_v9_ip_print(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t i, j, k, reg, index = 0; uint32_t reg_count = ARRAY_SIZE(gc_reg_list_9); @@ -7356,9 +7390,9 @@ static void gfx_v9_ip_print(void *handle, struct drm_printer *p) } -static void gfx_v9_ip_dump(void *handle) +static void gfx_v9_ip_dump(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t i, j, k, reg, index = 0; uint32_t reg_count = ARRAY_SIZE(gc_reg_list_9); diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0_cleaner_shader.h b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0_cleaner_shader.h index 36c0292b5110..0b6bd09b7529 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0_cleaner_shader.h +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0_cleaner_shader.h @@ -1,6 +1,6 @@ /* SPDX-License-Identifier: MIT */ /* - * Copyright 2018 Advanced Micro Devices, Inc. + * Copyright 2024 Advanced Micro Devices, Inc. * * Permission is hereby granted, free of charge, to any person obtaining a * copy of this software and associated documentation files (the "Software"), @@ -24,3 +24,45 @@ static const u32 __maybe_unused gfx_9_0_cleaner_shader_hex[] = { /* Add the cleaner shader code here */ }; + +/* Define the cleaner shader gfx_9_4_2 */ +static const u32 gfx_9_4_2_cleaner_shader_hex[] = { + 0xbf068100, 0xbf84003b, + 0xbf8a0000, 0xb07c0000, + 0xbe8200ff, 0x00000078, + 0xbf110802, 0x7e000280, + 0x7e020280, 0x7e040280, + 0x7e060280, 0x7e080280, + 0x7e0a0280, 0x7e0c0280, + 0x7e0e0280, 0x80828802, + 0xbe803202, 0xbf84fff5, + 0xbf9c0000, 0xbe8200ff, + 0x80000000, 0x86020102, + 0xbf840011, 0xbefe00c1, + 0xbeff00c1, 0xd28c0001, + 0x0001007f, 0xd28d0001, + 0x0002027e, 0x10020288, + 0xbe8200bf, 0xbefc00c1, + 0xd89c2000, 0x00020201, + 0xd89c6040, 0x00040401, + 0x320202ff, 0x00000400, + 0x80828102, 0xbf84fff8, + 0xbefc00ff, 0x0000005c, + 0xbf800000, 0xbe802c80, + 0xbe812c80, 0xbe822c80, + 0xbe832c80, 0x80fc847c, + 0xbf84fffa, 0xbee60080, + 0xbee70080, 0xbeea0180, + 0xbeec0180, 0xbeee0180, + 0xbef00180, 0xbef20180, + 0xbef40180, 0xbef60180, + 0xbef80180, 0xbefa0180, + 0xbf810000, 0xbf8d0001, + 0xbefc00ff, 0x0000005c, + 0xbf800000, 0xbe802c80, + 0xbe812c80, 0xbe822c80, + 0xbe832c80, 0x80fc847c, + 0xbf84fffa, 0xbee60080, + 0xbee70080, 0xbeea01ff, + 0x000000ee, 0xbf810000, +}; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2_cleaner_shader.asm b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2_cleaner_shader.asm new file mode 100644 index 000000000000..35b8cf9070bd --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2_cleaner_shader.asm @@ -0,0 +1,153 @@ +/* SPDX-License-Identifier: MIT */ +/* + * Copyright 2024 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +// This shader is to clean LDS, SGPRs and VGPRs. It is first 64 Dwords or 256 bytes of 192 Dwords cleaner shader. +//To turn this shader program on for complitaion change this to main and lower shader main to main_1 + +// MI200 : Clear SGPRs, VGPRs and LDS +// Uses two kernels launched separately: +// 1. Clean VGPRs, LDS, and lower SGPRs +// Launches one workgroup per CU, each workgroup with 4x wave64 per SIMD in the CU +// Waves are "wave64" and have 128 VGPRs each, which uses all 512 VGPRs per SIMD +// Waves in the workgroup share the 64KB of LDS +// Each wave clears SGPRs 0 - 95. Because there are 4 waves/SIMD, this is physical SGPRs 0-383 +// Each wave clears 128 VGPRs, so all 512 in the SIMD +// The first wave of the workgroup clears its 64KB of LDS +// The shader starts with "S_BARRIER" to ensure SPI has launched all waves of the workgroup +// before any wave in the workgroup could end. Without this, it is possible not all SGPRs get cleared. +// 2. Clean remaining SGPRs +// Launches a workgroup with 24 waves per workgroup, yielding 6 waves per SIMD in each CU +// Waves are allocating 96 SGPRs +// CP sets up SPI_RESOURCE_RESERVE_* registers to prevent these waves from allocating SGPRs 0-223. +// As such, these 6 waves per SIMD are allocated physical SGPRs 224-799 +// Barriers do not work for >16 waves per workgroup, so we cannot start with S_BARRIER +// Instead, the shader starts with an S_SETHALT 1. Once all waves are launched CP will send unhalt command +// The shader then clears all SGPRs allocated to it, cleaning out physical SGPRs 224-799 + +shader main + asic(MI200) + type(CS) + wave_size(64) +// Note: original source code from SQ team + +// (theorhetical fastest = ~512clks vgpr + 1536 lds + ~128 sgpr = 2176 clks) + + s_cmp_eq_u32 s0, 1 // Bit0 is set, sgpr0 is set then clear VGPRS and LDS as FW set COMPUTE_USER_DATA_3 + s_cbranch_scc0 label_0023 // Clean VGPRs and LDS if sgpr0 of wave is set, scc = (s3 == 1) + S_BARRIER + + s_movk_i32 m0, 0x0000 + s_mov_b32 s2, 0x00000078 // Loop 128/8=16 times (loop unrolled for performance) + // + // CLEAR VGPRs + // + s_set_gpr_idx_on s2, 0x8 // enable Dest VGPR indexing +label_0005: + v_mov_b32 v0, 0 + v_mov_b32 v1, 0 + v_mov_b32 v2, 0 + v_mov_b32 v3, 0 + v_mov_b32 v4, 0 + v_mov_b32 v5, 0 + v_mov_b32 v6, 0 + v_mov_b32 v7, 0 + s_sub_u32 s2, s2, 8 + s_set_gpr_idx_idx s2 + s_cbranch_scc0 label_0005 + s_set_gpr_idx_off + + // + // + + s_mov_b32 s2, 0x80000000 // Bit31 is first_wave + s_and_b32 s2, s2, s1 // sgpr0 has tg_size (first_wave) term as in ucode only COMPUTE_PGM_RSRC2.tg_size_en is set + s_cbranch_scc0 label_clean_sgpr_1 // Clean LDS if its first wave of ThreadGroup/WorkGroup + // CLEAR LDS + // + s_mov_b32 exec_lo, 0xffffffff + s_mov_b32 exec_hi, 0xffffffff + v_mbcnt_lo_u32_b32 v1, exec_hi, 0 // Set V1 to thread-ID (0..63) + v_mbcnt_hi_u32_b32 v1, exec_lo, v1 // Set V1 to thread-ID (0..63) + v_mul_u32_u24 v1, 0x00000008, v1 // * 8, so each thread is a double-dword address (8byte) + s_mov_b32 s2, 0x00000003f // 64 loop iterations + s_mov_b32 m0, 0xffffffff + // Clear all of LDS space + // Each FirstWave of WorkGroup clears 64kbyte block + +label_001F: + ds_write2_b64 v1, v[2:3], v[2:3] offset1:32 + ds_write2_b64 v1, v[4:5], v[4:5] offset0:64 offset1:96 + v_add_co_u32 v1, vcc, 0x00000400, v1 + s_sub_u32 s2, s2, 1 + s_cbranch_scc0 label_001F + // + // CLEAR SGPRs + // +label_clean_sgpr_1: + s_mov_b32 m0, 0x0000005c // Loop 96/4=24 times (loop unrolled for performance) + s_nop 0 +label_sgpr_loop: + s_movreld_b32 s0, 0 + s_movreld_b32 s1, 0 + s_movreld_b32 s2, 0 + s_movreld_b32 s3, 0 + s_sub_u32 m0, m0, 4 + s_cbranch_scc0 label_sgpr_loop + + //clear vcc, flat scratch + s_mov_b32 flat_scratch_lo, 0 //clear flat scratch lo SGPR + s_mov_b32 flat_scratch_hi, 0 //clear flat scratch hi SGPR + s_mov_b64 vcc, 0 //clear vcc + s_mov_b64 ttmp0, 0 //Clear ttmp0 and ttmp1 + s_mov_b64 ttmp2, 0 //Clear ttmp2 and ttmp3 + s_mov_b64 ttmp4, 0 //Clear ttmp4 and ttmp5 + s_mov_b64 ttmp6, 0 //Clear ttmp6 and ttmp7 + s_mov_b64 ttmp8, 0 //Clear ttmp8 and ttmp9 + s_mov_b64 ttmp10, 0 //Clear ttmp10 and ttmp11 + s_mov_b64 ttmp12, 0 //Clear ttmp12 and ttmp13 + s_mov_b64 ttmp14, 0 //Clear ttmp14 and ttmp15 +s_endpgm + +label_0023: + + s_sethalt 1 + + s_mov_b32 m0, 0x0000005c // Loop 96/4=24 times (loop unrolled for performance) + s_nop 0 +label_sgpr_loop1: + + s_movreld_b32 s0, 0 + s_movreld_b32 s1, 0 + s_movreld_b32 s2, 0 + s_movreld_b32 s3, 0 + s_sub_u32 m0, m0, 4 + s_cbranch_scc0 label_sgpr_loop1 + + //clear vcc, flat scratch + s_mov_b32 flat_scratch_lo, 0 //clear flat scratch lo SGPR + s_mov_b32 flat_scratch_hi, 0 //clear flat scratch hi SGPR + s_mov_b64 vcc, 0xee //clear vcc + +s_endpgm +end + diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c index c100845409f7..e2b3dda57030 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c @@ -1049,10 +1049,10 @@ static void gfx_v9_4_3_alloc_ip_dump(struct amdgpu_device *adev) } } -static int gfx_v9_4_3_sw_init(void *handle) +static int gfx_v9_4_3_sw_init(struct amdgpu_ip_block *ip_block) { int i, j, k, r, ring_id, xcc_id, num_xcc; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { case IP_VERSION(9, 4, 3): @@ -1157,6 +1157,19 @@ static int gfx_v9_4_3_sw_init(void *handle) return r; } + adev->gfx.compute_supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->gfx.compute_ring[0]); + switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { + case IP_VERSION(9, 4, 3): + case IP_VERSION(9, 4, 4): + if (adev->gfx.mec_fw_version >= 155) { + adev->gfx.compute_supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE; + adev->gfx.compute_supported_reset |= AMDGPU_RESET_TYPE_PER_PIPE; + } + break; + default: + break; + } r = gfx_v9_4_3_gpu_early_init(adev); if (r) return r; @@ -1165,26 +1178,19 @@ static int gfx_v9_4_3_sw_init(void *handle) if (r) return r; - - if (!amdgpu_sriov_vf(adev)) { - r = amdgpu_gfx_sysfs_init(adev); - if (r) - return r; - } - - gfx_v9_4_3_alloc_ip_dump(adev); - - r = amdgpu_gfx_sysfs_isolation_shader_init(adev); + r = amdgpu_gfx_sysfs_init(adev); if (r) return r; + gfx_v9_4_3_alloc_ip_dump(adev); + return 0; } -static int gfx_v9_4_3_sw_fini(void *handle) +static int gfx_v9_4_3_sw_fini(struct amdgpu_ip_block *ip_block) { int i, num_xcc; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; num_xcc = NUM_XCC(adev->gfx.xcc_mask); for (i = 0; i < adev->gfx.num_compute_rings * num_xcc; i++) @@ -1201,9 +1207,7 @@ static int gfx_v9_4_3_sw_fini(void *handle) gfx_v9_4_3_mec_fini(adev); amdgpu_bo_unref(&adev->gfx.rlc.clear_state_obj); gfx_v9_4_3_free_microcode(adev); - if (!amdgpu_sriov_vf(adev)) - amdgpu_gfx_sysfs_fini(adev); - amdgpu_gfx_sysfs_isolation_shader_fini(adev); + amdgpu_gfx_sysfs_fini(adev); kfree(adev->gfx.ip_dump_core); kfree(adev->gfx.ip_dump_compute_queues); @@ -1247,8 +1251,10 @@ static void gfx_v9_4_3_xcc_init_compute_vmid(struct amdgpu_device *adev, soc15_grbm_select(adev, 0, 0, 0, 0, GET_INST(GC, xcc_id)); mutex_unlock(&adev->srbm_mutex); - /* Initialize all compute VMIDs to have no GDS, GWS, or OA - acccess. These should be enabled by FW for target VMIDs. */ + /* + * Initialize all compute VMIDs to have no GDS, GWS, or OA + * access. These should be enabled by FW for target VMIDs. + */ for (i = adev->vm_manager.first_kfd_vmid; i < AMDGPU_NUM_VMID; i++) { WREG32_SOC15_OFFSET(GC, GET_INST(GC, xcc_id), regGDS_VMID0_BASE, 2 * i, 0); WREG32_SOC15_OFFSET(GC, GET_INST(GC, xcc_id), regGDS_VMID0_SIZE, 2 * i, 0); @@ -2343,10 +2349,10 @@ static void gfx_v9_4_3_xcc_fini(struct amdgpu_device *adev, int xcc_id) gfx_v9_4_3_xcc_cp_compute_enable(adev, false, xcc_id); } -static int gfx_v9_4_3_hw_init(void *handle) +static int gfx_v9_4_3_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_gfx_cleaner_shader_init(adev, adev->gfx.cleaner_shader_size, adev->gfx.cleaner_shader_ptr); @@ -2367,9 +2373,9 @@ static int gfx_v9_4_3_hw_init(void *handle) return r; } -static int gfx_v9_4_3_hw_fini(void *handle) +static int gfx_v9_4_3_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, num_xcc; amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0); @@ -2384,14 +2390,14 @@ static int gfx_v9_4_3_hw_fini(void *handle) return 0; } -static int gfx_v9_4_3_suspend(void *handle) +static int gfx_v9_4_3_suspend(struct amdgpu_ip_block *ip_block) { - return gfx_v9_4_3_hw_fini(handle); + return gfx_v9_4_3_hw_fini(ip_block); } -static int gfx_v9_4_3_resume(void *handle) +static int gfx_v9_4_3_resume(struct amdgpu_ip_block *ip_block) { - return gfx_v9_4_3_hw_init(handle); + return gfx_v9_4_3_hw_init(ip_block); } static bool gfx_v9_4_3_is_idle(void *handle) @@ -2408,24 +2414,24 @@ static bool gfx_v9_4_3_is_idle(void *handle) return true; } -static int gfx_v9_4_3_wait_for_idle(void *handle) +static int gfx_v9_4_3_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { - if (gfx_v9_4_3_is_idle(handle)) + if (gfx_v9_4_3_is_idle(adev)) return 0; udelay(1); } return -ETIMEDOUT; } -static int gfx_v9_4_3_soft_reset(void *handle) +static int gfx_v9_4_3_soft_reset(struct amdgpu_ip_block *ip_block) { u32 grbm_soft_reset = 0; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* GRBM_STATUS */ tmp = RREG32_SOC15(GC, GET_INST(GC, 0), regGRBM_STATUS); @@ -2509,9 +2515,9 @@ static void gfx_v9_4_3_ring_emit_gds_switch(struct amdgpu_ring *ring, (1 << (oa_size + oa_base)) - (1 << oa_base)); } -static int gfx_v9_4_3_early_init(void *handle) +static int gfx_v9_4_3_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->gfx.num_compute_rings = min(amdgpu_gfx_get_num_kcq(adev), AMDGPU_MAX_COMPUTE_RINGS); @@ -2527,9 +2533,9 @@ static int gfx_v9_4_3_early_init(void *handle) return gfx_v9_4_3_init_microcode(adev); } -static int gfx_v9_4_3_late_init(void *handle) +static int gfx_v9_4_3_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_irq_get(adev, &adev->gfx.priv_reg_irq, 0); @@ -3056,9 +3062,6 @@ static void gfx_v9_4_3_ring_soft_recovery(struct amdgpu_ring *ring, struct amdgpu_device *adev = ring->adev; uint32_t value = 0; - if (!adev->debug_exp_resets) - return; - value = REG_SET_FIELD(value, SQ_CMD, CMD, 0x03); value = REG_SET_FIELD(value, SQ_CMD, MODE, 0x01); value = REG_SET_FIELD(value, SQ_CMD, CHECK_VMID, 1); @@ -3574,9 +3577,6 @@ static int gfx_v9_4_3_reset_kcq(struct amdgpu_ring *ring, unsigned long flags; int r; - if (!adev->debug_exp_resets) - return -EINVAL; - if (amdgpu_sriov_vf(adev)) return -EINVAL; @@ -4567,8 +4567,6 @@ static void gfx_v9_4_3_enable_watchdog_timer(struct amdgpu_device *adev) static void gfx_v9_4_3_ring_insert_nop(struct amdgpu_ring *ring, uint32_t num_nop) { - int i; - /* Header itself is a NOP packet */ if (num_nop == 1) { amdgpu_ring_write(ring, ring->funcs->nop); @@ -4579,13 +4577,12 @@ static void gfx_v9_4_3_ring_insert_nop(struct amdgpu_ring *ring, uint32_t num_no amdgpu_ring_write(ring, PACKET3(PACKET3_NOP, min(num_nop - 2, 0x3ffe))); /* Header is at index 0, followed by num_nops - 1 NOP packet's */ - for (i = 1; i < num_nop; i++) - amdgpu_ring_write(ring, ring->funcs->nop); + amdgpu_ring_insert_nop(ring, num_nop - 1); } -static void gfx_v9_4_3_ip_print(void *handle, struct drm_printer *p) +static void gfx_v9_4_3_ip_print(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t i, j, k; uint32_t xcc_id, xcc_offset, inst_offset; uint32_t num_xcc, reg, num_inst; @@ -4643,9 +4640,9 @@ static void gfx_v9_4_3_ip_print(void *handle, struct drm_printer *p) } } -static void gfx_v9_4_3_ip_dump(void *handle) +static void gfx_v9_4_3_ip_dump(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t i, j, k; uint32_t num_xcc, reg, num_inst; uint32_t xcc_id, xcc_offset, inst_offset; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c index 9784a2892185..697599c46240 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c @@ -175,7 +175,10 @@ static int gmc_v10_0_process_interrupt(struct amdgpu_device *adev, addr, entry->client_id, soc15_ih_clientid_name[entry->client_id]); - if (!amdgpu_sriov_vf(adev)) + /* Only print L2 fault status if the status register could be read and + * contains useful information + */ + if (status != 0) hub->vmhub_funcs->print_l2_protection_fault_status(adev, status); @@ -630,9 +633,9 @@ static void gmc_v10_0_set_gfxhub_funcs(struct amdgpu_device *adev) } -static int gmc_v10_0_early_init(void *handle) +static int gmc_v10_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gmc_v10_0_set_mmhub_funcs(adev); gmc_v10_0_set_gfxhub_funcs(adev); @@ -651,9 +654,9 @@ static int gmc_v10_0_early_init(void *handle) return 0; } -static int gmc_v10_0_late_init(void *handle) +static int gmc_v10_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_gmc_allocate_vm_inv_eng(adev); @@ -769,10 +772,10 @@ static int gmc_v10_0_gart_init(struct amdgpu_device *adev) return amdgpu_gart_table_vram_alloc(adev); } -static int gmc_v10_0_sw_init(void *handle) +static int gmc_v10_0_sw_init(struct amdgpu_ip_block *ip_block) { int r, vram_width = 0, vram_type = 0, vram_vendor = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->gfxhub.funcs->init(adev); @@ -920,9 +923,9 @@ static void gmc_v10_0_gart_fini(struct amdgpu_device *adev) amdgpu_gart_table_vram_free(adev); } -static int gmc_v10_0_sw_fini(void *handle) +static int gmc_v10_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_vm_manager_fini(adev); gmc_v10_0_gart_fini(adev); @@ -985,9 +988,9 @@ static int gmc_v10_0_gart_enable(struct amdgpu_device *adev) return 0; } -static int gmc_v10_0_hw_init(void *handle) +static int gmc_v10_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; adev->gmc.flush_pasid_uses_kiq = !amdgpu_emu_mode; @@ -1032,9 +1035,9 @@ static void gmc_v10_0_gart_disable(struct amdgpu_device *adev) adev->mmhub.funcs->gart_disable(adev); } -static int gmc_v10_0_hw_fini(void *handle) +static int gmc_v10_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gmc_v10_0_gart_disable(adev); @@ -1053,25 +1056,22 @@ static int gmc_v10_0_hw_fini(void *handle) return 0; } -static int gmc_v10_0_suspend(void *handle) +static int gmc_v10_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - gmc_v10_0_hw_fini(adev); + gmc_v10_0_hw_fini(ip_block); return 0; } -static int gmc_v10_0_resume(void *handle) +static int gmc_v10_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = gmc_v10_0_hw_init(adev); + r = gmc_v10_0_hw_init(ip_block); if (r) return r; - amdgpu_vmid_reset_all(adev); + amdgpu_vmid_reset_all(ip_block->adev); return 0; } @@ -1082,17 +1082,12 @@ static bool gmc_v10_0_is_idle(void *handle) return true; } -static int gmc_v10_0_wait_for_idle(void *handle) +static int gmc_v10_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { /* There is no need to wait for MC idle in GMC v10.*/ return 0; } -static int gmc_v10_0_soft_reset(void *handle) -{ - return 0; -} - static int gmc_v10_0_set_clockgating_state(void *handle, enum amd_clockgating_state state) { @@ -1154,7 +1149,6 @@ const struct amd_ip_funcs gmc_v10_0_ip_funcs = { .resume = gmc_v10_0_resume, .is_idle = gmc_v10_0_is_idle, .wait_for_idle = gmc_v10_0_wait_for_idle, - .soft_reset = gmc_v10_0_soft_reset, .set_clockgating_state = gmc_v10_0_set_clockgating_state, .set_powergating_state = gmc_v10_0_set_powergating_state, .get_clockgating_state = gmc_v10_0_get_clockgating_state, diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c index 2797fd84432b..f893ab4c14df 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c @@ -144,7 +144,10 @@ static int gmc_v11_0_process_interrupt(struct amdgpu_device *adev, dev_err(adev->dev, " in page starting at address 0x%016llx from client %d\n", addr, entry->client_id); - if (!amdgpu_sriov_vf(adev)) + /* Only print L2 fault status if the status register could be read and + * contains useful information + */ + if (status != 0) hub->vmhub_funcs->print_l2_protection_fault_status(adev, status); } @@ -601,9 +604,9 @@ static void gmc_v11_0_set_gfxhub_funcs(struct amdgpu_device *adev) } } -static int gmc_v11_0_early_init(void *handle) +static int gmc_v11_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gmc_v11_0_set_gfxhub_funcs(adev); gmc_v11_0_set_mmhub_funcs(adev); @@ -622,9 +625,9 @@ static int gmc_v11_0_early_init(void *handle) return 0; } -static int gmc_v11_0_late_init(void *handle) +static int gmc_v11_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_gmc_allocate_vm_inv_eng(adev); @@ -729,10 +732,10 @@ static int gmc_v11_0_gart_init(struct amdgpu_device *adev) return amdgpu_gart_table_vram_alloc(adev); } -static int gmc_v11_0_sw_init(void *handle) +static int gmc_v11_0_sw_init(struct amdgpu_ip_block *ip_block) { int r, vram_width = 0, vram_type = 0, vram_vendor = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->mmhub.funcs->init(adev); @@ -849,9 +852,9 @@ static void gmc_v11_0_gart_fini(struct amdgpu_device *adev) amdgpu_gart_table_vram_free(adev); } -static int gmc_v11_0_sw_fini(void *handle) +static int gmc_v11_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_vm_manager_fini(adev); gmc_v11_0_gart_fini(adev); @@ -908,9 +911,9 @@ static int gmc_v11_0_gart_enable(struct amdgpu_device *adev) return 0; } -static int gmc_v11_0_hw_init(void *handle) +static int gmc_v11_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; adev->gmc.flush_pasid_uses_kiq = !amdgpu_emu_mode; @@ -940,9 +943,9 @@ static void gmc_v11_0_gart_disable(struct amdgpu_device *adev) adev->mmhub.funcs->gart_disable(adev); } -static int gmc_v11_0_hw_fini(void *handle) +static int gmc_v11_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) { /* full access mode, so don't touch any GMC register */ @@ -961,25 +964,22 @@ static int gmc_v11_0_hw_fini(void *handle) return 0; } -static int gmc_v11_0_suspend(void *handle) +static int gmc_v11_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - gmc_v11_0_hw_fini(adev); + gmc_v11_0_hw_fini(ip_block); return 0; } -static int gmc_v11_0_resume(void *handle) +static int gmc_v11_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = gmc_v11_0_hw_init(adev); + r = gmc_v11_0_hw_init(ip_block); if (r) return r; - amdgpu_vmid_reset_all(adev); + amdgpu_vmid_reset_all(ip_block->adev); return 0; } @@ -990,17 +990,12 @@ static bool gmc_v11_0_is_idle(void *handle) return true; } -static int gmc_v11_0_wait_for_idle(void *handle) +static int gmc_v11_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { /* There is no need to wait for MC idle in GMC v11.*/ return 0; } -static int gmc_v11_0_soft_reset(void *handle) -{ - return 0; -} - static int gmc_v11_0_set_clockgating_state(void *handle, enum amd_clockgating_state state) { @@ -1041,7 +1036,6 @@ const struct amd_ip_funcs gmc_v11_0_ip_funcs = { .resume = gmc_v11_0_resume, .is_idle = gmc_v11_0_is_idle, .wait_for_idle = gmc_v11_0_wait_for_idle, - .soft_reset = gmc_v11_0_soft_reset, .set_clockgating_state = gmc_v11_0_set_clockgating_state, .set_powergating_state = gmc_v11_0_set_powergating_state, .get_clockgating_state = gmc_v11_0_get_clockgating_state, diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v12_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v12_0.c index edcb5351f8cc..d22b027fd0bb 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v12_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v12_0.c @@ -137,7 +137,10 @@ static int gmc_v12_0_process_interrupt(struct amdgpu_device *adev, dev_err(adev->dev, " in page starting at address 0x%016llx from client %d\n", addr, entry->client_id); - if (!amdgpu_sriov_vf(adev)) + /* Only print L2 fault status if the status register could be read and + * contains useful information + */ + if (status != 0) hub->vmhub_funcs->print_l2_protection_fault_status(adev, status); } @@ -604,9 +607,9 @@ static void gmc_v12_0_set_gfxhub_funcs(struct amdgpu_device *adev) } } -static int gmc_v12_0_early_init(void *handle) +static int gmc_v12_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gmc_v12_0_set_gfxhub_funcs(adev); gmc_v12_0_set_mmhub_funcs(adev); @@ -624,9 +627,9 @@ static int gmc_v12_0_early_init(void *handle) return 0; } -static int gmc_v12_0_late_init(void *handle) +static int gmc_v12_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_gmc_allocate_vm_inv_eng(adev); @@ -731,10 +734,10 @@ static int gmc_v12_0_gart_init(struct amdgpu_device *adev) return amdgpu_gart_table_vram_alloc(adev); } -static int gmc_v12_0_sw_init(void *handle) +static int gmc_v12_0_sw_init(struct amdgpu_ip_block *ip_block) { int r, vram_width = 0, vram_type = 0, vram_vendor = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->mmhub.funcs->init(adev); @@ -841,9 +844,9 @@ static void gmc_v12_0_gart_fini(struct amdgpu_device *adev) amdgpu_gart_table_vram_free(adev); } -static int gmc_v12_0_sw_fini(void *handle) +static int gmc_v12_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_vm_manager_fini(adev); gmc_v12_0_gart_fini(adev); @@ -894,10 +897,10 @@ static int gmc_v12_0_gart_enable(struct amdgpu_device *adev) return 0; } -static int gmc_v12_0_hw_init(void *handle) +static int gmc_v12_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* The sequence of these two function calls matters.*/ gmc_v12_0_init_golden_registers(adev); @@ -924,9 +927,9 @@ static void gmc_v12_0_gart_disable(struct amdgpu_device *adev) adev->mmhub.funcs->gart_disable(adev); } -static int gmc_v12_0_hw_fini(void *handle) +static int gmc_v12_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) { /* full access mode, so don't touch any GMC register */ @@ -945,25 +948,22 @@ static int gmc_v12_0_hw_fini(void *handle) return 0; } -static int gmc_v12_0_suspend(void *handle) +static int gmc_v12_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - gmc_v12_0_hw_fini(adev); + gmc_v12_0_hw_fini(ip_block); return 0; } -static int gmc_v12_0_resume(void *handle) +static int gmc_v12_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = gmc_v12_0_hw_init(adev); + r = gmc_v12_0_hw_init(ip_block); if (r) return r; - amdgpu_vmid_reset_all(adev); + amdgpu_vmid_reset_all(ip_block->adev); return 0; } @@ -974,17 +974,12 @@ static bool gmc_v12_0_is_idle(void *handle) return true; } -static int gmc_v12_0_wait_for_idle(void *handle) +static int gmc_v12_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { /* There is no need to wait for MC idle in GMC v11.*/ return 0; } -static int gmc_v12_0_soft_reset(void *handle) -{ - return 0; -} - static int gmc_v12_0_set_clockgating_state(void *handle, enum amd_clockgating_state state) { @@ -1025,7 +1020,6 @@ const struct amd_ip_funcs gmc_v12_0_ip_funcs = { .resume = gmc_v12_0_resume, .is_idle = gmc_v12_0_is_idle, .wait_for_idle = gmc_v12_0_wait_for_idle, - .soft_reset = gmc_v12_0_soft_reset, .set_clockgating_state = gmc_v12_0_set_clockgating_state, .set_powergating_state = gmc_v12_0_set_powergating_state, .get_clockgating_state = gmc_v12_0_get_clockgating_state, diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c index d36725666b54..ca000b3d1afc 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c @@ -43,7 +43,7 @@ static void gmc_v6_0_set_gmc_funcs(struct amdgpu_device *adev); static void gmc_v6_0_set_irq_funcs(struct amdgpu_device *adev); -static int gmc_v6_0_wait_for_idle(void *handle); +static int gmc_v6_0_wait_for_idle(struct amdgpu_ip_block *ip_block); MODULE_FIRMWARE("amdgpu/tahiti_mc.bin"); MODULE_FIRMWARE("amdgpu/pitcairn_mc.bin"); @@ -64,8 +64,13 @@ MODULE_FIRMWARE("amdgpu/si58_mc.bin"); static void gmc_v6_0_mc_stop(struct amdgpu_device *adev) { u32 blackout; + struct amdgpu_ip_block *ip_block; - gmc_v6_0_wait_for_idle((void *)adev); + ip_block = amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_GMC); + if (!ip_block) + return; + + gmc_v6_0_wait_for_idle(ip_block); blackout = RREG32(mmMC_SHARED_BLACKOUT_CNTL); if (REG_GET_FIELD(blackout, MC_SHARED_BLACKOUT_CNTL, BLACKOUT_MODE) != 1) { @@ -213,6 +218,8 @@ static void gmc_v6_0_vram_gtt_location(struct amdgpu_device *adev, static void gmc_v6_0_mc_program(struct amdgpu_device *adev) { int i, j; + struct amdgpu_ip_block *ip_block; + /* Initialize HDP */ for (i = 0, j = 0; i < 32; i++, j += 0x6) { @@ -224,7 +231,11 @@ static void gmc_v6_0_mc_program(struct amdgpu_device *adev) } WREG32(mmHDP_REG_COHERENCY_FLUSH_CNTL, 0); - if (gmc_v6_0_wait_for_idle((void *)adev)) + ip_block = amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_GMC); + if (!ip_block) + return; + + if (gmc_v6_0_wait_for_idle(ip_block)) dev_warn(adev->dev, "Wait for MC idle timedout !\n"); if (adev->mode_info.num_crtc) { @@ -251,7 +262,7 @@ static void gmc_v6_0_mc_program(struct amdgpu_device *adev) WREG32(mmMC_VM_AGP_TOP, adev->gmc.agp_end >> 22); WREG32(mmMC_VM_AGP_BOT, adev->gmc.agp_start >> 22); - if (gmc_v6_0_wait_for_idle((void *)adev)) + if (gmc_v6_0_wait_for_idle(ip_block)) dev_warn(adev->dev, "Wait for MC idle timedout !\n"); } @@ -762,9 +773,9 @@ static int gmc_v6_0_convert_vram_type(int mc_seq_vram_type) } } -static int gmc_v6_0_early_init(void *handle) +static int gmc_v6_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gmc_v6_0_set_gmc_funcs(adev); gmc_v6_0_set_irq_funcs(adev); @@ -772,9 +783,9 @@ static int gmc_v6_0_early_init(void *handle) return 0; } -static int gmc_v6_0_late_init(void *handle) +static int gmc_v6_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_vm_fault_stop != AMDGPU_VM_FAULT_STOP_ALWAYS) return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0); @@ -799,10 +810,10 @@ static unsigned int gmc_v6_0_get_vbios_fb_size(struct amdgpu_device *adev) return size; } -static int gmc_v6_0_sw_init(void *handle) +static int gmc_v6_0_sw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; set_bit(AMDGPU_GFXHUB(0), adev->vmhubs_mask); @@ -876,9 +887,9 @@ static int gmc_v6_0_sw_init(void *handle) return 0; } -static int gmc_v6_0_sw_fini(void *handle) +static int gmc_v6_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_gem_force_release(adev); amdgpu_vm_manager_fini(adev); @@ -889,10 +900,10 @@ static int gmc_v6_0_sw_fini(void *handle) return 0; } -static int gmc_v6_0_hw_init(void *handle) +static int gmc_v6_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gmc_v6_0_mc_program(adev); @@ -914,9 +925,9 @@ static int gmc_v6_0_hw_init(void *handle) return 0; } -static int gmc_v6_0_hw_fini(void *handle) +static int gmc_v6_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0); gmc_v6_0_gart_disable(adev); @@ -924,21 +935,19 @@ static int gmc_v6_0_hw_fini(void *handle) return 0; } -static int gmc_v6_0_suspend(void *handle) +static int gmc_v6_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - gmc_v6_0_hw_fini(adev); + gmc_v6_0_hw_fini(ip_block); return 0; } -static int gmc_v6_0_resume(void *handle) +static int gmc_v6_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; - r = gmc_v6_0_hw_init(adev); + r = gmc_v6_0_hw_init(ip_block); if (r) return r; @@ -950,6 +959,7 @@ static int gmc_v6_0_resume(void *handle) static bool gmc_v6_0_is_idle(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + u32 tmp = RREG32(mmSRBM_STATUS); if (tmp & (SRBM_STATUS__MCB_BUSY_MASK | SRBM_STATUS__MCB_NON_DISPLAY_BUSY_MASK | @@ -959,13 +969,13 @@ static bool gmc_v6_0_is_idle(void *handle) return true; } -static int gmc_v6_0_wait_for_idle(void *handle) +static int gmc_v6_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { - if (gmc_v6_0_is_idle(handle)) + if (gmc_v6_0_is_idle(adev)) return 0; udelay(1); } @@ -973,9 +983,10 @@ static int gmc_v6_0_wait_for_idle(void *handle) } -static int gmc_v6_0_soft_reset(void *handle) +static int gmc_v6_0_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; + u32 srbm_soft_reset = 0; u32 tmp = RREG32(mmSRBM_STATUS); @@ -992,7 +1003,8 @@ static int gmc_v6_0_soft_reset(void *handle) if (srbm_soft_reset) { gmc_v6_0_mc_stop(adev); - if (gmc_v6_0_wait_for_idle(adev)) + + if (gmc_v6_0_wait_for_idle(ip_block)) dev_warn(adev->dev, "Wait for GMC idle timed out !\n"); tmp = RREG32(mmSRBM_SOFT_RESET); @@ -1109,8 +1121,6 @@ static const struct amd_ip_funcs gmc_v6_0_ip_funcs = { .soft_reset = gmc_v6_0_soft_reset, .set_clockgating_state = gmc_v6_0_set_clockgating_state, .set_powergating_state = gmc_v6_0_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_gmc_funcs gmc_v6_0_gmc_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c index 994432fb57ea..07f45f1a503a 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c @@ -52,7 +52,7 @@ static void gmc_v7_0_set_gmc_funcs(struct amdgpu_device *adev); static void gmc_v7_0_set_irq_funcs(struct amdgpu_device *adev); -static int gmc_v7_0_wait_for_idle(void *handle); +static int gmc_v7_0_wait_for_idle(struct amdgpu_ip_block *ip_block); MODULE_FIRMWARE("amdgpu/bonaire_mc.bin"); MODULE_FIRMWARE("amdgpu/hawaii_mc.bin"); @@ -921,9 +921,9 @@ static int gmc_v7_0_convert_vram_type(int mc_seq_vram_type) } } -static int gmc_v7_0_early_init(void *handle) +static int gmc_v7_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gmc_v7_0_set_gmc_funcs(adev); gmc_v7_0_set_irq_funcs(adev); @@ -940,9 +940,9 @@ static int gmc_v7_0_early_init(void *handle) return 0; } -static int gmc_v7_0_late_init(void *handle) +static int gmc_v7_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_vm_fault_stop != AMDGPU_VM_FAULT_STOP_ALWAYS) return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0); @@ -968,10 +968,10 @@ static unsigned int gmc_v7_0_get_vbios_fb_size(struct amdgpu_device *adev) return size; } -static int gmc_v7_0_sw_init(void *handle) +static int gmc_v7_0_sw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; set_bit(AMDGPU_GFXHUB(0), adev->vmhubs_mask); @@ -1060,9 +1060,9 @@ static int gmc_v7_0_sw_init(void *handle) return 0; } -static int gmc_v7_0_sw_fini(void *handle) +static int gmc_v7_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_gem_force_release(adev); amdgpu_vm_manager_fini(adev); @@ -1074,10 +1074,10 @@ static int gmc_v7_0_sw_fini(void *handle) return 0; } -static int gmc_v7_0_hw_init(void *handle) +static int gmc_v7_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gmc_v7_0_init_golden_registers(adev); @@ -1101,9 +1101,9 @@ static int gmc_v7_0_hw_init(void *handle) return 0; } -static int gmc_v7_0_hw_fini(void *handle) +static int gmc_v7_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0); gmc_v7_0_gart_disable(adev); @@ -1111,25 +1111,22 @@ static int gmc_v7_0_hw_fini(void *handle) return 0; } -static int gmc_v7_0_suspend(void *handle) +static int gmc_v7_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - gmc_v7_0_hw_fini(adev); + gmc_v7_0_hw_fini(ip_block); return 0; } -static int gmc_v7_0_resume(void *handle) +static int gmc_v7_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = gmc_v7_0_hw_init(adev); + r = gmc_v7_0_hw_init(ip_block); if (r) return r; - amdgpu_vmid_reset_all(adev); + amdgpu_vmid_reset_all(ip_block->adev); return 0; } @@ -1146,11 +1143,11 @@ static bool gmc_v7_0_is_idle(void *handle) return true; } -static int gmc_v7_0_wait_for_idle(void *handle) +static int gmc_v7_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned int i; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { /* read MC_STATUS */ @@ -1167,9 +1164,9 @@ static int gmc_v7_0_wait_for_idle(void *handle) } -static int gmc_v7_0_soft_reset(void *handle) +static int gmc_v7_0_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 srbm_soft_reset = 0; u32 tmp = RREG32(mmSRBM_STATUS); @@ -1351,8 +1348,6 @@ static const struct amd_ip_funcs gmc_v7_0_ip_funcs = { .soft_reset = gmc_v7_0_soft_reset, .set_clockgating_state = gmc_v7_0_set_clockgating_state, .set_powergating_state = gmc_v7_0_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_gmc_funcs gmc_v7_0_gmc_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c index 86488c052f82..12d5967ecd45 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c @@ -53,7 +53,7 @@ static void gmc_v8_0_set_gmc_funcs(struct amdgpu_device *adev); static void gmc_v8_0_set_irq_funcs(struct amdgpu_device *adev); -static int gmc_v8_0_wait_for_idle(void *handle); +static int gmc_v8_0_wait_for_idle(struct amdgpu_ip_block *ip_block); MODULE_FIRMWARE("amdgpu/tonga_mc.bin"); MODULE_FIRMWARE("amdgpu/polaris11_mc.bin"); @@ -170,8 +170,13 @@ static void gmc_v8_0_init_golden_registers(struct amdgpu_device *adev) static void gmc_v8_0_mc_stop(struct amdgpu_device *adev) { u32 blackout; + struct amdgpu_ip_block *ip_block; - gmc_v8_0_wait_for_idle(adev); + ip_block = amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_GMC); + if (!ip_block) + return; + + gmc_v8_0_wait_for_idle(ip_block); blackout = RREG32(mmMC_SHARED_BLACKOUT_CNTL); if (REG_GET_FIELD(blackout, MC_SHARED_BLACKOUT_CNTL, BLACKOUT_MODE) != 1) { @@ -426,6 +431,7 @@ static void gmc_v8_0_vram_gtt_location(struct amdgpu_device *adev, */ static void gmc_v8_0_mc_program(struct amdgpu_device *adev) { + struct amdgpu_ip_block *ip_block; u32 tmp; int i, j; @@ -439,7 +445,11 @@ static void gmc_v8_0_mc_program(struct amdgpu_device *adev) } WREG32(mmHDP_REG_COHERENCY_FLUSH_CNTL, 0); - if (gmc_v8_0_wait_for_idle((void *)adev)) + ip_block = amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_GMC); + if (!ip_block) + return; + + if (gmc_v8_0_wait_for_idle(ip_block)) dev_warn(adev->dev, "Wait for MC idle timedout !\n"); if (adev->mode_info.num_crtc) { @@ -474,7 +484,7 @@ static void gmc_v8_0_mc_program(struct amdgpu_device *adev) WREG32(mmMC_VM_AGP_BASE, 0); WREG32(mmMC_VM_AGP_TOP, adev->gmc.agp_end >> 22); WREG32(mmMC_VM_AGP_BOT, adev->gmc.agp_start >> 22); - if (gmc_v8_0_wait_for_idle((void *)adev)) + if (gmc_v8_0_wait_for_idle(ip_block)) dev_warn(adev->dev, "Wait for MC idle timedout !\n"); WREG32(mmBIF_FB_EN, BIF_FB_EN__FB_READ_EN_MASK | BIF_FB_EN__FB_WRITE_EN_MASK); @@ -1027,9 +1037,9 @@ static int gmc_v8_0_convert_vram_type(int mc_seq_vram_type) } } -static int gmc_v8_0_early_init(void *handle) +static int gmc_v8_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gmc_v8_0_set_gmc_funcs(adev); gmc_v8_0_set_irq_funcs(adev); @@ -1046,9 +1056,9 @@ static int gmc_v8_0_early_init(void *handle) return 0; } -static int gmc_v8_0_late_init(void *handle) +static int gmc_v8_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_vm_fault_stop != AMDGPU_VM_FAULT_STOP_ALWAYS) return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0); @@ -1076,10 +1086,10 @@ static unsigned int gmc_v8_0_get_vbios_fb_size(struct amdgpu_device *adev) #define mmMC_SEQ_MISC0_FIJI 0xA71 -static int gmc_v8_0_sw_init(void *handle) +static int gmc_v8_0_sw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; set_bit(AMDGPU_GFXHUB(0), adev->vmhubs_mask); @@ -1173,9 +1183,9 @@ static int gmc_v8_0_sw_init(void *handle) return 0; } -static int gmc_v8_0_sw_fini(void *handle) +static int gmc_v8_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_gem_force_release(adev); amdgpu_vm_manager_fini(adev); @@ -1187,10 +1197,10 @@ static int gmc_v8_0_sw_fini(void *handle) return 0; } -static int gmc_v8_0_hw_init(void *handle) +static int gmc_v8_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gmc_v8_0_init_golden_registers(adev); @@ -1222,9 +1232,9 @@ static int gmc_v8_0_hw_init(void *handle) return 0; } -static int gmc_v8_0_hw_fini(void *handle) +static int gmc_v8_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0); gmc_v8_0_gart_disable(adev); @@ -1232,25 +1242,22 @@ static int gmc_v8_0_hw_fini(void *handle) return 0; } -static int gmc_v8_0_suspend(void *handle) +static int gmc_v8_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - gmc_v8_0_hw_fini(adev); + gmc_v8_0_hw_fini(ip_block); return 0; } -static int gmc_v8_0_resume(void *handle) +static int gmc_v8_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = gmc_v8_0_hw_init(adev); + r = gmc_v8_0_hw_init(ip_block); if (r) return r; - amdgpu_vmid_reset_all(adev); + amdgpu_vmid_reset_all(ip_block->adev); return 0; } @@ -1267,11 +1274,11 @@ static bool gmc_v8_0_is_idle(void *handle) return true; } -static int gmc_v8_0_wait_for_idle(void *handle) +static int gmc_v8_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned int i; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { /* read MC_STATUS */ @@ -1289,10 +1296,10 @@ static int gmc_v8_0_wait_for_idle(void *handle) } -static bool gmc_v8_0_check_soft_reset(void *handle) +static bool gmc_v8_0_check_soft_reset(struct amdgpu_ip_block *ip_block) { u32 srbm_soft_reset = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 tmp = RREG32(mmSRBM_STATUS); if (tmp & SRBM_STATUS__VMC_BUSY_MASK) @@ -1316,23 +1323,23 @@ static bool gmc_v8_0_check_soft_reset(void *handle) return false; } -static int gmc_v8_0_pre_soft_reset(void *handle) +static int gmc_v8_0_pre_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (!adev->gmc.srbm_soft_reset) return 0; gmc_v8_0_mc_stop(adev); - if (gmc_v8_0_wait_for_idle(adev)) + if (gmc_v8_0_wait_for_idle(ip_block)) dev_warn(adev->dev, "Wait for GMC idle timed out !\n"); return 0; } -static int gmc_v8_0_soft_reset(void *handle) +static int gmc_v8_0_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 srbm_soft_reset; if (!adev->gmc.srbm_soft_reset) @@ -1361,9 +1368,9 @@ static int gmc_v8_0_soft_reset(void *handle) return 0; } -static int gmc_v8_0_post_soft_reset(void *handle) +static int gmc_v8_0_post_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (!adev->gmc.srbm_soft_reset) return 0; @@ -1715,8 +1722,6 @@ static const struct amd_ip_funcs gmc_v8_0_ip_funcs = { .set_clockgating_state = gmc_v8_0_set_clockgating_state, .set_powergating_state = gmc_v8_0_set_powergating_state, .get_clockgating_state = gmc_v8_0_get_clockgating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_gmc_funcs gmc_v8_0_gmc_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 7a45f3fdc734..50c5da3020cb 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -672,6 +672,12 @@ static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev, (amdgpu_ip_version(adev, GC_HWIP, 0) >= IP_VERSION(9, 4, 2))) return 0; + /* Only print L2 fault status if the status register could be read and + * contains useful information + */ + if (!status) + return 0; + if (!amdgpu_sriov_vf(adev)) WREG32_P(hub->vm_l2_pro_fault_cntl, 1, ~1); @@ -1389,15 +1395,45 @@ gmc_v9_0_get_memory_partition(struct amdgpu_device *adev, u32 *supp_modes) return mode; } +static enum amdgpu_memory_partition +gmc_v9_0_query_vf_memory_partition(struct amdgpu_device *adev) +{ + switch (adev->gmc.num_mem_partitions) { + case 0: + return UNKNOWN_MEMORY_PARTITION_MODE; + case 1: + return AMDGPU_NPS1_PARTITION_MODE; + case 2: + return AMDGPU_NPS2_PARTITION_MODE; + case 4: + return AMDGPU_NPS4_PARTITION_MODE; + default: + return AMDGPU_NPS1_PARTITION_MODE; + } + + return AMDGPU_NPS1_PARTITION_MODE; +} + static enum amdgpu_memory_partition gmc_v9_0_query_memory_partition(struct amdgpu_device *adev) { if (amdgpu_sriov_vf(adev)) - return AMDGPU_NPS1_PARTITION_MODE; + return gmc_v9_0_query_vf_memory_partition(adev); return gmc_v9_0_get_memory_partition(adev, NULL); } +static bool gmc_v9_0_need_reset_on_init(struct amdgpu_device *adev) +{ + if (adev->nbio.funcs && adev->nbio.funcs->is_nps_switch_requested && + adev->nbio.funcs->is_nps_switch_requested(adev)) { + adev->gmc.reset_flags |= AMDGPU_GMC_INIT_RESET_NPS; + return true; + } + + return false; +} + static const struct amdgpu_gmc_funcs gmc_v9_0_gmc_funcs = { .flush_gpu_tlb = gmc_v9_0_flush_gpu_tlb, .flush_gpu_tlb_pasid = gmc_v9_0_flush_gpu_tlb_pasid, @@ -1409,6 +1445,8 @@ static const struct amdgpu_gmc_funcs gmc_v9_0_gmc_funcs = { .override_vm_pte_flags = gmc_v9_0_override_vm_pte_flags, .get_vbios_fb_size = gmc_v9_0_get_vbios_fb_size, .query_mem_partition_mode = &gmc_v9_0_query_memory_partition, + .request_mem_partition_mode = &amdgpu_gmc_request_memory_partition, + .need_reset_on_init = &gmc_v9_0_need_reset_on_init, }; static void gmc_v9_0_set_gmc_funcs(struct amdgpu_device *adev) @@ -1548,9 +1586,31 @@ static void gmc_v9_0_set_xgmi_ras_funcs(struct amdgpu_device *adev) adev->gmc.xgmi.ras = &xgmi_ras; } -static int gmc_v9_0_early_init(void *handle) +static void gmc_v9_0_init_nps_details(struct amdgpu_device *adev) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + adev->gmc.supported_nps_modes = 0; + + if (amdgpu_sriov_vf(adev) || (adev->flags & AMD_IS_APU)) + return; + + /*TODO: Check PSP version also which supports NPS switch. Otherwise keep + * supported modes as 0. + */ + switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { + case IP_VERSION(9, 4, 3): + case IP_VERSION(9, 4, 4): + adev->gmc.supported_nps_modes = + BIT(AMDGPU_NPS1_PARTITION_MODE) | + BIT(AMDGPU_NPS4_PARTITION_MODE); + break; + default: + break; + } +} + +static int gmc_v9_0_early_init(struct amdgpu_ip_block *ip_block) +{ + struct amdgpu_device *adev = ip_block->adev; /* * 9.4.0, 9.4.1 and 9.4.3 don't have XGMI defined @@ -1604,9 +1664,9 @@ static int gmc_v9_0_early_init(void *handle) return 0; } -static int gmc_v9_0_late_init(void *handle) +static int gmc_v9_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_gmc_allocate_vm_inv_eng(adev); @@ -1903,6 +1963,8 @@ gmc_v9_0_init_sw_mem_ranges(struct amdgpu_device *adev, switch (mode) { case UNKNOWN_MEMORY_PARTITION_MODE: + adev->gmc.num_mem_partitions = 0; + break; case AMDGPU_NPS1_PARTITION_MODE: adev->gmc.num_mem_partitions = 1; break; @@ -1922,7 +1984,7 @@ gmc_v9_0_init_sw_mem_ranges(struct amdgpu_device *adev, /* Use NPS range info, if populated */ r = amdgpu_gmc_get_nps_memranges(adev, mem_ranges, - adev->gmc.num_mem_partitions); + &adev->gmc.num_mem_partitions); if (!r) { l = 0; for (i = 1; i < adev->gmc.num_mem_partitions; ++i) { @@ -1932,6 +1994,11 @@ gmc_v9_0_init_sw_mem_ranges(struct amdgpu_device *adev, } } else { + if (!adev->gmc.num_mem_partitions) { + dev_err(adev->dev, + "Not able to detect NPS mode, fall back to NPS1"); + adev->gmc.num_mem_partitions = 1; + } /* Fallback to sw based calculation */ size = (adev->gmc.real_vram_size + SZ_16M) >> AMDGPU_GPU_PAGE_SHIFT; size /= adev->gmc.num_mem_partitions; @@ -1990,10 +2057,10 @@ static void gmc_v9_4_3_init_vram_info(struct amdgpu_device *adev) adev->gmc.vram_width = 128 * 64; } -static int gmc_v9_0_sw_init(void *handle) +static int gmc_v9_0_sw_init(struct amdgpu_ip_block *ip_block) { int r, vram_width = 0, vram_type = 0, vram_vendor = 0, dma_addr_bits; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; unsigned long inst_mask = adev->aid_mask; adev->gfxhub.funcs->init(adev); @@ -2168,6 +2235,7 @@ static int gmc_v9_0_sw_init(void *handle) if (r) return r; + gmc_v9_0_init_nps_details(adev); /* * number of VMs * VMID 0 is reserved for System @@ -2201,9 +2269,9 @@ static int gmc_v9_0_sw_init(void *handle) return 0; } -static int gmc_v9_0_sw_fini(void *handle) +static int gmc_v9_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(9, 4, 3) || amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(9, 4, 4)) @@ -2311,9 +2379,9 @@ static int gmc_v9_0_gart_enable(struct amdgpu_device *adev) return 0; } -static int gmc_v9_0_hw_init(void *handle) +static int gmc_v9_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; bool value; int i, r; @@ -2396,9 +2464,9 @@ static void gmc_v9_0_gart_disable(struct amdgpu_device *adev) adev->mmhub.funcs->gart_disable(adev); } -static int gmc_v9_0_hw_fini(void *handle) +static int gmc_v9_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; gmc_v9_0_gart_disable(adev); @@ -2416,32 +2484,44 @@ static int gmc_v9_0_hw_fini(void *handle) if (adev->mmhub.funcs->update_power_gating) adev->mmhub.funcs->update_power_gating(adev, false); - amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0); + /* + * For minimal init, late_init is not called, hence VM fault/RAS irqs + * are not enabled. + */ + if (adev->init_lvl->level != AMDGPU_INIT_LEVEL_MINIMAL_XGMI) { + amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0); - if (adev->gmc.ecc_irq.funcs && - amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__UMC)) - amdgpu_irq_put(adev, &adev->gmc.ecc_irq, 0); + if (adev->gmc.ecc_irq.funcs && + amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__UMC)) + amdgpu_irq_put(adev, &adev->gmc.ecc_irq, 0); + } return 0; } -static int gmc_v9_0_suspend(void *handle) +static int gmc_v9_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return gmc_v9_0_hw_fini(adev); + return gmc_v9_0_hw_fini(ip_block); } -static int gmc_v9_0_resume(void *handle) +static int gmc_v9_0_resume(struct amdgpu_ip_block *ip_block) { + struct amdgpu_device *adev = ip_block->adev; int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = gmc_v9_0_hw_init(adev); + /* If a reset is done for NPS mode switch, read the memory range + * information again. + */ + if (adev->gmc.reset_flags & AMDGPU_GMC_INIT_RESET_NPS) { + gmc_v9_0_init_sw_mem_ranges(adev, adev->gmc.mem_partitions); + adev->gmc.reset_flags &= ~AMDGPU_GMC_INIT_RESET_NPS; + } + + r = gmc_v9_0_hw_init(ip_block); if (r) return r; - amdgpu_vmid_reset_all(adev); + amdgpu_vmid_reset_all(ip_block->adev); return 0; } @@ -2452,13 +2532,13 @@ static bool gmc_v9_0_is_idle(void *handle) return true; } -static int gmc_v9_0_wait_for_idle(void *handle) +static int gmc_v9_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { /* There is no need to wait for MC idle in GMC v9.*/ return 0; } -static int gmc_v9_0_soft_reset(void *handle) +static int gmc_v9_0_soft_reset(struct amdgpu_ip_block *ip_block) { /* XXX for emulation.*/ return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/iceland_ih.c b/drivers/gpu/drm/amd/amdgpu/iceland_ih.c index 07984f7c3ae7..7f45e93c0397 100644 --- a/drivers/gpu/drm/amd/amdgpu/iceland_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/iceland_ih.c @@ -273,9 +273,9 @@ static void iceland_ih_set_rptr(struct amdgpu_device *adev, WREG32(mmIH_RB_RPTR, ih->rptr); } -static int iceland_ih_early_init(void *handle) +static int iceland_ih_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int ret; ret = amdgpu_irq_add_domain(adev); @@ -287,10 +287,10 @@ static int iceland_ih_early_init(void *handle) return 0; } -static int iceland_ih_sw_init(void *handle) +static int iceland_ih_sw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_ih_ring_init(adev, &adev->irq.ih, 64 * 1024, false); if (r) @@ -301,9 +301,9 @@ static int iceland_ih_sw_init(void *handle) return r; } -static int iceland_ih_sw_fini(void *handle) +static int iceland_ih_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_fini_sw(adev); amdgpu_irq_remove_domain(adev); @@ -311,34 +311,28 @@ static int iceland_ih_sw_fini(void *handle) return 0; } -static int iceland_ih_hw_init(void *handle) +static int iceland_ih_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return iceland_ih_irq_init(adev); } -static int iceland_ih_hw_fini(void *handle) +static int iceland_ih_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - iceland_ih_irq_disable(adev); + iceland_ih_irq_disable(ip_block->adev); return 0; } -static int iceland_ih_suspend(void *handle) +static int iceland_ih_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return iceland_ih_hw_fini(adev); + return iceland_ih_hw_fini(ip_block); } -static int iceland_ih_resume(void *handle) +static int iceland_ih_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return iceland_ih_hw_init(adev); + return iceland_ih_hw_init(ip_block); } static bool iceland_ih_is_idle(void *handle) @@ -352,11 +346,11 @@ static bool iceland_ih_is_idle(void *handle) return true; } -static int iceland_ih_wait_for_idle(void *handle) +static int iceland_ih_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { /* read MC_STATUS */ @@ -368,10 +362,10 @@ static int iceland_ih_wait_for_idle(void *handle) return -ETIMEDOUT; } -static int iceland_ih_soft_reset(void *handle) +static int iceland_ih_soft_reset(struct amdgpu_ip_block *ip_block) { u32 srbm_soft_reset = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 tmp = RREG32(mmSRBM_STATUS); if (tmp & SRBM_STATUS__IH_BUSY_MASK) @@ -413,7 +407,6 @@ static int iceland_ih_set_powergating_state(void *handle, static const struct amd_ip_funcs iceland_ih_ip_funcs = { .name = "iceland_ih", .early_init = iceland_ih_early_init, - .late_init = NULL, .sw_init = iceland_ih_sw_init, .sw_fini = iceland_ih_sw_fini, .hw_init = iceland_ih_hw_init, @@ -425,8 +418,6 @@ static const struct amd_ip_funcs iceland_ih_ip_funcs = { .soft_reset = iceland_ih_soft_reset, .set_clockgating_state = iceland_ih_set_clockgating_state, .set_powergating_state = iceland_ih_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs iceland_ih_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/ih_v6_0.c b/drivers/gpu/drm/amd/amdgpu/ih_v6_0.c index 18a761d6ef33..38f953fd65d9 100644 --- a/drivers/gpu/drm/amd/amdgpu/ih_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/ih_v6_0.c @@ -559,19 +559,19 @@ static void ih_v6_0_set_self_irq_funcs(struct amdgpu_device *adev) adev->irq.self_irq.funcs = &ih_v6_0_self_irq_funcs; } -static int ih_v6_0_early_init(void *handle) +static int ih_v6_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; ih_v6_0_set_interrupt_funcs(adev); ih_v6_0_set_self_irq_funcs(adev); return 0; } -static int ih_v6_0_sw_init(void *handle) +static int ih_v6_0_sw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; bool use_bus_addr; r = amdgpu_irq_add_id(adev, SOC21_IH_CLIENTID_IH, 0, @@ -614,19 +614,19 @@ static int ih_v6_0_sw_init(void *handle) return r; } -static int ih_v6_0_sw_fini(void *handle) +static int ih_v6_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_fini_sw(adev); return 0; } -static int ih_v6_0_hw_init(void *handle) +static int ih_v6_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = ih_v6_0_irq_init(adev); if (r) @@ -635,27 +635,21 @@ static int ih_v6_0_hw_init(void *handle) return 0; } -static int ih_v6_0_hw_fini(void *handle) +static int ih_v6_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - ih_v6_0_irq_disable(adev); + ih_v6_0_irq_disable(ip_block->adev); return 0; } -static int ih_v6_0_suspend(void *handle) +static int ih_v6_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return ih_v6_0_hw_fini(adev); + return ih_v6_0_hw_fini(ip_block); } -static int ih_v6_0_resume(void *handle) +static int ih_v6_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return ih_v6_0_hw_init(adev); + return ih_v6_0_hw_init(ip_block); } static bool ih_v6_0_is_idle(void *handle) @@ -664,13 +658,13 @@ static bool ih_v6_0_is_idle(void *handle) return true; } -static int ih_v6_0_wait_for_idle(void *handle) +static int ih_v6_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { /* todo */ return -ETIMEDOUT; } -static int ih_v6_0_soft_reset(void *handle) +static int ih_v6_0_soft_reset(struct amdgpu_ip_block *ip_block) { /* todo */ return 0; @@ -785,7 +779,6 @@ static void ih_v6_0_get_clockgating_state(void *handle, u64 *flags) static const struct amd_ip_funcs ih_v6_0_ip_funcs = { .name = "ih_v6_0", .early_init = ih_v6_0_early_init, - .late_init = NULL, .sw_init = ih_v6_0_sw_init, .sw_fini = ih_v6_0_sw_fini, .hw_init = ih_v6_0_hw_init, @@ -798,8 +791,6 @@ static const struct amd_ip_funcs ih_v6_0_ip_funcs = { .set_clockgating_state = ih_v6_0_set_clockgating_state, .set_powergating_state = ih_v6_0_set_powergating_state, .get_clockgating_state = ih_v6_0_get_clockgating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs ih_v6_0_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/ih_v6_1.c b/drivers/gpu/drm/amd/amdgpu/ih_v6_1.c index 2e0469feca1e..61381e0c3795 100644 --- a/drivers/gpu/drm/amd/amdgpu/ih_v6_1.c +++ b/drivers/gpu/drm/amd/amdgpu/ih_v6_1.c @@ -532,9 +532,9 @@ static void ih_v6_1_set_self_irq_funcs(struct amdgpu_device *adev) adev->irq.self_irq.funcs = &ih_v6_1_self_irq_funcs; } -static int ih_v6_1_early_init(void *handle) +static int ih_v6_1_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int ret; ret = amdgpu_irq_add_domain(adev); @@ -547,10 +547,10 @@ static int ih_v6_1_early_init(void *handle) return 0; } -static int ih_v6_1_sw_init(void *handle) +static int ih_v6_1_sw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; bool use_bus_addr; r = amdgpu_irq_add_id(adev, SOC21_IH_CLIENTID_IH, 0, @@ -593,19 +593,19 @@ static int ih_v6_1_sw_init(void *handle) return r; } -static int ih_v6_1_sw_fini(void *handle) +static int ih_v6_1_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_fini_sw(adev); return 0; } -static int ih_v6_1_hw_init(void *handle) +static int ih_v6_1_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = ih_v6_1_irq_init(adev); if (r) @@ -614,27 +614,21 @@ static int ih_v6_1_hw_init(void *handle) return 0; } -static int ih_v6_1_hw_fini(void *handle) +static int ih_v6_1_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - ih_v6_1_irq_disable(adev); + ih_v6_1_irq_disable(ip_block->adev); return 0; } -static int ih_v6_1_suspend(void *handle) +static int ih_v6_1_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return ih_v6_1_hw_fini(adev); + return ih_v6_1_hw_fini(ip_block); } -static int ih_v6_1_resume(void *handle) +static int ih_v6_1_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return ih_v6_1_hw_init(adev); + return ih_v6_1_hw_init(ip_block); } static bool ih_v6_1_is_idle(void *handle) @@ -643,13 +637,13 @@ static bool ih_v6_1_is_idle(void *handle) return true; } -static int ih_v6_1_wait_for_idle(void *handle) +static int ih_v6_1_wait_for_idle(struct amdgpu_ip_block *ip_block) { /* todo */ return -ETIMEDOUT; } -static int ih_v6_1_soft_reset(void *handle) +static int ih_v6_1_soft_reset(struct amdgpu_ip_block *ip_block) { /* todo */ return 0; @@ -768,7 +762,6 @@ static void ih_v6_1_get_clockgating_state(void *handle, u64 *flags) static const struct amd_ip_funcs ih_v6_1_ip_funcs = { .name = "ih_v6_1", .early_init = ih_v6_1_early_init, - .late_init = NULL, .sw_init = ih_v6_1_sw_init, .sw_fini = ih_v6_1_sw_fini, .hw_init = ih_v6_1_hw_init, @@ -781,8 +774,6 @@ static const struct amd_ip_funcs ih_v6_1_ip_funcs = { .set_clockgating_state = ih_v6_1_set_clockgating_state, .set_powergating_state = ih_v6_1_set_powergating_state, .get_clockgating_state = ih_v6_1_get_clockgating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs ih_v6_1_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/ih_v7_0.c b/drivers/gpu/drm/amd/amdgpu/ih_v7_0.c index 6852081fcff2..d2428cf5d385 100644 --- a/drivers/gpu/drm/amd/amdgpu/ih_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/ih_v7_0.c @@ -528,19 +528,19 @@ static void ih_v7_0_set_self_irq_funcs(struct amdgpu_device *adev) adev->irq.self_irq.funcs = &ih_v7_0_self_irq_funcs; } -static int ih_v7_0_early_init(void *handle) +static int ih_v7_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; ih_v7_0_set_interrupt_funcs(adev); ih_v7_0_set_self_irq_funcs(adev); return 0; } -static int ih_v7_0_sw_init(void *handle) +static int ih_v7_0_sw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; bool use_bus_addr; r = amdgpu_irq_add_id(adev, SOC21_IH_CLIENTID_IH, 0, @@ -583,19 +583,19 @@ static int ih_v7_0_sw_init(void *handle) return r; } -static int ih_v7_0_sw_fini(void *handle) +static int ih_v7_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_fini_sw(adev); return 0; } -static int ih_v7_0_hw_init(void *handle) +static int ih_v7_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = ih_v7_0_irq_init(adev); if (r) @@ -604,27 +604,21 @@ static int ih_v7_0_hw_init(void *handle) return 0; } -static int ih_v7_0_hw_fini(void *handle) +static int ih_v7_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - ih_v7_0_irq_disable(adev); + ih_v7_0_irq_disable(ip_block->adev); return 0; } -static int ih_v7_0_suspend(void *handle) +static int ih_v7_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return ih_v7_0_hw_fini(adev); + return ih_v7_0_hw_fini(ip_block); } -static int ih_v7_0_resume(void *handle) +static int ih_v7_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return ih_v7_0_hw_init(adev); + return ih_v7_0_hw_init(ip_block); } static bool ih_v7_0_is_idle(void *handle) @@ -633,13 +627,13 @@ static bool ih_v7_0_is_idle(void *handle) return true; } -static int ih_v7_0_wait_for_idle(void *handle) +static int ih_v7_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { /* todo */ return -ETIMEDOUT; } -static int ih_v7_0_soft_reset(void *handle) +static int ih_v7_0_soft_reset(struct amdgpu_ip_block *ip_block) { /* todo */ return 0; @@ -758,7 +752,6 @@ static void ih_v7_0_get_clockgating_state(void *handle, u64 *flags) static const struct amd_ip_funcs ih_v7_0_ip_funcs = { .name = "ih_v7_0", .early_init = ih_v7_0_early_init, - .late_init = NULL, .sw_init = ih_v7_0_sw_init, .sw_fini = ih_v7_0_sw_fini, .hw_init = ih_v7_0_hw_init, @@ -771,8 +764,6 @@ static const struct amd_ip_funcs ih_v7_0_ip_funcs = { .set_clockgating_state = ih_v7_0_set_clockgating_state, .set_powergating_state = ih_v7_0_set_powergating_state, .get_clockgating_state = ih_v7_0_get_clockgating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs ih_v7_0_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v1_0.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v1_0.c index 6e0e88076224..03b8b7cd5229 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v1_0.c @@ -458,13 +458,13 @@ static int jpeg_v1_0_process_interrupt(struct amdgpu_device *adev, /** * jpeg_v1_0_early_init - set function pointers * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers */ -int jpeg_v1_0_early_init(void *handle) +int jpeg_v1_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->jpeg.num_jpeg_inst = 1; adev->jpeg.num_jpeg_rings = 1; @@ -478,12 +478,12 @@ int jpeg_v1_0_early_init(void *handle) /** * jpeg_v1_0_sw_init - sw init for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * */ -int jpeg_v1_0_sw_init(void *handle) +int jpeg_v1_0_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int r; @@ -509,13 +509,13 @@ int jpeg_v1_0_sw_init(void *handle) /** * jpeg_v1_0_sw_fini - sw fini for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * JPEG free up sw allocation */ -void jpeg_v1_0_sw_fini(void *handle) +void jpeg_v1_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_ring_fini(adev->jpeg.inst->ring_dec); } diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v1_0.h b/drivers/gpu/drm/amd/amdgpu/jpeg_v1_0.h index 9654d22e0376..097328635083 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v1_0.h +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v1_0.h @@ -24,9 +24,9 @@ #ifndef __JPEG_V1_0_H__ #define __JPEG_V1_0_H__ -int jpeg_v1_0_early_init(void *handle); -int jpeg_v1_0_sw_init(void *handle); -void jpeg_v1_0_sw_fini(void *handle); +int jpeg_v1_0_early_init(struct amdgpu_ip_block *ip_block); +int jpeg_v1_0_sw_init(struct amdgpu_ip_block *ip_block); +void jpeg_v1_0_sw_fini(struct amdgpu_ip_block *ip_block); void jpeg_v1_0_start(struct amdgpu_device *adev, int mode); #define JPEG_V1_REG_RANGE_START 0x8000 diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c index 41c0f8750dc1..d6823fb45d32 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c @@ -41,13 +41,13 @@ static int jpeg_v2_0_set_powergating_state(void *handle, /** * jpeg_v2_0_early_init - set function pointers * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers */ -static int jpeg_v2_0_early_init(void *handle) +static int jpeg_v2_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->jpeg.num_jpeg_inst = 1; adev->jpeg.num_jpeg_rings = 1; @@ -61,13 +61,13 @@ static int jpeg_v2_0_early_init(void *handle) /** * jpeg_v2_0_sw_init - sw init for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int jpeg_v2_0_sw_init(void *handle) +static int jpeg_v2_0_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int r; @@ -104,14 +104,14 @@ static int jpeg_v2_0_sw_init(void *handle) /** * jpeg_v2_0_sw_fini - sw fini for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * JPEG suspend and free up sw allocation */ -static int jpeg_v2_0_sw_fini(void *handle) +static int jpeg_v2_0_sw_fini(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_jpeg_suspend(adev); if (r) @@ -125,12 +125,12 @@ static int jpeg_v2_0_sw_fini(void *handle) /** * jpeg_v2_0_hw_init - start and test JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * */ -static int jpeg_v2_0_hw_init(void *handle) +static int jpeg_v2_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring = adev->jpeg.inst->ring_dec; adev->nbio.funcs->vcn_doorbell_range(adev, ring->use_doorbell, @@ -142,13 +142,13 @@ static int jpeg_v2_0_hw_init(void *handle) /** * jpeg_v2_0_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the JPEG block, mark ring as not ready any more */ -static int jpeg_v2_0_hw_fini(void *handle) +static int jpeg_v2_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->vcn.idle_work); @@ -162,20 +162,19 @@ static int jpeg_v2_0_hw_fini(void *handle) /** * jpeg_v2_0_suspend - suspend JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend JPEG block */ -static int jpeg_v2_0_suspend(void *handle) +static int jpeg_v2_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = jpeg_v2_0_hw_fini(adev); + r = jpeg_v2_0_hw_fini(ip_block); if (r) return r; - r = amdgpu_jpeg_suspend(adev); + r = amdgpu_jpeg_suspend(ip_block->adev); return r; } @@ -183,20 +182,19 @@ static int jpeg_v2_0_suspend(void *handle) /** * jpeg_v2_0_resume - resume JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init JPEG block */ -static int jpeg_v2_0_resume(void *handle) +static int jpeg_v2_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_jpeg_resume(adev); + r = amdgpu_jpeg_resume(ip_block->adev); if (r) return r; - r = jpeg_v2_0_hw_init(adev); + r = jpeg_v2_0_hw_init(ip_block); return r; } @@ -666,9 +664,9 @@ static bool jpeg_v2_0_is_idle(void *handle) UVD_JRBC_STATUS__RB_JOB_DONE_MASK); } -static int jpeg_v2_0_wait_for_idle(void *handle) +static int jpeg_v2_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int ret; ret = SOC15_WAIT_ON_RREG(JPEG, 0, mmUVD_JRBC_STATUS, UVD_JRBC_STATUS__RB_JOB_DONE_MASK, @@ -744,7 +742,6 @@ static int jpeg_v2_0_process_interrupt(struct amdgpu_device *adev, static const struct amd_ip_funcs jpeg_v2_0_ip_funcs = { .name = "jpeg_v2_0", .early_init = jpeg_v2_0_early_init, - .late_init = NULL, .sw_init = jpeg_v2_0_sw_init, .sw_fini = jpeg_v2_0_sw_fini, .hw_init = jpeg_v2_0_hw_init, @@ -753,14 +750,8 @@ static const struct amd_ip_funcs jpeg_v2_0_ip_funcs = { .resume = jpeg_v2_0_resume, .is_idle = jpeg_v2_0_is_idle, .wait_for_idle = jpeg_v2_0_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = jpeg_v2_0_set_clockgating_state, .set_powergating_state = jpeg_v2_0_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs jpeg_v2_0_dec_ring_vm_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c index eedb9a829d95..5063a38801d6 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c @@ -50,13 +50,13 @@ static int amdgpu_ih_clientid_jpeg[] = { /** * jpeg_v2_5_early_init - set function pointers * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers */ -static int jpeg_v2_5_early_init(void *handle) +static int jpeg_v2_5_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 harvest; int i; @@ -81,15 +81,15 @@ static int jpeg_v2_5_early_init(void *handle) /** * jpeg_v2_5_sw_init - sw init for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int jpeg_v2_5_sw_init(void *handle) +static int jpeg_v2_5_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int i, r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) { if (adev->jpeg.harvest_config & (1 << i)) @@ -153,14 +153,14 @@ static int jpeg_v2_5_sw_init(void *handle) /** * jpeg_v2_5_sw_fini - sw fini for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * JPEG suspend and free up sw allocation */ -static int jpeg_v2_5_sw_fini(void *handle) +static int jpeg_v2_5_sw_fini(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_jpeg_suspend(adev); if (r) @@ -174,12 +174,12 @@ static int jpeg_v2_5_sw_fini(void *handle) /** * jpeg_v2_5_hw_init - start and test JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * */ -static int jpeg_v2_5_hw_init(void *handle) +static int jpeg_v2_5_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int i, r; @@ -202,13 +202,13 @@ static int jpeg_v2_5_hw_init(void *handle) /** * jpeg_v2_5_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the JPEG block, mark ring as not ready any more */ -static int jpeg_v2_5_hw_fini(void *handle) +static int jpeg_v2_5_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; cancel_delayed_work_sync(&adev->vcn.idle_work); @@ -231,20 +231,19 @@ static int jpeg_v2_5_hw_fini(void *handle) /** * jpeg_v2_5_suspend - suspend JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend JPEG block */ -static int jpeg_v2_5_suspend(void *handle) +static int jpeg_v2_5_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = jpeg_v2_5_hw_fini(adev); + r = jpeg_v2_5_hw_fini(ip_block); if (r) return r; - r = amdgpu_jpeg_suspend(adev); + r = amdgpu_jpeg_suspend(ip_block->adev); return r; } @@ -252,20 +251,19 @@ static int jpeg_v2_5_suspend(void *handle) /** * jpeg_v2_5_resume - resume JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init JPEG block */ -static int jpeg_v2_5_resume(void *handle) +static int jpeg_v2_5_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = amdgpu_jpeg_resume(adev); + r = amdgpu_jpeg_resume(ip_block->adev); if (r) return r; - r = jpeg_v2_5_hw_init(adev); + r = jpeg_v2_5_hw_init(ip_block); return r; } @@ -501,9 +499,9 @@ static bool jpeg_v2_5_is_idle(void *handle) return ret; } -static int jpeg_v2_5_wait_for_idle(void *handle) +static int jpeg_v2_5_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, ret; for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) { @@ -615,7 +613,6 @@ static int jpeg_v2_5_process_interrupt(struct amdgpu_device *adev, static const struct amd_ip_funcs jpeg_v2_5_ip_funcs = { .name = "jpeg_v2_5", .early_init = jpeg_v2_5_early_init, - .late_init = NULL, .sw_init = jpeg_v2_5_sw_init, .sw_fini = jpeg_v2_5_sw_fini, .hw_init = jpeg_v2_5_hw_init, @@ -624,20 +621,13 @@ static const struct amd_ip_funcs jpeg_v2_5_ip_funcs = { .resume = jpeg_v2_5_resume, .is_idle = jpeg_v2_5_is_idle, .wait_for_idle = jpeg_v2_5_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = jpeg_v2_5_set_clockgating_state, .set_powergating_state = jpeg_v2_5_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amd_ip_funcs jpeg_v2_6_ip_funcs = { .name = "jpeg_v2_6", .early_init = jpeg_v2_5_early_init, - .late_init = NULL, .sw_init = jpeg_v2_5_sw_init, .sw_fini = jpeg_v2_5_sw_fini, .hw_init = jpeg_v2_5_hw_init, @@ -646,14 +636,8 @@ static const struct amd_ip_funcs jpeg_v2_6_ip_funcs = { .resume = jpeg_v2_5_resume, .is_idle = jpeg_v2_5_is_idle, .wait_for_idle = jpeg_v2_5_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = jpeg_v2_5_set_clockgating_state, .set_powergating_state = jpeg_v2_5_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs jpeg_v2_5_dec_ring_vm_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v3_0.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v3_0.c index b1e7fd25afbc..10adbb7cbf53 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v3_0.c @@ -42,13 +42,13 @@ static int jpeg_v3_0_set_powergating_state(void *handle, /** * jpeg_v3_0_early_init - set function pointers * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers */ -static int jpeg_v3_0_early_init(void *handle) +static int jpeg_v3_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 harvest; @@ -75,13 +75,13 @@ static int jpeg_v3_0_early_init(void *handle) /** * jpeg_v3_0_sw_init - sw init for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int jpeg_v3_0_sw_init(void *handle) +static int jpeg_v3_0_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int r; @@ -118,13 +118,13 @@ static int jpeg_v3_0_sw_init(void *handle) /** * jpeg_v3_0_sw_fini - sw fini for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * JPEG suspend and free up sw allocation */ -static int jpeg_v3_0_sw_fini(void *handle) +static int jpeg_v3_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_jpeg_suspend(adev); @@ -139,12 +139,12 @@ static int jpeg_v3_0_sw_fini(void *handle) /** * jpeg_v3_0_hw_init - start and test JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * */ -static int jpeg_v3_0_hw_init(void *handle) +static int jpeg_v3_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring = adev->jpeg.inst->ring_dec; adev->nbio.funcs->vcn_doorbell_range(adev, ring->use_doorbell, @@ -156,13 +156,13 @@ static int jpeg_v3_0_hw_init(void *handle) /** * jpeg_v3_0_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the JPEG block, mark ring as not ready any more */ -static int jpeg_v3_0_hw_fini(void *handle) +static int jpeg_v3_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->vcn.idle_work); @@ -176,20 +176,19 @@ static int jpeg_v3_0_hw_fini(void *handle) /** * jpeg_v3_0_suspend - suspend JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend JPEG block */ -static int jpeg_v3_0_suspend(void *handle) +static int jpeg_v3_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = jpeg_v3_0_hw_fini(adev); + r = jpeg_v3_0_hw_fini(ip_block); if (r) return r; - r = amdgpu_jpeg_suspend(adev); + r = amdgpu_jpeg_suspend(ip_block->adev); return r; } @@ -197,20 +196,19 @@ static int jpeg_v3_0_suspend(void *handle) /** * jpeg_v3_0_resume - resume JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init JPEG block */ -static int jpeg_v3_0_resume(void *handle) +static int jpeg_v3_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = amdgpu_jpeg_resume(adev); + r = amdgpu_jpeg_resume(ip_block->adev); if (r) return r; - r = jpeg_v3_0_hw_init(adev); + r = jpeg_v3_0_hw_init(ip_block); return r; } @@ -459,9 +457,9 @@ static bool jpeg_v3_0_is_idle(void *handle) return ret; } -static int jpeg_v3_0_wait_for_idle(void *handle) +static int jpeg_v3_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return SOC15_WAIT_ON_RREG(JPEG, 0, mmUVD_JRBC_STATUS, UVD_JRBC_STATUS__RB_JOB_DONE_MASK, @@ -535,7 +533,6 @@ static int jpeg_v3_0_process_interrupt(struct amdgpu_device *adev, static const struct amd_ip_funcs jpeg_v3_0_ip_funcs = { .name = "jpeg_v3_0", .early_init = jpeg_v3_0_early_init, - .late_init = NULL, .sw_init = jpeg_v3_0_sw_init, .sw_fini = jpeg_v3_0_sw_fini, .hw_init = jpeg_v3_0_hw_init, @@ -544,14 +541,8 @@ static const struct amd_ip_funcs jpeg_v3_0_ip_funcs = { .resume = jpeg_v3_0_resume, .is_idle = jpeg_v3_0_is_idle, .wait_for_idle = jpeg_v3_0_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = jpeg_v3_0_set_clockgating_state, .set_powergating_state = jpeg_v3_0_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs jpeg_v3_0_dec_ring_vm_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c index 6c5c1a68a9b7..193dfac5dc76 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c @@ -48,13 +48,13 @@ static void jpeg_v4_0_dec_ring_set_wptr(struct amdgpu_ring *ring); /** * jpeg_v4_0_early_init - set function pointers * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers */ -static int jpeg_v4_0_early_init(void *handle) +static int jpeg_v4_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->jpeg.num_jpeg_inst = 1; @@ -70,13 +70,13 @@ static int jpeg_v4_0_early_init(void *handle) /** * jpeg_v4_0_sw_init - sw init for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int jpeg_v4_0_sw_init(void *handle) +static int jpeg_v4_0_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int r; @@ -121,6 +121,12 @@ static int jpeg_v4_0_sw_init(void *handle) adev->jpeg.inst->external.jpeg_pitch[0] = SOC15_REG_OFFSET(JPEG, 0, regUVD_JPEG_PITCH); r = amdgpu_jpeg_ras_sw_init(adev); + if (r) + return r; + /* TODO: Add queue reset mask when FW fully supports it */ + adev->jpeg.supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->jpeg.inst[0].ring_dec[0]); + r = amdgpu_jpeg_sysfs_reset_mask_init(adev); if (r) return r; @@ -130,19 +136,20 @@ static int jpeg_v4_0_sw_init(void *handle) /** * jpeg_v4_0_sw_fini - sw fini for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * JPEG suspend and free up sw allocation */ -static int jpeg_v4_0_sw_fini(void *handle) +static int jpeg_v4_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_jpeg_suspend(adev); if (r) return r; + amdgpu_jpeg_sysfs_reset_mask_fini(adev); r = amdgpu_jpeg_sw_fini(adev); return r; @@ -151,12 +158,12 @@ static int jpeg_v4_0_sw_fini(void *handle) /** * jpeg_v4_0_hw_init - start and test JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * */ -static int jpeg_v4_0_hw_init(void *handle) +static int jpeg_v4_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring = adev->jpeg.inst->ring_dec; int r; @@ -187,13 +194,13 @@ static int jpeg_v4_0_hw_init(void *handle) /** * jpeg_v4_0_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the JPEG block, mark ring as not ready any more */ -static int jpeg_v4_0_hw_fini(void *handle) +static int jpeg_v4_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->vcn.idle_work); if (!amdgpu_sriov_vf(adev)) { @@ -210,20 +217,19 @@ static int jpeg_v4_0_hw_fini(void *handle) /** * jpeg_v4_0_suspend - suspend JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend JPEG block */ -static int jpeg_v4_0_suspend(void *handle) +static int jpeg_v4_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = jpeg_v4_0_hw_fini(adev); + r = jpeg_v4_0_hw_fini(ip_block); if (r) return r; - r = amdgpu_jpeg_suspend(adev); + r = amdgpu_jpeg_suspend(ip_block->adev); return r; } @@ -231,20 +237,19 @@ static int jpeg_v4_0_suspend(void *handle) /** * jpeg_v4_0_resume - resume JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init JPEG block */ -static int jpeg_v4_0_resume(void *handle) +static int jpeg_v4_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = amdgpu_jpeg_resume(adev); + r = amdgpu_jpeg_resume(ip_block->adev); if (r) return r; - r = jpeg_v4_0_hw_init(adev); + r = jpeg_v4_0_hw_init(ip_block); return r; } @@ -621,9 +626,9 @@ static bool jpeg_v4_0_is_idle(void *handle) return ret; } -static int jpeg_v4_0_wait_for_idle(void *handle) +static int jpeg_v4_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return SOC15_WAIT_ON_RREG(JPEG, 0, regUVD_JRBC_STATUS, UVD_JRBC_STATUS__RB_JOB_DONE_MASK, @@ -702,7 +707,6 @@ static int jpeg_v4_0_process_interrupt(struct amdgpu_device *adev, static const struct amd_ip_funcs jpeg_v4_0_ip_funcs = { .name = "jpeg_v4_0", .early_init = jpeg_v4_0_early_init, - .late_init = NULL, .sw_init = jpeg_v4_0_sw_init, .sw_fini = jpeg_v4_0_sw_fini, .hw_init = jpeg_v4_0_hw_init, @@ -711,14 +715,8 @@ static const struct amd_ip_funcs jpeg_v4_0_ip_funcs = { .resume = jpeg_v4_0_resume, .is_idle = jpeg_v4_0_is_idle, .wait_for_idle = jpeg_v4_0_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = jpeg_v4_0_set_clockgating_state, .set_powergating_state = jpeg_v4_0_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs jpeg_v4_0_dec_ring_vm_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c index 86958cb2c2ab..67b51bcbacd1 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c @@ -68,13 +68,13 @@ static inline bool jpeg_v4_0_3_normalizn_reqd(struct amdgpu_device *adev) /** * jpeg_v4_0_3_early_init - set function pointers * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers */ -static int jpeg_v4_0_3_early_init(void *handle) +static int jpeg_v4_0_3_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->jpeg.num_jpeg_rings = AMDGPU_MAX_JPEG_RINGS; @@ -88,13 +88,13 @@ static int jpeg_v4_0_3_early_init(void *handle) /** * jpeg_v4_0_3_sw_init - sw init for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int jpeg_v4_0_3_sw_init(void *handle) +static int jpeg_v4_0_3_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int i, j, r, jpeg_inst; @@ -159,25 +159,33 @@ static int jpeg_v4_0_3_sw_init(void *handle) } } + /* TODO: Add queue reset mask when FW fully supports it */ + adev->jpeg.supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->jpeg.inst[0].ring_dec[0]); + r = amdgpu_jpeg_sysfs_reset_mask_init(adev); + if (r) + return r; + return 0; } /** * jpeg_v4_0_3_sw_fini - sw fini for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * JPEG suspend and free up sw allocation */ -static int jpeg_v4_0_3_sw_fini(void *handle) +static int jpeg_v4_0_3_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_jpeg_suspend(adev); if (r) return r; + amdgpu_jpeg_sysfs_reset_mask_fini(adev); r = amdgpu_jpeg_sw_fini(adev); return r; @@ -299,12 +307,12 @@ static int jpeg_v4_0_3_start_sriov(struct amdgpu_device *adev) /** * jpeg_v4_0_3_hw_init - start and test JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * */ -static int jpeg_v4_0_3_hw_init(void *handle) +static int jpeg_v4_0_3_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int i, j, r, jpeg_inst; @@ -358,13 +366,13 @@ static int jpeg_v4_0_3_hw_init(void *handle) /** * jpeg_v4_0_3_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the JPEG block, mark ring as not ready any more */ -static int jpeg_v4_0_3_hw_fini(void *handle) +static int jpeg_v4_0_3_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int ret = 0; cancel_delayed_work_sync(&adev->jpeg.idle_work); @@ -380,20 +388,19 @@ static int jpeg_v4_0_3_hw_fini(void *handle) /** * jpeg_v4_0_3_suspend - suspend JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend JPEG block */ -static int jpeg_v4_0_3_suspend(void *handle) +static int jpeg_v4_0_3_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = jpeg_v4_0_3_hw_fini(adev); + r = jpeg_v4_0_3_hw_fini(ip_block); if (r) return r; - r = amdgpu_jpeg_suspend(adev); + r = amdgpu_jpeg_suspend(ip_block->adev); return r; } @@ -401,20 +408,19 @@ static int jpeg_v4_0_3_suspend(void *handle) /** * jpeg_v4_0_3_resume - resume JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init JPEG block */ -static int jpeg_v4_0_3_resume(void *handle) +static int jpeg_v4_0_3_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = amdgpu_jpeg_resume(adev); + r = amdgpu_jpeg_resume(ip_block->adev); if (r) return r; - r = jpeg_v4_0_3_hw_init(adev); + r = jpeg_v4_0_3_hw_init(ip_block); return r; } @@ -674,11 +680,12 @@ void jpeg_v4_0_3_dec_ring_insert_start(struct amdgpu_ring *ring) amdgpu_ring_write(ring, PACKETJ(regUVD_JRBC_EXTERNAL_REG_INTERNAL_OFFSET, 0, 0, PACKETJ_TYPE0)); amdgpu_ring_write(ring, 0x62a04); /* PCTL0_MMHUB_DEEPSLEEP_IB */ - } - amdgpu_ring_write(ring, PACKETJ(JRBC_DEC_EXTERNAL_REG_WRITE_ADDR, - 0, 0, PACKETJ_TYPE0)); - amdgpu_ring_write(ring, 0x80004000); + amdgpu_ring_write(ring, + PACKETJ(JRBC_DEC_EXTERNAL_REG_WRITE_ADDR, 0, + 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, 0x80004000); + } } /** @@ -694,11 +701,12 @@ void jpeg_v4_0_3_dec_ring_insert_end(struct amdgpu_ring *ring) amdgpu_ring_write(ring, PACKETJ(regUVD_JRBC_EXTERNAL_REG_INTERNAL_OFFSET, 0, 0, PACKETJ_TYPE0)); amdgpu_ring_write(ring, 0x62a04); - } - amdgpu_ring_write(ring, PACKETJ(JRBC_DEC_EXTERNAL_REG_WRITE_ADDR, - 0, 0, PACKETJ_TYPE0)); - amdgpu_ring_write(ring, 0x00004000); + amdgpu_ring_write(ring, + PACKETJ(JRBC_DEC_EXTERNAL_REG_WRITE_ADDR, 0, + 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, 0x00004000); + } } /** @@ -743,14 +751,6 @@ void jpeg_v4_0_3_dec_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq amdgpu_ring_write(ring, PACKETJ(0, 0, 0, PACKETJ_TYPE6)); amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, PACKETJ(regUVD_JRBC_EXTERNAL_REG_INTERNAL_OFFSET, - 0, 0, PACKETJ_TYPE0)); - amdgpu_ring_write(ring, 0x3fbc); - - amdgpu_ring_write(ring, PACKETJ(JRBC_DEC_EXTERNAL_REG_WRITE_ADDR, - 0, 0, PACKETJ_TYPE0)); - amdgpu_ring_write(ring, 0x1); - amdgpu_ring_write(ring, PACKETJ(0, 0, 0, PACKETJ_TYPE6)); amdgpu_ring_write(ring, 0); @@ -929,9 +929,9 @@ static bool jpeg_v4_0_3_is_idle(void *handle) return ret; } -static int jpeg_v4_0_3_wait_for_idle(void *handle) +static int jpeg_v4_0_3_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int ret = 0; int i, j; @@ -1058,7 +1058,6 @@ static int jpeg_v4_0_3_process_interrupt(struct amdgpu_device *adev, static const struct amd_ip_funcs jpeg_v4_0_3_ip_funcs = { .name = "jpeg_v4_0_3", .early_init = jpeg_v4_0_3_early_init, - .late_init = NULL, .sw_init = jpeg_v4_0_3_sw_init, .sw_fini = jpeg_v4_0_3_sw_fini, .hw_init = jpeg_v4_0_3_hw_init, @@ -1067,14 +1066,8 @@ static const struct amd_ip_funcs jpeg_v4_0_3_ip_funcs = { .resume = jpeg_v4_0_3_resume, .is_idle = jpeg_v4_0_3_is_idle, .wait_for_idle = jpeg_v4_0_3_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = jpeg_v4_0_3_set_clockgating_state, .set_powergating_state = jpeg_v4_0_3_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs jpeg_v4_0_3_dec_ring_vm_funcs = { @@ -1088,7 +1081,7 @@ static const struct amdgpu_ring_funcs jpeg_v4_0_3_dec_ring_vm_funcs = { SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 + 8 + /* jpeg_v4_0_3_dec_ring_emit_vm_flush */ - 22 + 22 + /* jpeg_v4_0_3_dec_ring_emit_fence x2 vm fence */ + 18 + 18 + /* jpeg_v4_0_3_dec_ring_emit_fence x2 vm fence */ 8 + 16, .emit_ib_size = 22, /* jpeg_v4_0_3_dec_ring_emit_ib */ .emit_ib = jpeg_v4_0_3_dec_ring_emit_ib, diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_5.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_5.c index 44eeed445ea9..b48e2412e6cc 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_5.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_5.c @@ -61,13 +61,13 @@ static int amdgpu_ih_clientid_jpeg[] = { /** * jpeg_v4_0_5_early_init - set function pointers * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers */ -static int jpeg_v4_0_5_early_init(void *handle) +static int jpeg_v4_0_5_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; switch (amdgpu_ip_version(adev, UVD_HWIP, 0)) { case IP_VERSION(4, 0, 5): @@ -94,13 +94,13 @@ static int jpeg_v4_0_5_early_init(void *handle) /** * jpeg_v4_0_5_sw_init - sw init for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int jpeg_v4_0_5_sw_init(void *handle) +static int jpeg_v4_0_5_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int r, i; @@ -153,25 +153,33 @@ static int jpeg_v4_0_5_sw_init(void *handle) adev->jpeg.inst[i].external.jpeg_pitch[0] = SOC15_REG_OFFSET(JPEG, i, regUVD_JPEG_PITCH); } + /* TODO: Add queue reset mask when FW fully supports it */ + adev->jpeg.supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->jpeg.inst[0].ring_dec[0]); + r = amdgpu_jpeg_sysfs_reset_mask_init(adev); + if (r) + return r; + return 0; } /** * jpeg_v4_0_5_sw_fini - sw fini for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * JPEG suspend and free up sw allocation */ -static int jpeg_v4_0_5_sw_fini(void *handle) +static int jpeg_v4_0_5_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_jpeg_suspend(adev); if (r) return r; + amdgpu_jpeg_sysfs_reset_mask_fini(adev); r = amdgpu_jpeg_sw_fini(adev); return r; @@ -180,12 +188,12 @@ static int jpeg_v4_0_5_sw_fini(void *handle) /** * jpeg_v4_0_5_hw_init - start and test JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * */ -static int jpeg_v4_0_5_hw_init(void *handle) +static int jpeg_v4_0_5_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int i, r = 0; @@ -210,13 +218,13 @@ static int jpeg_v4_0_5_hw_init(void *handle) /** * jpeg_v4_0_5_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the JPEG block, mark ring as not ready any more */ -static int jpeg_v4_0_5_hw_fini(void *handle) +static int jpeg_v4_0_5_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; cancel_delayed_work_sync(&adev->vcn.idle_work); @@ -237,20 +245,19 @@ static int jpeg_v4_0_5_hw_fini(void *handle) /** * jpeg_v4_0_5_suspend - suspend JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend JPEG block */ -static int jpeg_v4_0_5_suspend(void *handle) +static int jpeg_v4_0_5_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = jpeg_v4_0_5_hw_fini(adev); + r = jpeg_v4_0_5_hw_fini(ip_block); if (r) return r; - r = amdgpu_jpeg_suspend(adev); + r = amdgpu_jpeg_suspend(ip_block->adev); return r; } @@ -258,20 +265,19 @@ static int jpeg_v4_0_5_suspend(void *handle) /** * jpeg_v4_0_5_resume - resume JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init JPEG block */ -static int jpeg_v4_0_5_resume(void *handle) +static int jpeg_v4_0_5_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = amdgpu_jpeg_resume(adev); + r = amdgpu_jpeg_resume(ip_block->adev); if (r) return r; - r = jpeg_v4_0_5_hw_init(adev); + r = jpeg_v4_0_5_hw_init(ip_block); return r; } @@ -637,9 +643,9 @@ static bool jpeg_v4_0_5_is_idle(void *handle) return ret; } -static int jpeg_v4_0_5_wait_for_idle(void *handle) +static int jpeg_v4_0_5_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) { @@ -743,7 +749,6 @@ static int jpeg_v4_0_5_process_interrupt(struct amdgpu_device *adev, static const struct amd_ip_funcs jpeg_v4_0_5_ip_funcs = { .name = "jpeg_v4_0_5", .early_init = jpeg_v4_0_5_early_init, - .late_init = NULL, .sw_init = jpeg_v4_0_5_sw_init, .sw_fini = jpeg_v4_0_5_sw_fini, .hw_init = jpeg_v4_0_5_hw_init, @@ -752,14 +757,8 @@ static const struct amd_ip_funcs jpeg_v4_0_5_ip_funcs = { .resume = jpeg_v4_0_5_resume, .is_idle = jpeg_v4_0_5_is_idle, .wait_for_idle = jpeg_v4_0_5_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = jpeg_v4_0_5_set_clockgating_state, .set_powergating_state = jpeg_v4_0_5_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs jpeg_v4_0_5_dec_ring_vm_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_0.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_0.c index d662aa841f97..686f9605239d 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_0.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_0.c @@ -42,13 +42,13 @@ static int jpeg_v5_0_0_set_powergating_state(void *handle, /** * jpeg_v5_0_0_early_init - set function pointers * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers */ -static int jpeg_v5_0_0_early_init(void *handle) +static int jpeg_v5_0_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->jpeg.num_jpeg_inst = 1; adev->jpeg.num_jpeg_rings = 1; @@ -62,13 +62,13 @@ static int jpeg_v5_0_0_early_init(void *handle) /** * jpeg_v5_0_0_sw_init - sw init for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int jpeg_v5_0_0_sw_init(void *handle) +static int jpeg_v5_0_0_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int r; @@ -100,25 +100,32 @@ static int jpeg_v5_0_0_sw_init(void *handle) adev->jpeg.internal.jpeg_pitch[0] = regUVD_JPEG_PITCH_INTERNAL_OFFSET; adev->jpeg.inst->external.jpeg_pitch[0] = SOC15_REG_OFFSET(JPEG, 0, regUVD_JPEG_PITCH); + /* TODO: Add queue reset mask when FW fully supports it */ + adev->jpeg.supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->jpeg.inst[0].ring_dec[0]); + r = amdgpu_jpeg_sysfs_reset_mask_init(adev); + if (r) + return r; return 0; } /** * jpeg_v5_0_0_sw_fini - sw fini for JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * JPEG suspend and free up sw allocation */ -static int jpeg_v5_0_0_sw_fini(void *handle) +static int jpeg_v5_0_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_jpeg_suspend(adev); if (r) return r; + amdgpu_jpeg_sysfs_reset_mask_fini(adev); r = amdgpu_jpeg_sw_fini(adev); return r; @@ -127,12 +134,12 @@ static int jpeg_v5_0_0_sw_fini(void *handle) /** * jpeg_v5_0_0_hw_init - start and test JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * */ -static int jpeg_v5_0_0_hw_init(void *handle) +static int jpeg_v5_0_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring = adev->jpeg.inst->ring_dec; int r; @@ -153,13 +160,13 @@ static int jpeg_v5_0_0_hw_init(void *handle) /** * jpeg_v5_0_0_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the JPEG block, mark ring as not ready any more */ -static int jpeg_v5_0_0_hw_fini(void *handle) +static int jpeg_v5_0_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->vcn.idle_work); @@ -173,20 +180,19 @@ static int jpeg_v5_0_0_hw_fini(void *handle) /** * jpeg_v5_0_0_suspend - suspend JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend JPEG block */ -static int jpeg_v5_0_0_suspend(void *handle) +static int jpeg_v5_0_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = jpeg_v5_0_0_hw_fini(adev); + r = jpeg_v5_0_0_hw_fini(ip_block); if (r) return r; - r = amdgpu_jpeg_suspend(adev); + r = amdgpu_jpeg_suspend(ip_block->adev); return r; } @@ -194,20 +200,19 @@ static int jpeg_v5_0_0_suspend(void *handle) /** * jpeg_v5_0_0_resume - resume JPEG block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init JPEG block */ -static int jpeg_v5_0_0_resume(void *handle) +static int jpeg_v5_0_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = amdgpu_jpeg_resume(adev); + r = amdgpu_jpeg_resume(ip_block->adev); if (r) return r; - r = jpeg_v5_0_0_hw_init(adev); + r = jpeg_v5_0_0_hw_init(ip_block); return r; } @@ -546,9 +551,9 @@ static bool jpeg_v5_0_0_is_idle(void *handle) return ret; } -static int jpeg_v5_0_0_wait_for_idle(void *handle) +static int jpeg_v5_0_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return SOC15_WAIT_ON_RREG(JPEG, 0, regUVD_JRBC_STATUS, UVD_JRBC_STATUS__RB_JOB_DONE_MASK, @@ -622,7 +627,6 @@ static int jpeg_v5_0_0_process_interrupt(struct amdgpu_device *adev, static const struct amd_ip_funcs jpeg_v5_0_0_ip_funcs = { .name = "jpeg_v5_0_0", .early_init = jpeg_v5_0_0_early_init, - .late_init = NULL, .sw_init = jpeg_v5_0_0_sw_init, .sw_fini = jpeg_v5_0_0_sw_fini, .hw_init = jpeg_v5_0_0_hw_init, @@ -631,14 +635,8 @@ static const struct amd_ip_funcs jpeg_v5_0_0_ip_funcs = { .resume = jpeg_v5_0_0_resume, .is_idle = jpeg_v5_0_0_is_idle, .wait_for_idle = jpeg_v5_0_0_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = jpeg_v5_0_0_set_clockgating_state, .set_powergating_state = jpeg_v5_0_0_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs jpeg_v5_0_0_dec_ring_vm_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c index 231a3d490ea8..9c905b9e9376 100644 --- a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c @@ -55,8 +55,8 @@ MODULE_FIRMWARE("amdgpu/gc_11_5_1_mes1.bin"); MODULE_FIRMWARE("amdgpu/gc_11_5_2_mes_2.bin"); MODULE_FIRMWARE("amdgpu/gc_11_5_2_mes1.bin"); -static int mes_v11_0_hw_init(void *handle); -static int mes_v11_0_hw_fini(void *handle); +static int mes_v11_0_hw_init(struct amdgpu_ip_block *ip_block); +static int mes_v11_0_hw_fini(struct amdgpu_ip_block *ip_block); static int mes_v11_0_kiq_hw_init(struct amdgpu_device *adev); static int mes_v11_0_kiq_hw_fini(struct amdgpu_device *adev); @@ -366,7 +366,7 @@ static int mes_v11_0_reset_queue_mmio(struct amdgpu_mes *mes, uint32_t queue_typ uint32_t queue_id, uint32_t vmid) { struct amdgpu_device *adev = mes->adev; - uint32_t value; + uint32_t value, reg; int i, r = 0; amdgpu_gfx_rlc_enter_safe_mode(adev, 0); @@ -424,6 +424,31 @@ static int mes_v11_0_reset_queue_mmio(struct amdgpu_mes *mes, uint32_t queue_typ } soc21_grbm_select(adev, 0, 0, 0, 0); mutex_unlock(&adev->srbm_mutex); + } else if (queue_type == AMDGPU_RING_TYPE_SDMA) { + dev_info(adev->dev, "reset sdma queue (%d:%d:%d)\n", + me_id, pipe_id, queue_id); + switch (me_id) { + case 1: + reg = SOC15_REG_OFFSET(GC, 0, regSDMA1_QUEUE_RESET_REQ); + break; + case 0: + default: + reg = SOC15_REG_OFFSET(GC, 0, regSDMA0_QUEUE_RESET_REQ); + break; + } + + value = 1 << queue_id; + WREG32(reg, value); + /* wait for queue reset done */ + for (i = 0; i < adev->usec_timeout; i++) { + if (!(RREG32(reg) & value)) + break; + udelay(1); + } + if (i >= adev->usec_timeout) { + dev_err(adev->dev, "failed to wait on sdma queue reset done\n"); + r = -ETIMEDOUT; + } } amdgpu_gfx_rlc_exit_safe_mode(adev, 0); @@ -619,6 +644,18 @@ static int mes_v11_0_misc_op(struct amdgpu_mes *mes, sizeof(misc_pkt.set_shader_debugger.tcp_watch_cntl)); misc_pkt.set_shader_debugger.trap_en = input->set_shader_debugger.trap_en; break; + case MES_MISC_OP_CHANGE_CONFIG: + if ((mes->adev->mes.sched_version & AMDGPU_MES_VERSION_MASK) < 0x63) { + dev_err(mes->adev->dev, "MES FW versoin must be larger than 0x63 to support limit single process feature.\n"); + return -EINVAL; + } + misc_pkt.opcode = MESAPI_MISC__CHANGE_CONFIG; + misc_pkt.change_config.opcode = + MESAPI_MISC__CHANGE_CONFIG_OPTION_LIMIT_SINGLE_PROCESS; + misc_pkt.change_config.option.bits.limit_single_process = + input->change_config.option.limit_single_process; + break; + default: DRM_ERROR("unsupported misc op (%d) \n", input->op); return -EINVAL; @@ -683,6 +720,9 @@ static int mes_v11_0_set_hw_resources(struct amdgpu_mes *mes) mes->event_log_gpu_addr; } + if (enforce_isolation) + mes_set_hw_res_pkt.limit_single_process = 1; + return mes_v11_0_submit_pkt_and_poll_completion(mes, &mes_set_hw_res_pkt, sizeof(mes_set_hw_res_pkt), offsetof(union MESAPI_SET_HW_RESOURCES, api_status)); @@ -883,6 +923,16 @@ static void mes_v11_0_enable(struct amdgpu_device *adev, bool enable) uint32_t pipe, data = 0; if (enable) { + if (amdgpu_mes_log_enable) { + WREG32_SOC15(GC, 0, regCP_MES_MSCRATCH_LO, + lower_32_bits(adev->mes.event_log_gpu_addr + AMDGPU_MES_LOG_BUFFER_SIZE)); + WREG32_SOC15(GC, 0, regCP_MES_MSCRATCH_HI, + upper_32_bits(adev->mes.event_log_gpu_addr + AMDGPU_MES_LOG_BUFFER_SIZE)); + dev_info(adev->dev, "Setup CP MES MSCRATCH address : 0x%x. 0x%x\n", + RREG32_SOC15(GC, 0, regCP_MES_MSCRATCH_HI), + RREG32_SOC15(GC, 0, regCP_MES_MSCRATCH_LO)); + } + data = RREG32_SOC15(GC, 0, regCP_MES_CNTL); data = REG_SET_FIELD(data, CP_MES_CNTL, MES_PIPE0_RESET, 1); data = REG_SET_FIELD(data, CP_MES_CNTL, @@ -1336,16 +1386,16 @@ static int mes_v11_0_mqd_sw_init(struct amdgpu_device *adev, return 0; } -static int mes_v11_0_sw_init(void *handle) +static int mes_v11_0_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int pipe, r; adev->mes.funcs = &mes_v11_0_funcs; adev->mes.kiq_hw_init = &mes_v11_0_kiq_hw_init; adev->mes.kiq_hw_fini = &mes_v11_0_kiq_hw_fini; - adev->mes.event_log_size = AMDGPU_MES_LOG_BUFFER_SIZE; + adev->mes.event_log_size = AMDGPU_MES_LOG_BUFFER_SIZE + AMDGPU_MES_MSCRATCH_SIZE; r = amdgpu_mes_init(adev); if (r) @@ -1377,9 +1427,9 @@ static int mes_v11_0_sw_init(void *handle) return 0; } -static int mes_v11_0_sw_fini(void *handle) +static int mes_v11_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int pipe; for (pipe = 0; pipe < AMDGPU_MAX_MES_PIPES; pipe++) { @@ -1473,6 +1523,7 @@ static void mes_v11_0_kiq_clear(struct amdgpu_device *adev) static int mes_v11_0_kiq_hw_init(struct amdgpu_device *adev) { int r = 0; + struct amdgpu_ip_block *ip_block; if (adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT) { @@ -1496,6 +1547,12 @@ static int mes_v11_0_kiq_hw_init(struct amdgpu_device *adev) mes_v11_0_kiq_setting(&adev->gfx.kiq[0].ring); + ip_block = amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_MES); + if (unlikely(!ip_block)) { + dev_err(adev->dev, "Failed to get MES handle\n"); + return -EINVAL; + } + r = mes_v11_0_queue_init(adev, AMDGPU_MES_KIQ_PIPE); if (r) goto failure; @@ -1506,7 +1563,7 @@ static int mes_v11_0_kiq_hw_init(struct amdgpu_device *adev) adev->mes.enable_legacy_queue_map = false; if (adev->mes.enable_legacy_queue_map) { - r = mes_v11_0_hw_init(adev); + r = mes_v11_0_hw_init(ip_block); if (r) goto failure; } @@ -1514,7 +1571,7 @@ static int mes_v11_0_kiq_hw_init(struct amdgpu_device *adev) return r; failure: - mes_v11_0_hw_fini(adev); + mes_v11_0_hw_fini(ip_block); return r; } @@ -1535,10 +1592,10 @@ static int mes_v11_0_kiq_hw_fini(struct amdgpu_device *adev) return 0; } -static int mes_v11_0_hw_init(void *handle) +static int mes_v11_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->mes.ring[0].sched.ready) goto out; @@ -1590,13 +1647,13 @@ out: return 0; failure: - mes_v11_0_hw_fini(adev); + mes_v11_0_hw_fini(ip_block); return r; } -static int mes_v11_0_hw_fini(void *handle) +static int mes_v11_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_is_mes_info_enable(adev)) { amdgpu_bo_free_kernel(&adev->mes.resource_1, &adev->mes.resource_1_gpu_addr, &adev->mes.resource_1_addr); @@ -1604,33 +1661,31 @@ static int mes_v11_0_hw_fini(void *handle) return 0; } -static int mes_v11_0_suspend(void *handle) +static int mes_v11_0_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_mes_suspend(adev); + r = amdgpu_mes_suspend(ip_block->adev); if (r) return r; - return mes_v11_0_hw_fini(adev); + return mes_v11_0_hw_fini(ip_block); } -static int mes_v11_0_resume(void *handle) +static int mes_v11_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = mes_v11_0_hw_init(adev); + r = mes_v11_0_hw_init(ip_block); if (r) return r; - return amdgpu_mes_resume(adev); + return amdgpu_mes_resume(ip_block->adev); } -static int mes_v11_0_early_init(void *handle) +static int mes_v11_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int pipe, r; for (pipe = 0; pipe < AMDGPU_MAX_MES_PIPES; pipe++) { @@ -1644,9 +1699,9 @@ static int mes_v11_0_early_init(void *handle) return 0; } -static int mes_v11_0_late_init(void *handle) +static int mes_v11_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* it's only intended for use in mes_self_test case, not for s0ix and reset */ if (!amdgpu_in_reset(adev) && !adev->in_s0ix && !adev->in_suspend && @@ -1666,8 +1721,6 @@ static const struct amd_ip_funcs mes_v11_0_ip_funcs = { .hw_fini = mes_v11_0_hw_fini, .suspend = mes_v11_0_suspend, .resume = mes_v11_0_resume, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; const struct amdgpu_ip_block_version mes_v11_0_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/mes_v12_0.c b/drivers/gpu/drm/amd/amdgpu/mes_v12_0.c index b3175ff676f3..9ecc5d61e49b 100644 --- a/drivers/gpu/drm/amd/amdgpu/mes_v12_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mes_v12_0.c @@ -39,8 +39,8 @@ MODULE_FIRMWARE("amdgpu/gc_12_0_1_mes.bin"); MODULE_FIRMWARE("amdgpu/gc_12_0_1_mes1.bin"); MODULE_FIRMWARE("amdgpu/gc_12_0_1_uni_mes.bin"); -static int mes_v12_0_hw_init(void *handle); -static int mes_v12_0_hw_fini(void *handle); +static int mes_v12_0_hw_init(struct amdgpu_ip_block *ip_block); +static int mes_v12_0_hw_fini(struct amdgpu_ip_block *ip_block); static int mes_v12_0_kiq_hw_init(struct amdgpu_device *adev); static int mes_v12_0_kiq_hw_fini(struct amdgpu_device *adev); @@ -531,6 +531,14 @@ static int mes_v12_0_misc_op(struct amdgpu_mes *mes, sizeof(misc_pkt.set_shader_debugger.tcp_watch_cntl)); misc_pkt.set_shader_debugger.trap_en = input->set_shader_debugger.trap_en; break; + case MES_MISC_OP_CHANGE_CONFIG: + misc_pkt.opcode = MESAPI_MISC__CHANGE_CONFIG; + misc_pkt.change_config.opcode = + MESAPI_MISC__CHANGE_CONFIG_OPTION_LIMIT_SINGLE_PROCESS; + misc_pkt.change_config.option.bits.limit_single_process = + input->change_config.option.limit_single_process; + break; + default: DRM_ERROR("unsupported misc op (%d) \n", input->op); return -EINVAL; @@ -624,6 +632,9 @@ static int mes_v12_0_set_hw_resources(struct amdgpu_mes *mes, int pipe) mes_set_hw_res_pkt.event_intr_history_gpu_mc_ptr = mes->event_log_gpu_addr + pipe * AMDGPU_MES_LOG_BUFFER_SIZE; } + if (enforce_isolation) + mes_set_hw_res_pkt.limit_single_process = 1; + return mes_v12_0_submit_pkt_and_poll_completion(mes, pipe, &mes_set_hw_res_pkt, sizeof(mes_set_hw_res_pkt), offsetof(union MESAPI_SET_HW_RESOURCES, api_status)); @@ -1326,9 +1337,9 @@ static int mes_v12_0_mqd_sw_init(struct amdgpu_device *adev, return 0; } -static int mes_v12_0_sw_init(void *handle) +static int mes_v12_0_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int pipe, r; adev->mes.funcs = &mes_v12_0_funcs; @@ -1362,9 +1373,9 @@ static int mes_v12_0_sw_init(void *handle) return 0; } -static int mes_v12_0_sw_fini(void *handle) +static int mes_v12_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int pipe; for (pipe = 0; pipe < AMDGPU_MAX_MES_PIPES; pipe++) { @@ -1452,6 +1463,7 @@ static void mes_v12_0_kiq_setting(struct amdgpu_ring *ring) static int mes_v12_0_kiq_hw_init(struct amdgpu_device *adev) { int r = 0; + struct amdgpu_ip_block *ip_block; if (adev->enable_uni_mes) mes_v12_0_kiq_setting(&adev->mes.ring[AMDGPU_MES_KIQ_PIPE]); @@ -1479,6 +1491,12 @@ static int mes_v12_0_kiq_hw_init(struct amdgpu_device *adev) mes_v12_0_enable(adev, true); + ip_block = amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_MES); + if (unlikely(!ip_block)) { + dev_err(adev->dev, "Failed to get MES handle\n"); + return -EINVAL; + } + r = mes_v12_0_queue_init(adev, AMDGPU_MES_KIQ_PIPE); if (r) goto failure; @@ -1492,7 +1510,7 @@ static int mes_v12_0_kiq_hw_init(struct amdgpu_device *adev) } if (adev->mes.enable_legacy_queue_map) { - r = mes_v12_0_hw_init(adev); + r = mes_v12_0_hw_init(ip_block); if (r) goto failure; } @@ -1500,7 +1518,7 @@ static int mes_v12_0_kiq_hw_init(struct amdgpu_device *adev) return r; failure: - mes_v12_0_hw_fini(adev); + mes_v12_0_hw_fini(ip_block); return r; } @@ -1522,10 +1540,10 @@ static int mes_v12_0_kiq_hw_fini(struct amdgpu_device *adev) return 0; } -static int mes_v12_0_hw_init(void *handle) +static int mes_v12_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->mes.ring[0].sched.ready) goto out; @@ -1584,42 +1602,40 @@ out: return 0; failure: - mes_v12_0_hw_fini(adev); + mes_v12_0_hw_fini(ip_block); return r; } -static int mes_v12_0_hw_fini(void *handle) +static int mes_v12_0_hw_fini(struct amdgpu_ip_block *ip_block) { return 0; } -static int mes_v12_0_suspend(void *handle) +static int mes_v12_0_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_mes_suspend(adev); + r = amdgpu_mes_suspend(ip_block->adev); if (r) return r; - return mes_v12_0_hw_fini(adev); + return mes_v12_0_hw_fini(ip_block); } -static int mes_v12_0_resume(void *handle) +static int mes_v12_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = mes_v12_0_hw_init(adev); + r = mes_v12_0_hw_init(ip_block); if (r) return r; - return amdgpu_mes_resume(adev); + return amdgpu_mes_resume(ip_block->adev); } -static int mes_v12_0_early_init(void *handle) +static int mes_v12_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int pipe, r; for (pipe = 0; pipe < AMDGPU_MAX_MES_PIPES; pipe++) { @@ -1631,9 +1647,9 @@ static int mes_v12_0_early_init(void *handle) return 0; } -static int mes_v12_0_late_init(void *handle) +static int mes_v12_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* it's only intended for use in mes_self_test case, not for s0ix and reset */ if (!amdgpu_in_reset(adev) && !adev->in_s0ix && !adev->in_suspend) diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c index e3ddd22aa172..e9a6f33ca710 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c @@ -229,6 +229,52 @@ static void mmhub_v1_0_disable_identity_aperture(struct amdgpu_device *adev) 0); } +static void mmhub_v1_0_init_saw(struct amdgpu_device *adev) +{ + uint64_t pt_base = amdgpu_gmc_pd_addr(adev->gart.bo); + uint32_t tmp; + + /* VM_9_X_REGISTER_VM_L2_SAW_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32 */ + WREG32_SOC15(MMHUB, 0, mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32, + lower_32_bits(pt_base >> 12)); + + /* VM_9_X_REGISTER_VM_L2_SAW_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32 */ + WREG32_SOC15(MMHUB, 0, mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32, + upper_32_bits(pt_base >> 12)); + + /* VM_9_X_REGISTER_VM_L2_SAW_CONTEXT0_PAGE_TABLE_START_ADDR_LO32 */ + WREG32_SOC15(MMHUB, 0, mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_START_ADDR_LO32, + (u32)(adev->gmc.gart_start >> 12)); + + /* VM_9_X_REGISTER_VM_L2_SAW_CONTEXT0_PAGE_TABLE_START_ADDR_HI32 */ + WREG32_SOC15(MMHUB, 0, mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_START_ADDR_HI32, + (u32)(adev->gmc.gart_start >> 44)); + + /* VM_9_X_REGISTER_VM_L2_SAW_CONTEXT0_PAGE_TABLE_END_ADDR_LO32 */ + WREG32_SOC15(MMHUB, 0, mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_END_ADDR_LO32, + (u32)(adev->gmc.gart_end >> 12)); + + /* VM_9_X_REGISTER_VM_L2_SAW_CONTEXT0_PAGE_TABLE_END_ADDR_HI32 */ + WREG32_SOC15(MMHUB, 0, mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_END_ADDR_HI32, + (u32)(adev->gmc.gart_end >> 44)); + + /* Program SAW CONTEXT0 CNTL */ + tmp = RREG32_SOC15(MMHUB, 0, mmVM_L2_SAW_CONTEXT0_CNTL); + tmp |= 1 << CONTEXT0_CNTL_ENABLE_OFFSET; + tmp &= ~(3 << CONTEXT0_CNTL_PAGE_TABLE_DEPTH_OFFSET); + WREG32_SOC15(MMHUB, 0, mmVM_L2_SAW_CONTEXT0_CNTL, tmp); + + /* Disable all Contexts except Context0 */ + tmp = 0xfffe; + WREG32_SOC15(MMHUB, 0, mmVM_L2_SAW_CONTEXTS_DISABLE, tmp); + + /* Program SAW CNTL4 */ + tmp = RREG32_SOC15(MMHUB, 0, mmVM_L2_SAW_CNTL4); + tmp |= 1 << VMC_TAP_PDE_REQUEST_SNOOP_OFFSET; + tmp |= 1 << VMC_TAP_PTE_REQUEST_SNOOP_OFFSET; + WREG32_SOC15(MMHUB, 0, mmVM_L2_SAW_CNTL4, tmp); +} + static void mmhub_v1_0_setup_vmid_config(struct amdgpu_device *adev) { struct amdgpu_vmhub *hub = &adev->vmhub[AMDGPU_MMHUB0(0)]; @@ -283,6 +329,9 @@ static void mmhub_v1_0_setup_vmid_config(struct amdgpu_device *adev) i * hub->ctx_addr_distance, upper_32_bits(adev->vm_manager.max_pfn - 1)); } + + if (amdgpu_ip_version(adev, ISP_HWIP, 0)) + mmhub_v1_0_init_saw(adev); } static void mmhub_v1_0_program_invalidation(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c index f47bd7ada4d7..4dcb72d1bdda 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c @@ -61,15 +61,18 @@ static enum idh_event xgpu_nv_mailbox_peek_msg(struct amdgpu_device *adev) static int xgpu_nv_mailbox_rcv_msg(struct amdgpu_device *adev, enum idh_event event) { + int r = 0; u32 reg; reg = RREG32_NO_KIQ(mmMAILBOX_MSGBUF_RCV_DW0); - if (reg != event) + if (reg == IDH_FAIL) + r = -EINVAL; + else if (reg != event) return -ENOENT; xgpu_nv_mailbox_send_ack(adev); - return 0; + return r; } static uint8_t xgpu_nv_peek_ack(struct amdgpu_device *adev) @@ -178,6 +181,9 @@ send_request: if (data1 != 0) event = IDH_RAS_POISON_READY; break; + case IDH_REQ_RAS_ERROR_COUNT: + event = IDH_RAS_ERROR_COUNT_READY; + break; default: break; } @@ -456,6 +462,11 @@ static bool xgpu_nv_rcvd_ras_intr(struct amdgpu_device *adev) return (msg == IDH_RAS_ERROR_DETECTED || msg == 0xFFFFFFFF); } +static int xgpu_nv_req_ras_err_count(struct amdgpu_device *adev) +{ + return xgpu_nv_send_access_requests(adev, IDH_REQ_RAS_ERROR_COUNT); +} + const struct amdgpu_virt_ops xgpu_nv_virt_ops = { .req_full_gpu = xgpu_nv_request_full_gpu_access, .rel_full_gpu = xgpu_nv_release_full_gpu_access, @@ -466,4 +477,5 @@ const struct amdgpu_virt_ops xgpu_nv_virt_ops = { .trans_msg = xgpu_nv_mailbox_trans_msg, .ras_poison_handler = xgpu_nv_ras_poison_handler, .rcvd_ras_intr = xgpu_nv_rcvd_ras_intr, + .req_ras_err_count = xgpu_nv_req_ras_err_count, }; diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.h b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.h index 1d099ffb3a5a..9d61d76e1bf9 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.h +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.h @@ -40,6 +40,7 @@ enum idh_request { IDH_LOG_VF_ERROR = 200, IDH_READY_TO_RESET = 201, IDH_RAS_POISON = 202, + IDH_REQ_RAS_ERROR_COUNT = 203, }; enum idh_event { @@ -54,6 +55,8 @@ enum idh_event { IDH_RAS_POISON_READY, IDH_PF_SOFT_FLR_NOTIFICATION, IDH_RAS_ERROR_DETECTED, + IDH_RAS_ERROR_COUNT_READY = 11, + IDH_TEXT_MESSAGE = 255, }; diff --git a/drivers/gpu/drm/amd/amdgpu/navi10_ih.c b/drivers/gpu/drm/amd/amdgpu/navi10_ih.c index b281462093f1..0820ed62e2e8 100644 --- a/drivers/gpu/drm/amd/amdgpu/navi10_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/navi10_ih.c @@ -542,19 +542,19 @@ static void navi10_ih_set_self_irq_funcs(struct amdgpu_device *adev) adev->irq.self_irq.funcs = &navi10_ih_self_irq_funcs; } -static int navi10_ih_early_init(void *handle) +static int navi10_ih_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; navi10_ih_set_interrupt_funcs(adev); navi10_ih_set_self_irq_funcs(adev); return 0; } -static int navi10_ih_sw_init(void *handle) +static int navi10_ih_sw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; bool use_bus_addr; r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_IH, 0, @@ -593,43 +593,37 @@ static int navi10_ih_sw_init(void *handle) return r; } -static int navi10_ih_sw_fini(void *handle) +static int navi10_ih_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_fini_sw(adev); return 0; } -static int navi10_ih_hw_init(void *handle) +static int navi10_ih_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return navi10_ih_irq_init(adev); } -static int navi10_ih_hw_fini(void *handle) +static int navi10_ih_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - navi10_ih_irq_disable(adev); + navi10_ih_irq_disable(ip_block->adev); return 0; } -static int navi10_ih_suspend(void *handle) +static int navi10_ih_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return navi10_ih_hw_fini(adev); + return navi10_ih_hw_fini(ip_block); } -static int navi10_ih_resume(void *handle) +static int navi10_ih_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return navi10_ih_hw_init(adev); + return navi10_ih_hw_init(ip_block); } static bool navi10_ih_is_idle(void *handle) @@ -638,13 +632,13 @@ static bool navi10_ih_is_idle(void *handle) return true; } -static int navi10_ih_wait_for_idle(void *handle) +static int navi10_ih_wait_for_idle(struct amdgpu_ip_block *ip_block) { /* todo */ return -ETIMEDOUT; } -static int navi10_ih_soft_reset(void *handle) +static int navi10_ih_soft_reset(struct amdgpu_ip_block *ip_block) { /* todo */ return 0; @@ -700,7 +694,6 @@ static void navi10_ih_get_clockgating_state(void *handle, u64 *flags) static const struct amd_ip_funcs navi10_ih_ip_funcs = { .name = "navi10_ih", .early_init = navi10_ih_early_init, - .late_init = NULL, .sw_init = navi10_ih_sw_init, .sw_fini = navi10_ih_sw_fini, .hw_init = navi10_ih_hw_init, @@ -713,8 +706,6 @@ static const struct amd_ip_funcs navi10_ih_ip_funcs = { .set_clockgating_state = navi10_ih_set_clockgating_state, .set_powergating_state = navi10_ih_set_powergating_state, .get_clockgating_state = navi10_ih_get_clockgating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs navi10_ih_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/navi10_sdma_pkt_open.h b/drivers/gpu/drm/amd/amdgpu/navi10_sdma_pkt_open.h index a5b60c9a2418..c88284ff92d8 100644 --- a/drivers/gpu/drm/amd/amdgpu/navi10_sdma_pkt_open.h +++ b/drivers/gpu/drm/amd/amdgpu/navi10_sdma_pkt_open.h @@ -68,6 +68,7 @@ #define SDMA_SUBOP_POLL_REG_WRITE_MEM 1 #define SDMA_SUBOP_POLL_DBIT_WRITE_MEM 2 #define SDMA_SUBOP_POLL_MEM_VERIFY 3 +#define SDMA_SUBOP_VM_INVALIDATION 4 #define HEADER_AGENT_DISPATCH 4 #define HEADER_BARRIER 5 #define SDMA_OP_AQL_COPY 0 @@ -4040,6 +4041,69 @@ #define SDMA_PKT_POLL_MEM_VERIFY_RESERVED_RESERVED(x) (((x) & SDMA_PKT_POLL_MEM_VERIFY_RESERVED_reserved_mask) << SDMA_PKT_POLL_MEM_VERIFY_RESERVED_reserved_shift) +/* +** Definitions for SDMA_PKT_VM_INVALIDATION packet +*/ + +/*define for HEADER word*/ +/*define for op field*/ +#define SDMA_PKT_VM_INVALIDATION_HEADER_op_offset 0 +#define SDMA_PKT_VM_INVALIDATION_HEADER_op_mask 0x000000FF +#define SDMA_PKT_VM_INVALIDATION_HEADER_op_shift 0 +#define SDMA_PKT_VM_INVALIDATION_HEADER_OP(x) (((x) & SDMA_PKT_VM_INVALIDATION_HEADER_op_mask) << SDMA_PKT_VM_INVALIDATION_HEADER_op_shift) + +/*define for sub_op field*/ +#define SDMA_PKT_VM_INVALIDATION_HEADER_sub_op_offset 0 +#define SDMA_PKT_VM_INVALIDATION_HEADER_sub_op_mask 0x000000FF +#define SDMA_PKT_VM_INVALIDATION_HEADER_sub_op_shift 8 +#define SDMA_PKT_VM_INVALIDATION_HEADER_SUB_OP(x) (((x) & SDMA_PKT_VM_INVALIDATION_HEADER_sub_op_mask) << SDMA_PKT_VM_INVALIDATION_HEADER_sub_op_shift) + +/*define for gfx_eng_id field*/ +#define SDMA_PKT_VM_INVALIDATION_HEADER_gfx_eng_id_offset 0 +#define SDMA_PKT_VM_INVALIDATION_HEADER_gfx_eng_id_mask 0x0000001F +#define SDMA_PKT_VM_INVALIDATION_HEADER_gfx_eng_id_shift 16 +#define SDMA_PKT_VM_INVALIDATION_HEADER_GFX_ENG_ID(x) (((x) & SDMA_PKT_VM_INVALIDATION_HEADER_gfx_eng_id_mask) << SDMA_PKT_VM_INVALIDATION_HEADER_gfx_eng_id_shift) + +/*define for mm_eng_id field*/ +#define SDMA_PKT_VM_INVALIDATION_HEADER_mm_eng_id_offset 0 +#define SDMA_PKT_VM_INVALIDATION_HEADER_mm_eng_id_mask 0x0000001F +#define SDMA_PKT_VM_INVALIDATION_HEADER_mm_eng_id_shift 24 +#define SDMA_PKT_VM_INVALIDATION_HEADER_MM_ENG_ID(x) (((x) & SDMA_PKT_VM_INVALIDATION_HEADER_mm_eng_id_mask) << SDMA_PKT_VM_INVALIDATION_HEADER_mm_eng_id_shift) + +/*define for INVALIDATEREQ word*/ +/*define for invalidatereq field*/ +#define SDMA_PKT_VM_INVALIDATION_INVALIDATEREQ_invalidatereq_offset 1 +#define SDMA_PKT_VM_INVALIDATION_INVALIDATEREQ_invalidatereq_mask 0xFFFFFFFF +#define SDMA_PKT_VM_INVALIDATION_INVALIDATEREQ_invalidatereq_shift 0 +#define SDMA_PKT_VM_INVALIDATION_INVALIDATEREQ_INVALIDATEREQ(x) (((x) & SDMA_PKT_VM_INVALIDATION_INVALIDATEREQ_invalidatereq_mask) << SDMA_PKT_VM_INVALIDATION_INVALIDATEREQ_invalidatereq_shift) + +/*define for ADDRESSRANGELO word*/ +/*define for addressrangelo field*/ +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGELO_addressrangelo_offset 2 +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGELO_addressrangelo_mask 0xFFFFFFFF +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGELO_addressrangelo_shift 0 +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGELO_ADDRESSRANGELO(x) (((x) & SDMA_PKT_VM_INVALIDATION_ADDRESSRANGELO_addressrangelo_mask) << SDMA_PKT_VM_INVALIDATION_ADDRESSRANGELO_addressrangelo_shift) + +/*define for ADDRESSRANGEHI word*/ +/*define for invalidateack field*/ +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_invalidateack_offset 3 +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_invalidateack_mask 0x0000FFFF +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_invalidateack_shift 0 +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_INVALIDATEACK(x) (((x) & SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_invalidateack_mask) << SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_invalidateack_shift) + +/*define for addressrangehi field*/ +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_addressrangehi_offset 3 +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_addressrangehi_mask 0x0000001F +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_addressrangehi_shift 16 +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_ADDRESSRANGEHI(x) (((x) & SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_addressrangehi_mask) << SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_addressrangehi_shift) + +/*define for reserved field*/ +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_reserved_offset 3 +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_reserved_mask 0x000001FF +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_reserved_shift 23 +#define SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_RESERVED(x) (((x) & SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_reserved_mask) << SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_reserved_shift) + + /* ** Definitions for SDMA_PKT_ATOMIC packet */ diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c index 8d80df94bd8b..a26a9be58eac 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c @@ -414,8 +414,7 @@ static void nbio_v7_4_handle_ras_controller_intr_no_bifring(struct amdgpu_device /* ras_controller_int is dedicated for nbif ras error, * not the global interrupt for sync flood */ - amdgpu_ras_set_fed(adev, true); - amdgpu_ras_reset_gpu(adev); + amdgpu_ras_global_ras_isr(adev); } amdgpu_ras_error_data_fini(&err_data); diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_9.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_9.c index d1bd79bbae53..8a0a63ac88d2 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_9.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_9.c @@ -401,6 +401,17 @@ static int nbio_v7_9_get_compute_partition_mode(struct amdgpu_device *adev) return px; } +static bool nbio_v7_9_is_nps_switch_requested(struct amdgpu_device *adev) +{ + u32 tmp; + + tmp = RREG32_SOC15(NBIO, 0, regBIF_BX_PF0_PARTITION_MEM_STATUS); + tmp = REG_GET_FIELD(tmp, BIF_BX_PF0_PARTITION_MEM_STATUS, + CHANGE_STATUE); + + /* 0x8 - NPS switch requested */ + return (tmp == 0x8); +} static u32 nbio_v7_9_get_memory_partition_mode(struct amdgpu_device *adev, u32 *supp_modes) { @@ -508,6 +519,7 @@ const struct amdgpu_nbio_funcs nbio_v7_9_funcs = { .remap_hdp_registers = nbio_v7_9_remap_hdp_registers, .get_compute_partition_mode = nbio_v7_9_get_compute_partition_mode, .get_memory_partition_mode = nbio_v7_9_get_memory_partition_mode, + .is_nps_switch_requested = nbio_v7_9_is_nps_switch_requested, .init_registers = nbio_v7_9_init_registers, .get_pcie_replay_count = nbio_v7_9_get_pcie_replay_count, .set_reg_remap = nbio_v7_9_set_reg_remap, diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c index 73065a85e0d2..3bad565ded73 100644 --- a/drivers/gpu/drm/amd/amdgpu/nv.c +++ b/drivers/gpu/drm/amd/amdgpu/nv.c @@ -634,9 +634,9 @@ static const struct amdgpu_asic_funcs nv_asic_funcs = { .query_video_codecs = &nv_query_video_codecs, }; -static int nv_common_early_init(void *handle) +static int nv_common_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->nbio.funcs->set_reg_remap(adev); adev->smc_rreg = NULL; @@ -944,9 +944,9 @@ static int nv_common_early_init(void *handle) return 0; } -static int nv_common_late_init(void *handle) +static int nv_common_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) { xgpu_nv_mailbox_get_irq(adev); @@ -973,9 +973,9 @@ static int nv_common_late_init(void *handle) return 0; } -static int nv_common_sw_init(void *handle) +static int nv_common_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) xgpu_nv_mailbox_add_irq_id(adev); @@ -983,14 +983,9 @@ static int nv_common_sw_init(void *handle) return 0; } -static int nv_common_sw_fini(void *handle) +static int nv_common_hw_init(struct amdgpu_ip_block *ip_block) { - return 0; -} - -static int nv_common_hw_init(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->nbio.funcs->apply_lc_spc_mode_wa) adev->nbio.funcs->apply_lc_spc_mode_wa(adev); @@ -1014,9 +1009,9 @@ static int nv_common_hw_init(void *handle) return 0; } -static int nv_common_hw_fini(void *handle) +static int nv_common_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* Disable the doorbell aperture and selfring doorbell aperture * separately in hw_fini because nv_enable_doorbell_aperture @@ -1029,18 +1024,14 @@ static int nv_common_hw_fini(void *handle) return 0; } -static int nv_common_suspend(void *handle) +static int nv_common_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return nv_common_hw_fini(adev); + return nv_common_hw_fini(ip_block); } -static int nv_common_resume(void *handle) +static int nv_common_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return nv_common_hw_init(adev); + return nv_common_hw_init(ip_block); } static bool nv_common_is_idle(void *handle) @@ -1048,16 +1039,6 @@ static bool nv_common_is_idle(void *handle) return true; } -static int nv_common_wait_for_idle(void *handle) -{ - return 0; -} - -static int nv_common_soft_reset(void *handle) -{ - return 0; -} - static int nv_common_set_clockgating_state(void *handle, enum amd_clockgating_state state) { @@ -1115,17 +1096,12 @@ static const struct amd_ip_funcs nv_common_ip_funcs = { .early_init = nv_common_early_init, .late_init = nv_common_late_init, .sw_init = nv_common_sw_init, - .sw_fini = nv_common_sw_fini, .hw_init = nv_common_hw_init, .hw_fini = nv_common_hw_fini, .suspend = nv_common_suspend, .resume = nv_common_resume, .is_idle = nv_common_is_idle, - .wait_for_idle = nv_common_wait_for_idle, - .soft_reset = nv_common_soft_reset, .set_clockgating_state = nv_common_set_clockgating_state, .set_powergating_state = nv_common_set_powergating_state, .get_clockgating_state = nv_common_get_clockgating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; diff --git a/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h b/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h index 37b5ddd6f13b..f4a91b126c73 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h +++ b/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h @@ -103,6 +103,10 @@ enum psp_gfx_cmd_id GFX_CMD_ID_AUTOLOAD_RLC = 0x00000021, /* Indicates all graphics fw loaded, start RLC autoload */ GFX_CMD_ID_BOOT_CFG = 0x00000022, /* Boot Config */ GFX_CMD_ID_SRIOV_SPATIAL_PART = 0x00000027, /* Configure spatial partitioning mode */ + /*IDs of performance monitoring/profiling*/ + GFX_CMD_ID_CONFIG_SQ_PERFMON = 0x00000046, /* Config CGTT_SQ_CLK_CTRL */ + /* Dynamic memory partitioninig (NPS mode change)*/ + GFX_CMD_ID_FB_NPS_MODE = 0x00000048, /* Configure memory partitioning mode */ }; /* PSP boot config sub-commands */ @@ -351,6 +355,20 @@ struct psp_gfx_cmd_sriov_spatial_part { uint32_t override_this_aid; }; +/*Structure for sq performance monitoring/profiling enable/disable*/ +struct psp_gfx_cmd_config_sq_perfmon { + uint32_t gfx_xcp_mask; + uint8_t core_override; + uint8_t reg_override; + uint8_t perfmon_override; + uint8_t reserved[5]; +}; + +struct psp_gfx_cmd_fb_memory_part { + uint32_t mode; /* requested NPS mode */ + uint32_t resvd; +}; + /* All GFX ring buffer commands. */ union psp_gfx_commands { @@ -365,6 +383,8 @@ union psp_gfx_commands struct psp_gfx_cmd_load_toc cmd_load_toc; struct psp_gfx_cmd_boot_cfg boot_cfg; struct psp_gfx_cmd_sriov_spatial_part cmd_spatial_part; + struct psp_gfx_cmd_config_sq_perfmon config_sq_perfmon; + struct psp_gfx_cmd_fb_memory_part cmd_memory_part; }; struct psp_gfx_uresp_reserved diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v13_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v13_0.c index 51e470e8d67d..c4b775aaee9f 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v13_0.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v13_0.c @@ -823,6 +823,30 @@ static bool psp_v13_0_is_aux_sos_load_required(struct psp_context *psp) return (pmfw_ver < 0x557300); } +static bool psp_v13_0_is_reload_needed(struct psp_context *psp) +{ + uint32_t ucode_ver; + + if (!psp_v13_0_is_sos_alive(psp)) + return false; + + /* Restrict reload support only to specific IP versions */ + switch (amdgpu_ip_version(psp->adev, MP0_HWIP, 0)) { + case IP_VERSION(13, 0, 2): + case IP_VERSION(13, 0, 6): + case IP_VERSION(13, 0, 14): + /* TOS version read from microcode header */ + ucode_ver = psp->sos.fw_version; + /* Read TOS version from hardware */ + psp_v13_0_init_sos_version(psp); + return (ucode_ver != psp->sos.fw_version); + default: + return false; + } + + return false; +} + static const struct psp_funcs psp_v13_0_funcs = { .init_microcode = psp_v13_0_init_microcode, .wait_for_bootloader = psp_v13_0_wait_for_bootloader_steady_state, @@ -847,6 +871,7 @@ static const struct psp_funcs psp_v13_0_funcs = { .fatal_error_recovery_quirk = psp_v13_0_fatal_error_recovery_quirk, .get_ras_capability = psp_v13_0_get_ras_capability, .is_aux_sos_load_required = psp_v13_0_is_aux_sos_load_required, + .is_reload_needed = psp_v13_0_is_reload_needed, }; void psp_v13_0_set_psp_funcs(struct psp_context *psp) diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c index 725392522267..7948d74f8722 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c @@ -807,9 +807,9 @@ static void sdma_v2_4_ring_emit_wreg(struct amdgpu_ring *ring, amdgpu_ring_write(ring, val); } -static int sdma_v2_4_early_init(void *handle) +static int sdma_v2_4_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; adev->sdma.num_instances = SDMA_MAX_INSTANCE; @@ -826,11 +826,11 @@ static int sdma_v2_4_early_init(void *handle) return 0; } -static int sdma_v2_4_sw_init(void *handle) +static int sdma_v2_4_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* SDMA trap event */ r = amdgpu_irq_add_id(adev, AMDGPU_IRQ_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_SDMA_TRAP, @@ -866,9 +866,9 @@ static int sdma_v2_4_sw_init(void *handle) return r; } -static int sdma_v2_4_sw_fini(void *handle) +static int sdma_v2_4_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; for (i = 0; i < adev->sdma.num_instances; i++) @@ -878,10 +878,10 @@ static int sdma_v2_4_sw_fini(void *handle) return 0; } -static int sdma_v2_4_hw_init(void *handle) +static int sdma_v2_4_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; sdma_v2_4_init_golden_registers(adev); @@ -892,27 +892,21 @@ static int sdma_v2_4_hw_init(void *handle) return r; } -static int sdma_v2_4_hw_fini(void *handle) +static int sdma_v2_4_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - sdma_v2_4_enable(adev, false); + sdma_v2_4_enable(ip_block->adev, false); return 0; } -static int sdma_v2_4_suspend(void *handle) +static int sdma_v2_4_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return sdma_v2_4_hw_fini(adev); + return sdma_v2_4_hw_fini(ip_block); } -static int sdma_v2_4_resume(void *handle) +static int sdma_v2_4_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return sdma_v2_4_hw_init(adev); + return sdma_v2_4_hw_init(ip_block); } static bool sdma_v2_4_is_idle(void *handle) @@ -927,11 +921,11 @@ static bool sdma_v2_4_is_idle(void *handle) return true; } -static int sdma_v2_4_wait_for_idle(void *handle) +static int sdma_v2_4_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { tmp = RREG32(mmSRBM_STATUS2) & (SRBM_STATUS2__SDMA_BUSY_MASK | @@ -944,10 +938,10 @@ static int sdma_v2_4_wait_for_idle(void *handle) return -ETIMEDOUT; } -static int sdma_v2_4_soft_reset(void *handle) +static int sdma_v2_4_soft_reset(struct amdgpu_ip_block *ip_block) { u32 srbm_soft_reset = 0; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 tmp = RREG32(mmSRBM_STATUS2); if (tmp & SRBM_STATUS2__SDMA_BUSY_MASK) { @@ -1102,7 +1096,6 @@ static int sdma_v2_4_set_powergating_state(void *handle, static const struct amd_ip_funcs sdma_v2_4_ip_funcs = { .name = "sdma_v2_4", .early_init = sdma_v2_4_early_init, - .late_init = NULL, .sw_init = sdma_v2_4_sw_init, .sw_fini = sdma_v2_4_sw_fini, .hw_init = sdma_v2_4_hw_init, @@ -1114,8 +1107,6 @@ static const struct amd_ip_funcs sdma_v2_4_ip_funcs = { .soft_reset = sdma_v2_4_soft_reset, .set_clockgating_state = sdma_v2_4_set_clockgating_state, .set_powergating_state = sdma_v2_4_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs sdma_v2_4_ring_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c index e65194fe94af..9a3d729545a7 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c @@ -1080,9 +1080,9 @@ static void sdma_v3_0_ring_emit_wreg(struct amdgpu_ring *ring, amdgpu_ring_write(ring, val); } -static int sdma_v3_0_early_init(void *handle) +static int sdma_v3_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; switch (adev->asic_type) { @@ -1106,11 +1106,11 @@ static int sdma_v3_0_early_init(void *handle) return 0; } -static int sdma_v3_0_sw_init(void *handle) +static int sdma_v3_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* SDMA trap event */ r = amdgpu_irq_add_id(adev, AMDGPU_IRQ_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_SDMA_TRAP, @@ -1152,9 +1152,9 @@ static int sdma_v3_0_sw_init(void *handle) return r; } -static int sdma_v3_0_sw_fini(void *handle) +static int sdma_v3_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; for (i = 0; i < adev->sdma.num_instances; i++) @@ -1164,10 +1164,10 @@ static int sdma_v3_0_sw_fini(void *handle) return 0; } -static int sdma_v3_0_hw_init(void *handle) +static int sdma_v3_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; sdma_v3_0_init_golden_registers(adev); @@ -1178,9 +1178,9 @@ static int sdma_v3_0_hw_init(void *handle) return r; } -static int sdma_v3_0_hw_fini(void *handle) +static int sdma_v3_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; sdma_v3_0_ctx_switch_enable(adev, false); sdma_v3_0_enable(adev, false); @@ -1188,18 +1188,14 @@ static int sdma_v3_0_hw_fini(void *handle) return 0; } -static int sdma_v3_0_suspend(void *handle) +static int sdma_v3_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return sdma_v3_0_hw_fini(adev); + return sdma_v3_0_hw_fini(ip_block); } -static int sdma_v3_0_resume(void *handle) +static int sdma_v3_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return sdma_v3_0_hw_init(adev); + return sdma_v3_0_hw_init(ip_block); } static bool sdma_v3_0_is_idle(void *handle) @@ -1214,11 +1210,11 @@ static bool sdma_v3_0_is_idle(void *handle) return true; } -static int sdma_v3_0_wait_for_idle(void *handle) +static int sdma_v3_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { tmp = RREG32(mmSRBM_STATUS2) & (SRBM_STATUS2__SDMA_BUSY_MASK | @@ -1231,9 +1227,9 @@ static int sdma_v3_0_wait_for_idle(void *handle) return -ETIMEDOUT; } -static bool sdma_v3_0_check_soft_reset(void *handle) +static bool sdma_v3_0_check_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 srbm_soft_reset = 0; u32 tmp = RREG32(mmSRBM_STATUS2); @@ -1252,9 +1248,9 @@ static bool sdma_v3_0_check_soft_reset(void *handle) } } -static int sdma_v3_0_pre_soft_reset(void *handle) +static int sdma_v3_0_pre_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 srbm_soft_reset = 0; if (!adev->sdma.srbm_soft_reset) @@ -1271,9 +1267,9 @@ static int sdma_v3_0_pre_soft_reset(void *handle) return 0; } -static int sdma_v3_0_post_soft_reset(void *handle) +static int sdma_v3_0_post_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 srbm_soft_reset = 0; if (!adev->sdma.srbm_soft_reset) @@ -1290,9 +1286,9 @@ static int sdma_v3_0_post_soft_reset(void *handle) return 0; } -static int sdma_v3_0_soft_reset(void *handle) +static int sdma_v3_0_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 srbm_soft_reset = 0; u32 tmp; @@ -1538,7 +1534,6 @@ static void sdma_v3_0_get_clockgating_state(void *handle, u64 *flags) static const struct amd_ip_funcs sdma_v3_0_ip_funcs = { .name = "sdma_v3_0", .early_init = sdma_v3_0_early_init, - .late_init = NULL, .sw_init = sdma_v3_0_sw_init, .sw_fini = sdma_v3_0_sw_fini, .hw_init = sdma_v3_0_hw_init, @@ -1554,8 +1549,6 @@ static const struct amd_ip_funcs sdma_v3_0_ip_funcs = { .set_clockgating_state = sdma_v3_0_set_clockgating_state, .set_powergating_state = sdma_v3_0_set_powergating_state, .get_clockgating_state = sdma_v3_0_get_clockgating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs sdma_v3_0_ring_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c index 23ef4eb36b40..c1f98f6cf20d 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c @@ -1751,9 +1751,9 @@ static bool sdma_v4_0_fw_support_paging_queue(struct amdgpu_device *adev) } } -static int sdma_v4_0_early_init(void *handle) +static int sdma_v4_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = sdma_v4_0_init_microcode(adev); @@ -1780,9 +1780,9 @@ static int sdma_v4_0_process_ras_data_cb(struct amdgpu_device *adev, void *err_data, struct amdgpu_iv_entry *entry); -static int sdma_v4_0_late_init(void *handle) +static int sdma_v4_0_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; sdma_v4_0_setup_ulv(adev); @@ -1792,11 +1792,11 @@ static int sdma_v4_0_late_init(void *handle) return 0; } -static int sdma_v4_0_sw_init(void *handle) +static int sdma_v4_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_4_0); uint32_t *ptr; @@ -1929,9 +1929,9 @@ static int sdma_v4_0_sw_init(void *handle) return r; } -static int sdma_v4_0_sw_fini(void *handle) +static int sdma_v4_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; for (i = 0; i < adev->sdma.num_instances; i++) { @@ -1951,9 +1951,9 @@ static int sdma_v4_0_sw_fini(void *handle) return 0; } -static int sdma_v4_0_hw_init(void *handle) +static int sdma_v4_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->flags & AMD_IS_APU) amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_SDMA, false); @@ -1964,9 +1964,9 @@ static int sdma_v4_0_hw_init(void *handle) return sdma_v4_0_start(adev); } -static int sdma_v4_0_hw_fini(void *handle) +static int sdma_v4_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; if (amdgpu_sriov_vf(adev)) @@ -1988,9 +1988,9 @@ static int sdma_v4_0_hw_fini(void *handle) return 0; } -static int sdma_v4_0_suspend(void *handle) +static int sdma_v4_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* SMU saves SDMA state for us */ if (adev->in_s0ix) { @@ -1998,12 +1998,12 @@ static int sdma_v4_0_suspend(void *handle) return 0; } - return sdma_v4_0_hw_fini(adev); + return sdma_v4_0_hw_fini(ip_block); } -static int sdma_v4_0_resume(void *handle) +static int sdma_v4_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* SMU restores SDMA state for us */ if (adev->in_s0ix) { @@ -2012,7 +2012,7 @@ static int sdma_v4_0_resume(void *handle) return 0; } - return sdma_v4_0_hw_init(adev); + return sdma_v4_0_hw_init(ip_block); } static bool sdma_v4_0_is_idle(void *handle) @@ -2030,11 +2030,11 @@ static bool sdma_v4_0_is_idle(void *handle) return true; } -static int sdma_v4_0_wait_for_idle(void *handle) +static int sdma_v4_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i, j; u32 sdma[AMDGPU_MAX_SDMA_INSTANCES]; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { for (j = 0; j < adev->sdma.num_instances; j++) { @@ -2049,7 +2049,7 @@ static int sdma_v4_0_wait_for_idle(void *handle) return -ETIMEDOUT; } -static int sdma_v4_0_soft_reset(void *handle) +static int sdma_v4_0_soft_reset(struct amdgpu_ip_block *ip_block) { /* todo */ @@ -2350,9 +2350,9 @@ static void sdma_v4_0_get_clockgating_state(void *handle, u64 *flags) *flags |= AMD_CG_SUPPORT_SDMA_LS; } -static void sdma_v4_0_print_ip_state(void *handle, struct drm_printer *p) +static void sdma_v4_0_print_ip_state(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_4_0); uint32_t instance_offset; @@ -2371,9 +2371,9 @@ static void sdma_v4_0_print_ip_state(void *handle, struct drm_printer *p) } } -static void sdma_v4_0_dump_ip_state(void *handle) +static void sdma_v4_0_dump_ip_state(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t instance_offset; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_4_0); diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c index c77889040760..a38553f38fdc 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c @@ -1290,9 +1290,9 @@ static bool sdma_v4_4_2_fw_support_paging_queue(struct amdgpu_device *adev) } } -static int sdma_v4_4_2_early_init(void *handle) +static int sdma_v4_4_2_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = sdma_v4_4_2_init_microcode(adev); @@ -1318,9 +1318,9 @@ static int sdma_v4_4_2_process_ras_data_cb(struct amdgpu_device *adev, struct amdgpu_iv_entry *entry); #endif -static int sdma_v4_4_2_late_init(void *handle) +static int sdma_v4_4_2_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; #if 0 struct ras_ih_if ih_info = { .cb = sdma_v4_4_2_process_ras_data_cb, @@ -1332,11 +1332,11 @@ static int sdma_v4_4_2_late_init(void *handle) return 0; } -static int sdma_v4_4_2_sw_init(void *handle) +static int sdma_v4_4_2_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 aid_id; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_4_4_2); uint32_t *ptr; @@ -1430,6 +1430,10 @@ static int sdma_v4_4_2_sw_init(void *handle) } } + /* TODO: Add queue reset mask when FW fully supports it */ + adev->sdma.supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->sdma.instance[0].ring); + if (amdgpu_sdma_ras_sw_init(adev)) { dev_err(adev->dev, "fail to initialize sdma ras block\n"); return -EINVAL; @@ -1442,12 +1446,16 @@ static int sdma_v4_4_2_sw_init(void *handle) else DRM_ERROR("Failed to allocated memory for SDMA IP Dump\n"); + r = amdgpu_sdma_sysfs_reset_mask_init(adev); + if (r) + return r; + return r; } -static int sdma_v4_4_2_sw_fini(void *handle) +static int sdma_v4_4_2_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; for (i = 0; i < adev->sdma.num_instances; i++) { @@ -1456,6 +1464,7 @@ static int sdma_v4_4_2_sw_fini(void *handle) amdgpu_ring_fini(&adev->sdma.instance[i].page); } + amdgpu_sdma_sysfs_reset_mask_fini(adev); if (amdgpu_ip_version(adev, SDMA0_HWIP, 0) == IP_VERSION(4, 4, 2) || amdgpu_ip_version(adev, SDMA0_HWIP, 0) == IP_VERSION(4, 4, 5)) amdgpu_sdma_destroy_inst_ctx(adev, true); @@ -1467,10 +1476,10 @@ static int sdma_v4_4_2_sw_fini(void *handle) return 0; } -static int sdma_v4_4_2_hw_init(void *handle) +static int sdma_v4_4_2_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t inst_mask; inst_mask = GENMASK(adev->sdma.num_instances - 1, 0); @@ -1482,9 +1491,9 @@ static int sdma_v4_4_2_hw_init(void *handle) return r; } -static int sdma_v4_4_2_hw_fini(void *handle) +static int sdma_v4_4_2_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t inst_mask; int i; @@ -1508,21 +1517,19 @@ static int sdma_v4_4_2_hw_fini(void *handle) static int sdma_v4_4_2_set_clockgating_state(void *handle, enum amd_clockgating_state state); -static int sdma_v4_4_2_suspend(void *handle) +static int sdma_v4_4_2_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_in_reset(adev)) sdma_v4_4_2_set_clockgating_state(adev, AMD_CG_STATE_UNGATE); - return sdma_v4_4_2_hw_fini(adev); + return sdma_v4_4_2_hw_fini(ip_block); } -static int sdma_v4_4_2_resume(void *handle) +static int sdma_v4_4_2_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return sdma_v4_4_2_hw_init(adev); + return sdma_v4_4_2_hw_init(ip_block); } static bool sdma_v4_4_2_is_idle(void *handle) @@ -1540,11 +1547,11 @@ static bool sdma_v4_4_2_is_idle(void *handle) return true; } -static int sdma_v4_4_2_wait_for_idle(void *handle) +static int sdma_v4_4_2_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i, j; u32 sdma[AMDGPU_MAX_SDMA_INSTANCES]; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { for (j = 0; j < adev->sdma.num_instances; j++) { @@ -1559,7 +1566,7 @@ static int sdma_v4_4_2_wait_for_idle(void *handle) return -ETIMEDOUT; } -static int sdma_v4_4_2_soft_reset(void *handle) +static int sdma_v4_4_2_soft_reset(struct amdgpu_ip_block *ip_block) { /* todo */ @@ -1857,9 +1864,9 @@ static void sdma_v4_4_2_get_clockgating_state(void *handle, u64 *flags) *flags |= AMD_CG_SUPPORT_SDMA_LS; } -static void sdma_v4_4_2_print_ip_state(void *handle, struct drm_printer *p) +static void sdma_v4_4_2_print_ip_state(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_4_4_2); uint32_t instance_offset; @@ -1878,9 +1885,9 @@ static void sdma_v4_4_2_print_ip_state(void *handle, struct drm_printer *p) } } -static void sdma_v4_4_2_dump_ip_state(void *handle) +static void sdma_v4_4_2_dump_ip_state(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t instance_offset; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_4_4_2); diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c index 3e48ea38385d..fa9b40934957 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c @@ -705,14 +705,16 @@ static void sdma_v5_0_enable(struct amdgpu_device *adev, bool enable) } /** - * sdma_v5_0_gfx_resume - setup and start the async dma engines + * sdma_v5_0_gfx_resume_instance - start/restart a certain sdma engine * * @adev: amdgpu_device pointer + * @i: instance + * @restore: used to restore wptr when restart * - * Set up the gfx DMA ring buffers and enable them (NAVI10). - * Returns 0 for success, error for failure. + * Set up the gfx DMA ring buffers and enable them. On restart, we will restore wptr and rptr. + * Return 0 for success. */ -static int sdma_v5_0_gfx_resume(struct amdgpu_device *adev) +static int sdma_v5_0_gfx_resume_instance(struct amdgpu_device *adev, int i, bool restore) { struct amdgpu_ring *ring; u32 rb_cntl, ib_cntl; @@ -722,142 +724,163 @@ static int sdma_v5_0_gfx_resume(struct amdgpu_device *adev) u32 temp; u32 wptr_poll_cntl; u64 wptr_gpu_addr; - int i, r; - for (i = 0; i < adev->sdma.num_instances; i++) { - ring = &adev->sdma.instance[i].ring; + ring = &adev->sdma.instance[i].ring; - if (!amdgpu_sriov_vf(adev)) - WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_SEM_WAIT_FAIL_TIMER_CNTL), 0); + if (!amdgpu_sriov_vf(adev)) + WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_SEM_WAIT_FAIL_TIMER_CNTL), 0); - /* Set ring buffer size in dwords */ - rb_bufsz = order_base_2(ring->ring_size / 4); - rb_cntl = RREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL)); - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_SIZE, rb_bufsz); + /* Set ring buffer size in dwords */ + rb_bufsz = order_base_2(ring->ring_size / 4); + rb_cntl = RREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL)); + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_SIZE, rb_bufsz); #ifdef __BIG_ENDIAN - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_SWAP_ENABLE, 1); - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, - RPTR_WRITEBACK_SWAP_ENABLE, 1); + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_SWAP_ENABLE, 1); + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, + RPTR_WRITEBACK_SWAP_ENABLE, 1); #endif - WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); - /* Initialize the ring buffer's read and write pointers */ + /* Initialize the ring buffer's read and write pointers */ + if (restore) { + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR), lower_32_bits(ring->wptr << 2)); + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_HI), upper_32_bits(ring->wptr << 2)); + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR), lower_32_bits(ring->wptr << 2)); + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_HI), upper_32_bits(ring->wptr << 2)); + } else { WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR), 0); WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_HI), 0); WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR), 0); WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_HI), 0); + } + /* setup the wptr shadow polling */ + wptr_gpu_addr = ring->wptr_gpu_addr; + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_LO), + lower_32_bits(wptr_gpu_addr)); + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_HI), + upper_32_bits(wptr_gpu_addr)); + wptr_poll_cntl = RREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, + mmSDMA0_GFX_RB_WPTR_POLL_CNTL)); + wptr_poll_cntl = REG_SET_FIELD(wptr_poll_cntl, + SDMA0_GFX_RB_WPTR_POLL_CNTL, + F32_POLL_ENABLE, 1); + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_CNTL), + wptr_poll_cntl); - /* setup the wptr shadow polling */ - wptr_gpu_addr = ring->wptr_gpu_addr; - WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_LO), - lower_32_bits(wptr_gpu_addr)); - WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_HI), - upper_32_bits(wptr_gpu_addr)); - wptr_poll_cntl = RREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, - mmSDMA0_GFX_RB_WPTR_POLL_CNTL)); - wptr_poll_cntl = REG_SET_FIELD(wptr_poll_cntl, - SDMA0_GFX_RB_WPTR_POLL_CNTL, - F32_POLL_ENABLE, 1); - WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_CNTL), - wptr_poll_cntl); + /* set the wb address whether it's enabled or not */ + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_ADDR_HI), + upper_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFF); + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_ADDR_LO), + lower_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFC); - /* set the wb address whether it's enabled or not */ - WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_ADDR_HI), - upper_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFF); - WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_ADDR_LO), - lower_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFC); + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RPTR_WRITEBACK_ENABLE, 1); - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RPTR_WRITEBACK_ENABLE, 1); - - WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_BASE), - ring->gpu_addr >> 8); - WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_BASE_HI), - ring->gpu_addr >> 40); + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_BASE), + ring->gpu_addr >> 8); + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_BASE_HI), + ring->gpu_addr >> 40); + if (!restore) ring->wptr = 0; - /* before programing wptr to a less value, need set minor_ptr_update first */ - WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 1); + /* before programing wptr to a less value, need set minor_ptr_update first */ + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 1); - if (!amdgpu_sriov_vf(adev)) { /* only bare-metal use register write for wptr */ - WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR), - lower_32_bits(ring->wptr << 2)); - WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_HI), - upper_32_bits(ring->wptr << 2)); - } + if (!amdgpu_sriov_vf(adev)) { /* only bare-metal use register write for wptr */ + WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR), + lower_32_bits(ring->wptr << 2)); + WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_HI), + upper_32_bits(ring->wptr << 2)); + } - doorbell = RREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL)); - doorbell_offset = RREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, - mmSDMA0_GFX_DOORBELL_OFFSET)); + doorbell = RREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL)); + doorbell_offset = RREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, + mmSDMA0_GFX_DOORBELL_OFFSET)); - if (ring->use_doorbell) { - doorbell = REG_SET_FIELD(doorbell, SDMA0_GFX_DOORBELL, ENABLE, 1); - doorbell_offset = REG_SET_FIELD(doorbell_offset, SDMA0_GFX_DOORBELL_OFFSET, - OFFSET, ring->doorbell_index); - } else { - doorbell = REG_SET_FIELD(doorbell, SDMA0_GFX_DOORBELL, ENABLE, 0); - } - WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL), doorbell); - WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL_OFFSET), - doorbell_offset); + if (ring->use_doorbell) { + doorbell = REG_SET_FIELD(doorbell, SDMA0_GFX_DOORBELL, ENABLE, 1); + doorbell_offset = REG_SET_FIELD(doorbell_offset, SDMA0_GFX_DOORBELL_OFFSET, + OFFSET, ring->doorbell_index); + } else { + doorbell = REG_SET_FIELD(doorbell, SDMA0_GFX_DOORBELL, ENABLE, 0); + } + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL), doorbell); + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL_OFFSET), + doorbell_offset); - adev->nbio.funcs->sdma_doorbell_range(adev, i, ring->use_doorbell, - ring->doorbell_index, 20); + adev->nbio.funcs->sdma_doorbell_range(adev, i, ring->use_doorbell, + ring->doorbell_index, 20); - if (amdgpu_sriov_vf(adev)) - sdma_v5_0_ring_set_wptr(ring); + if (amdgpu_sriov_vf(adev)) + sdma_v5_0_ring_set_wptr(ring); - /* set minor_ptr_update to 0 after wptr programed */ - WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 0); + /* set minor_ptr_update to 0 after wptr programed */ + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 0); - if (!amdgpu_sriov_vf(adev)) { - /* set utc l1 enable flag always to 1 */ - temp = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_CNTL)); - temp = REG_SET_FIELD(temp, SDMA0_CNTL, UTC_L1_ENABLE, 1); + if (!amdgpu_sriov_vf(adev)) { + /* set utc l1 enable flag always to 1 */ + temp = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_CNTL)); + temp = REG_SET_FIELD(temp, SDMA0_CNTL, UTC_L1_ENABLE, 1); - /* enable MCBP */ - temp = REG_SET_FIELD(temp, SDMA0_CNTL, MIDCMD_PREEMPT_ENABLE, 1); - WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_CNTL), temp); + /* enable MCBP */ + temp = REG_SET_FIELD(temp, SDMA0_CNTL, MIDCMD_PREEMPT_ENABLE, 1); + WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_CNTL), temp); - /* Set up RESP_MODE to non-copy addresses */ - temp = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_UTCL1_CNTL)); - temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, RESP_MODE, 3); - temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, REDO_DELAY, 9); - WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_UTCL1_CNTL), temp); + /* Set up RESP_MODE to non-copy addresses */ + temp = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_UTCL1_CNTL)); + temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, RESP_MODE, 3); + temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, REDO_DELAY, 9); + WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_UTCL1_CNTL), temp); - /* program default cache read and write policy */ - temp = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_UTCL1_PAGE)); - /* clean read policy and write policy bits */ - temp &= 0xFF0FFF; - temp |= ((CACHE_READ_POLICY_L2__DEFAULT << 12) | (CACHE_WRITE_POLICY_L2__DEFAULT << 14)); - WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_UTCL1_PAGE), temp); - } + /* program default cache read and write policy */ + temp = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_UTCL1_PAGE)); + /* clean read policy and write policy bits */ + temp &= 0xFF0FFF; + temp |= ((CACHE_READ_POLICY_L2__DEFAULT << 12) | (CACHE_WRITE_POLICY_L2__DEFAULT << 14)); + WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_UTCL1_PAGE), temp); + } - if (!amdgpu_sriov_vf(adev)) { - /* unhalt engine */ - temp = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_F32_CNTL)); - temp = REG_SET_FIELD(temp, SDMA0_F32_CNTL, HALT, 0); - WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_F32_CNTL), temp); - } + if (!amdgpu_sriov_vf(adev)) { + /* unhalt engine */ + temp = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_F32_CNTL)); + temp = REG_SET_FIELD(temp, SDMA0_F32_CNTL, HALT, 0); + WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_F32_CNTL), temp); + } - /* enable DMA RB */ - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_ENABLE, 1); - WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); + /* enable DMA RB */ + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_ENABLE, 1); + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); - ib_cntl = RREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL)); - ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_ENABLE, 1); + ib_cntl = RREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL)); + ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_ENABLE, 1); #ifdef __BIG_ENDIAN - ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_SWAP_ENABLE, 1); + ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_SWAP_ENABLE, 1); #endif - /* enable DMA IBs */ - WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL), ib_cntl); + /* enable DMA IBs */ + WREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL), ib_cntl); - if (amdgpu_sriov_vf(adev)) { /* bare-metal sequence doesn't need below to lines */ - sdma_v5_0_ctx_switch_enable(adev, true); - sdma_v5_0_enable(adev, true); - } + if (amdgpu_sriov_vf(adev)) { /* bare-metal sequence doesn't need below to lines */ + sdma_v5_0_ctx_switch_enable(adev, true); + sdma_v5_0_enable(adev, true); + } - r = amdgpu_ring_test_helper(ring); + return amdgpu_ring_test_helper(ring); +} + +/** + * sdma_v5_0_gfx_resume - setup and start the async dma engines + * + * @adev: amdgpu_device pointer + * + * Set up the gfx DMA ring buffers and enable them (NAVI10). + * Returns 0 for success, error for failure. + */ +static int sdma_v5_0_gfx_resume(struct amdgpu_device *adev) +{ + int i, r; + + for (i = 0; i < adev->sdma.num_instances; i++) { + r = sdma_v5_0_gfx_resume_instance(adev, i, false); if (r) return r; } @@ -1366,9 +1389,9 @@ static void sdma_v5_0_ring_emit_reg_write_reg_wait(struct amdgpu_ring *ring, amdgpu_ring_emit_reg_wait(ring, reg1, mask, mask); } -static int sdma_v5_0_early_init(void *handle) +static int sdma_v5_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = sdma_v5_0_init_microcode(adev); @@ -1385,11 +1408,11 @@ static int sdma_v5_0_early_init(void *handle) } -static int sdma_v5_0_sw_init(void *handle) +static int sdma_v5_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_5_0); uint32_t *ptr; @@ -1429,6 +1452,19 @@ static int sdma_v5_0_sw_init(void *handle) return r; } + adev->sdma.supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->sdma.instance[0].ring); + switch (amdgpu_ip_version(adev, SDMA0_HWIP, 0)) { + case IP_VERSION(5, 0, 0): + case IP_VERSION(5, 0, 2): + case IP_VERSION(5, 0, 5): + if (adev->sdma.instance[0].fw_version >= 35) + adev->sdma.supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE; + break; + default: + break; + } + /* Allocate memory for SDMA IP Dump buffer */ ptr = kcalloc(adev->sdma.num_instances * reg_count, sizeof(uint32_t), GFP_KERNEL); if (ptr) @@ -1436,17 +1472,22 @@ static int sdma_v5_0_sw_init(void *handle) else DRM_ERROR("Failed to allocated memory for SDMA IP Dump\n"); + r = amdgpu_sdma_sysfs_reset_mask_init(adev); + if (r) + return r; + return r; } -static int sdma_v5_0_sw_fini(void *handle) +static int sdma_v5_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; for (i = 0; i < adev->sdma.num_instances; i++) amdgpu_ring_fini(&adev->sdma.instance[i].ring); + amdgpu_sdma_sysfs_reset_mask_fini(adev); amdgpu_sdma_destroy_inst_ctx(adev, false); kfree(adev->sdma.ip_dump); @@ -1454,10 +1495,10 @@ static int sdma_v5_0_sw_fini(void *handle) return 0; } -static int sdma_v5_0_hw_init(void *handle) +static int sdma_v5_0_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; sdma_v5_0_init_golden_registers(adev); @@ -1466,9 +1507,9 @@ static int sdma_v5_0_hw_init(void *handle) return r; } -static int sdma_v5_0_hw_fini(void *handle) +static int sdma_v5_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) return 0; @@ -1479,18 +1520,14 @@ static int sdma_v5_0_hw_fini(void *handle) return 0; } -static int sdma_v5_0_suspend(void *handle) +static int sdma_v5_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return sdma_v5_0_hw_fini(adev); + return sdma_v5_0_hw_fini(ip_block); } -static int sdma_v5_0_resume(void *handle) +static int sdma_v5_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return sdma_v5_0_hw_init(adev); + return sdma_v5_0_hw_init(ip_block); } static bool sdma_v5_0_is_idle(void *handle) @@ -1508,11 +1545,11 @@ static bool sdma_v5_0_is_idle(void *handle) return true; } -static int sdma_v5_0_wait_for_idle(void *handle) +static int sdma_v5_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; u32 sdma0, sdma1; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { sdma0 = RREG32(sdma_v5_0_get_reg_offset(adev, 0, mmSDMA0_STATUS_REG)); @@ -1525,13 +1562,100 @@ static int sdma_v5_0_wait_for_idle(void *handle) return -ETIMEDOUT; } -static int sdma_v5_0_soft_reset(void *handle) +static int sdma_v5_0_soft_reset(struct amdgpu_ip_block *ip_block) { /* todo */ return 0; } +static int sdma_v5_0_reset_queue(struct amdgpu_ring *ring, unsigned int vmid) +{ + struct amdgpu_device *adev = ring->adev; + int i, j, r; + u32 rb_cntl, ib_cntl, f32_cntl, freeze, cntl, preempt, soft_reset, stat1_reg; + + if (amdgpu_sriov_vf(adev)) + return -EINVAL; + + for (i = 0; i < adev->sdma.num_instances; i++) { + if (ring == &adev->sdma.instance[i].ring) + break; + } + + if (i == adev->sdma.num_instances) { + DRM_ERROR("sdma instance not found\n"); + return -EINVAL; + } + + amdgpu_gfx_rlc_enter_safe_mode(adev, 0); + + /* stop queue */ + ib_cntl = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL)); + ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_ENABLE, 0); + WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL), ib_cntl); + + rb_cntl = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL)); + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_ENABLE, 0); + WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); + + /* engine stop SDMA1_F32_CNTL.HALT to 1 and SDMAx_FREEZE freeze bit to 1 */ + freeze = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_FREEZE)); + freeze = REG_SET_FIELD(freeze, SDMA0_FREEZE, FREEZE, 1); + WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_FREEZE), freeze); + + for (j = 0; j < adev->usec_timeout; j++) { + freeze = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_FREEZE)); + if (REG_GET_FIELD(freeze, SDMA0_FREEZE, FROZEN) & 1) + break; + udelay(1); + } + + /* check sdma copy engine all idle if frozen not received*/ + if (j == adev->usec_timeout) { + stat1_reg = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_STATUS1_REG)); + if ((stat1_reg & 0x3FF) != 0x3FF) { + DRM_ERROR("cannot soft reset as sdma not idle\n"); + r = -ETIMEDOUT; + goto err0; + } + } + + f32_cntl = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_F32_CNTL)); + f32_cntl = REG_SET_FIELD(f32_cntl, SDMA0_F32_CNTL, HALT, 1); + WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_F32_CNTL), f32_cntl); + + cntl = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_CNTL)); + cntl = REG_SET_FIELD(cntl, SDMA0_CNTL, UTC_L1_ENABLE, 0); + WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_CNTL), cntl); + + /* soft reset SDMA_GFX_PREEMPT.IB_PREEMPT = 0 mmGRBM_SOFT_RESET.SOFT_RESET_SDMA0/1 = 1 */ + preempt = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_PREEMPT)); + preempt = REG_SET_FIELD(preempt, SDMA0_GFX_PREEMPT, IB_PREEMPT, 0); + WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_PREEMPT), preempt); + + soft_reset = RREG32_SOC15(GC, 0, mmGRBM_SOFT_RESET); + soft_reset |= 1 << GRBM_SOFT_RESET__SOFT_RESET_SDMA0__SHIFT << i; + + WREG32_SOC15(GC, 0, mmGRBM_SOFT_RESET, soft_reset); + + udelay(50); + + soft_reset &= ~(1 << GRBM_SOFT_RESET__SOFT_RESET_SDMA0__SHIFT << i); + WREG32_SOC15(GC, 0, mmGRBM_SOFT_RESET, soft_reset); + + /* unfreeze*/ + freeze = RREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_FREEZE)); + freeze = REG_SET_FIELD(freeze, SDMA0_FREEZE, FREEZE, 0); + WREG32(sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_FREEZE), freeze); + + r = sdma_v5_0_gfx_resume_instance(adev, i, true); + +err0: + amdgpu_gfx_rlc_exit_safe_mode(adev, 0); + return r; +} + static int sdma_v5_0_ring_preempt_ib(struct amdgpu_ring *ring) { int i, r = 0; @@ -1778,9 +1902,9 @@ static void sdma_v5_0_get_clockgating_state(void *handle, u64 *flags) *flags |= AMD_CG_SUPPORT_SDMA_LS; } -static void sdma_v5_0_print_ip_state(void *handle, struct drm_printer *p) +static void sdma_v5_0_print_ip_state(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_5_0); uint32_t instance_offset; @@ -1799,9 +1923,9 @@ static void sdma_v5_0_print_ip_state(void *handle, struct drm_printer *p) } } -static void sdma_v5_0_dump_ip_state(void *handle) +static void sdma_v5_0_dump_ip_state(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t instance_offset; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_5_0); @@ -1823,7 +1947,6 @@ static void sdma_v5_0_dump_ip_state(void *handle) static const struct amd_ip_funcs sdma_v5_0_ip_funcs = { .name = "sdma_v5_0", .early_init = sdma_v5_0_early_init, - .late_init = NULL, .sw_init = sdma_v5_0_sw_init, .sw_fini = sdma_v5_0_sw_fini, .hw_init = sdma_v5_0_hw_init, @@ -1874,6 +1997,7 @@ static const struct amdgpu_ring_funcs sdma_v5_0_ring_funcs = { .emit_reg_write_reg_wait = sdma_v5_0_ring_emit_reg_write_reg_wait, .init_cond_exec = sdma_v5_0_ring_init_cond_exec, .preempt_ib = sdma_v5_0_ring_preempt_ib, + .reset = sdma_v5_0_reset_queue, }; static void sdma_v5_0_set_ring_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c index bc9b240a3488..ba5160399ab2 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c @@ -522,14 +522,17 @@ static void sdma_v5_2_enable(struct amdgpu_device *adev, bool enable) } /** - * sdma_v5_2_gfx_resume - setup and start the async dma engines + * sdma_v5_2_gfx_resume_instance - start/restart a certain sdma engine * * @adev: amdgpu_device pointer + * @i: instance + * @restore: used to restore wptr when restart * - * Set up the gfx DMA ring buffers and enable them. - * Returns 0 for success, error for failure. + * Set up the gfx DMA ring buffers and enable them. On restart, we will restore wptr and rptr. + * Return 0 for success. */ -static int sdma_v5_2_gfx_resume(struct amdgpu_device *adev) + +static int sdma_v5_2_gfx_resume_instance(struct amdgpu_device *adev, int i, bool restore) { struct amdgpu_ring *ring; u32 rb_cntl, ib_cntl; @@ -539,139 +542,161 @@ static int sdma_v5_2_gfx_resume(struct amdgpu_device *adev) u32 temp; u32 wptr_poll_cntl; u64 wptr_gpu_addr; - int i, r; - for (i = 0; i < adev->sdma.num_instances; i++) { - ring = &adev->sdma.instance[i].ring; + ring = &adev->sdma.instance[i].ring; - if (!amdgpu_sriov_vf(adev)) - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_SEM_WAIT_FAIL_TIMER_CNTL), 0); + if (!amdgpu_sriov_vf(adev)) + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_SEM_WAIT_FAIL_TIMER_CNTL), 0); - /* Set ring buffer size in dwords */ - rb_bufsz = order_base_2(ring->ring_size / 4); - rb_cntl = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL)); - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_SIZE, rb_bufsz); + /* Set ring buffer size in dwords */ + rb_bufsz = order_base_2(ring->ring_size / 4); + rb_cntl = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL)); + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_SIZE, rb_bufsz); #ifdef __BIG_ENDIAN - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_SWAP_ENABLE, 1); - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, - RPTR_WRITEBACK_SWAP_ENABLE, 1); + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_SWAP_ENABLE, 1); + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, + RPTR_WRITEBACK_SWAP_ENABLE, 1); #endif - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); - /* Initialize the ring buffer's read and write pointers */ + /* Initialize the ring buffer's read and write pointers */ + if (restore) { + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR), lower_32_bits(ring->wptr << 2)); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_HI), upper_32_bits(ring->wptr << 2)); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR), lower_32_bits(ring->wptr << 2)); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_HI), upper_32_bits(ring->wptr << 2)); + } else { WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR), 0); WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_HI), 0); WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR), 0); WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_HI), 0); + } - /* setup the wptr shadow polling */ - wptr_gpu_addr = ring->wptr_gpu_addr; - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_LO), - lower_32_bits(wptr_gpu_addr)); - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_HI), - upper_32_bits(wptr_gpu_addr)); - wptr_poll_cntl = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, - mmSDMA0_GFX_RB_WPTR_POLL_CNTL)); - wptr_poll_cntl = REG_SET_FIELD(wptr_poll_cntl, - SDMA0_GFX_RB_WPTR_POLL_CNTL, - F32_POLL_ENABLE, 1); - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_CNTL), - wptr_poll_cntl); + /* setup the wptr shadow polling */ + wptr_gpu_addr = ring->wptr_gpu_addr; + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_LO), + lower_32_bits(wptr_gpu_addr)); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_HI), + upper_32_bits(wptr_gpu_addr)); + wptr_poll_cntl = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, + mmSDMA0_GFX_RB_WPTR_POLL_CNTL)); + wptr_poll_cntl = REG_SET_FIELD(wptr_poll_cntl, + SDMA0_GFX_RB_WPTR_POLL_CNTL, + F32_POLL_ENABLE, 1); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_CNTL), + wptr_poll_cntl); - /* set the wb address whether it's enabled or not */ - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_ADDR_HI), - upper_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFF); - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_ADDR_LO), - lower_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFC); + /* set the wb address whether it's enabled or not */ + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_ADDR_HI), + upper_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFF); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_ADDR_LO), + lower_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFC); - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RPTR_WRITEBACK_ENABLE, 1); + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RPTR_WRITEBACK_ENABLE, 1); - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_BASE), ring->gpu_addr >> 8); - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_BASE_HI), ring->gpu_addr >> 40); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_BASE), ring->gpu_addr >> 8); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_BASE_HI), ring->gpu_addr >> 40); + if (!restore) ring->wptr = 0; - /* before programing wptr to a less value, need set minor_ptr_update first */ - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 1); + /* before programing wptr to a less value, need set minor_ptr_update first */ + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 1); - if (!amdgpu_sriov_vf(adev)) { /* only bare-metal use register write for wptr */ - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR), lower_32_bits(ring->wptr << 2)); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_HI), upper_32_bits(ring->wptr << 2)); - } + if (!amdgpu_sriov_vf(adev)) { /* only bare-metal use register write for wptr */ + WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR), lower_32_bits(ring->wptr << 2)); + WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_HI), upper_32_bits(ring->wptr << 2)); + } - doorbell = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL)); - doorbell_offset = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL_OFFSET)); + doorbell = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL)); + doorbell_offset = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL_OFFSET)); - if (ring->use_doorbell) { - doorbell = REG_SET_FIELD(doorbell, SDMA0_GFX_DOORBELL, ENABLE, 1); - doorbell_offset = REG_SET_FIELD(doorbell_offset, SDMA0_GFX_DOORBELL_OFFSET, - OFFSET, ring->doorbell_index); - } else { - doorbell = REG_SET_FIELD(doorbell, SDMA0_GFX_DOORBELL, ENABLE, 0); - } - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL), doorbell); - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL_OFFSET), doorbell_offset); + if (ring->use_doorbell) { + doorbell = REG_SET_FIELD(doorbell, SDMA0_GFX_DOORBELL, ENABLE, 1); + doorbell_offset = REG_SET_FIELD(doorbell_offset, SDMA0_GFX_DOORBELL_OFFSET, + OFFSET, ring->doorbell_index); + } else { + doorbell = REG_SET_FIELD(doorbell, SDMA0_GFX_DOORBELL, ENABLE, 0); + } + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL), doorbell); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL_OFFSET), doorbell_offset); - adev->nbio.funcs->sdma_doorbell_range(adev, i, ring->use_doorbell, - ring->doorbell_index, - adev->doorbell_index.sdma_doorbell_range); + adev->nbio.funcs->sdma_doorbell_range(adev, i, ring->use_doorbell, + ring->doorbell_index, + adev->doorbell_index.sdma_doorbell_range); - if (amdgpu_sriov_vf(adev)) - sdma_v5_2_ring_set_wptr(ring); + if (amdgpu_sriov_vf(adev)) + sdma_v5_2_ring_set_wptr(ring); - /* set minor_ptr_update to 0 after wptr programed */ + /* set minor_ptr_update to 0 after wptr programed */ - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 0); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 0); - /* SRIOV VF has no control of any of registers below */ - if (!amdgpu_sriov_vf(adev)) { - /* set utc l1 enable flag always to 1 */ - temp = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_CNTL)); - temp = REG_SET_FIELD(temp, SDMA0_CNTL, UTC_L1_ENABLE, 1); + /* SRIOV VF has no control of any of registers below */ + if (!amdgpu_sriov_vf(adev)) { + /* set utc l1 enable flag always to 1 */ + temp = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_CNTL)); + temp = REG_SET_FIELD(temp, SDMA0_CNTL, UTC_L1_ENABLE, 1); - /* enable MCBP */ - temp = REG_SET_FIELD(temp, SDMA0_CNTL, MIDCMD_PREEMPT_ENABLE, 1); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_CNTL), temp); + /* enable MCBP */ + temp = REG_SET_FIELD(temp, SDMA0_CNTL, MIDCMD_PREEMPT_ENABLE, 1); + WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_CNTL), temp); - /* Set up RESP_MODE to non-copy addresses */ - temp = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_CNTL)); - temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, RESP_MODE, 3); - temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, REDO_DELAY, 9); - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_CNTL), temp); + /* Set up RESP_MODE to non-copy addresses */ + temp = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_CNTL)); + temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, RESP_MODE, 3); + temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, REDO_DELAY, 9); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_CNTL), temp); - /* program default cache read and write policy */ - temp = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_PAGE)); - /* clean read policy and write policy bits */ - temp &= 0xFF0FFF; - temp |= ((CACHE_READ_POLICY_L2__DEFAULT << 12) | - (CACHE_WRITE_POLICY_L2__DEFAULT << 14) | - SDMA0_UTCL1_PAGE__LLC_NOALLOC_MASK); - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_PAGE), temp); + /* program default cache read and write policy */ + temp = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_PAGE)); + /* clean read policy and write policy bits */ + temp &= 0xFF0FFF; + temp |= ((CACHE_READ_POLICY_L2__DEFAULT << 12) | + (CACHE_WRITE_POLICY_L2__DEFAULT << 14) | + SDMA0_UTCL1_PAGE__LLC_NOALLOC_MASK); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_PAGE), temp); - /* unhalt engine */ - temp = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_F32_CNTL)); - temp = REG_SET_FIELD(temp, SDMA0_F32_CNTL, HALT, 0); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_F32_CNTL), temp); - } + /* unhalt engine */ + temp = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_F32_CNTL)); + temp = REG_SET_FIELD(temp, SDMA0_F32_CNTL, HALT, 0); + WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_F32_CNTL), temp); + } - /* enable DMA RB */ - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_ENABLE, 1); - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); + /* enable DMA RB */ + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_ENABLE, 1); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); - ib_cntl = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL)); - ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_ENABLE, 1); + ib_cntl = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL)); + ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_ENABLE, 1); #ifdef __BIG_ENDIAN - ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_SWAP_ENABLE, 1); + ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_SWAP_ENABLE, 1); #endif - /* enable DMA IBs */ - WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL), ib_cntl); + /* enable DMA IBs */ + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL), ib_cntl); - if (amdgpu_sriov_vf(adev)) { /* bare-metal sequence doesn't need below to lines */ - sdma_v5_2_ctx_switch_enable(adev, true); - sdma_v5_2_enable(adev, true); - } + if (amdgpu_sriov_vf(adev)) { /* bare-metal sequence doesn't need below to lines */ + sdma_v5_2_ctx_switch_enable(adev, true); + sdma_v5_2_enable(adev, true); + } - r = amdgpu_ring_test_helper(ring); + return amdgpu_ring_test_helper(ring); +} + +/** + * sdma_v5_2_gfx_resume - setup and start the async dma engines + * + * @adev: amdgpu_device pointer + * + * Set up the gfx DMA ring buffers and enable them. + * Returns 0 for success, error for failure. + */ +static int sdma_v5_2_gfx_resume(struct amdgpu_device *adev) +{ + int i, r; + + for (i = 0; i < adev->sdma.num_instances; i++) { + r = sdma_v5_2_gfx_resume_instance(adev, i, false); if (r) return r; } @@ -736,9 +761,9 @@ static int sdma_v5_2_load_microcode(struct amdgpu_device *adev) return 0; } -static int sdma_v5_2_soft_reset(void *handle) +static int sdma_v5_2_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 grbm_soft_reset; u32 tmp; int i; @@ -778,6 +803,7 @@ static int sdma_v5_2_soft_reset(void *handle) static int sdma_v5_2_start(struct amdgpu_device *adev) { int r = 0; + struct amdgpu_ip_block *ip_block; if (amdgpu_sriov_vf(adev)) { sdma_v5_2_ctx_switch_enable(adev, false); @@ -798,7 +824,11 @@ static int sdma_v5_2_start(struct amdgpu_device *adev) msleep(1000); } - sdma_v5_2_soft_reset(adev); + ip_block = amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_SDMA); + if (!ip_block) + return -EINVAL; + + sdma_v5_2_soft_reset(ip_block); /* unhalt the MEs */ sdma_v5_2_enable(adev, true); /* enable sdma ring preemption */ @@ -1180,7 +1210,28 @@ static void sdma_v5_2_ring_emit_pipeline_sync(struct amdgpu_ring *ring) static void sdma_v5_2_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vmid, uint64_t pd_addr) { - amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); + struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->vm_hub]; + uint32_t req = hub->vmhub_funcs->get_invalidate_req(vmid, 0); + + /* Update the PD address for this VMID. */ + amdgpu_ring_emit_wreg(ring, hub->ctx0_ptb_addr_lo32 + + (hub->ctx_addr_distance * vmid), + lower_32_bits(pd_addr)); + amdgpu_ring_emit_wreg(ring, hub->ctx0_ptb_addr_hi32 + + (hub->ctx_addr_distance * vmid), + upper_32_bits(pd_addr)); + + /* Trigger invalidation. */ + amdgpu_ring_write(ring, + SDMA_PKT_VM_INVALIDATION_HEADER_OP(SDMA_OP_POLL_REGMEM) | + SDMA_PKT_VM_INVALIDATION_HEADER_SUB_OP(SDMA_SUBOP_VM_INVALIDATION) | + SDMA_PKT_VM_INVALIDATION_HEADER_GFX_ENG_ID(ring->vm_inv_eng) | + SDMA_PKT_VM_INVALIDATION_HEADER_MM_ENG_ID(0x1f)); + amdgpu_ring_write(ring, req); + amdgpu_ring_write(ring, 0xFFFFFFFF); + amdgpu_ring_write(ring, + SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_INVALIDATEACK(1 << vmid) | + SDMA_PKT_VM_INVALIDATION_ADDRESSRANGEHI_ADDRESSRANGEHI(0x1F)); } static void sdma_v5_2_ring_emit_wreg(struct amdgpu_ring *ring, @@ -1216,9 +1267,9 @@ static void sdma_v5_2_ring_emit_reg_write_reg_wait(struct amdgpu_ring *ring, amdgpu_ring_emit_reg_wait(ring, reg1, mask, mask); } -static int sdma_v5_2_early_init(void *handle) +static int sdma_v5_2_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_sdma_init_microcode(adev, 0, true); @@ -1268,11 +1319,11 @@ static unsigned sdma_v5_2_seq_to_trap_id(int seq_num) return -EINVAL; } -static int sdma_v5_2_sw_init(void *handle) +static int sdma_v5_2_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_5_2); uint32_t *ptr; @@ -1306,6 +1357,24 @@ static int sdma_v5_2_sw_init(void *handle) return r; } + adev->sdma.supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->sdma.instance[0].ring); + switch (amdgpu_ip_version(adev, SDMA0_HWIP, 0)) { + case IP_VERSION(5, 2, 0): + case IP_VERSION(5, 2, 2): + case IP_VERSION(5, 2, 3): + case IP_VERSION(5, 2, 4): + if (adev->sdma.instance[0].fw_version >= 76) + adev->sdma.supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE; + break; + case IP_VERSION(5, 2, 5): + if (adev->sdma.instance[0].fw_version >= 34) + adev->sdma.supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE; + break; + default: + break; + } + /* Allocate memory for SDMA IP Dump buffer */ ptr = kcalloc(adev->sdma.num_instances * reg_count, sizeof(uint32_t), GFP_KERNEL); if (ptr) @@ -1313,17 +1382,22 @@ static int sdma_v5_2_sw_init(void *handle) else DRM_ERROR("Failed to allocated memory for SDMA IP Dump\n"); + r = amdgpu_sdma_sysfs_reset_mask_init(adev); + if (r) + return r; + return r; } -static int sdma_v5_2_sw_fini(void *handle) +static int sdma_v5_2_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; for (i = 0; i < adev->sdma.num_instances; i++) amdgpu_ring_fini(&adev->sdma.instance[i].ring); + amdgpu_sdma_sysfs_reset_mask_fini(adev); amdgpu_sdma_destroy_inst_ctx(adev, true); kfree(adev->sdma.ip_dump); @@ -1331,16 +1405,16 @@ static int sdma_v5_2_sw_fini(void *handle) return 0; } -static int sdma_v5_2_hw_init(void *handle) +static int sdma_v5_2_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return sdma_v5_2_start(adev); } -static int sdma_v5_2_hw_fini(void *handle) +static int sdma_v5_2_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) return 0; @@ -1351,18 +1425,14 @@ static int sdma_v5_2_hw_fini(void *handle) return 0; } -static int sdma_v5_2_suspend(void *handle) +static int sdma_v5_2_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return sdma_v5_2_hw_fini(adev); + return sdma_v5_2_hw_fini(ip_block); } -static int sdma_v5_2_resume(void *handle) +static int sdma_v5_2_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return sdma_v5_2_hw_init(adev); + return sdma_v5_2_hw_init(ip_block); } static bool sdma_v5_2_is_idle(void *handle) @@ -1380,11 +1450,11 @@ static bool sdma_v5_2_is_idle(void *handle) return true; } -static int sdma_v5_2_wait_for_idle(void *handle) +static int sdma_v5_2_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; u32 sdma0, sdma1, sdma2, sdma3; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { sdma0 = RREG32(sdma_v5_2_get_reg_offset(adev, 0, mmSDMA0_STATUS_REG)); @@ -1399,6 +1469,96 @@ static int sdma_v5_2_wait_for_idle(void *handle) return -ETIMEDOUT; } +static int sdma_v5_2_reset_queue(struct amdgpu_ring *ring, unsigned int vmid) +{ + struct amdgpu_device *adev = ring->adev; + int i, j, r; + u32 rb_cntl, ib_cntl, f32_cntl, freeze, cntl, preempt, soft_reset, stat1_reg; + + if (amdgpu_sriov_vf(adev)) + return -EINVAL; + + for (i = 0; i < adev->sdma.num_instances; i++) { + if (ring == &adev->sdma.instance[i].ring) + break; + } + + if (i == adev->sdma.num_instances) { + DRM_ERROR("sdma instance not found\n"); + return -EINVAL; + } + + amdgpu_gfx_rlc_enter_safe_mode(adev, 0); + + /* stop queue */ + ib_cntl = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL)); + ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_ENABLE, 0); + WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL), ib_cntl); + + rb_cntl = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL)); + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_ENABLE, 0); + WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); + + /*engine stop SDMA1_F32_CNTL.HALT to 1 and SDMAx_FREEZE freeze bit to 1 */ + freeze = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_FREEZE)); + freeze = REG_SET_FIELD(freeze, SDMA0_FREEZE, FREEZE, 1); + WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_FREEZE), freeze); + + for (j = 0; j < adev->usec_timeout; j++) { + freeze = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_FREEZE)); + + if (REG_GET_FIELD(freeze, SDMA0_FREEZE, FROZEN) & 1) + break; + udelay(1); + } + + + if (j == adev->usec_timeout) { + stat1_reg = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_STATUS1_REG)); + if ((stat1_reg & 0x3FF) != 0x3FF) { + DRM_ERROR("cannot soft reset as sdma not idle\n"); + r = -ETIMEDOUT; + goto err0; + } + } + + f32_cntl = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_F32_CNTL)); + f32_cntl = REG_SET_FIELD(f32_cntl, SDMA0_F32_CNTL, HALT, 1); + WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_F32_CNTL), f32_cntl); + + cntl = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_CNTL)); + cntl = REG_SET_FIELD(cntl, SDMA0_CNTL, UTC_L1_ENABLE, 0); + WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_CNTL), cntl); + + /* soft reset SDMA_GFX_PREEMPT.IB_PREEMPT = 0 mmGRBM_SOFT_RESET.SOFT_RESET_SDMA0/1 = 1 */ + preempt = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_PREEMPT)); + preempt = REG_SET_FIELD(preempt, SDMA0_GFX_PREEMPT, IB_PREEMPT, 0); + WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_PREEMPT), preempt); + + soft_reset = RREG32_SOC15(GC, 0, mmGRBM_SOFT_RESET); + soft_reset |= 1 << GRBM_SOFT_RESET__SOFT_RESET_SDMA0__SHIFT << i; + + + WREG32_SOC15(GC, 0, mmGRBM_SOFT_RESET, soft_reset); + + udelay(50); + + soft_reset &= ~(1 << GRBM_SOFT_RESET__SOFT_RESET_SDMA0__SHIFT << i); + + WREG32_SOC15(GC, 0, mmGRBM_SOFT_RESET, soft_reset); + + /* unfreeze and unhalt */ + freeze = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_FREEZE)); + freeze = REG_SET_FIELD(freeze, SDMA0_FREEZE, FREEZE, 0); + WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_FREEZE), freeze); + + r = sdma_v5_2_gfx_resume_instance(adev, i, true); + +err0: + amdgpu_gfx_rlc_exit_safe_mode(adev, 0); + return r; +} + static int sdma_v5_2_ring_preempt_ib(struct amdgpu_ring *ring) { int i, r = 0; @@ -1736,9 +1896,9 @@ static void sdma_v5_2_ring_end_use(struct amdgpu_ring *ring) amdgpu_gfx_off_ctrl(adev, true); } -static void sdma_v5_2_print_ip_state(void *handle, struct drm_printer *p) +static void sdma_v5_2_print_ip_state(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_5_2); uint32_t instance_offset; @@ -1757,9 +1917,9 @@ static void sdma_v5_2_print_ip_state(void *handle, struct drm_printer *p) } } -static void sdma_v5_2_dump_ip_state(void *handle) +static void sdma_v5_2_dump_ip_state(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t instance_offset; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_5_2); @@ -1781,7 +1941,6 @@ static void sdma_v5_2_dump_ip_state(void *handle) static const struct amd_ip_funcs sdma_v5_2_ip_funcs = { .name = "sdma_v5_2", .early_init = sdma_v5_2_early_init, - .late_init = NULL, .sw_init = sdma_v5_2_sw_init, .sw_fini = sdma_v5_2_sw_fini, .hw_init = sdma_v5_2_hw_init, @@ -1834,6 +1993,7 @@ static const struct amdgpu_ring_funcs sdma_v5_2_ring_funcs = { .emit_reg_write_reg_wait = sdma_v5_2_ring_emit_reg_write_reg_wait, .init_cond_exec = sdma_v5_2_ring_init_cond_exec, .preempt_ib = sdma_v5_2_ring_preempt_ib, + .reset = sdma_v5_2_reset_queue, }; static void sdma_v5_2_set_ring_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c index 208a1fa9d4e7..d46128b0ec92 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c @@ -468,6 +468,157 @@ static void sdma_v6_0_enable(struct amdgpu_device *adev, bool enable) } } +/** + * sdma_v6_0_gfx_resume_instance - start/restart a certain sdma engine + * + * @adev: amdgpu_device pointer + * @i: instance + * @restore: used to restore wptr when restart + * + * Set up the gfx DMA ring buffers and enable them. On restart, we will restore wptr and rptr. + * Return 0 for success. + */ +static int sdma_v6_0_gfx_resume_instance(struct amdgpu_device *adev, int i, bool restore) +{ + struct amdgpu_ring *ring; + u32 rb_cntl, ib_cntl; + u32 rb_bufsz; + u32 doorbell; + u32 doorbell_offset; + u32 temp; + u64 wptr_gpu_addr; + + ring = &adev->sdma.instance[i].ring; + if (!amdgpu_sriov_vf(adev)) + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_SEM_WAIT_FAIL_TIMER_CNTL), 0); + + /* Set ring buffer size in dwords */ + rb_bufsz = order_base_2(ring->ring_size / 4); + rb_cntl = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_CNTL)); + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RB_SIZE, rb_bufsz); +#ifdef __BIG_ENDIAN + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RB_SWAP_ENABLE, 1); + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, + RPTR_WRITEBACK_SWAP_ENABLE, 1); +#endif + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RB_PRIV, 1); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_CNTL), rb_cntl); + + /* Initialize the ring buffer's read and write pointers */ + if (restore) { + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_RPTR), lower_32_bits(ring->wptr << 2)); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_RPTR_HI), upper_32_bits(ring->wptr << 2)); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR), lower_32_bits(ring->wptr << 2)); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR_HI), upper_32_bits(ring->wptr << 2)); + } else { + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_RPTR), 0); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_RPTR_HI), 0); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR), 0); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR_HI), 0); + } + /* setup the wptr shadow polling */ + wptr_gpu_addr = ring->wptr_gpu_addr; + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR_POLL_ADDR_LO), + lower_32_bits(wptr_gpu_addr)); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR_POLL_ADDR_HI), + upper_32_bits(wptr_gpu_addr)); + + /* set the wb address whether it's enabled or not */ + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_RPTR_ADDR_HI), + upper_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFF); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_RPTR_ADDR_LO), + lower_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFC); + + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RPTR_WRITEBACK_ENABLE, 1); + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, WPTR_POLL_ENABLE, 0); + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, F32_WPTR_POLL_ENABLE, 1); + + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_BASE), ring->gpu_addr >> 8); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_BASE_HI), ring->gpu_addr >> 40); + + if (!restore) + ring->wptr = 0; + + /* before programing wptr to a less value, need set minor_ptr_update first */ + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_MINOR_PTR_UPDATE), 1); + + if (!amdgpu_sriov_vf(adev)) { /* only bare-metal use register write for wptr */ + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR), lower_32_bits(ring->wptr) << 2); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR_HI), upper_32_bits(ring->wptr) << 2); + } + + doorbell = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_DOORBELL)); + doorbell_offset = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_DOORBELL_OFFSET)); + + if (ring->use_doorbell) { + doorbell = REG_SET_FIELD(doorbell, SDMA0_QUEUE0_DOORBELL, ENABLE, 1); + doorbell_offset = REG_SET_FIELD(doorbell_offset, SDMA0_QUEUE0_DOORBELL_OFFSET, + OFFSET, ring->doorbell_index); + } else { + doorbell = REG_SET_FIELD(doorbell, SDMA0_QUEUE0_DOORBELL, ENABLE, 0); + } + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_DOORBELL), doorbell); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_DOORBELL_OFFSET), doorbell_offset); + + if (i == 0) + adev->nbio.funcs->sdma_doorbell_range(adev, i, ring->use_doorbell, + ring->doorbell_index, + adev->doorbell_index.sdma_doorbell_range * adev->sdma.num_instances); + + if (amdgpu_sriov_vf(adev)) + sdma_v6_0_ring_set_wptr(ring); + + /* set minor_ptr_update to 0 after wptr programed */ + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_MINOR_PTR_UPDATE), 0); + + /* Set up sdma hang watchdog */ + temp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_WATCHDOG_CNTL)); + /* 100ms per unit */ + temp = REG_SET_FIELD(temp, SDMA0_WATCHDOG_CNTL, QUEUE_HANG_COUNT, + max(adev->usec_timeout/100000, 1)); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_WATCHDOG_CNTL), temp); + + /* Set up RESP_MODE to non-copy addresses */ + temp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UTCL1_CNTL)); + temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, RESP_MODE, 3); + temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, REDO_DELAY, 9); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UTCL1_CNTL), temp); + + /* program default cache read and write policy */ + temp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UTCL1_PAGE)); + /* clean read policy and write policy bits */ + temp &= 0xFF0FFF; + temp |= ((CACHE_READ_POLICY_L2__DEFAULT << 12) | + (CACHE_WRITE_POLICY_L2__DEFAULT << 14) | + SDMA0_UTCL1_PAGE__LLC_NOALLOC_MASK); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UTCL1_PAGE), temp); + + if (!amdgpu_sriov_vf(adev)) { + /* unhalt engine */ + temp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_F32_CNTL)); + temp = REG_SET_FIELD(temp, SDMA0_F32_CNTL, HALT, 0); + temp = REG_SET_FIELD(temp, SDMA0_F32_CNTL, TH1_RESET, 0); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_F32_CNTL), temp); + } + + /* enable DMA RB */ + rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RB_ENABLE, 1); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_CNTL), rb_cntl); + + ib_cntl = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_IB_CNTL)); + ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_QUEUE0_IB_CNTL, IB_ENABLE, 1); +#ifdef __BIG_ENDIAN + ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_QUEUE0_IB_CNTL, IB_SWAP_ENABLE, 1); +#endif + /* enable DMA IBs */ + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_IB_CNTL), ib_cntl); + + if (amdgpu_sriov_vf(adev)) + sdma_v6_0_enable(adev, true); + + return amdgpu_ring_test_helper(ring); +} + /** * sdma_v6_0_gfx_resume - setup and start the async dma engines * @@ -478,139 +629,10 @@ static void sdma_v6_0_enable(struct amdgpu_device *adev, bool enable) */ static int sdma_v6_0_gfx_resume(struct amdgpu_device *adev) { - struct amdgpu_ring *ring; - u32 rb_cntl, ib_cntl; - u32 rb_bufsz; - u32 doorbell; - u32 doorbell_offset; - u32 temp; - u64 wptr_gpu_addr; int i, r; for (i = 0; i < adev->sdma.num_instances; i++) { - ring = &adev->sdma.instance[i].ring; - - if (!amdgpu_sriov_vf(adev)) - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_SEM_WAIT_FAIL_TIMER_CNTL), 0); - - /* Set ring buffer size in dwords */ - rb_bufsz = order_base_2(ring->ring_size / 4); - rb_cntl = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_CNTL)); - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RB_SIZE, rb_bufsz); -#ifdef __BIG_ENDIAN - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RB_SWAP_ENABLE, 1); - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, - RPTR_WRITEBACK_SWAP_ENABLE, 1); -#endif - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RB_PRIV, 1); - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_CNTL), rb_cntl); - - /* Initialize the ring buffer's read and write pointers */ - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_RPTR), 0); - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_RPTR_HI), 0); - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR), 0); - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR_HI), 0); - - /* setup the wptr shadow polling */ - wptr_gpu_addr = ring->wptr_gpu_addr; - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR_POLL_ADDR_LO), - lower_32_bits(wptr_gpu_addr)); - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR_POLL_ADDR_HI), - upper_32_bits(wptr_gpu_addr)); - - /* set the wb address whether it's enabled or not */ - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_RPTR_ADDR_HI), - upper_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFF); - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_RPTR_ADDR_LO), - lower_32_bits(ring->rptr_gpu_addr) & 0xFFFFFFFC); - - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RPTR_WRITEBACK_ENABLE, 1); - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, WPTR_POLL_ENABLE, 0); - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, F32_WPTR_POLL_ENABLE, 1); - - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_BASE), ring->gpu_addr >> 8); - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_BASE_HI), ring->gpu_addr >> 40); - - ring->wptr = 0; - - /* before programing wptr to a less value, need set minor_ptr_update first */ - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_MINOR_PTR_UPDATE), 1); - - if (!amdgpu_sriov_vf(adev)) { /* only bare-metal use register write for wptr */ - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR), lower_32_bits(ring->wptr) << 2); - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_WPTR_HI), upper_32_bits(ring->wptr) << 2); - } - - doorbell = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_DOORBELL)); - doorbell_offset = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_DOORBELL_OFFSET)); - - if (ring->use_doorbell) { - doorbell = REG_SET_FIELD(doorbell, SDMA0_QUEUE0_DOORBELL, ENABLE, 1); - doorbell_offset = REG_SET_FIELD(doorbell_offset, SDMA0_QUEUE0_DOORBELL_OFFSET, - OFFSET, ring->doorbell_index); - } else { - doorbell = REG_SET_FIELD(doorbell, SDMA0_QUEUE0_DOORBELL, ENABLE, 0); - } - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_DOORBELL), doorbell); - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_DOORBELL_OFFSET), doorbell_offset); - - if (i == 0) - adev->nbio.funcs->sdma_doorbell_range(adev, i, ring->use_doorbell, - ring->doorbell_index, - adev->doorbell_index.sdma_doorbell_range * adev->sdma.num_instances); - - if (amdgpu_sriov_vf(adev)) - sdma_v6_0_ring_set_wptr(ring); - - /* set minor_ptr_update to 0 after wptr programed */ - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_MINOR_PTR_UPDATE), 0); - - /* Set up sdma hang watchdog */ - temp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_WATCHDOG_CNTL)); - /* 100ms per unit */ - temp = REG_SET_FIELD(temp, SDMA0_WATCHDOG_CNTL, QUEUE_HANG_COUNT, - max(adev->usec_timeout/100000, 1)); - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_WATCHDOG_CNTL), temp); - - /* Set up RESP_MODE to non-copy addresses */ - temp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UTCL1_CNTL)); - temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, RESP_MODE, 3); - temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, REDO_DELAY, 9); - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UTCL1_CNTL), temp); - - /* program default cache read and write policy */ - temp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UTCL1_PAGE)); - /* clean read policy and write policy bits */ - temp &= 0xFF0FFF; - temp |= ((CACHE_READ_POLICY_L2__DEFAULT << 12) | - (CACHE_WRITE_POLICY_L2__DEFAULT << 14) | - SDMA0_UTCL1_PAGE__LLC_NOALLOC_MASK); - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UTCL1_PAGE), temp); - - if (!amdgpu_sriov_vf(adev)) { - /* unhalt engine */ - temp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_F32_CNTL)); - temp = REG_SET_FIELD(temp, SDMA0_F32_CNTL, HALT, 0); - temp = REG_SET_FIELD(temp, SDMA0_F32_CNTL, TH1_RESET, 0); - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_F32_CNTL), temp); - } - - /* enable DMA RB */ - rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_QUEUE0_RB_CNTL, RB_ENABLE, 1); - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_CNTL), rb_cntl); - - ib_cntl = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_IB_CNTL)); - ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_QUEUE0_IB_CNTL, IB_ENABLE, 1); -#ifdef __BIG_ENDIAN - ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_QUEUE0_IB_CNTL, IB_SWAP_ENABLE, 1); -#endif - /* enable DMA IBs */ - WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_IB_CNTL), ib_cntl); - - if (amdgpu_sriov_vf(adev)) - sdma_v6_0_enable(adev, true); - - r = amdgpu_ring_test_helper(ring); + r = sdma_v6_0_gfx_resume_instance(adev, i, false); if (r) return r; } @@ -733,9 +755,9 @@ static int sdma_v6_0_load_microcode(struct amdgpu_device *adev) return 0; } -static int sdma_v6_0_soft_reset(void *handle) +static int sdma_v6_0_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 tmp; int i; @@ -769,9 +791,9 @@ static int sdma_v6_0_soft_reset(void *handle) return sdma_v6_0_start(adev); } -static bool sdma_v6_0_check_soft_reset(void *handle) +static bool sdma_v6_0_check_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int i, r; long tmo = msecs_to_jiffies(1000); @@ -1272,9 +1294,9 @@ static void sdma_v6_0_set_ras_funcs(struct amdgpu_device *adev) } } -static int sdma_v6_0_early_init(void *handle) +static int sdma_v6_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_sdma_init_microcode(adev, 0, true); @@ -1291,11 +1313,11 @@ static int sdma_v6_0_early_init(void *handle) return 0; } -static int sdma_v6_0_sw_init(void *handle) +static int sdma_v6_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_6_0); uint32_t *ptr; @@ -1328,6 +1350,19 @@ static int sdma_v6_0_sw_init(void *handle) return r; } + adev->sdma.supported_reset = + amdgpu_get_soft_full_reset_mask(&adev->sdma.instance[0].ring); + switch (amdgpu_ip_version(adev, SDMA0_HWIP, 0)) { + case IP_VERSION(6, 0, 0): + case IP_VERSION(6, 0, 2): + case IP_VERSION(6, 0, 3): + if (adev->sdma.instance[0].fw_version >= 21) + adev->sdma.supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE; + break; + default: + break; + } + if (amdgpu_sdma_ras_sw_init(adev)) { dev_err(adev->dev, "Failed to initialize sdma ras block!\n"); return -EINVAL; @@ -1340,17 +1375,22 @@ static int sdma_v6_0_sw_init(void *handle) else DRM_ERROR("Failed to allocated memory for SDMA IP Dump\n"); + r = amdgpu_sdma_sysfs_reset_mask_init(adev); + if (r) + return r; + return r; } -static int sdma_v6_0_sw_fini(void *handle) +static int sdma_v6_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; for (i = 0; i < adev->sdma.num_instances; i++) amdgpu_ring_fini(&adev->sdma.instance[i].ring); + amdgpu_sdma_sysfs_reset_mask_fini(adev); amdgpu_sdma_destroy_inst_ctx(adev, true); kfree(adev->sdma.ip_dump); @@ -1358,16 +1398,16 @@ static int sdma_v6_0_sw_fini(void *handle) return 0; } -static int sdma_v6_0_hw_init(void *handle) +static int sdma_v6_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return sdma_v6_0_start(adev); } -static int sdma_v6_0_hw_fini(void *handle) +static int sdma_v6_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) return 0; @@ -1378,18 +1418,14 @@ static int sdma_v6_0_hw_fini(void *handle) return 0; } -static int sdma_v6_0_suspend(void *handle) +static int sdma_v6_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return sdma_v6_0_hw_fini(adev); + return sdma_v6_0_hw_fini(ip_block); } -static int sdma_v6_0_resume(void *handle) +static int sdma_v6_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return sdma_v6_0_hw_init(adev); + return sdma_v6_0_hw_init(ip_block); } static bool sdma_v6_0_is_idle(void *handle) @@ -1407,11 +1443,11 @@ static bool sdma_v6_0_is_idle(void *handle) return true; } -static int sdma_v6_0_wait_for_idle(void *handle) +static int sdma_v6_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; u32 sdma0, sdma1; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { sdma0 = RREG32(sdma_v6_0_get_reg_offset(adev, 0, regSDMA0_STATUS_REG)); @@ -1469,6 +1505,31 @@ static int sdma_v6_0_ring_preempt_ib(struct amdgpu_ring *ring) return r; } +static int sdma_v6_0_reset_queue(struct amdgpu_ring *ring, unsigned int vmid) +{ + struct amdgpu_device *adev = ring->adev; + int i, r; + + if (amdgpu_sriov_vf(adev)) + return -EINVAL; + + for (i = 0; i < adev->sdma.num_instances; i++) { + if (ring == &adev->sdma.instance[i].ring) + break; + } + + if (i == adev->sdma.num_instances) { + DRM_ERROR("sdma instance not found\n"); + return -EINVAL; + } + + r = amdgpu_mes_reset_legacy_queue(adev, ring, vmid, true); + if (r) + return r; + + return sdma_v6_0_gfx_resume_instance(adev, i, true); +} + static int sdma_v6_0_set_trap_irq_state(struct amdgpu_device *adev, struct amdgpu_irq_src *source, unsigned type, @@ -1556,9 +1617,9 @@ static void sdma_v6_0_get_clockgating_state(void *handle, u64 *flags) { } -static void sdma_v6_0_print_ip_state(void *handle, struct drm_printer *p) +static void sdma_v6_0_print_ip_state(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_6_0); uint32_t instance_offset; @@ -1577,9 +1638,9 @@ static void sdma_v6_0_print_ip_state(void *handle, struct drm_printer *p) } } -static void sdma_v6_0_dump_ip_state(void *handle) +static void sdma_v6_0_dump_ip_state(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t instance_offset; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_6_0); @@ -1601,7 +1662,6 @@ static void sdma_v6_0_dump_ip_state(void *handle) const struct amd_ip_funcs sdma_v6_0_ip_funcs = { .name = "sdma_v6_0", .early_init = sdma_v6_0_early_init, - .late_init = NULL, .sw_init = sdma_v6_0_sw_init, .sw_fini = sdma_v6_0_sw_fini, .hw_init = sdma_v6_0_hw_init, @@ -1652,6 +1712,7 @@ static const struct amdgpu_ring_funcs sdma_v6_0_ring_funcs = { .emit_reg_write_reg_wait = sdma_v6_0_ring_emit_reg_write_reg_wait, .init_cond_exec = sdma_v6_0_ring_init_cond_exec, .preempt_ib = sdma_v6_0_ring_preempt_ib, + .reset = sdma_v6_0_reset_queue, }; static void sdma_v6_0_set_ring_funcs(struct amdgpu_device *adev) @@ -1726,7 +1787,7 @@ static void sdma_v6_0_emit_fill_buffer(struct amdgpu_ib *ib, uint64_t dst_offset, uint32_t byte_count) { - ib->ptr[ib->length_dw++] = SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_CONST_FILL); + ib->ptr[ib->length_dw++] = SDMA_PKT_CONSTANT_FILL_HEADER_OP(SDMA_OP_CONST_FILL); ib->ptr[ib->length_dw++] = lower_32_bits(dst_offset); ib->ptr[ib->length_dw++] = upper_32_bits(dst_offset); ib->ptr[ib->length_dw++] = src_data; diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v7_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v7_0.c index 9288f37a3cc5..d2ce6b6a7ff6 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v7_0.c @@ -753,9 +753,9 @@ static int sdma_v7_0_load_microcode(struct amdgpu_device *adev) return 0; } -static int sdma_v7_0_soft_reset(void *handle) +static int sdma_v7_0_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 tmp; int i; @@ -789,9 +789,9 @@ static int sdma_v7_0_soft_reset(void *handle) return sdma_v7_0_start(adev); } -static bool sdma_v7_0_check_soft_reset(void *handle) +static bool sdma_v7_0_check_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int i, r; long tmo = msecs_to_jiffies(1000); @@ -1259,9 +1259,9 @@ static void sdma_v7_0_ring_emit_reg_write_reg_wait(struct amdgpu_ring *ring, amdgpu_ring_emit_reg_wait(ring, reg1, mask, mask); } -static int sdma_v7_0_early_init(void *handle) +static int sdma_v7_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_sdma_init_microcode(adev, 0, true); @@ -1279,11 +1279,11 @@ static int sdma_v7_0_early_init(void *handle) return 0; } -static int sdma_v7_0_sw_init(void *handle) +static int sdma_v7_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_7_0); uint32_t *ptr; @@ -1326,9 +1326,9 @@ static int sdma_v7_0_sw_init(void *handle) return r; } -static int sdma_v7_0_sw_fini(void *handle) +static int sdma_v7_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; for (i = 0; i < adev->sdma.num_instances; i++) @@ -1344,16 +1344,16 @@ static int sdma_v7_0_sw_fini(void *handle) return 0; } -static int sdma_v7_0_hw_init(void *handle) +static int sdma_v7_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return sdma_v7_0_start(adev); } -static int sdma_v7_0_hw_fini(void *handle) +static int sdma_v7_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) return 0; @@ -1364,18 +1364,14 @@ static int sdma_v7_0_hw_fini(void *handle) return 0; } -static int sdma_v7_0_suspend(void *handle) +static int sdma_v7_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return sdma_v7_0_hw_fini(adev); + return sdma_v7_0_hw_fini(ip_block); } -static int sdma_v7_0_resume(void *handle) +static int sdma_v7_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return sdma_v7_0_hw_init(adev); + return sdma_v7_0_hw_init(ip_block); } static bool sdma_v7_0_is_idle(void *handle) @@ -1393,11 +1389,11 @@ static bool sdma_v7_0_is_idle(void *handle) return true; } -static int sdma_v7_0_wait_for_idle(void *handle) +static int sdma_v7_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; u32 sdma0, sdma1; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { sdma0 = RREG32(sdma_v7_0_get_reg_offset(adev, 0, regSDMA0_STATUS_REG)); @@ -1544,9 +1540,9 @@ static void sdma_v7_0_get_clockgating_state(void *handle, u64 *flags) { } -static void sdma_v7_0_print_ip_state(void *handle, struct drm_printer *p) +static void sdma_v7_0_print_ip_state(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_7_0); uint32_t instance_offset; @@ -1565,9 +1561,9 @@ static void sdma_v7_0_print_ip_state(void *handle, struct drm_printer *p) } } -static void sdma_v7_0_dump_ip_state(void *handle) +static void sdma_v7_0_dump_ip_state(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t instance_offset; uint32_t reg_count = ARRAY_SIZE(sdma_reg_list_7_0); diff --git a/drivers/gpu/drm/amd/amdgpu/si.c b/drivers/gpu/drm/amd/amdgpu/si.c index 85235470e872..00f63d3fbea7 100644 --- a/drivers/gpu/drm/amd/amdgpu/si.c +++ b/drivers/gpu/drm/amd/amdgpu/si.c @@ -2022,9 +2022,9 @@ static uint32_t si_get_rev_id(struct amdgpu_device *adev) >> CC_DRM_ID_STRAPS__ATI_REV_ID__SHIFT; } -static int si_common_early_init(void *handle) +static int si_common_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->smc_rreg = &si_smc_rreg; adev->smc_wreg = &si_smc_wreg; @@ -2148,17 +2148,6 @@ static int si_common_early_init(void *handle) return 0; } -static int si_common_sw_init(void *handle) -{ - return 0; -} - -static int si_common_sw_fini(void *handle) -{ - return 0; -} - - static void si_init_golden_registers(struct amdgpu_device *adev) { switch (adev->asic_type) { @@ -2633,9 +2622,9 @@ static void si_fix_pci_max_read_req_size(struct amdgpu_device *adev) pcie_set_readrq(adev->pdev, 512); } -static int si_common_hw_init(void *handle) +static int si_common_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; si_fix_pci_max_read_req_size(adev); si_init_golden_registers(adev); @@ -2645,23 +2634,14 @@ static int si_common_hw_init(void *handle) return 0; } -static int si_common_hw_fini(void *handle) +static int si_common_hw_fini(struct amdgpu_ip_block *ip_block) { return 0; } -static int si_common_suspend(void *handle) +static int si_common_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return si_common_hw_fini(adev); -} - -static int si_common_resume(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return si_common_hw_init(adev); + return si_common_hw_init(ip_block); } static bool si_common_is_idle(void *handle) @@ -2669,16 +2649,6 @@ static bool si_common_is_idle(void *handle) return true; } -static int si_common_wait_for_idle(void *handle) -{ - return 0; -} - -static int si_common_soft_reset(void *handle) -{ - return 0; -} - static int si_common_set_clockgating_state(void *handle, enum amd_clockgating_state state) { @@ -2694,20 +2664,12 @@ static int si_common_set_powergating_state(void *handle, static const struct amd_ip_funcs si_common_ip_funcs = { .name = "si_common", .early_init = si_common_early_init, - .late_init = NULL, - .sw_init = si_common_sw_init, - .sw_fini = si_common_sw_fini, .hw_init = si_common_hw_init, .hw_fini = si_common_hw_fini, - .suspend = si_common_suspend, .resume = si_common_resume, .is_idle = si_common_is_idle, - .wait_for_idle = si_common_wait_for_idle, - .soft_reset = si_common_soft_reset, .set_clockgating_state = si_common_set_clockgating_state, .set_powergating_state = si_common_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ip_block_version si_common_ip_block = diff --git a/drivers/gpu/drm/amd/amdgpu/si_dma.c b/drivers/gpu/drm/amd/amdgpu/si_dma.c index 11db5b755832..47647a6083e8 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_dma.c +++ b/drivers/gpu/drm/amd/amdgpu/si_dma.c @@ -457,9 +457,9 @@ static void si_dma_ring_emit_wreg(struct amdgpu_ring *ring, amdgpu_ring_write(ring, val); } -static int si_dma_early_init(void *handle) +static int si_dma_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->sdma.num_instances = 2; @@ -471,11 +471,11 @@ static int si_dma_early_init(void *handle) return 0; } -static int si_dma_sw_init(void *handle) +static int si_dma_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* DMA0 trap event */ r = amdgpu_irq_add_id(adev, AMDGPU_IRQ_CLIENTID_LEGACY, 224, @@ -506,9 +506,9 @@ static int si_dma_sw_init(void *handle) return r; } -static int si_dma_sw_fini(void *handle) +static int si_dma_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; for (i = 0; i < adev->sdma.num_instances; i++) @@ -517,39 +517,34 @@ static int si_dma_sw_fini(void *handle) return 0; } -static int si_dma_hw_init(void *handle) +static int si_dma_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return si_dma_start(adev); } -static int si_dma_hw_fini(void *handle) +static int si_dma_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - si_dma_stop(adev); + si_dma_stop(ip_block->adev); return 0; } -static int si_dma_suspend(void *handle) +static int si_dma_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return si_dma_hw_fini(adev); + return si_dma_hw_fini(ip_block); } -static int si_dma_resume(void *handle) +static int si_dma_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return si_dma_hw_init(adev); + return si_dma_hw_init(ip_block); } static bool si_dma_is_idle(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + u32 tmp = RREG32(SRBM_STATUS2); if (tmp & (DMA_BUSY_MASK | DMA1_BUSY_MASK)) @@ -558,20 +553,20 @@ static bool si_dma_is_idle(void *handle) return true; } -static int si_dma_wait_for_idle(void *handle) +static int si_dma_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { - if (si_dma_is_idle(handle)) + if (si_dma_is_idle(adev)) return 0; udelay(1); } return -ETIMEDOUT; } -static int si_dma_soft_reset(void *handle) +static int si_dma_soft_reset(struct amdgpu_ip_block *ip_block) { DRM_INFO("si_dma_soft_reset --- not implemented !!!!!!!\n"); return 0; @@ -696,7 +691,6 @@ static int si_dma_set_powergating_state(void *handle, static const struct amd_ip_funcs si_dma_ip_funcs = { .name = "si_dma", .early_init = si_dma_early_init, - .late_init = NULL, .sw_init = si_dma_sw_init, .sw_fini = si_dma_sw_fini, .hw_init = si_dma_hw_init, @@ -708,8 +702,6 @@ static const struct amd_ip_funcs si_dma_ip_funcs = { .soft_reset = si_dma_soft_reset, .set_clockgating_state = si_dma_set_clockgating_state, .set_powergating_state = si_dma_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs si_dma_ring_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/si_ih.c b/drivers/gpu/drm/amd/amdgpu/si_ih.c index 5237395e4fab..2ec1ebe4db11 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/si_ih.c @@ -156,19 +156,19 @@ static void si_ih_set_rptr(struct amdgpu_device *adev, WREG32(IH_RB_RPTR, ih->rptr); } -static int si_ih_early_init(void *handle) +static int si_ih_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; si_ih_set_interrupt_funcs(adev); return 0; } -static int si_ih_sw_init(void *handle) +static int si_ih_sw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_ih_ring_init(adev, &adev->irq.ih, 64 * 1024, false); if (r) @@ -177,43 +177,37 @@ static int si_ih_sw_init(void *handle) return amdgpu_irq_init(adev); } -static int si_ih_sw_fini(void *handle) +static int si_ih_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_fini_sw(adev); return 0; } -static int si_ih_hw_init(void *handle) +static int si_ih_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return si_ih_irq_init(adev); } -static int si_ih_hw_fini(void *handle) +static int si_ih_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - si_ih_irq_disable(adev); + si_ih_irq_disable(ip_block->adev); return 0; } -static int si_ih_suspend(void *handle) +static int si_ih_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return si_ih_hw_fini(adev); + return si_ih_hw_fini(ip_block); } -static int si_ih_resume(void *handle) +static int si_ih_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return si_ih_hw_init(adev); + return si_ih_hw_init(ip_block); } static bool si_ih_is_idle(void *handle) @@ -227,22 +221,22 @@ static bool si_ih_is_idle(void *handle) return true; } -static int si_ih_wait_for_idle(void *handle) +static int si_ih_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { - if (si_ih_is_idle(handle)) + if (si_ih_is_idle(adev)) return 0; udelay(1); } return -ETIMEDOUT; } -static int si_ih_soft_reset(void *handle) +static int si_ih_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 srbm_soft_reset = 0; u32 tmp = RREG32(SRBM_STATUS); @@ -284,7 +278,6 @@ static int si_ih_set_powergating_state(void *handle, static const struct amd_ip_funcs si_ih_ip_funcs = { .name = "si_ih", .early_init = si_ih_early_init, - .late_init = NULL, .sw_init = si_ih_sw_init, .sw_fini = si_ih_sw_fini, .hw_init = si_ih_hw_init, @@ -296,8 +289,6 @@ static const struct amd_ip_funcs si_ih_ip_funcs = { .soft_reset = si_ih_soft_reset, .set_clockgating_state = si_ih_set_clockgating_state, .set_powergating_state = si_ih_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs si_ih_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c b/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c index 481217c32d85..9b01e074af47 100644 --- a/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c +++ b/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c @@ -81,15 +81,9 @@ static int sienna_cichlid_mode2_suspend_ip(struct amdgpu_device *adev) AMD_IP_BLOCK_TYPE_SDMA)) continue; - r = adev->ip_blocks[i].version->funcs->suspend(adev); - - if (r) { - dev_err(adev->dev, - "suspend of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); + r = amdgpu_ip_block_suspend(&adev->ip_blocks[i]); + if (r) return r; - } - adev->ip_blocks[i].status.hw = false; } return 0; @@ -175,15 +169,9 @@ static int sienna_cichlid_mode2_restore_ip(struct amdgpu_device *adev) for (i = 0; i < adev->num_ip_blocks; i++) { if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH) { - r = adev->ip_blocks[i].version->funcs->resume(adev); - if (r) { - dev_err(adev->dev, - "resume of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); + r = amdgpu_ip_block_resume(&adev->ip_blocks[i]); + if (r) return r; - } - - adev->ip_blocks[i].status.hw = true; } } @@ -193,15 +181,9 @@ static int sienna_cichlid_mode2_restore_ip(struct amdgpu_device *adev) adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_SDMA)) continue; - r = adev->ip_blocks[i].version->funcs->resume(adev); - if (r) { - dev_err(adev->dev, - "resume of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); + r = amdgpu_ip_block_resume(&adev->ip_blocks[i]); + if (r) return r; - } - - adev->ip_blocks[i].status.hw = true; } for (i = 0; i < adev->num_ip_blocks; i++) { @@ -213,7 +195,7 @@ static int sienna_cichlid_mode2_restore_ip(struct amdgpu_device *adev) if (adev->ip_blocks[i].version->funcs->late_init) { r = adev->ip_blocks[i].version->funcs->late_init( - (void *)adev); + &adev->ip_blocks[i]); if (r) { dev_err(adev->dev, "late_init of IP block <%s> failed %d after reset\n", diff --git a/drivers/gpu/drm/amd/amdgpu/smu_v13_0_10.c b/drivers/gpu/drm/amd/amdgpu/smu_v13_0_10.c index 0af648931df5..e70ebad3f9fa 100644 --- a/drivers/gpu/drm/amd/amdgpu/smu_v13_0_10.c +++ b/drivers/gpu/drm/amd/amdgpu/smu_v13_0_10.c @@ -80,15 +80,9 @@ static int smu_v13_0_10_mode2_suspend_ip(struct amdgpu_device *adev) AMD_IP_BLOCK_TYPE_MES)) continue; - r = adev->ip_blocks[i].version->funcs->suspend(adev); - - if (r) { - dev_err(adev->dev, - "suspend of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); + r = amdgpu_ip_block_suspend(&adev->ip_blocks[i]); + if (r) return r; - } - adev->ip_blocks[i].status.hw = false; } return 0; @@ -186,15 +180,9 @@ static int smu_v13_0_10_mode2_restore_ip(struct amdgpu_device *adev) adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_SDMA)) continue; - r = adev->ip_blocks[i].version->funcs->resume(adev); - if (r) { - dev_err(adev->dev, - "resume of IP block <%s> failed %d\n", - adev->ip_blocks[i].version->funcs->name, r); + r = amdgpu_ip_block_resume(&adev->ip_blocks[i]); + if (r) return r; - } - - adev->ip_blocks[i].status.hw = true; } for (i = 0; i < adev->num_ip_blocks; i++) { @@ -208,7 +196,7 @@ static int smu_v13_0_10_mode2_restore_ip(struct amdgpu_device *adev) if (adev->ip_blocks[i].version->funcs->late_init) { r = adev->ip_blocks[i].version->funcs->late_init( - (void *)adev); + &adev->ip_blocks[i]); if (r) { dev_err(adev->dev, "late_init of IP block <%s> failed %d after reset\n", diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index 307185c0e1b8..ede072758dab 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -578,20 +578,16 @@ soc15_asic_reset_method(struct amdgpu_device *adev) static bool soc15_need_reset_on_resume(struct amdgpu_device *adev) { - u32 sol_reg; - - sol_reg = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_81); - /* Will reset for the following suspend abort cases. - * 1) Only reset limit on APU side, dGPU hasn't checked yet. - * 2) S3 suspend abort and TOS already launched. + * 1) Only reset on APU side, dGPU hasn't checked yet. + * 2) S3 suspend aborted in the normal S3 suspend or + * performing pm core test. */ if (adev->flags & AMD_IS_APU && adev->in_s3 && - !adev->suspend_complete && - sol_reg) + !pm_resume_via_firmware()) return true; - - return false; + else + return false; } static int soc15_asic_reset(struct amdgpu_device *adev) @@ -601,11 +597,17 @@ static int soc15_asic_reset(struct amdgpu_device *adev) * successfully. So now, temporarily enable it for the * S3 suspend abort case. */ - if (((adev->apu_flags & AMD_APU_IS_RAVEN) || - (adev->apu_flags & AMD_APU_IS_RAVEN2)) && - !soc15_need_reset_on_resume(adev)) + + if ((adev->apu_flags & AMD_APU_IS_PICASSO || + !(adev->apu_flags & AMD_APU_IS_RAVEN)) && + soc15_need_reset_on_resume(adev)) + goto asic_reset; + + if ((adev->apu_flags & AMD_APU_IS_RAVEN) || + (adev->apu_flags & AMD_APU_IS_RAVEN2)) return 0; +asic_reset: switch (soc15_asic_reset_method(adev)) { case AMD_RESET_METHOD_PCI: dev_info(adev->dev, "PCI reset\n"); @@ -829,6 +831,10 @@ static bool soc15_need_reset_on_init(struct amdgpu_device *adev) if (adev->asic_type == CHIP_RENOIR) return true; + if (amdgpu_gmc_need_reset_on_init(adev)) + return true; + if (amdgpu_psp_tos_reload_needed(adev)) + return true; /* Just return false for soc15 GPUs. Reset does not seem to * be necessary. */ @@ -929,9 +935,9 @@ static const struct amdgpu_asic_funcs aqua_vanjaram_asic_funcs = .get_reg_state = &aqua_vanjaram_get_reg_state, }; -static int soc15_common_early_init(void *handle) +static int soc15_common_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->nbio.funcs->set_reg_remap(adev); adev->smc_rreg = NULL; @@ -1198,9 +1204,9 @@ static int soc15_common_early_init(void *handle) return 0; } -static int soc15_common_late_init(void *handle) +static int soc15_common_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) xgpu_ai_mailbox_get_irq(adev); @@ -1213,9 +1219,9 @@ static int soc15_common_late_init(void *handle) return 0; } -static int soc15_common_sw_init(void *handle) +static int soc15_common_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) xgpu_ai_mailbox_add_irq_id(adev); @@ -1227,9 +1233,9 @@ static int soc15_common_sw_init(void *handle) return 0; } -static int soc15_common_sw_fini(void *handle) +static int soc15_common_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->df.funcs && adev->df.funcs->sw_fini) @@ -1251,9 +1257,9 @@ static void soc15_sdma_doorbell_range_init(struct amdgpu_device *adev) } } -static int soc15_common_hw_init(void *handle) +static int soc15_common_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* enable aspm */ soc15_program_aspm(adev); @@ -1280,9 +1286,9 @@ static int soc15_common_hw_init(void *handle) return 0; } -static int soc15_common_hw_fini(void *handle) +static int soc15_common_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* Disable the doorbell aperture and selfring doorbell aperture * separately in hw_fini because soc15_enable_doorbell_aperture @@ -1295,7 +1301,12 @@ static int soc15_common_hw_fini(void *handle) if (amdgpu_sriov_vf(adev)) xgpu_ai_mailbox_put_irq(adev); + /* + * For minimal init, late_init is not called, hence RAS irqs are not + * enabled. + */ if ((!amdgpu_sriov_vf(adev)) && + (adev->init_lvl->level != AMDGPU_INIT_LEVEL_MINIMAL_XGMI) && adev->nbio.ras_if && amdgpu_ras_is_supported(adev, adev->nbio.ras_if->block)) { if (adev->nbio.ras && @@ -1309,22 +1320,20 @@ static int soc15_common_hw_fini(void *handle) return 0; } -static int soc15_common_suspend(void *handle) +static int soc15_common_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return soc15_common_hw_fini(adev); + return soc15_common_hw_fini(ip_block); } -static int soc15_common_resume(void *handle) +static int soc15_common_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (soc15_need_reset_on_resume(adev)) { dev_info(adev->dev, "S3 suspend abort case, let's reset ASIC.\n"); soc15_asic_reset(adev); } - return soc15_common_hw_init(adev); + return soc15_common_hw_init(ip_block); } static bool soc15_common_is_idle(void *handle) @@ -1332,16 +1341,6 @@ static bool soc15_common_is_idle(void *handle) return true; } -static int soc15_common_wait_for_idle(void *handle) -{ - return 0; -} - -static int soc15_common_soft_reset(void *handle) -{ - return 0; -} - static void soc15_update_drm_clock_gating(struct amdgpu_device *adev, bool enable) { uint32_t def, data; @@ -1492,11 +1491,7 @@ static const struct amd_ip_funcs soc15_common_ip_funcs = { .suspend = soc15_common_suspend, .resume = soc15_common_resume, .is_idle = soc15_common_is_idle, - .wait_for_idle = soc15_common_wait_for_idle, - .soft_reset = soc15_common_soft_reset, .set_clockgating_state = soc15_common_set_clockgating_state, .set_powergating_state = soc15_common_set_powergating_state, .get_clockgating_state= soc15_common_get_clockgating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; diff --git a/drivers/gpu/drm/amd/amdgpu/soc21.c b/drivers/gpu/drm/amd/amdgpu/soc21.c index bba35880badb..d6999835918f 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc21.c +++ b/drivers/gpu/drm/amd/amdgpu/soc21.c @@ -556,9 +556,9 @@ static const struct amdgpu_asic_funcs soc21_asic_funcs = { .update_umd_stable_pstate = &soc21_update_umd_stable_pstate, }; -static int soc21_common_early_init(void *handle) +static int soc21_common_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->nbio.funcs->set_reg_remap(adev); adev->smc_rreg = NULL; @@ -794,9 +794,9 @@ static int soc21_common_early_init(void *handle) return 0; } -static int soc21_common_late_init(void *handle) +static int soc21_common_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) { xgpu_nv_mailbox_get_irq(adev); @@ -832,9 +832,9 @@ static int soc21_common_late_init(void *handle) return 0; } -static int soc21_common_sw_init(void *handle) +static int soc21_common_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) xgpu_nv_mailbox_add_irq_id(adev); @@ -842,14 +842,9 @@ static int soc21_common_sw_init(void *handle) return 0; } -static int soc21_common_sw_fini(void *handle) +static int soc21_common_hw_init(struct amdgpu_ip_block *ip_block) { - return 0; -} - -static int soc21_common_hw_init(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* enable aspm */ soc21_program_aspm(adev); @@ -867,9 +862,9 @@ static int soc21_common_hw_init(void *handle) return 0; } -static int soc21_common_hw_fini(void *handle) +static int soc21_common_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* Disable the doorbell aperture and selfring doorbell aperture * separately in hw_fini because soc21_enable_doorbell_aperture @@ -890,11 +885,9 @@ static int soc21_common_hw_fini(void *handle) return 0; } -static int soc21_common_suspend(void *handle) +static int soc21_common_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return soc21_common_hw_fini(adev); + return soc21_common_hw_fini(ip_block); } static bool soc21_need_reset_on_resume(struct amdgpu_device *adev) @@ -904,9 +897,10 @@ static bool soc21_need_reset_on_resume(struct amdgpu_device *adev) /* Will reset for the following suspend abort cases. * 1) Only reset dGPU side. * 2) S3 suspend got aborted and TOS is active. + * As for dGPU suspend abort cases the SOL value + * will be kept as zero at this resume point. */ - if (!(adev->flags & AMD_IS_APU) && adev->in_s3 && - !adev->suspend_complete) { + if (!(adev->flags & AMD_IS_APU) && adev->in_s3) { sol_reg1 = RREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_81); msleep(100); sol_reg2 = RREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_81); @@ -917,16 +911,16 @@ static bool soc21_need_reset_on_resume(struct amdgpu_device *adev) return false; } -static int soc21_common_resume(void *handle) +static int soc21_common_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (soc21_need_reset_on_resume(adev)) { dev_info(adev->dev, "S3 suspend aborted, resetting..."); soc21_asic_reset(adev); } - return soc21_common_hw_init(adev); + return soc21_common_hw_init(ip_block); } static bool soc21_common_is_idle(void *handle) @@ -934,16 +928,6 @@ static bool soc21_common_is_idle(void *handle) return true; } -static int soc21_common_wait_for_idle(void *handle) -{ - return 0; -} - -static int soc21_common_soft_reset(void *handle) -{ - return 0; -} - static int soc21_common_set_clockgating_state(void *handle, enum amd_clockgating_state state) { @@ -1002,17 +986,12 @@ static const struct amd_ip_funcs soc21_common_ip_funcs = { .early_init = soc21_common_early_init, .late_init = soc21_common_late_init, .sw_init = soc21_common_sw_init, - .sw_fini = soc21_common_sw_fini, .hw_init = soc21_common_hw_init, .hw_fini = soc21_common_hw_fini, .suspend = soc21_common_suspend, .resume = soc21_common_resume, .is_idle = soc21_common_is_idle, - .wait_for_idle = soc21_common_wait_for_idle, - .soft_reset = soc21_common_soft_reset, .set_clockgating_state = soc21_common_set_clockgating_state, .set_powergating_state = soc21_common_set_powergating_state, .get_clockgating_state = soc21_common_get_clockgating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; diff --git a/drivers/gpu/drm/amd/amdgpu/soc24.c b/drivers/gpu/drm/amd/amdgpu/soc24.c index 29a848f2466b..be96de92b2f5 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc24.c +++ b/drivers/gpu/drm/amd/amdgpu/soc24.c @@ -363,9 +363,9 @@ static const struct amdgpu_asic_funcs soc24_asic_funcs = { .update_umd_stable_pstate = &soc24_update_umd_stable_pstate, }; -static int soc24_common_early_init(void *handle) +static int soc24_common_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->nbio.funcs->set_reg_remap(adev); adev->smc_rreg = NULL; @@ -440,9 +440,9 @@ static int soc24_common_early_init(void *handle) return 0; } -static int soc24_common_late_init(void *handle) +static int soc24_common_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) xgpu_nv_mailbox_get_irq(adev); @@ -455,9 +455,9 @@ static int soc24_common_late_init(void *handle) return 0; } -static int soc24_common_sw_init(void *handle) +static int soc24_common_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) xgpu_nv_mailbox_add_irq_id(adev); @@ -465,14 +465,9 @@ static int soc24_common_sw_init(void *handle) return 0; } -static int soc24_common_sw_fini(void *handle) +static int soc24_common_hw_init(struct amdgpu_ip_block *ip_block) { - return 0; -} - -static int soc24_common_hw_init(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* enable aspm */ soc24_program_aspm(adev); @@ -494,9 +489,9 @@ static int soc24_common_hw_init(void *handle) return 0; } -static int soc24_common_hw_fini(void *handle) +static int soc24_common_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* Disable the doorbell aperture and selfring doorbell aperture * separately in hw_fini because soc21_enable_doorbell_aperture @@ -512,18 +507,14 @@ static int soc24_common_hw_fini(void *handle) return 0; } -static int soc24_common_suspend(void *handle) +static int soc24_common_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return soc24_common_hw_fini(adev); + return soc24_common_hw_fini(ip_block); } -static int soc24_common_resume(void *handle) +static int soc24_common_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return soc24_common_hw_init(adev); + return soc24_common_hw_init(ip_block); } static bool soc24_common_is_idle(void *handle) @@ -531,16 +522,6 @@ static bool soc24_common_is_idle(void *handle) return true; } -static int soc24_common_wait_for_idle(void *handle) -{ - return 0; -} - -static int soc24_common_soft_reset(void *handle) -{ - return 0; -} - static int soc24_common_set_clockgating_state(void *handle, enum amd_clockgating_state state) { @@ -595,14 +576,11 @@ static const struct amd_ip_funcs soc24_common_ip_funcs = { .early_init = soc24_common_early_init, .late_init = soc24_common_late_init, .sw_init = soc24_common_sw_init, - .sw_fini = soc24_common_sw_fini, .hw_init = soc24_common_hw_init, .hw_fini = soc24_common_hw_fini, .suspend = soc24_common_suspend, .resume = soc24_common_resume, .is_idle = soc24_common_is_idle, - .wait_for_idle = soc24_common_wait_for_idle, - .soft_reset = soc24_common_soft_reset, .set_clockgating_state = soc24_common_set_clockgating_state, .set_powergating_state = soc24_common_set_powergating_state, .get_clockgating_state = soc24_common_get_clockgating_state, diff --git a/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h b/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h index 3ac56a9645eb..21b71a427b1f 100644 --- a/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h +++ b/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h @@ -113,6 +113,14 @@ enum ta_ras_address_type { TA_RAS_PA_TO_MCA, }; +enum ta_ras_nps_mode { + TA_RAS_UNKNOWN_MODE = 0, + TA_RAS_NPS1_MODE = 1, + TA_RAS_NPS2_MODE = 2, + TA_RAS_NPS4_MODE = 4, + TA_RAS_NPS8_MODE = 8, +}; + /* Input/output structures for RAS commands */ /**********************************************************/ @@ -139,6 +147,7 @@ struct ta_ras_init_flags { uint8_t dgpu_mode; uint16_t xcc_mask; uint8_t channel_dis_num; + uint8_t nps_mode; }; struct ta_ras_mca_addr { diff --git a/drivers/gpu/drm/amd/amdgpu/tonga_ih.c b/drivers/gpu/drm/amd/amdgpu/tonga_ih.c index 24d49d813607..5a04a6770138 100644 --- a/drivers/gpu/drm/amd/amdgpu/tonga_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/tonga_ih.c @@ -283,9 +283,9 @@ static void tonga_ih_set_rptr(struct amdgpu_device *adev, } } -static int tonga_ih_early_init(void *handle) +static int tonga_ih_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int ret; ret = amdgpu_irq_add_domain(adev); @@ -297,10 +297,10 @@ static int tonga_ih_early_init(void *handle) return 0; } -static int tonga_ih_sw_init(void *handle) +static int tonga_ih_sw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_ih_ring_init(adev, &adev->irq.ih, 64 * 1024, true); if (r) @@ -314,9 +314,9 @@ static int tonga_ih_sw_init(void *handle) return r; } -static int tonga_ih_sw_fini(void *handle) +static int tonga_ih_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_fini_sw(adev); amdgpu_irq_remove_domain(adev); @@ -324,10 +324,10 @@ static int tonga_ih_sw_fini(void *handle) return 0; } -static int tonga_ih_hw_init(void *handle) +static int tonga_ih_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = tonga_ih_irq_init(adev); if (r) @@ -336,27 +336,21 @@ static int tonga_ih_hw_init(void *handle) return 0; } -static int tonga_ih_hw_fini(void *handle) +static int tonga_ih_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - tonga_ih_irq_disable(adev); + tonga_ih_irq_disable(ip_block->adev); return 0; } -static int tonga_ih_suspend(void *handle) +static int tonga_ih_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return tonga_ih_hw_fini(adev); + return tonga_ih_hw_fini(ip_block); } -static int tonga_ih_resume(void *handle) +static int tonga_ih_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return tonga_ih_hw_init(adev); + return tonga_ih_hw_init(ip_block); } static bool tonga_ih_is_idle(void *handle) @@ -370,11 +364,11 @@ static bool tonga_ih_is_idle(void *handle) return true; } -static int tonga_ih_wait_for_idle(void *handle) +static int tonga_ih_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; u32 tmp; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { /* read MC_STATUS */ @@ -386,9 +380,9 @@ static int tonga_ih_wait_for_idle(void *handle) return -ETIMEDOUT; } -static bool tonga_ih_check_soft_reset(void *handle) +static bool tonga_ih_check_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 srbm_soft_reset = 0; u32 tmp = RREG32(mmSRBM_STATUS); @@ -405,29 +399,27 @@ static bool tonga_ih_check_soft_reset(void *handle) } } -static int tonga_ih_pre_soft_reset(void *handle) +static int tonga_ih_pre_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + if (!ip_block->adev->irq.srbm_soft_reset) + return 0; + + return tonga_ih_hw_fini(ip_block); +} + +static int tonga_ih_post_soft_reset(struct amdgpu_ip_block *ip_block) +{ + struct amdgpu_device *adev = ip_block->adev; if (!adev->irq.srbm_soft_reset) return 0; - return tonga_ih_hw_fini(adev); + return tonga_ih_hw_init(ip_block); } -static int tonga_ih_post_soft_reset(void *handle) +static int tonga_ih_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (!adev->irq.srbm_soft_reset) - return 0; - - return tonga_ih_hw_init(adev); -} - -static int tonga_ih_soft_reset(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 srbm_soft_reset; if (!adev->irq.srbm_soft_reset) @@ -471,7 +463,6 @@ static int tonga_ih_set_powergating_state(void *handle, static const struct amd_ip_funcs tonga_ih_ip_funcs = { .name = "tonga_ih", .early_init = tonga_ih_early_init, - .late_init = NULL, .sw_init = tonga_ih_sw_init, .sw_fini = tonga_ih_sw_fini, .hw_init = tonga_ih_hw_init, @@ -486,8 +477,6 @@ static const struct amd_ip_funcs tonga_ih_ip_funcs = { .post_soft_reset = tonga_ih_post_soft_reset, .set_clockgating_state = tonga_ih_set_clockgating_state, .set_powergating_state = tonga_ih_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs tonga_ih_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c b/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c index 805d6662c88b..bdbca25d80c4 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c @@ -531,9 +531,9 @@ static void uvd_v3_1_set_irq_funcs(struct amdgpu_device *adev) } -static int uvd_v3_1_early_init(void *handle) +static int uvd_v3_1_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->uvd.num_uvd_inst = 1; uvd_v3_1_set_ring_funcs(adev); @@ -542,10 +542,10 @@ static int uvd_v3_1_early_init(void *handle) return 0; } -static int uvd_v3_1_sw_init(void *handle) +static int uvd_v3_1_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; void *ptr; uint32_t ucode_len; @@ -580,10 +580,10 @@ static int uvd_v3_1_sw_init(void *handle) return r; } -static int uvd_v3_1_sw_fini(void *handle) +static int uvd_v3_1_sw_fini(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_uvd_suspend(adev); if (r) @@ -621,13 +621,13 @@ static void uvd_v3_1_enable_mgcg(struct amdgpu_device *adev, /** * uvd_v3_1_hw_init - start and test UVD block * - * @handle: handle used to pass amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Initialize the hardware, boot up the VCPU and do some testing */ -static int uvd_v3_1_hw_init(void *handle) +static int uvd_v3_1_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring = &adev->uvd.inst->ring; uint32_t tmp; int r; @@ -688,13 +688,13 @@ done: /** * uvd_v3_1_hw_fini - stop the hardware block * - * @handle: handle used to pass amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the UVD block, mark ring as not ready any more */ -static int uvd_v3_1_hw_fini(void *handle) +static int uvd_v3_1_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->uvd.idle_work); @@ -704,17 +704,17 @@ static int uvd_v3_1_hw_fini(void *handle) return 0; } -static int uvd_v3_1_prepare_suspend(void *handle) +static int uvd_v3_1_prepare_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return amdgpu_uvd_prepare_suspend(adev); } -static int uvd_v3_1_suspend(void *handle) +static int uvd_v3_1_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* * Proper cleanups before halting the HW engine: @@ -740,23 +740,22 @@ static int uvd_v3_1_suspend(void *handle) AMD_CG_STATE_GATE); } - r = uvd_v3_1_hw_fini(adev); + r = uvd_v3_1_hw_fini(ip_block); if (r) return r; return amdgpu_uvd_suspend(adev); } -static int uvd_v3_1_resume(void *handle) +static int uvd_v3_1_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_uvd_resume(adev); + r = amdgpu_uvd_resume(ip_block->adev); if (r) return r; - return uvd_v3_1_hw_init(adev); + return uvd_v3_1_hw_init(ip_block); } static bool uvd_v3_1_is_idle(void *handle) @@ -766,10 +765,10 @@ static bool uvd_v3_1_is_idle(void *handle) return !(RREG32(mmSRBM_STATUS) & SRBM_STATUS__UVD_BUSY_MASK); } -static int uvd_v3_1_wait_for_idle(void *handle) +static int uvd_v3_1_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { if (!(RREG32(mmSRBM_STATUS) & SRBM_STATUS__UVD_BUSY_MASK)) @@ -778,9 +777,9 @@ static int uvd_v3_1_wait_for_idle(void *handle) return -ETIMEDOUT; } -static int uvd_v3_1_soft_reset(void *handle) +static int uvd_v3_1_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uvd_v3_1_stop(adev); @@ -806,7 +805,6 @@ static int uvd_v3_1_set_powergating_state(void *handle, static const struct amd_ip_funcs uvd_v3_1_ip_funcs = { .name = "uvd_v3_1", .early_init = uvd_v3_1_early_init, - .late_init = NULL, .sw_init = uvd_v3_1_sw_init, .sw_fini = uvd_v3_1_sw_fini, .hw_init = uvd_v3_1_hw_init, @@ -819,8 +817,6 @@ static const struct amd_ip_funcs uvd_v3_1_ip_funcs = { .soft_reset = uvd_v3_1_soft_reset, .set_clockgating_state = uvd_v3_1_set_clockgating_state, .set_powergating_state = uvd_v3_1_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; const struct amdgpu_ip_block_version uvd_v3_1_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c index 3f19c606f4de..a836dc9cfcad 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c @@ -90,9 +90,9 @@ static void uvd_v4_2_ring_set_wptr(struct amdgpu_ring *ring) WREG32(mmUVD_RBC_RB_WPTR, lower_32_bits(ring->wptr)); } -static int uvd_v4_2_early_init(void *handle) +static int uvd_v4_2_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->uvd.num_uvd_inst = 1; uvd_v4_2_set_ring_funcs(adev); @@ -101,10 +101,10 @@ static int uvd_v4_2_early_init(void *handle) return 0; } -static int uvd_v4_2_sw_init(void *handle) +static int uvd_v4_2_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; /* UVD TRAP */ @@ -130,10 +130,10 @@ static int uvd_v4_2_sw_init(void *handle) return r; } -static int uvd_v4_2_sw_fini(void *handle) +static int uvd_v4_2_sw_fini(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_uvd_suspend(adev); if (r) @@ -147,13 +147,13 @@ static void uvd_v4_2_enable_mgcg(struct amdgpu_device *adev, /** * uvd_v4_2_hw_init - start and test UVD block * - * @handle: handle used to pass amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Initialize the hardware, boot up the VCPU and do some testing */ -static int uvd_v4_2_hw_init(void *handle) +static int uvd_v4_2_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring = &adev->uvd.inst->ring; uint32_t tmp; int r; @@ -202,13 +202,13 @@ done: /** * uvd_v4_2_hw_fini - stop the hardware block * - * @handle: handle used to pass amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the UVD block, mark ring as not ready any more */ -static int uvd_v4_2_hw_fini(void *handle) +static int uvd_v4_2_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->uvd.idle_work); @@ -218,17 +218,17 @@ static int uvd_v4_2_hw_fini(void *handle) return 0; } -static int uvd_v4_2_prepare_suspend(void *handle) +static int uvd_v4_2_prepare_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return amdgpu_uvd_prepare_suspend(adev); } -static int uvd_v4_2_suspend(void *handle) +static int uvd_v4_2_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* * Proper cleanups before halting the HW engine: @@ -254,23 +254,22 @@ static int uvd_v4_2_suspend(void *handle) AMD_CG_STATE_GATE); } - r = uvd_v4_2_hw_fini(adev); + r = uvd_v4_2_hw_fini(ip_block); if (r) return r; return amdgpu_uvd_suspend(adev); } -static int uvd_v4_2_resume(void *handle) +static int uvd_v4_2_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_uvd_resume(adev); + r = amdgpu_uvd_resume(ip_block->adev); if (r) return r; - return uvd_v4_2_hw_init(adev); + return uvd_v4_2_hw_init(ip_block); } /** @@ -666,10 +665,10 @@ static bool uvd_v4_2_is_idle(void *handle) return !(RREG32(mmSRBM_STATUS) & SRBM_STATUS__UVD_BUSY_MASK); } -static int uvd_v4_2_wait_for_idle(void *handle) +static int uvd_v4_2_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { if (!(RREG32(mmSRBM_STATUS) & SRBM_STATUS__UVD_BUSY_MASK)) @@ -678,9 +677,9 @@ static int uvd_v4_2_wait_for_idle(void *handle) return -ETIMEDOUT; } -static int uvd_v4_2_soft_reset(void *handle) +static int uvd_v4_2_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uvd_v4_2_stop(adev); @@ -756,7 +755,6 @@ static int uvd_v4_2_set_powergating_state(void *handle, static const struct amd_ip_funcs uvd_v4_2_ip_funcs = { .name = "uvd_v4_2", .early_init = uvd_v4_2_early_init, - .late_init = NULL, .sw_init = uvd_v4_2_sw_init, .sw_fini = uvd_v4_2_sw_fini, .hw_init = uvd_v4_2_hw_init, @@ -769,8 +767,6 @@ static const struct amd_ip_funcs uvd_v4_2_ip_funcs = { .soft_reset = uvd_v4_2_soft_reset, .set_clockgating_state = uvd_v4_2_set_clockgating_state, .set_powergating_state = uvd_v4_2_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs uvd_v4_2_ring_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c index efd903c21d48..ab55fae3569e 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c @@ -88,9 +88,9 @@ static void uvd_v5_0_ring_set_wptr(struct amdgpu_ring *ring) WREG32(mmUVD_RBC_RB_WPTR, lower_32_bits(ring->wptr)); } -static int uvd_v5_0_early_init(void *handle) +static int uvd_v5_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->uvd.num_uvd_inst = 1; uvd_v5_0_set_ring_funcs(adev); @@ -99,10 +99,10 @@ static int uvd_v5_0_early_init(void *handle) return 0; } -static int uvd_v5_0_sw_init(void *handle) +static int uvd_v5_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; /* UVD TRAP */ @@ -128,10 +128,10 @@ static int uvd_v5_0_sw_init(void *handle) return r; } -static int uvd_v5_0_sw_fini(void *handle) +static int uvd_v5_0_sw_fini(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_uvd_suspend(adev); if (r) @@ -143,13 +143,13 @@ static int uvd_v5_0_sw_fini(void *handle) /** * uvd_v5_0_hw_init - start and test UVD block * - * @handle: handle used to pass amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Initialize the hardware, boot up the VCPU and do some testing */ -static int uvd_v5_0_hw_init(void *handle) +static int uvd_v5_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring = &adev->uvd.inst->ring; uint32_t tmp; int r; @@ -200,13 +200,13 @@ done: /** * uvd_v5_0_hw_fini - stop the hardware block * - * @handle: handle used to pass amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the UVD block, mark ring as not ready any more */ -static int uvd_v5_0_hw_fini(void *handle) +static int uvd_v5_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->uvd.idle_work); @@ -216,17 +216,17 @@ static int uvd_v5_0_hw_fini(void *handle) return 0; } -static int uvd_v5_0_prepare_suspend(void *handle) +static int uvd_v5_0_prepare_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return amdgpu_uvd_prepare_suspend(adev); } -static int uvd_v5_0_suspend(void *handle) +static int uvd_v5_0_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* * Proper cleanups before halting the HW engine: @@ -252,23 +252,22 @@ static int uvd_v5_0_suspend(void *handle) AMD_CG_STATE_GATE); } - r = uvd_v5_0_hw_fini(adev); + r = uvd_v5_0_hw_fini(ip_block); if (r) return r; return amdgpu_uvd_suspend(adev); } -static int uvd_v5_0_resume(void *handle) +static int uvd_v5_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_uvd_resume(adev); + r = amdgpu_uvd_resume(ip_block->adev); if (r) return r; - return uvd_v5_0_hw_init(adev); + return uvd_v5_0_hw_init(ip_block); } /** @@ -588,10 +587,10 @@ static bool uvd_v5_0_is_idle(void *handle) return !(RREG32(mmSRBM_STATUS) & SRBM_STATUS__UVD_BUSY_MASK); } -static int uvd_v5_0_wait_for_idle(void *handle) +static int uvd_v5_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { if (!(RREG32(mmSRBM_STATUS) & SRBM_STATUS__UVD_BUSY_MASK)) @@ -600,9 +599,9 @@ static int uvd_v5_0_wait_for_idle(void *handle) return -ETIMEDOUT; } -static int uvd_v5_0_soft_reset(void *handle) +static int uvd_v5_0_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; uvd_v5_0_stop(adev); @@ -796,10 +795,15 @@ static int uvd_v5_0_set_clockgating_state(void *handle, { struct amdgpu_device *adev = (struct amdgpu_device *)handle; bool enable = (state == AMD_CG_STATE_GATE); + struct amdgpu_ip_block *ip_block; + + ip_block = amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_UVD); + if (!ip_block) + return -EINVAL; if (enable) { /* wait for STATUS to clear */ - if (uvd_v5_0_wait_for_idle(handle)) + if (uvd_v5_0_wait_for_idle(ip_block)) return -EBUSY; uvd_v5_0_enable_clock_gating(adev, true); @@ -863,7 +867,6 @@ out: static const struct amd_ip_funcs uvd_v5_0_ip_funcs = { .name = "uvd_v5_0", .early_init = uvd_v5_0_early_init, - .late_init = NULL, .sw_init = uvd_v5_0_sw_init, .sw_fini = uvd_v5_0_sw_fini, .hw_init = uvd_v5_0_hw_init, @@ -877,8 +880,6 @@ static const struct amd_ip_funcs uvd_v5_0_ip_funcs = { .set_clockgating_state = uvd_v5_0_set_clockgating_state, .set_powergating_state = uvd_v5_0_set_powergating_state, .get_clockgating_state = uvd_v5_0_get_clockgating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs uvd_v5_0_ring_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c index 495de5068455..39f8c3d3a135 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c @@ -354,9 +354,9 @@ error: return r; } -static int uvd_v6_0_early_init(void *handle) +static int uvd_v6_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->uvd.num_uvd_inst = 1; if (!(adev->flags & AMD_IS_APU) && @@ -375,11 +375,11 @@ static int uvd_v6_0_early_init(void *handle) return 0; } -static int uvd_v6_0_sw_init(void *handle) +static int uvd_v6_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int i, r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* UVD TRAP */ r = amdgpu_irq_add_id(adev, AMDGPU_IRQ_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_UVD_SYSTEM_MESSAGE, &adev->uvd.inst->irq); @@ -435,10 +435,10 @@ static int uvd_v6_0_sw_init(void *handle) return r; } -static int uvd_v6_0_sw_fini(void *handle) +static int uvd_v6_0_sw_fini(struct amdgpu_ip_block *ip_block) { int i, r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_uvd_suspend(adev); if (r) @@ -455,13 +455,13 @@ static int uvd_v6_0_sw_fini(void *handle) /** * uvd_v6_0_hw_init - start and test UVD block * - * @handle: handle used to pass amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Initialize the hardware, boot up the VCPU and do some testing */ -static int uvd_v6_0_hw_init(void *handle) +static int uvd_v6_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring = &adev->uvd.inst->ring; uint32_t tmp; int i, r; @@ -524,13 +524,13 @@ done: /** * uvd_v6_0_hw_fini - stop the hardware block * - * @handle: handle used to pass amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the UVD block, mark ring as not ready any more */ -static int uvd_v6_0_hw_fini(void *handle) +static int uvd_v6_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->uvd.idle_work); @@ -540,17 +540,17 @@ static int uvd_v6_0_hw_fini(void *handle) return 0; } -static int uvd_v6_0_prepare_suspend(void *handle) +static int uvd_v6_0_prepare_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return amdgpu_uvd_prepare_suspend(adev); } -static int uvd_v6_0_suspend(void *handle) +static int uvd_v6_0_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* * Proper cleanups before halting the HW engine: @@ -576,23 +576,22 @@ static int uvd_v6_0_suspend(void *handle) AMD_CG_STATE_GATE); } - r = uvd_v6_0_hw_fini(adev); + r = uvd_v6_0_hw_fini(ip_block); if (r) return r; return amdgpu_uvd_suspend(adev); } -static int uvd_v6_0_resume(void *handle) +static int uvd_v6_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_uvd_resume(adev); + r = amdgpu_uvd_resume(ip_block->adev); if (r) return r; - return uvd_v6_0_hw_init(adev); + return uvd_v6_0_hw_init(ip_block); } /** @@ -1151,22 +1150,22 @@ static bool uvd_v6_0_is_idle(void *handle) return !(RREG32(mmSRBM_STATUS) & SRBM_STATUS__UVD_BUSY_MASK); } -static int uvd_v6_0_wait_for_idle(void *handle) +static int uvd_v6_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) { - if (uvd_v6_0_is_idle(handle)) + if (uvd_v6_0_is_idle(adev)) return 0; } return -ETIMEDOUT; } #define AMDGPU_UVD_STATUS_BUSY_MASK 0xfd -static bool uvd_v6_0_check_soft_reset(void *handle) +static bool uvd_v6_0_check_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 srbm_soft_reset = 0; u32 tmp = RREG32(mmSRBM_STATUS); @@ -1184,9 +1183,9 @@ static bool uvd_v6_0_check_soft_reset(void *handle) } } -static int uvd_v6_0_pre_soft_reset(void *handle) +static int uvd_v6_0_pre_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (!adev->uvd.inst->srbm_soft_reset) return 0; @@ -1195,9 +1194,9 @@ static int uvd_v6_0_pre_soft_reset(void *handle) return 0; } -static int uvd_v6_0_soft_reset(void *handle) +static int uvd_v6_0_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 srbm_soft_reset; if (!adev->uvd.inst->srbm_soft_reset) @@ -1226,9 +1225,9 @@ static int uvd_v6_0_soft_reset(void *handle) return 0; } -static int uvd_v6_0_post_soft_reset(void *handle) +static int uvd_v6_0_post_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (!adev->uvd.inst->srbm_soft_reset) return 0; @@ -1455,11 +1454,16 @@ static int uvd_v6_0_set_clockgating_state(void *handle, enum amd_clockgating_state state) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_ip_block *ip_block; bool enable = (state == AMD_CG_STATE_GATE); + ip_block = amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_UVD); + if (!ip_block) + return -EINVAL; + if (enable) { /* wait for STATUS to clear */ - if (uvd_v6_0_wait_for_idle(handle)) + if (uvd_v6_0_wait_for_idle(ip_block)) return -EBUSY; uvd_v6_0_enable_clock_gating(adev, true); /* enable HW gates because UVD is idle */ @@ -1528,7 +1532,6 @@ out: static const struct amd_ip_funcs uvd_v6_0_ip_funcs = { .name = "uvd_v6_0", .early_init = uvd_v6_0_early_init, - .late_init = NULL, .sw_init = uvd_v6_0_sw_init, .sw_fini = uvd_v6_0_sw_fini, .hw_init = uvd_v6_0_hw_init, @@ -1545,8 +1548,6 @@ static const struct amd_ip_funcs uvd_v6_0_ip_funcs = { .set_clockgating_state = uvd_v6_0_set_clockgating_state, .set_powergating_state = uvd_v6_0_set_powergating_state, .get_clockgating_state = uvd_v6_0_get_clockgating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs uvd_v6_0_ring_phys_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c index 6068b784dc69..079131aeb2f7 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c @@ -361,9 +361,9 @@ error: return r; } -static int uvd_v7_0_early_init(void *handle) +static int uvd_v7_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->asic_type == CHIP_VEGA20) { u32 harvest; @@ -395,12 +395,12 @@ static int uvd_v7_0_early_init(void *handle) return 0; } -static int uvd_v7_0_sw_init(void *handle) +static int uvd_v7_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int i, j, r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (j = 0; j < adev->uvd.num_uvd_inst; j++) { if (adev->uvd.harvest_config & (1 << j)) @@ -487,10 +487,10 @@ static int uvd_v7_0_sw_init(void *handle) return r; } -static int uvd_v7_0_sw_fini(void *handle) +static int uvd_v7_0_sw_fini(struct amdgpu_ip_block *ip_block) { int i, j, r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_virt_free_mm_table(adev); @@ -510,13 +510,13 @@ static int uvd_v7_0_sw_fini(void *handle) /** * uvd_v7_0_hw_init - start and test UVD block * - * @handle: handle used to pass amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Initialize the hardware, boot up the VCPU and do some testing */ -static int uvd_v7_0_hw_init(void *handle) +static int uvd_v7_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; uint32_t tmp; int i, j, r; @@ -588,13 +588,13 @@ done: /** * uvd_v7_0_hw_fini - stop the hardware block * - * @handle: handle used to pass amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the UVD block, mark ring as not ready any more */ -static int uvd_v7_0_hw_fini(void *handle) +static int uvd_v7_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->uvd.idle_work); @@ -608,17 +608,17 @@ static int uvd_v7_0_hw_fini(void *handle) return 0; } -static int uvd_v7_0_prepare_suspend(void *handle) +static int uvd_v7_0_prepare_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; return amdgpu_uvd_prepare_suspend(adev); } -static int uvd_v7_0_suspend(void *handle) +static int uvd_v7_0_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* * Proper cleanups before halting the HW engine: @@ -644,23 +644,22 @@ static int uvd_v7_0_suspend(void *handle) AMD_CG_STATE_GATE); } - r = uvd_v7_0_hw_fini(adev); + r = uvd_v7_0_hw_fini(ip_block); if (r) return r; return amdgpu_uvd_suspend(adev); } -static int uvd_v7_0_resume(void *handle) +static int uvd_v7_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_uvd_resume(adev); + r = amdgpu_uvd_resume(ip_block->adev); if (r) return r; - return uvd_v7_0_hw_init(adev); + return uvd_v7_0_hw_init(ip_block); } /** @@ -1463,104 +1462,6 @@ static void uvd_v7_0_enc_ring_emit_wreg(struct amdgpu_ring *ring, amdgpu_ring_write(ring, val); } -#if 0 -static bool uvd_v7_0_is_idle(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return !(RREG32(mmSRBM_STATUS) & SRBM_STATUS__UVD_BUSY_MASK); -} - -static int uvd_v7_0_wait_for_idle(void *handle) -{ - unsigned i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - for (i = 0; i < adev->usec_timeout; i++) { - if (uvd_v7_0_is_idle(handle)) - return 0; - } - return -ETIMEDOUT; -} - -#define AMDGPU_UVD_STATUS_BUSY_MASK 0xfd -static bool uvd_v7_0_check_soft_reset(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - u32 srbm_soft_reset = 0; - u32 tmp = RREG32(mmSRBM_STATUS); - - if (REG_GET_FIELD(tmp, SRBM_STATUS, UVD_RQ_PENDING) || - REG_GET_FIELD(tmp, SRBM_STATUS, UVD_BUSY) || - (RREG32_SOC15(UVD, ring->me, mmUVD_STATUS) & - AMDGPU_UVD_STATUS_BUSY_MASK)) - srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, - SRBM_SOFT_RESET, SOFT_RESET_UVD, 1); - - if (srbm_soft_reset) { - adev->uvd.inst[ring->me].srbm_soft_reset = srbm_soft_reset; - return true; - } else { - adev->uvd.inst[ring->me].srbm_soft_reset = 0; - return false; - } -} - -static int uvd_v7_0_pre_soft_reset(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (!adev->uvd.inst[ring->me].srbm_soft_reset) - return 0; - - uvd_v7_0_stop(adev); - return 0; -} - -static int uvd_v7_0_soft_reset(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - u32 srbm_soft_reset; - - if (!adev->uvd.inst[ring->me].srbm_soft_reset) - return 0; - srbm_soft_reset = adev->uvd.inst[ring->me].srbm_soft_reset; - - if (srbm_soft_reset) { - u32 tmp; - - tmp = RREG32(mmSRBM_SOFT_RESET); - tmp |= srbm_soft_reset; - dev_info(adev->dev, "SRBM_SOFT_RESET=0x%08X\n", tmp); - WREG32(mmSRBM_SOFT_RESET, tmp); - tmp = RREG32(mmSRBM_SOFT_RESET); - - udelay(50); - - tmp &= ~srbm_soft_reset; - WREG32(mmSRBM_SOFT_RESET, tmp); - tmp = RREG32(mmSRBM_SOFT_RESET); - - /* Wait a little for things to settle down */ - udelay(50); - } - - return 0; -} - -static int uvd_v7_0_post_soft_reset(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (!adev->uvd.inst[ring->me].srbm_soft_reset) - return 0; - - mdelay(5); - - return uvd_v7_0_start(adev); -} -#endif - static int uvd_v7_0_set_interrupt_state(struct amdgpu_device *adev, struct amdgpu_irq_src *source, unsigned type, @@ -1610,171 +1511,6 @@ static int uvd_v7_0_process_interrupt(struct amdgpu_device *adev, return 0; } -#if 0 -static void uvd_v7_0_set_sw_clock_gating(struct amdgpu_device *adev) -{ - uint32_t data, data1, data2, suvd_flags; - - data = RREG32_SOC15(UVD, ring->me, mmUVD_CGC_CTRL); - data1 = RREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_GATE); - data2 = RREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_CTRL); - - data &= ~(UVD_CGC_CTRL__CLK_OFF_DELAY_MASK | - UVD_CGC_CTRL__CLK_GATE_DLY_TIMER_MASK); - - suvd_flags = UVD_SUVD_CGC_GATE__SRE_MASK | - UVD_SUVD_CGC_GATE__SIT_MASK | - UVD_SUVD_CGC_GATE__SMP_MASK | - UVD_SUVD_CGC_GATE__SCM_MASK | - UVD_SUVD_CGC_GATE__SDB_MASK; - - data |= UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK | - (1 << REG_FIELD_SHIFT(UVD_CGC_CTRL, CLK_GATE_DLY_TIMER)) | - (4 << REG_FIELD_SHIFT(UVD_CGC_CTRL, CLK_OFF_DELAY)); - - data &= ~(UVD_CGC_CTRL__UDEC_RE_MODE_MASK | - UVD_CGC_CTRL__UDEC_CM_MODE_MASK | - UVD_CGC_CTRL__UDEC_IT_MODE_MASK | - UVD_CGC_CTRL__UDEC_DB_MODE_MASK | - UVD_CGC_CTRL__UDEC_MP_MODE_MASK | - UVD_CGC_CTRL__SYS_MODE_MASK | - UVD_CGC_CTRL__UDEC_MODE_MASK | - UVD_CGC_CTRL__MPEG2_MODE_MASK | - UVD_CGC_CTRL__REGS_MODE_MASK | - UVD_CGC_CTRL__RBC_MODE_MASK | - UVD_CGC_CTRL__LMI_MC_MODE_MASK | - UVD_CGC_CTRL__LMI_UMC_MODE_MASK | - UVD_CGC_CTRL__IDCT_MODE_MASK | - UVD_CGC_CTRL__MPRD_MODE_MASK | - UVD_CGC_CTRL__MPC_MODE_MASK | - UVD_CGC_CTRL__LBSI_MODE_MASK | - UVD_CGC_CTRL__LRBBM_MODE_MASK | - UVD_CGC_CTRL__WCB_MODE_MASK | - UVD_CGC_CTRL__VCPU_MODE_MASK | - UVD_CGC_CTRL__JPEG_MODE_MASK | - UVD_CGC_CTRL__JPEG2_MODE_MASK | - UVD_CGC_CTRL__SCPU_MODE_MASK); - data2 &= ~(UVD_SUVD_CGC_CTRL__SRE_MODE_MASK | - UVD_SUVD_CGC_CTRL__SIT_MODE_MASK | - UVD_SUVD_CGC_CTRL__SMP_MODE_MASK | - UVD_SUVD_CGC_CTRL__SCM_MODE_MASK | - UVD_SUVD_CGC_CTRL__SDB_MODE_MASK); - data1 |= suvd_flags; - - WREG32_SOC15(UVD, ring->me, mmUVD_CGC_CTRL, data); - WREG32_SOC15(UVD, ring->me, mmUVD_CGC_GATE, 0); - WREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_GATE, data1); - WREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_CTRL, data2); -} - -static void uvd_v7_0_set_hw_clock_gating(struct amdgpu_device *adev) -{ - uint32_t data, data1, cgc_flags, suvd_flags; - - data = RREG32_SOC15(UVD, ring->me, mmUVD_CGC_GATE); - data1 = RREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_GATE); - - cgc_flags = UVD_CGC_GATE__SYS_MASK | - UVD_CGC_GATE__UDEC_MASK | - UVD_CGC_GATE__MPEG2_MASK | - UVD_CGC_GATE__RBC_MASK | - UVD_CGC_GATE__LMI_MC_MASK | - UVD_CGC_GATE__IDCT_MASK | - UVD_CGC_GATE__MPRD_MASK | - UVD_CGC_GATE__MPC_MASK | - UVD_CGC_GATE__LBSI_MASK | - UVD_CGC_GATE__LRBBM_MASK | - UVD_CGC_GATE__UDEC_RE_MASK | - UVD_CGC_GATE__UDEC_CM_MASK | - UVD_CGC_GATE__UDEC_IT_MASK | - UVD_CGC_GATE__UDEC_DB_MASK | - UVD_CGC_GATE__UDEC_MP_MASK | - UVD_CGC_GATE__WCB_MASK | - UVD_CGC_GATE__VCPU_MASK | - UVD_CGC_GATE__SCPU_MASK | - UVD_CGC_GATE__JPEG_MASK | - UVD_CGC_GATE__JPEG2_MASK; - - suvd_flags = UVD_SUVD_CGC_GATE__SRE_MASK | - UVD_SUVD_CGC_GATE__SIT_MASK | - UVD_SUVD_CGC_GATE__SMP_MASK | - UVD_SUVD_CGC_GATE__SCM_MASK | - UVD_SUVD_CGC_GATE__SDB_MASK; - - data |= cgc_flags; - data1 |= suvd_flags; - - WREG32_SOC15(UVD, ring->me, mmUVD_CGC_GATE, data); - WREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_GATE, data1); -} - -static void uvd_v7_0_set_bypass_mode(struct amdgpu_device *adev, bool enable) -{ - u32 tmp = RREG32_SMC(ixGCK_DFS_BYPASS_CNTL); - - if (enable) - tmp |= (GCK_DFS_BYPASS_CNTL__BYPASSDCLK_MASK | - GCK_DFS_BYPASS_CNTL__BYPASSVCLK_MASK); - else - tmp &= ~(GCK_DFS_BYPASS_CNTL__BYPASSDCLK_MASK | - GCK_DFS_BYPASS_CNTL__BYPASSVCLK_MASK); - - WREG32_SMC(ixGCK_DFS_BYPASS_CNTL, tmp); -} - - -static int uvd_v7_0_set_clockgating_state(void *handle, - enum amd_clockgating_state state) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - bool enable = (state == AMD_CG_STATE_GATE); - - uvd_v7_0_set_bypass_mode(adev, enable); - - if (!(adev->cg_flags & AMD_CG_SUPPORT_UVD_MGCG)) - return 0; - - if (enable) { - /* disable HW gating and enable Sw gating */ - uvd_v7_0_set_sw_clock_gating(adev); - } else { - /* wait for STATUS to clear */ - if (uvd_v7_0_wait_for_idle(handle)) - return -EBUSY; - - /* enable HW gates because UVD is idle */ - /* uvd_v7_0_set_hw_clock_gating(adev); */ - } - - return 0; -} - -static int uvd_v7_0_set_powergating_state(void *handle, - enum amd_powergating_state state) -{ - /* This doesn't actually powergate the UVD block. - * That's done in the dpm code via the SMC. This - * just re-inits the block as necessary. The actual - * gating still happens in the dpm code. We should - * revisit this when there is a cleaner line between - * the smc and the hw blocks - */ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (!(adev->pg_flags & AMD_PG_SUPPORT_UVD)) - return 0; - - WREG32_SOC15(UVD, ring->me, mmUVD_POWER_STATUS, UVD_POWER_STATUS__UVD_PG_EN_MASK); - - if (state == AMD_PG_STATE_GATE) { - uvd_v7_0_stop(adev); - return 0; - } else { - return uvd_v7_0_start(adev); - } -} -#endif - static int uvd_v7_0_set_clockgating_state(void *handle, enum amd_clockgating_state state) { @@ -1785,7 +1521,6 @@ static int uvd_v7_0_set_clockgating_state(void *handle, const struct amd_ip_funcs uvd_v7_0_ip_funcs = { .name = "uvd_v7_0", .early_init = uvd_v7_0_early_init, - .late_init = NULL, .sw_init = uvd_v7_0_sw_init, .sw_fini = uvd_v7_0_sw_fini, .hw_init = uvd_v7_0_hw_init, @@ -1793,12 +1528,6 @@ const struct amd_ip_funcs uvd_v7_0_ip_funcs = { .prepare_suspend = uvd_v7_0_prepare_suspend, .suspend = uvd_v7_0_suspend, .resume = uvd_v7_0_resume, - .is_idle = NULL /* uvd_v7_0_is_idle */, - .wait_for_idle = NULL /* uvd_v7_0_wait_for_idle */, - .check_soft_reset = NULL /* uvd_v7_0_check_soft_reset */, - .pre_soft_reset = NULL /* uvd_v7_0_pre_soft_reset */, - .soft_reset = NULL /* uvd_v7_0_soft_reset */, - .post_soft_reset = NULL /* uvd_v7_0_post_soft_reset */, .set_clockgating_state = uvd_v7_0_set_clockgating_state, .set_powergating_state = NULL /* uvd_v7_0_set_powergating_state */, }; diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c index 66fada199bda..c1ed91b39415 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c @@ -208,13 +208,13 @@ static bool vce_v2_0_is_idle(void *handle) return !(RREG32(mmSRBM_STATUS2) & SRBM_STATUS2__VCE_BUSY_MASK); } -static int vce_v2_0_wait_for_idle(void *handle) +static int vce_v2_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; unsigned i; for (i = 0; i < adev->usec_timeout; i++) { - if (vce_v2_0_is_idle(handle)) + if (vce_v2_0_is_idle(adev)) return 0; } return -ETIMEDOUT; @@ -274,15 +274,21 @@ static int vce_v2_0_start(struct amdgpu_device *adev) static int vce_v2_0_stop(struct amdgpu_device *adev) { + struct amdgpu_ip_block *ip_block; int i; int status; + if (vce_v2_0_lmi_clean(adev)) { DRM_INFO("vce is not idle \n"); return 0; } - if (vce_v2_0_wait_for_idle(adev)) { + ip_block = amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_VCN); + if (!ip_block) + return -EINVAL; + + if (vce_v2_0_wait_for_idle(ip_block)) { DRM_INFO("VCE is busy, Can't set clock gating"); return 0; } @@ -398,9 +404,9 @@ static void vce_v2_0_enable_mgcg(struct amdgpu_device *adev, bool enable, } } -static int vce_v2_0_early_init(void *handle) +static int vce_v2_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->vce.num_rings = 2; @@ -410,11 +416,11 @@ static int vce_v2_0_early_init(void *handle) return 0; } -static int vce_v2_0_sw_init(void *handle) +static int vce_v2_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* VCE */ r = amdgpu_irq_add_id(adev, AMDGPU_IRQ_CLIENTID_LEGACY, 167, &adev->vce.irq); @@ -444,10 +450,10 @@ static int vce_v2_0_sw_init(void *handle) return r; } -static int vce_v2_0_sw_fini(void *handle) +static int vce_v2_0_sw_fini(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_vce_suspend(adev); if (r) @@ -456,10 +462,10 @@ static int vce_v2_0_sw_fini(void *handle) return amdgpu_vce_sw_fini(adev); } -static int vce_v2_0_hw_init(void *handle) +static int vce_v2_0_hw_init(struct amdgpu_ip_block *ip_block) { int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_asic_set_vce_clocks(adev, 10000, 10000); vce_v2_0_enable_mgcg(adev, true, false); @@ -475,19 +481,17 @@ static int vce_v2_0_hw_init(void *handle) return 0; } -static int vce_v2_0_hw_fini(void *handle) +static int vce_v2_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - cancel_delayed_work_sync(&adev->vce.idle_work); + cancel_delayed_work_sync(&ip_block->adev->vce.idle_work); return 0; } -static int vce_v2_0_suspend(void *handle) +static int vce_v2_0_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* @@ -513,28 +517,27 @@ static int vce_v2_0_suspend(void *handle) AMD_CG_STATE_GATE); } - r = vce_v2_0_hw_fini(adev); + r = vce_v2_0_hw_fini(ip_block); if (r) return r; return amdgpu_vce_suspend(adev); } -static int vce_v2_0_resume(void *handle) +static int vce_v2_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_vce_resume(adev); + r = amdgpu_vce_resume(ip_block->adev); if (r) return r; - return vce_v2_0_hw_init(adev); + return vce_v2_0_hw_init(ip_block); } -static int vce_v2_0_soft_reset(void *handle) +static int vce_v2_0_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; WREG32_FIELD(SRBM_SOFT_RESET, SOFT_RESET_VCE, 1); mdelay(5); @@ -614,7 +617,6 @@ static int vce_v2_0_set_powergating_state(void *handle, static const struct amd_ip_funcs vce_v2_0_ip_funcs = { .name = "vce_v2_0", .early_init = vce_v2_0_early_init, - .late_init = NULL, .sw_init = vce_v2_0_sw_init, .sw_fini = vce_v2_0_sw_fini, .hw_init = vce_v2_0_hw_init, @@ -626,8 +628,6 @@ static const struct amd_ip_funcs vce_v2_0_ip_funcs = { .soft_reset = vce_v2_0_soft_reset, .set_clockgating_state = vce_v2_0_set_clockgating_state, .set_powergating_state = vce_v2_0_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs vce_v2_0_ring_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c index 4bfba2931b08..6bb318a06f19 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c @@ -64,7 +64,7 @@ static void vce_v3_0_mc_resume(struct amdgpu_device *adev, int idx); static void vce_v3_0_set_ring_funcs(struct amdgpu_device *adev); static void vce_v3_0_set_irq_funcs(struct amdgpu_device *adev); -static int vce_v3_0_wait_for_idle(void *handle); +static int vce_v3_0_wait_for_idle(struct amdgpu_ip_block *ip_block); static int vce_v3_0_set_clockgating_state(void *handle, enum amd_clockgating_state state); /** @@ -396,9 +396,9 @@ static unsigned vce_v3_0_get_harvest_config(struct amdgpu_device *adev) } } -static int vce_v3_0_early_init(void *handle) +static int vce_v3_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->vce.harvest_config = vce_v3_0_get_harvest_config(adev); @@ -415,9 +415,9 @@ static int vce_v3_0_early_init(void *handle) return 0; } -static int vce_v3_0_sw_init(void *handle) +static int vce_v3_0_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int r, i; @@ -453,10 +453,10 @@ static int vce_v3_0_sw_init(void *handle) return r; } -static int vce_v3_0_sw_fini(void *handle) +static int vce_v3_0_sw_fini(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_vce_suspend(adev); if (r) @@ -465,10 +465,10 @@ static int vce_v3_0_sw_fini(void *handle) return amdgpu_vce_sw_fini(adev); } -static int vce_v3_0_hw_init(void *handle) +static int vce_v3_0_hw_init(struct amdgpu_ip_block *ip_block) { int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; vce_v3_0_override_vce_clock_gating(adev, true); @@ -485,14 +485,14 @@ static int vce_v3_0_hw_init(void *handle) return 0; } -static int vce_v3_0_hw_fini(void *handle) +static int vce_v3_0_hw_fini(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->vce.idle_work); - r = vce_v3_0_wait_for_idle(handle); + r = vce_v3_0_wait_for_idle(ip_block); if (r) return r; @@ -500,10 +500,10 @@ static int vce_v3_0_hw_fini(void *handle) return vce_v3_0_set_clockgating_state(adev, AMD_CG_STATE_GATE); } -static int vce_v3_0_suspend(void *handle) +static int vce_v3_0_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* * Proper cleanups before halting the HW engine: @@ -528,23 +528,22 @@ static int vce_v3_0_suspend(void *handle) AMD_CG_STATE_GATE); } - r = vce_v3_0_hw_fini(adev); + r = vce_v3_0_hw_fini(ip_block); if (r) return r; return amdgpu_vce_suspend(adev); } -static int vce_v3_0_resume(void *handle) +static int vce_v3_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_vce_resume(adev); + r = amdgpu_vce_resume(ip_block->adev); if (r) return r; - return vce_v3_0_hw_init(adev); + return vce_v3_0_hw_init(ip_block); } static void vce_v3_0_mc_resume(struct amdgpu_device *adev, int idx) @@ -609,13 +608,13 @@ static bool vce_v3_0_is_idle(void *handle) return !(RREG32(mmSRBM_STATUS2) & mask); } -static int vce_v3_0_wait_for_idle(void *handle) +static int vce_v3_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { unsigned i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (i = 0; i < adev->usec_timeout; i++) - if (vce_v3_0_is_idle(handle)) + if (vce_v3_0_is_idle(adev)) return 0; return -ETIMEDOUT; @@ -627,9 +626,9 @@ static int vce_v3_0_wait_for_idle(void *handle) #define AMDGPU_VCE_STATUS_BUSY_MASK (VCE_STATUS_VCPU_REPORT_AUTO_BUSY_MASK | \ VCE_STATUS_VCPU_REPORT_RB0_BUSY_MASK) -static bool vce_v3_0_check_soft_reset(void *handle) +static bool vce_v3_0_check_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 srbm_soft_reset = 0; /* According to VCE team , we should use VCE_STATUS instead @@ -668,9 +667,9 @@ static bool vce_v3_0_check_soft_reset(void *handle) } } -static int vce_v3_0_soft_reset(void *handle) +static int vce_v3_0_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; u32 srbm_soft_reset; if (!adev->vce.srbm_soft_reset) @@ -699,29 +698,29 @@ static int vce_v3_0_soft_reset(void *handle) return 0; } -static int vce_v3_0_pre_soft_reset(void *handle) +static int vce_v3_0_pre_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (!adev->vce.srbm_soft_reset) return 0; mdelay(5); - return vce_v3_0_suspend(adev); + return vce_v3_0_suspend(ip_block); } -static int vce_v3_0_post_soft_reset(void *handle) +static int vce_v3_0_post_soft_reset(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (!adev->vce.srbm_soft_reset) return 0; mdelay(5); - return vce_v3_0_resume(adev); + return vce_v3_0_resume(ip_block); } static int vce_v3_0_set_interrupt_state(struct amdgpu_device *adev, @@ -897,7 +896,6 @@ static void vce_v3_0_emit_pipeline_sync(struct amdgpu_ring *ring) static const struct amd_ip_funcs vce_v3_0_ip_funcs = { .name = "vce_v3_0", .early_init = vce_v3_0_early_init, - .late_init = NULL, .sw_init = vce_v3_0_sw_init, .sw_fini = vce_v3_0_sw_fini, .hw_init = vce_v3_0_hw_init, @@ -913,8 +911,6 @@ static const struct amd_ip_funcs vce_v3_0_ip_funcs = { .set_clockgating_state = vce_v3_0_set_clockgating_state, .set_powergating_state = vce_v3_0_set_powergating_state, .get_clockgating_state = vce_v3_0_get_clockgating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs vce_v3_0_ring_phys_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c index 0748bf44c880..79ee555768a5 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c @@ -407,9 +407,9 @@ static int vce_v4_0_stop(struct amdgpu_device *adev) return 0; } -static int vce_v4_0_early_init(void *handle) +static int vce_v4_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) /* currently only VCN0 support SRIOV */ adev->vce.num_rings = 1; @@ -422,9 +422,9 @@ static int vce_v4_0_early_init(void *handle) return 0; } -static int vce_v4_0_sw_init(void *handle) +static int vce_v4_0_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; unsigned size; @@ -493,10 +493,10 @@ static int vce_v4_0_sw_init(void *handle) return r; } -static int vce_v4_0_sw_fini(void *handle) +static int vce_v4_0_sw_fini(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* free MM table */ amdgpu_virt_free_mm_table(adev); @@ -513,10 +513,10 @@ static int vce_v4_0_sw_fini(void *handle) return amdgpu_vce_sw_fini(adev); } -static int vce_v4_0_hw_init(void *handle) +static int vce_v4_0_hw_init(struct amdgpu_ip_block *ip_block) { int r, i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) r = vce_v4_0_sriov_start(adev); @@ -536,14 +536,14 @@ static int vce_v4_0_hw_init(void *handle) return 0; } -static int vce_v4_0_hw_fini(void *handle) +static int vce_v4_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->vce.idle_work); if (!amdgpu_sriov_vf(adev)) { - /* vce_v4_0_wait_for_idle(handle); */ + /* vce_v4_0_wait_for_idle(ip_block); */ vce_v4_0_stop(adev); } else { /* full access mode, so don't touch any VCE register */ @@ -553,9 +553,9 @@ static int vce_v4_0_hw_fini(void *handle) return 0; } -static int vce_v4_0_suspend(void *handle) +static int vce_v4_0_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r, idx; if (adev->vce.vcpu_bo == NULL) @@ -594,16 +594,16 @@ static int vce_v4_0_suspend(void *handle) AMD_CG_STATE_GATE); } - r = vce_v4_0_hw_fini(adev); + r = vce_v4_0_hw_fini(ip_block); if (r) return r; return amdgpu_vce_suspend(adev); } -static int vce_v4_0_resume(void *handle) +static int vce_v4_0_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r, idx; if (adev->vce.vcpu_bo == NULL) @@ -624,7 +624,7 @@ static int vce_v4_0_resume(void *handle) return r; } - return vce_v4_0_hw_init(adev); + return vce_v4_0_hw_init(ip_block); } static void vce_v4_0_mc_resume(struct amdgpu_device *adev) @@ -691,273 +691,6 @@ static int vce_v4_0_set_clockgating_state(void *handle, return 0; } -#if 0 -static bool vce_v4_0_is_idle(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - u32 mask = 0; - - mask |= (adev->vce.harvest_config & AMDGPU_VCE_HARVEST_VCE0) ? 0 : SRBM_STATUS2__VCE0_BUSY_MASK; - mask |= (adev->vce.harvest_config & AMDGPU_VCE_HARVEST_VCE1) ? 0 : SRBM_STATUS2__VCE1_BUSY_MASK; - - return !(RREG32(mmSRBM_STATUS2) & mask); -} - -static int vce_v4_0_wait_for_idle(void *handle) -{ - unsigned i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - for (i = 0; i < adev->usec_timeout; i++) - if (vce_v4_0_is_idle(handle)) - return 0; - - return -ETIMEDOUT; -} - -#define VCE_STATUS_VCPU_REPORT_AUTO_BUSY_MASK 0x00000008L /* AUTO_BUSY */ -#define VCE_STATUS_VCPU_REPORT_RB0_BUSY_MASK 0x00000010L /* RB0_BUSY */ -#define VCE_STATUS_VCPU_REPORT_RB1_BUSY_MASK 0x00000020L /* RB1_BUSY */ -#define AMDGPU_VCE_STATUS_BUSY_MASK (VCE_STATUS_VCPU_REPORT_AUTO_BUSY_MASK | \ - VCE_STATUS_VCPU_REPORT_RB0_BUSY_MASK) - -static bool vce_v4_0_check_soft_reset(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - u32 srbm_soft_reset = 0; - - /* According to VCE team , we should use VCE_STATUS instead - * SRBM_STATUS.VCE_BUSY bit for busy status checking. - * GRBM_GFX_INDEX.INSTANCE_INDEX is used to specify which VCE - * instance's registers are accessed - * (0 for 1st instance, 10 for 2nd instance). - * - *VCE_STATUS - *|UENC|ACPI|AUTO ACTIVE|RB1 |RB0 |RB2 | |FW_LOADED|JOB | - *|----+----+-----------+----+----+----+----------+---------+----| - *|bit8|bit7| bit6 |bit5|bit4|bit3| bit2 | bit1 |bit0| - * - * VCE team suggest use bit 3--bit 6 for busy status check - */ - mutex_lock(&adev->grbm_idx_mutex); - WREG32_FIELD(GRBM_GFX_INDEX, INSTANCE_INDEX, 0); - if (RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_STATUS) & AMDGPU_VCE_STATUS_BUSY_MASK) { - srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_VCE0, 1); - srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_VCE1, 1); - } - WREG32_FIELD(GRBM_GFX_INDEX, INSTANCE_INDEX, 0x10); - if (RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_STATUS) & AMDGPU_VCE_STATUS_BUSY_MASK) { - srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_VCE0, 1); - srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_VCE1, 1); - } - WREG32_FIELD(GRBM_GFX_INDEX, INSTANCE_INDEX, 0); - mutex_unlock(&adev->grbm_idx_mutex); - - if (srbm_soft_reset) { - adev->vce.srbm_soft_reset = srbm_soft_reset; - return true; - } else { - adev->vce.srbm_soft_reset = 0; - return false; - } -} - -static int vce_v4_0_soft_reset(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - u32 srbm_soft_reset; - - if (!adev->vce.srbm_soft_reset) - return 0; - srbm_soft_reset = adev->vce.srbm_soft_reset; - - if (srbm_soft_reset) { - u32 tmp; - - tmp = RREG32(mmSRBM_SOFT_RESET); - tmp |= srbm_soft_reset; - dev_info(adev->dev, "SRBM_SOFT_RESET=0x%08X\n", tmp); - WREG32(mmSRBM_SOFT_RESET, tmp); - tmp = RREG32(mmSRBM_SOFT_RESET); - - udelay(50); - - tmp &= ~srbm_soft_reset; - WREG32(mmSRBM_SOFT_RESET, tmp); - tmp = RREG32(mmSRBM_SOFT_RESET); - - /* Wait a little for things to settle down */ - udelay(50); - } - - return 0; -} - -static int vce_v4_0_pre_soft_reset(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (!adev->vce.srbm_soft_reset) - return 0; - - mdelay(5); - - return vce_v4_0_suspend(adev); -} - - -static int vce_v4_0_post_soft_reset(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - if (!adev->vce.srbm_soft_reset) - return 0; - - mdelay(5); - - return vce_v4_0_resume(adev); -} - -static void vce_v4_0_override_vce_clock_gating(struct amdgpu_device *adev, bool override) -{ - u32 tmp, data; - - tmp = data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_ARB_CTRL)); - if (override) - data |= VCE_RB_ARB_CTRL__VCE_CGTT_OVERRIDE_MASK; - else - data &= ~VCE_RB_ARB_CTRL__VCE_CGTT_OVERRIDE_MASK; - - if (tmp != data) - WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_ARB_CTRL), data); -} - -static void vce_v4_0_set_vce_sw_clock_gating(struct amdgpu_device *adev, - bool gated) -{ - u32 data; - - /* Set Override to disable Clock Gating */ - vce_v4_0_override_vce_clock_gating(adev, true); - - /* This function enables MGCG which is controlled by firmware. - With the clocks in the gated state the core is still - accessible but the firmware will throttle the clocks on the - fly as necessary. - */ - if (gated) { - data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_CLOCK_GATING_B)); - data |= 0x1ff; - data &= ~0xef0000; - WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_CLOCK_GATING_B), data); - - data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_CLOCK_GATING)); - data |= 0x3ff000; - data &= ~0xffc00000; - WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_CLOCK_GATING), data); - - data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_CLOCK_GATING_2)); - data |= 0x2; - data &= ~0x00010000; - WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_CLOCK_GATING_2), data); - - data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_REG_CLOCK_GATING)); - data |= 0x37f; - WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_REG_CLOCK_GATING), data); - - data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_DMA_DCLK_CTRL)); - data |= VCE_UENC_DMA_DCLK_CTRL__WRDMCLK_FORCEON_MASK | - VCE_UENC_DMA_DCLK_CTRL__RDDMCLK_FORCEON_MASK | - VCE_UENC_DMA_DCLK_CTRL__REGCLK_FORCEON_MASK | - 0x8; - WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_DMA_DCLK_CTRL), data); - } else { - data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_CLOCK_GATING_B)); - data &= ~0x80010; - data |= 0xe70008; - WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_CLOCK_GATING_B), data); - - data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_CLOCK_GATING)); - data |= 0xffc00000; - WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_CLOCK_GATING), data); - - data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_CLOCK_GATING_2)); - data |= 0x10000; - WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_CLOCK_GATING_2), data); - - data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_REG_CLOCK_GATING)); - data &= ~0xffc00000; - WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_REG_CLOCK_GATING), data); - - data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_DMA_DCLK_CTRL)); - data &= ~(VCE_UENC_DMA_DCLK_CTRL__WRDMCLK_FORCEON_MASK | - VCE_UENC_DMA_DCLK_CTRL__RDDMCLK_FORCEON_MASK | - VCE_UENC_DMA_DCLK_CTRL__REGCLK_FORCEON_MASK | - 0x8); - WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_DMA_DCLK_CTRL), data); - } - vce_v4_0_override_vce_clock_gating(adev, false); -} - -static void vce_v4_0_set_bypass_mode(struct amdgpu_device *adev, bool enable) -{ - u32 tmp = RREG32_SMC(ixGCK_DFS_BYPASS_CNTL); - - if (enable) - tmp |= GCK_DFS_BYPASS_CNTL__BYPASSECLK_MASK; - else - tmp &= ~GCK_DFS_BYPASS_CNTL__BYPASSECLK_MASK; - - WREG32_SMC(ixGCK_DFS_BYPASS_CNTL, tmp); -} - -static int vce_v4_0_set_clockgating_state(void *handle, - enum amd_clockgating_state state) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - bool enable = (state == AMD_CG_STATE_GATE); - int i; - - if ((adev->asic_type == CHIP_POLARIS10) || - (adev->asic_type == CHIP_TONGA) || - (adev->asic_type == CHIP_FIJI)) - vce_v4_0_set_bypass_mode(adev, enable); - - if (!(adev->cg_flags & AMD_CG_SUPPORT_VCE_MGCG)) - return 0; - - mutex_lock(&adev->grbm_idx_mutex); - for (i = 0; i < 2; i++) { - /* Program VCE Instance 0 or 1 if not harvested */ - if (adev->vce.harvest_config & (1 << i)) - continue; - - WREG32_FIELD(GRBM_GFX_INDEX, VCE_INSTANCE, i); - - if (enable) { - /* initialize VCE_CLOCK_GATING_A: Clock ON/OFF delay */ - uint32_t data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_CLOCK_GATING_A); - data &= ~(0xf | 0xff0); - data |= ((0x0 << 0) | (0x04 << 4)); - WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_CLOCK_GATING_A, data); - - /* initialize VCE_UENC_CLOCK_GATING: Clock ON/OFF delay */ - data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_CLOCK_GATING); - data &= ~(0xf | 0xff0); - data |= ((0x0 << 0) | (0x04 << 4)); - WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_CLOCK_GATING, data); - } - - vce_v4_0_set_vce_sw_clock_gating(adev, enable); - } - - WREG32_FIELD(GRBM_GFX_INDEX, VCE_INSTANCE, 0); - mutex_unlock(&adev->grbm_idx_mutex); - - return 0; -} -#endif - static int vce_v4_0_set_powergating_state(void *handle, enum amd_powergating_state state) { @@ -1076,19 +809,12 @@ static int vce_v4_0_process_interrupt(struct amdgpu_device *adev, const struct amd_ip_funcs vce_v4_0_ip_funcs = { .name = "vce_v4_0", .early_init = vce_v4_0_early_init, - .late_init = NULL, .sw_init = vce_v4_0_sw_init, .sw_fini = vce_v4_0_sw_fini, .hw_init = vce_v4_0_hw_init, .hw_fini = vce_v4_0_hw_fini, .suspend = vce_v4_0_suspend, .resume = vce_v4_0_resume, - .is_idle = NULL /* vce_v4_0_is_idle */, - .wait_for_idle = NULL /* vce_v4_0_wait_for_idle */, - .check_soft_reset = NULL /* vce_v4_0_check_soft_reset */, - .pre_soft_reset = NULL /* vce_v4_0_pre_soft_reset */, - .soft_reset = NULL /* vce_v4_0_soft_reset */, - .post_soft_reset = NULL /* vce_v4_0_post_soft_reset */, .set_clockgating_state = vce_v4_0_set_clockgating_state, .set_powergating_state = vce_v4_0_set_powergating_state, }; diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c index ecdfbfefd66a..10e99c926fb8 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c @@ -95,14 +95,14 @@ static void vcn_v1_0_ring_begin_use(struct amdgpu_ring *ring); /** * vcn_v1_0_early_init - set function pointers and load microcode * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers * Load microcode from filesystem */ -static int vcn_v1_0_early_init(void *handle) +static int vcn_v1_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->vcn.num_enc_rings = 2; @@ -110,7 +110,7 @@ static int vcn_v1_0_early_init(void *handle) vcn_v1_0_set_enc_ring_funcs(adev); vcn_v1_0_set_irq_funcs(adev); - jpeg_v1_0_early_init(handle); + jpeg_v1_0_early_init(ip_block); return amdgpu_vcn_early_init(adev); } @@ -118,17 +118,17 @@ static int vcn_v1_0_early_init(void *handle) /** * vcn_v1_0_sw_init - sw init for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int vcn_v1_0_sw_init(void *handle) +static int vcn_v1_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int i, r; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_1_0); uint32_t *ptr; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* VCN DEC TRAP */ r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VCN, @@ -197,7 +197,7 @@ static int vcn_v1_0_sw_init(void *handle) amdgpu_vcn_fwlog_init(adev->vcn.inst); } - r = jpeg_v1_0_sw_init(handle); + r = jpeg_v1_0_sw_init(ip_block); /* Allocate memory for VCN IP Dump buffer */ ptr = kcalloc(adev->vcn.num_vcn_inst * reg_count, sizeof(uint32_t), GFP_KERNEL); @@ -213,20 +213,20 @@ static int vcn_v1_0_sw_init(void *handle) /** * vcn_v1_0_sw_fini - sw fini for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * VCN suspend and free up sw allocation */ -static int vcn_v1_0_sw_fini(void *handle) +static int vcn_v1_0_sw_fini(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_vcn_suspend(adev); if (r) return r; - jpeg_v1_0_sw_fini(handle); + jpeg_v1_0_sw_fini(ip_block); r = amdgpu_vcn_sw_fini(adev); @@ -238,13 +238,13 @@ static int vcn_v1_0_sw_fini(void *handle) /** * vcn_v1_0_hw_init - start and test VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Initialize the hardware, boot up the VCPU and do some testing */ -static int vcn_v1_0_hw_init(void *handle) +static int vcn_v1_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring = &adev->vcn.inst->ring_dec; int i, r; @@ -268,13 +268,13 @@ static int vcn_v1_0_hw_init(void *handle) /** * vcn_v1_0_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the VCN block, mark ring as not ready any more */ -static int vcn_v1_0_hw_fini(void *handle) +static int vcn_v1_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->vcn.idle_work); @@ -290,14 +290,14 @@ static int vcn_v1_0_hw_fini(void *handle) /** * vcn_v1_0_suspend - suspend VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend VCN block */ -static int vcn_v1_0_suspend(void *handle) +static int vcn_v1_0_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; bool idle_work_unexecuted; idle_work_unexecuted = cancel_delayed_work_sync(&adev->vcn.idle_work); @@ -306,7 +306,7 @@ static int vcn_v1_0_suspend(void *handle) amdgpu_dpm_enable_uvd(adev, false); } - r = vcn_v1_0_hw_fini(adev); + r = vcn_v1_0_hw_fini(ip_block); if (r) return r; @@ -318,20 +318,19 @@ static int vcn_v1_0_suspend(void *handle) /** * vcn_v1_0_resume - resume VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init VCN block */ -static int vcn_v1_0_resume(void *handle) +static int vcn_v1_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_vcn_resume(adev); + r = amdgpu_vcn_resume(ip_block->adev); if (r) return r; - r = vcn_v1_0_hw_init(adev); + r = vcn_v1_0_hw_init(ip_block); return r; } @@ -1384,9 +1383,9 @@ static bool vcn_v1_0_is_idle(void *handle) return (RREG32_SOC15(VCN, 0, mmUVD_STATUS) == UVD_STATUS__IDLE); } -static int vcn_v1_0_wait_for_idle(void *handle) +static int vcn_v1_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int ret; ret = SOC15_WAIT_ON_RREG(VCN, 0, mmUVD_STATUS, UVD_STATUS__IDLE, @@ -1925,9 +1924,9 @@ void vcn_v1_0_ring_end_use(struct amdgpu_ring *ring) mutex_unlock(&ring->adev->vcn.vcn1_jpeg1_workaround); } -static void vcn_v1_0_print_ip_state(void *handle, struct drm_printer *p) +static void vcn_v1_0_print_ip_state(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_1_0); uint32_t inst_off, is_powered; @@ -1957,9 +1956,9 @@ static void vcn_v1_0_print_ip_state(void *handle, struct drm_printer *p) } } -static void vcn_v1_0_dump_ip_state(void *handle) +static void vcn_v1_0_dump_ip_state(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; bool is_powered; uint32_t inst_off; @@ -1988,7 +1987,6 @@ static void vcn_v1_0_dump_ip_state(void *handle) static const struct amd_ip_funcs vcn_v1_0_ip_funcs = { .name = "vcn_v1_0", .early_init = vcn_v1_0_early_init, - .late_init = NULL, .sw_init = vcn_v1_0_sw_init, .sw_fini = vcn_v1_0_sw_fini, .hw_init = vcn_v1_0_hw_init, @@ -1997,10 +1995,6 @@ static const struct amd_ip_funcs vcn_v1_0_ip_funcs = { .resume = vcn_v1_0_resume, .is_idle = vcn_v1_0_is_idle, .wait_for_idle = vcn_v1_0_wait_for_idle, - .check_soft_reset = NULL /* vcn_v1_0_check_soft_reset */, - .pre_soft_reset = NULL /* vcn_v1_0_pre_soft_reset */, - .soft_reset = NULL /* vcn_v1_0_soft_reset */, - .post_soft_reset = NULL /* vcn_v1_0_post_soft_reset */, .set_clockgating_state = vcn_v1_0_set_clockgating_state, .set_powergating_state = vcn_v1_0_set_powergating_state, .dump_ip_state = vcn_v1_0_dump_ip_state, diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c index bfd067e2d2f1..e0322cbca3ec 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c @@ -100,14 +100,14 @@ static int vcn_v2_0_start_sriov(struct amdgpu_device *adev); /** * vcn_v2_0_early_init - set function pointers and load microcode * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers * Load microcode from filesystem */ -static int vcn_v2_0_early_init(void *handle) +static int vcn_v2_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) adev->vcn.num_enc_rings = 1; @@ -124,17 +124,17 @@ static int vcn_v2_0_early_init(void *handle) /** * vcn_v2_0_sw_init - sw init for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int vcn_v2_0_sw_init(void *handle) +static int vcn_v2_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int i, r; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_2_0); uint32_t *ptr; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; volatile struct amdgpu_fw_shared *fw_shared; /* VCN DEC TRAP */ @@ -237,14 +237,14 @@ static int vcn_v2_0_sw_init(void *handle) /** * vcn_v2_0_sw_fini - sw fini for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * VCN suspend and free up sw allocation */ -static int vcn_v2_0_sw_fini(void *handle) +static int vcn_v2_0_sw_fini(struct amdgpu_ip_block *ip_block) { int r, idx; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; volatile struct amdgpu_fw_shared *fw_shared = adev->vcn.inst->fw_shared.cpu_addr; if (drm_dev_enter(adev_to_drm(adev), &idx)) { @@ -268,13 +268,13 @@ static int vcn_v2_0_sw_fini(void *handle) /** * vcn_v2_0_hw_init - start and test VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Initialize the hardware, boot up the VCPU and do some testing */ -static int vcn_v2_0_hw_init(void *handle) +static int vcn_v2_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring = &adev->vcn.inst->ring_dec; int i, r; @@ -305,13 +305,13 @@ static int vcn_v2_0_hw_init(void *handle) /** * vcn_v2_0_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the VCN block, mark ring as not ready any more */ -static int vcn_v2_0_hw_fini(void *handle) +static int vcn_v2_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->vcn.idle_work); @@ -326,20 +326,19 @@ static int vcn_v2_0_hw_fini(void *handle) /** * vcn_v2_0_suspend - suspend VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend VCN block */ -static int vcn_v2_0_suspend(void *handle) +static int vcn_v2_0_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = vcn_v2_0_hw_fini(adev); + r = vcn_v2_0_hw_fini(ip_block); if (r) return r; - r = amdgpu_vcn_suspend(adev); + r = amdgpu_vcn_suspend(ip_block->adev); return r; } @@ -347,20 +346,19 @@ static int vcn_v2_0_suspend(void *handle) /** * vcn_v2_0_resume - resume VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init VCN block */ -static int vcn_v2_0_resume(void *handle) +static int vcn_v2_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_vcn_resume(adev); + r = amdgpu_vcn_resume(ip_block->adev); if (r) return r; - r = vcn_v2_0_hw_init(adev); + r = vcn_v2_0_hw_init(ip_block); return r; } @@ -1326,9 +1324,9 @@ static bool vcn_v2_0_is_idle(void *handle) return (RREG32_SOC15(VCN, 0, mmUVD_STATUS) == UVD_STATUS__IDLE); } -static int vcn_v2_0_wait_for_idle(void *handle) +static int vcn_v2_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int ret; ret = SOC15_WAIT_ON_RREG(VCN, 0, mmUVD_STATUS, UVD_STATUS__IDLE, @@ -2034,9 +2032,9 @@ static int vcn_v2_0_start_sriov(struct amdgpu_device *adev) return vcn_v2_0_start_mmsch(adev, &adev->virt.mm_table); } -static void vcn_v2_0_print_ip_state(void *handle, struct drm_printer *p) +static void vcn_v2_0_print_ip_state(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_2_0); uint32_t inst_off, is_powered; @@ -2066,9 +2064,9 @@ static void vcn_v2_0_print_ip_state(void *handle, struct drm_printer *p) } } -static void vcn_v2_0_dump_ip_state(void *handle) +static void vcn_v2_0_dump_ip_state(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; bool is_powered; uint32_t inst_off; @@ -2097,7 +2095,6 @@ static void vcn_v2_0_dump_ip_state(void *handle) static const struct amd_ip_funcs vcn_v2_0_ip_funcs = { .name = "vcn_v2_0", .early_init = vcn_v2_0_early_init, - .late_init = NULL, .sw_init = vcn_v2_0_sw_init, .sw_fini = vcn_v2_0_sw_fini, .hw_init = vcn_v2_0_hw_init, @@ -2106,10 +2103,6 @@ static const struct amd_ip_funcs vcn_v2_0_ip_funcs = { .resume = vcn_v2_0_resume, .is_idle = vcn_v2_0_is_idle, .wait_for_idle = vcn_v2_0_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = vcn_v2_0_set_clockgating_state, .set_powergating_state = vcn_v2_0_set_powergating_state, .dump_ip_state = vcn_v2_0_dump_ip_state, diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c index 04e9e806e318..6aa08281d094 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c @@ -110,14 +110,14 @@ static int amdgpu_ih_clientid_vcns[] = { /** * vcn_v2_5_early_init - set function pointers and load microcode * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers * Load microcode from filesystem */ -static int vcn_v2_5_early_init(void *handle) +static int vcn_v2_5_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) { adev->vcn.num_vcn_inst = 2; @@ -151,17 +151,17 @@ static int vcn_v2_5_early_init(void *handle) /** * vcn_v2_5_sw_init - sw init for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int vcn_v2_5_sw_init(void *handle) +static int vcn_v2_5_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int i, j, r; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_2_5); uint32_t *ptr; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; for (j = 0; j < adev->vcn.num_vcn_inst; j++) { if (adev->vcn.harvest_config & (1 << j)) @@ -295,14 +295,14 @@ static int vcn_v2_5_sw_init(void *handle) /** * vcn_v2_5_sw_fini - sw fini for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * VCN suspend and free up sw allocation */ -static int vcn_v2_5_sw_fini(void *handle) +static int vcn_v2_5_sw_fini(struct amdgpu_ip_block *ip_block) { int i, r, idx; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; volatile struct amdgpu_fw_shared *fw_shared; if (drm_dev_enter(adev_to_drm(adev), &idx)) { @@ -333,13 +333,13 @@ static int vcn_v2_5_sw_fini(void *handle) /** * vcn_v2_5_hw_init - start and test VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Initialize the hardware, boot up the VCPU and do some testing */ -static int vcn_v2_5_hw_init(void *handle) +static int vcn_v2_5_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int i, j, r = 0; @@ -381,13 +381,13 @@ static int vcn_v2_5_hw_init(void *handle) /** * vcn_v2_5_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the VCN block, mark ring as not ready any more */ -static int vcn_v2_5_hw_fini(void *handle) +static int vcn_v2_5_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; cancel_delayed_work_sync(&adev->vcn.idle_work); @@ -411,20 +411,19 @@ static int vcn_v2_5_hw_fini(void *handle) /** * vcn_v2_5_suspend - suspend VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend VCN block */ -static int vcn_v2_5_suspend(void *handle) +static int vcn_v2_5_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = vcn_v2_5_hw_fini(adev); + r = vcn_v2_5_hw_fini(ip_block); if (r) return r; - r = amdgpu_vcn_suspend(adev); + r = amdgpu_vcn_suspend(ip_block->adev); return r; } @@ -432,20 +431,19 @@ static int vcn_v2_5_suspend(void *handle) /** * vcn_v2_5_resume - resume VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init VCN block */ -static int vcn_v2_5_resume(void *handle) +static int vcn_v2_5_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_vcn_resume(adev); + r = amdgpu_vcn_resume(ip_block->adev); if (r) return r; - r = vcn_v2_5_hw_init(adev); + r = vcn_v2_5_hw_init(ip_block); return r; } @@ -1786,9 +1784,9 @@ static bool vcn_v2_5_is_idle(void *handle) return ret; } -static int vcn_v2_5_wait_for_idle(void *handle) +static int vcn_v2_5_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, ret = 0; for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { @@ -1926,9 +1924,9 @@ static void vcn_v2_5_set_irq_funcs(struct amdgpu_device *adev) } } -static void vcn_v2_5_print_ip_state(void *handle, struct drm_printer *p) +static void vcn_v2_5_print_ip_state(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_2_5); uint32_t inst_off, is_powered; @@ -1958,9 +1956,9 @@ static void vcn_v2_5_print_ip_state(void *handle, struct drm_printer *p) } } -static void vcn_v2_5_dump_ip_state(void *handle) +static void vcn_v2_5_dump_ip_state(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; bool is_powered; uint32_t inst_off; @@ -1989,7 +1987,6 @@ static void vcn_v2_5_dump_ip_state(void *handle) static const struct amd_ip_funcs vcn_v2_5_ip_funcs = { .name = "vcn_v2_5", .early_init = vcn_v2_5_early_init, - .late_init = NULL, .sw_init = vcn_v2_5_sw_init, .sw_fini = vcn_v2_5_sw_fini, .hw_init = vcn_v2_5_hw_init, @@ -1998,10 +1995,6 @@ static const struct amd_ip_funcs vcn_v2_5_ip_funcs = { .resume = vcn_v2_5_resume, .is_idle = vcn_v2_5_is_idle, .wait_for_idle = vcn_v2_5_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = vcn_v2_5_set_clockgating_state, .set_powergating_state = vcn_v2_5_set_powergating_state, .dump_ip_state = vcn_v2_5_dump_ip_state, @@ -2011,7 +2004,6 @@ static const struct amd_ip_funcs vcn_v2_5_ip_funcs = { static const struct amd_ip_funcs vcn_v2_6_ip_funcs = { .name = "vcn_v2_6", .early_init = vcn_v2_5_early_init, - .late_init = NULL, .sw_init = vcn_v2_5_sw_init, .sw_fini = vcn_v2_5_sw_fini, .hw_init = vcn_v2_5_hw_init, @@ -2020,10 +2012,6 @@ static const struct amd_ip_funcs vcn_v2_6_ip_funcs = { .resume = vcn_v2_5_resume, .is_idle = vcn_v2_5_is_idle, .wait_for_idle = vcn_v2_5_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = vcn_v2_5_set_clockgating_state, .set_powergating_state = vcn_v2_5_set_powergating_state, .dump_ip_state = vcn_v2_5_dump_ip_state, diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c index 65dd68b32280..6732ad7f16f5 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c @@ -116,14 +116,14 @@ static void vcn_v3_0_enc_ring_set_wptr(struct amdgpu_ring *ring); /** * vcn_v3_0_early_init - set function pointers and load microcode * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers * Load microcode from filesystem */ -static int vcn_v3_0_early_init(void *handle) +static int vcn_v3_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) { adev->vcn.num_vcn_inst = VCN_INSTANCES_SIENNA_CICHLID; @@ -153,18 +153,18 @@ static int vcn_v3_0_early_init(void *handle) /** * vcn_v3_0_sw_init - sw init for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int vcn_v3_0_sw_init(void *handle) +static int vcn_v3_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; int i, j, r; int vcn_doorbell_index = 0; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_3_0); uint32_t *ptr; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = amdgpu_vcn_sw_init(adev); if (r) @@ -299,13 +299,13 @@ static int vcn_v3_0_sw_init(void *handle) /** * vcn_v3_0_sw_fini - sw fini for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * VCN suspend and free up sw allocation */ -static int vcn_v3_0_sw_fini(void *handle) +static int vcn_v3_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, r, idx; if (drm_dev_enter(adev_to_drm(adev), &idx)) { @@ -338,13 +338,13 @@ static int vcn_v3_0_sw_fini(void *handle) /** * vcn_v3_0_hw_init - start and test VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Initialize the hardware, boot up the VCPU and do some testing */ -static int vcn_v3_0_hw_init(void *handle) +static int vcn_v3_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int i, j, r; @@ -413,13 +413,13 @@ static int vcn_v3_0_hw_init(void *handle) /** * vcn_v3_0_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the VCN block, mark ring as not ready any more */ -static int vcn_v3_0_hw_fini(void *handle) +static int vcn_v3_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; cancel_delayed_work_sync(&adev->vcn.idle_work); @@ -443,20 +443,19 @@ static int vcn_v3_0_hw_fini(void *handle) /** * vcn_v3_0_suspend - suspend VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend VCN block */ -static int vcn_v3_0_suspend(void *handle) +static int vcn_v3_0_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = vcn_v3_0_hw_fini(adev); + r = vcn_v3_0_hw_fini(ip_block); if (r) return r; - r = amdgpu_vcn_suspend(adev); + r = amdgpu_vcn_suspend(ip_block->adev); return r; } @@ -464,20 +463,19 @@ static int vcn_v3_0_suspend(void *handle) /** * vcn_v3_0_resume - resume VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init VCN block */ -static int vcn_v3_0_resume(void *handle) +static int vcn_v3_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_vcn_resume(adev); + r = amdgpu_vcn_resume(ip_block->adev); if (r) return r; - r = vcn_v3_0_hw_init(adev); + r = vcn_v3_0_hw_init(ip_block); return r; } @@ -2116,9 +2114,9 @@ static bool vcn_v3_0_is_idle(void *handle) return ret; } -static int vcn_v3_0_wait_for_idle(void *handle) +static int vcn_v3_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, ret = 0; for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { @@ -2251,9 +2249,9 @@ static void vcn_v3_0_set_irq_funcs(struct amdgpu_device *adev) } } -static void vcn_v3_0_print_ip_state(void *handle, struct drm_printer *p) +static void vcn_v3_0_print_ip_state(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_3_0); uint32_t inst_off; @@ -2284,9 +2282,9 @@ static void vcn_v3_0_print_ip_state(void *handle, struct drm_printer *p) } } -static void vcn_v3_0_dump_ip_state(void *handle) +static void vcn_v3_0_dump_ip_state(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; bool is_powered; uint32_t inst_off; @@ -2315,7 +2313,6 @@ static void vcn_v3_0_dump_ip_state(void *handle) static const struct amd_ip_funcs vcn_v3_0_ip_funcs = { .name = "vcn_v3_0", .early_init = vcn_v3_0_early_init, - .late_init = NULL, .sw_init = vcn_v3_0_sw_init, .sw_fini = vcn_v3_0_sw_fini, .hw_init = vcn_v3_0_hw_init, @@ -2324,10 +2321,6 @@ static const struct amd_ip_funcs vcn_v3_0_ip_funcs = { .resume = vcn_v3_0_resume, .is_idle = vcn_v3_0_is_idle, .wait_for_idle = vcn_v3_0_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = vcn_v3_0_set_clockgating_state, .set_powergating_state = vcn_v3_0_set_powergating_state, .dump_ip_state = vcn_v3_0_dump_ip_state, diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c index 26c6f10a8c8f..5512259cac79 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c @@ -106,14 +106,14 @@ static void vcn_v4_0_set_ras_funcs(struct amdgpu_device *adev); /** * vcn_v4_0_early_init - set function pointers and load microcode * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers * Load microcode from filesystem */ -static int vcn_v4_0_early_init(void *handle) +static int vcn_v4_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; if (amdgpu_sriov_vf(adev)) { @@ -164,14 +164,14 @@ static int vcn_v4_0_fw_shared_init(struct amdgpu_device *adev, int inst_idx) /** * vcn_v4_0_sw_init - sw init for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int vcn_v4_0_sw_init(void *handle) +static int vcn_v4_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, r; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_4_0); uint32_t *ptr; @@ -253,13 +253,13 @@ static int vcn_v4_0_sw_init(void *handle) /** * vcn_v4_0_sw_fini - sw fini for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * VCN suspend and free up sw allocation */ -static int vcn_v4_0_sw_fini(void *handle) +static int vcn_v4_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, r, idx; if (drm_dev_enter(adev_to_drm(adev), &idx)) { @@ -294,13 +294,13 @@ static int vcn_v4_0_sw_fini(void *handle) /** * vcn_v4_0_hw_init - start and test VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Initialize the hardware, boot up the VCPU and do some testing */ -static int vcn_v4_0_hw_init(void *handle) +static int vcn_v4_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int i, r; @@ -341,13 +341,13 @@ static int vcn_v4_0_hw_init(void *handle) /** * vcn_v4_0_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the VCN block, mark ring as not ready any more */ -static int vcn_v4_0_hw_fini(void *handle) +static int vcn_v4_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; cancel_delayed_work_sync(&adev->vcn.idle_work); @@ -372,20 +372,19 @@ static int vcn_v4_0_hw_fini(void *handle) /** * vcn_v4_0_suspend - suspend VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend VCN block */ -static int vcn_v4_0_suspend(void *handle) +static int vcn_v4_0_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = vcn_v4_0_hw_fini(adev); + r = vcn_v4_0_hw_fini(ip_block); if (r) return r; - r = amdgpu_vcn_suspend(adev); + r = amdgpu_vcn_suspend(ip_block->adev); return r; } @@ -393,20 +392,19 @@ static int vcn_v4_0_suspend(void *handle) /** * vcn_v4_0_resume - resume VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init VCN block */ -static int vcn_v4_0_resume(void *handle) +static int vcn_v4_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_vcn_resume(adev); + r = amdgpu_vcn_resume(ip_block->adev); if (r) return r; - r = vcn_v4_0_hw_init(adev); + r = vcn_v4_0_hw_init(ip_block); return r; } @@ -1975,13 +1973,13 @@ static bool vcn_v4_0_is_idle(void *handle) /** * vcn_v4_0_wait_for_idle - wait for VCN block idle * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Wait for VCN block idle */ -static int vcn_v4_0_wait_for_idle(void *handle) +static int vcn_v4_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, ret = 0; for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { @@ -2158,9 +2156,9 @@ static void vcn_v4_0_set_irq_funcs(struct amdgpu_device *adev) } } -static void vcn_v4_0_print_ip_state(void *handle, struct drm_printer *p) +static void vcn_v4_0_print_ip_state(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_4_0); uint32_t inst_off, is_powered; @@ -2190,9 +2188,9 @@ static void vcn_v4_0_print_ip_state(void *handle, struct drm_printer *p) } } -static void vcn_v4_0_dump_ip_state(void *handle) +static void vcn_v4_0_dump_ip_state(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; bool is_powered; uint32_t inst_off; @@ -2222,7 +2220,6 @@ static void vcn_v4_0_dump_ip_state(void *handle) static const struct amd_ip_funcs vcn_v4_0_ip_funcs = { .name = "vcn_v4_0", .early_init = vcn_v4_0_early_init, - .late_init = NULL, .sw_init = vcn_v4_0_sw_init, .sw_fini = vcn_v4_0_sw_fini, .hw_init = vcn_v4_0_hw_init, @@ -2231,10 +2228,6 @@ static const struct amd_ip_funcs vcn_v4_0_ip_funcs = { .resume = vcn_v4_0_resume, .is_idle = vcn_v4_0_is_idle, .wait_for_idle = vcn_v4_0_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = vcn_v4_0_set_clockgating_state, .set_powergating_state = vcn_v4_0_set_powergating_state, .dump_ip_state = vcn_v4_0_dump_ip_state, diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c index 0fda70336300..cf808a153fce 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c @@ -95,16 +95,23 @@ static void vcn_v4_0_3_unified_ring_set_wptr(struct amdgpu_ring *ring); static void vcn_v4_0_3_set_ras_funcs(struct amdgpu_device *adev); static void vcn_v4_0_3_enable_ras(struct amdgpu_device *adev, int inst_idx, bool indirect); + +static inline bool vcn_v4_0_3_normalizn_reqd(struct amdgpu_device *adev) +{ + return (amdgpu_sriov_vf(adev) || + (amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(9, 4, 4))); +} + /** * vcn_v4_0_3_early_init - set function pointers * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers */ -static int vcn_v4_0_3_early_init(void *handle) +static int vcn_v4_0_3_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* re-use enc ring as unified ring */ adev->vcn.num_enc_rings = 1; @@ -119,13 +126,13 @@ static int vcn_v4_0_3_early_init(void *handle) /** * vcn_v4_0_3_sw_init - sw init for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int vcn_v4_0_3_sw_init(void *handle) +static int vcn_v4_0_3_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int i, r, vcn_inst; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_4_0_3); @@ -212,13 +219,13 @@ static int vcn_v4_0_3_sw_init(void *handle) /** * vcn_v4_0_3_sw_fini - sw fini for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * VCN suspend and free up sw allocation */ -static int vcn_v4_0_3_sw_fini(void *handle) +static int vcn_v4_0_3_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, r, idx; if (drm_dev_enter(&adev->ddev, &idx)) { @@ -249,13 +256,13 @@ static int vcn_v4_0_3_sw_fini(void *handle) /** * vcn_v4_0_3_hw_init - start and test VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Initialize the hardware, boot up the VCPU and do some testing */ -static int vcn_v4_0_3_hw_init(void *handle) +static int vcn_v4_0_3_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int i, r, vcn_inst; @@ -308,13 +315,13 @@ static int vcn_v4_0_3_hw_init(void *handle) /** * vcn_v4_0_3_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the VCN block, mark ring as not ready any more */ -static int vcn_v4_0_3_hw_fini(void *handle) +static int vcn_v4_0_3_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; cancel_delayed_work_sync(&adev->vcn.idle_work); @@ -327,20 +334,19 @@ static int vcn_v4_0_3_hw_fini(void *handle) /** * vcn_v4_0_3_suspend - suspend VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend VCN block */ -static int vcn_v4_0_3_suspend(void *handle) +static int vcn_v4_0_3_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = vcn_v4_0_3_hw_fini(adev); + r = vcn_v4_0_3_hw_fini(ip_block); if (r) return r; - r = amdgpu_vcn_suspend(adev); + r = amdgpu_vcn_suspend(ip_block->adev); return r; } @@ -348,20 +354,19 @@ static int vcn_v4_0_3_suspend(void *handle) /** * vcn_v4_0_3_resume - resume VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init VCN block */ -static int vcn_v4_0_3_resume(void *handle) +static int vcn_v4_0_3_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; - r = amdgpu_vcn_resume(adev); + r = amdgpu_vcn_resume(ip_block->adev); if (r) return r; - r = vcn_v4_0_3_hw_init(adev); + r = vcn_v4_0_3_hw_init(ip_block); return r; } @@ -1430,8 +1435,8 @@ static uint64_t vcn_v4_0_3_unified_ring_get_wptr(struct amdgpu_ring *ring) static void vcn_v4_0_3_enc_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg, uint32_t val, uint32_t mask) { - /* For VF, only local offsets should be used */ - if (amdgpu_sriov_vf(ring->adev)) + /* Use normalized offsets when required */ + if (vcn_v4_0_3_normalizn_reqd(ring->adev)) reg = NORMALIZE_VCN_REG_OFFSET(reg); amdgpu_ring_write(ring, VCN_ENC_CMD_REG_WAIT); @@ -1442,8 +1447,8 @@ static void vcn_v4_0_3_enc_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t static void vcn_v4_0_3_enc_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg, uint32_t val) { - /* For VF, only local offsets should be used */ - if (amdgpu_sriov_vf(ring->adev)) + /* Use normalized offsets when required */ + if (vcn_v4_0_3_normalizn_reqd(ring->adev)) reg = NORMALIZE_VCN_REG_OFFSET(reg); amdgpu_ring_write(ring, VCN_ENC_CMD_REG_WRITE); @@ -1567,13 +1572,13 @@ static bool vcn_v4_0_3_is_idle(void *handle) /** * vcn_v4_0_3_wait_for_idle - wait for VCN block idle * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Wait for VCN block idle */ -static int vcn_v4_0_3_wait_for_idle(void *handle) +static int vcn_v4_0_3_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, ret = 0; for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { @@ -1733,9 +1738,9 @@ static void vcn_v4_0_3_set_irq_funcs(struct amdgpu_device *adev) adev->vcn.inst->irq.funcs = &vcn_v4_0_3_irq_funcs; } -static void vcn_v4_0_3_print_ip_state(void *handle, struct drm_printer *p) +static void vcn_v4_0_3_print_ip_state(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_4_0_3); uint32_t inst_off, is_powered; @@ -1765,9 +1770,9 @@ static void vcn_v4_0_3_print_ip_state(void *handle, struct drm_printer *p) } } -static void vcn_v4_0_3_dump_ip_state(void *handle) +static void vcn_v4_0_3_dump_ip_state(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; bool is_powered; uint32_t inst_off, inst_id; @@ -1798,7 +1803,6 @@ static void vcn_v4_0_3_dump_ip_state(void *handle) static const struct amd_ip_funcs vcn_v4_0_3_ip_funcs = { .name = "vcn_v4_0_3", .early_init = vcn_v4_0_3_early_init, - .late_init = NULL, .sw_init = vcn_v4_0_3_sw_init, .sw_fini = vcn_v4_0_3_sw_fini, .hw_init = vcn_v4_0_3_hw_init, @@ -1807,10 +1811,6 @@ static const struct amd_ip_funcs vcn_v4_0_3_ip_funcs = { .resume = vcn_v4_0_3_resume, .is_idle = vcn_v4_0_3_is_idle, .wait_for_idle = vcn_v4_0_3_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = vcn_v4_0_3_set_clockgating_state, .set_powergating_state = vcn_v4_0_3_set_powergating_state, .dump_ip_state = vcn_v4_0_3_dump_ip_state, diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c index 9d4f5352a62c..71961fb3f7ff 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c @@ -104,14 +104,14 @@ static void vcn_v4_0_5_unified_ring_set_wptr(struct amdgpu_ring *ring); /** * vcn_v4_0_5_early_init - set function pointers and load microcode * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers * Load microcode from filesystem */ -static int vcn_v4_0_5_early_init(void *handle) +static int vcn_v4_0_5_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* re-use enc ring as unified ring */ adev->vcn.num_enc_rings = 1; @@ -124,14 +124,14 @@ static int vcn_v4_0_5_early_init(void *handle) /** * vcn_v4_0_5_sw_init - sw init for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int vcn_v4_0_5_sw_init(void *handle) +static int vcn_v4_0_5_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, r; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_4_0_5); uint32_t *ptr; @@ -220,13 +220,13 @@ static int vcn_v4_0_5_sw_init(void *handle) /** * vcn_v4_0_5_sw_fini - sw fini for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * VCN suspend and free up sw allocation */ -static int vcn_v4_0_5_sw_fini(void *handle) +static int vcn_v4_0_5_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, r, idx; if (drm_dev_enter(adev_to_drm(adev), &idx)) { @@ -261,13 +261,13 @@ static int vcn_v4_0_5_sw_fini(void *handle) /** * vcn_v4_0_5_hw_init - start and test VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Initialize the hardware, boot up the VCPU and do some testing */ -static int vcn_v4_0_5_hw_init(void *handle) +static int vcn_v4_0_5_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int i, r; @@ -291,13 +291,13 @@ static int vcn_v4_0_5_hw_init(void *handle) /** * vcn_v4_0_5_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the VCN block, mark ring as not ready any more */ -static int vcn_v4_0_5_hw_fini(void *handle) +static int vcn_v4_0_5_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; cancel_delayed_work_sync(&adev->vcn.idle_work); @@ -320,20 +320,19 @@ static int vcn_v4_0_5_hw_fini(void *handle) /** * vcn_v4_0_5_suspend - suspend VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend VCN block */ -static int vcn_v4_0_5_suspend(void *handle) +static int vcn_v4_0_5_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = vcn_v4_0_5_hw_fini(adev); + r = vcn_v4_0_5_hw_fini(ip_block); if (r) return r; - r = amdgpu_vcn_suspend(adev); + r = amdgpu_vcn_suspend(ip_block->adev); return r; } @@ -341,20 +340,19 @@ static int vcn_v4_0_5_suspend(void *handle) /** * vcn_v4_0_5_resume - resume VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init VCN block */ -static int vcn_v4_0_5_resume(void *handle) +static int vcn_v4_0_5_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_vcn_resume(adev); + r = amdgpu_vcn_resume(ip_block->adev); if (r) return r; - r = vcn_v4_0_5_hw_init(adev); + r = vcn_v4_0_5_hw_init(ip_block); return r; } @@ -1469,13 +1467,13 @@ static bool vcn_v4_0_5_is_idle(void *handle) /** * vcn_v4_0_5_wait_for_idle - wait for VCN block idle * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Wait for VCN block idle */ -static int vcn_v4_0_5_wait_for_idle(void *handle) +static int vcn_v4_0_5_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, ret = 0; for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { @@ -1616,9 +1614,9 @@ static void vcn_v4_0_5_set_irq_funcs(struct amdgpu_device *adev) } } -static void vcn_v4_0_5_print_ip_state(void *handle, struct drm_printer *p) +static void vcn_v4_0_5_print_ip_state(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_4_0_5); uint32_t inst_off, is_powered; @@ -1648,9 +1646,9 @@ static void vcn_v4_0_5_print_ip_state(void *handle, struct drm_printer *p) } } -static void vcn_v4_0_5_dump_ip_state(void *handle) +static void vcn_v4_0_5_dump_ip_state(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; bool is_powered; uint32_t inst_off; @@ -1680,7 +1678,6 @@ static void vcn_v4_0_5_dump_ip_state(void *handle) static const struct amd_ip_funcs vcn_v4_0_5_ip_funcs = { .name = "vcn_v4_0_5", .early_init = vcn_v4_0_5_early_init, - .late_init = NULL, .sw_init = vcn_v4_0_5_sw_init, .sw_fini = vcn_v4_0_5_sw_fini, .hw_init = vcn_v4_0_5_hw_init, @@ -1689,10 +1686,6 @@ static const struct amd_ip_funcs vcn_v4_0_5_ip_funcs = { .resume = vcn_v4_0_5_resume, .is_idle = vcn_v4_0_5_is_idle, .wait_for_idle = vcn_v4_0_5_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = vcn_v4_0_5_set_clockgating_state, .set_powergating_state = vcn_v4_0_5_set_powergating_state, .dump_ip_state = vcn_v4_0_5_dump_ip_state, diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_0.c index c305386358b4..fe2cc1a80c13 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_0.c @@ -87,14 +87,14 @@ static void vcn_v5_0_0_unified_ring_set_wptr(struct amdgpu_ring *ring); /** * vcn_v5_0_0_early_init - set function pointers and load microcode * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Set ring and irq function pointers * Load microcode from filesystem */ -static int vcn_v5_0_0_early_init(void *handle) +static int vcn_v5_0_0_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* re-use enc ring as unified ring */ adev->vcn.num_enc_rings = 1; @@ -108,14 +108,14 @@ static int vcn_v5_0_0_early_init(void *handle) /** * vcn_v5_0_0_sw_init - sw init for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Load firmware and sw initialization */ -static int vcn_v5_0_0_sw_init(void *handle) +static int vcn_v5_0_0_sw_init(struct amdgpu_ip_block *ip_block) { struct amdgpu_ring *ring; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, r; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_5_0); uint32_t *ptr; @@ -187,13 +187,13 @@ static int vcn_v5_0_0_sw_init(void *handle) /** * vcn_v5_0_0_sw_fini - sw fini for VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * VCN suspend and free up sw allocation */ -static int vcn_v5_0_0_sw_fini(void *handle) +static int vcn_v5_0_0_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, r, idx; if (drm_dev_enter(adev_to_drm(adev), &idx)) { @@ -225,13 +225,13 @@ static int vcn_v5_0_0_sw_fini(void *handle) /** * vcn_v5_0_0_hw_init - start and test VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Initialize the hardware, boot up the VCPU and do some testing */ -static int vcn_v5_0_0_hw_init(void *handle) +static int vcn_v5_0_0_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_ring *ring; int i, r; @@ -255,13 +255,13 @@ static int vcn_v5_0_0_hw_init(void *handle) /** * vcn_v5_0_0_hw_fini - stop the hardware block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Stop the VCN block, mark ring as not ready any more */ -static int vcn_v5_0_0_hw_fini(void *handle) +static int vcn_v5_0_0_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i; cancel_delayed_work_sync(&adev->vcn.idle_work); @@ -284,20 +284,19 @@ static int vcn_v5_0_0_hw_fini(void *handle) /** * vcn_v5_0_0_suspend - suspend VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * HW fini and suspend VCN block */ -static int vcn_v5_0_0_suspend(void *handle) +static int vcn_v5_0_0_suspend(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = vcn_v5_0_0_hw_fini(adev); + r = vcn_v5_0_0_hw_fini(ip_block); if (r) return r; - r = amdgpu_vcn_suspend(adev); + r = amdgpu_vcn_suspend(ip_block->adev); return r; } @@ -305,20 +304,19 @@ static int vcn_v5_0_0_suspend(void *handle) /** * vcn_v5_0_0_resume - resume VCN block * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Resume firmware and hw init VCN block */ -static int vcn_v5_0_0_resume(void *handle) +static int vcn_v5_0_0_resume(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_vcn_resume(adev); + r = amdgpu_vcn_resume(ip_block->adev); if (r) return r; - r = vcn_v5_0_0_hw_init(adev); + r = vcn_v5_0_0_hw_init(ip_block); return r; } @@ -1196,13 +1194,13 @@ static bool vcn_v5_0_0_is_idle(void *handle) /** * vcn_v5_0_0_wait_for_idle - wait for VCN block idle * - * @handle: amdgpu_device pointer + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Wait for VCN block idle */ -static int vcn_v5_0_0_wait_for_idle(void *handle) +static int vcn_v5_0_0_wait_for_idle(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, ret = 0; for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { @@ -1343,9 +1341,9 @@ static void vcn_v5_0_0_set_irq_funcs(struct amdgpu_device *adev) } } -static void vcn_v5_0_print_ip_state(void *handle, struct drm_printer *p) +static void vcn_v5_0_print_ip_state(struct amdgpu_ip_block *ip_block, struct drm_printer *p) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_5_0); uint32_t inst_off, is_powered; @@ -1375,9 +1373,9 @@ static void vcn_v5_0_print_ip_state(void *handle, struct drm_printer *p) } } -static void vcn_v5_0_dump_ip_state(void *handle) +static void vcn_v5_0_dump_ip_state(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int i, j; bool is_powered; uint32_t inst_off; @@ -1406,7 +1404,6 @@ static void vcn_v5_0_dump_ip_state(void *handle) static const struct amd_ip_funcs vcn_v5_0_0_ip_funcs = { .name = "vcn_v5_0_0", .early_init = vcn_v5_0_0_early_init, - .late_init = NULL, .sw_init = vcn_v5_0_0_sw_init, .sw_fini = vcn_v5_0_0_sw_fini, .hw_init = vcn_v5_0_0_hw_init, @@ -1415,10 +1412,6 @@ static const struct amd_ip_funcs vcn_v5_0_0_ip_funcs = { .resume = vcn_v5_0_0_resume, .is_idle = vcn_v5_0_0_is_idle, .wait_for_idle = vcn_v5_0_0_wait_for_idle, - .check_soft_reset = NULL, - .pre_soft_reset = NULL, - .soft_reset = NULL, - .post_soft_reset = NULL, .set_clockgating_state = vcn_v5_0_0_set_clockgating_state, .set_powergating_state = vcn_v5_0_0_set_powergating_state, .dump_ip_state = vcn_v5_0_dump_ip_state, diff --git a/drivers/gpu/drm/amd/amdgpu/vega10_ih.c b/drivers/gpu/drm/amd/amdgpu/vega10_ih.c index bf68e18e3824..0fedadd0a6a4 100644 --- a/drivers/gpu/drm/amd/amdgpu/vega10_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/vega10_ih.c @@ -472,18 +472,18 @@ static void vega10_ih_set_self_irq_funcs(struct amdgpu_device *adev) adev->irq.self_irq.funcs = &vega10_ih_self_irq_funcs; } -static int vega10_ih_early_init(void *handle) +static int vega10_ih_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; vega10_ih_set_interrupt_funcs(adev); vega10_ih_set_self_irq_funcs(adev); return 0; } -static int vega10_ih_sw_init(void *handle) +static int vega10_ih_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_IH, 0, @@ -525,43 +525,35 @@ static int vega10_ih_sw_init(void *handle) return r; } -static int vega10_ih_sw_fini(void *handle) +static int vega10_ih_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_fini_sw(adev); return 0; } -static int vega10_ih_hw_init(void *handle) +static int vega10_ih_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return vega10_ih_irq_init(adev); + return vega10_ih_irq_init(ip_block->adev); } -static int vega10_ih_hw_fini(void *handle) +static int vega10_ih_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - vega10_ih_irq_disable(adev); + vega10_ih_irq_disable(ip_block->adev); return 0; } -static int vega10_ih_suspend(void *handle) +static int vega10_ih_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return vega10_ih_hw_fini(adev); + return vega10_ih_hw_fini(ip_block); } -static int vega10_ih_resume(void *handle) +static int vega10_ih_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return vega10_ih_hw_init(adev); + return vega10_ih_hw_init(ip_block); } static bool vega10_ih_is_idle(void *handle) @@ -570,13 +562,13 @@ static bool vega10_ih_is_idle(void *handle) return true; } -static int vega10_ih_wait_for_idle(void *handle) +static int vega10_ih_wait_for_idle(struct amdgpu_ip_block *ip_block) { /* todo */ return -ETIMEDOUT; } -static int vega10_ih_soft_reset(void *handle) +static int vega10_ih_soft_reset(struct amdgpu_ip_block *ip_block) { /* todo */ @@ -633,7 +625,6 @@ static int vega10_ih_set_powergating_state(void *handle, const struct amd_ip_funcs vega10_ih_ip_funcs = { .name = "vega10_ih", .early_init = vega10_ih_early_init, - .late_init = NULL, .sw_init = vega10_ih_sw_init, .sw_fini = vega10_ih_sw_fini, .hw_init = vega10_ih_hw_init, diff --git a/drivers/gpu/drm/amd/amdgpu/vega20_ih.c b/drivers/gpu/drm/amd/amdgpu/vega20_ih.c index ac439f0565e3..1c9aff742e43 100644 --- a/drivers/gpu/drm/amd/amdgpu/vega20_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/vega20_ih.c @@ -114,6 +114,33 @@ static int vega20_ih_toggle_ring_interrupts(struct amdgpu_device *adev, tmp = REG_SET_FIELD(tmp, IH_RB_CNTL, RB_ENABLE, (enable ? 1 : 0)); tmp = REG_SET_FIELD(tmp, IH_RB_CNTL, RB_GPU_TS_ENABLE, 1); + if (enable) { + /* Unset the CLEAR_OVERFLOW bit to make sure the next step + * is switching the bit from 0 to 1 + */ + tmp = REG_SET_FIELD(tmp, IH_RB_CNTL, WPTR_OVERFLOW_CLEAR, 0); + if (amdgpu_sriov_vf(adev) && amdgpu_sriov_reg_indirect_ih(adev)) { + if (psp_reg_program(&adev->psp, ih_regs->psp_reg_id, tmp)) + return -ETIMEDOUT; + } else { + WREG32_NO_KIQ(ih_regs->ih_rb_cntl, tmp); + } + + /* Clear RB_OVERFLOW bit */ + tmp = REG_SET_FIELD(tmp, IH_RB_CNTL, WPTR_OVERFLOW_CLEAR, 1); + if (amdgpu_sriov_vf(adev) && amdgpu_sriov_reg_indirect_ih(adev)) { + if (psp_reg_program(&adev->psp, ih_regs->psp_reg_id, tmp)) + return -ETIMEDOUT; + } else { + WREG32_NO_KIQ(ih_regs->ih_rb_cntl, tmp); + } + + /* Unset the CLEAR_OVERFLOW bit immediately so new overflows + * can be detected. + */ + tmp = REG_SET_FIELD(tmp, IH_RB_CNTL, WPTR_OVERFLOW_CLEAR, 0); + } + /* enable_intr field is only valid in ring0 */ if (ih == &adev->irq.ih) tmp = REG_SET_FIELD(tmp, IH_RB_CNTL, ENABLE_INTR, (enable ? 1 : 0)); @@ -526,18 +553,18 @@ static void vega20_ih_set_self_irq_funcs(struct amdgpu_device *adev) adev->irq.self_irq.funcs = &vega20_ih_self_irq_funcs; } -static int vega20_ih_early_init(void *handle) +static int vega20_ih_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; vega20_ih_set_interrupt_funcs(adev); vega20_ih_set_self_irq_funcs(adev); return 0; } -static int vega20_ih_sw_init(void *handle) +static int vega20_ih_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; bool use_bus_addr = true; int r; @@ -586,19 +613,19 @@ static int vega20_ih_sw_init(void *handle) return r; } -static int vega20_ih_sw_fini(void *handle) +static int vega20_ih_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_irq_fini_sw(adev); return 0; } -static int vega20_ih_hw_init(void *handle) +static int vega20_ih_hw_init(struct amdgpu_ip_block *ip_block) { int r; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; r = vega20_ih_irq_init(adev); if (r) @@ -607,27 +634,21 @@ static int vega20_ih_hw_init(void *handle) return 0; } -static int vega20_ih_hw_fini(void *handle) +static int vega20_ih_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - vega20_ih_irq_disable(adev); + vega20_ih_irq_disable(ip_block->adev); return 0; } -static int vega20_ih_suspend(void *handle) +static int vega20_ih_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return vega20_ih_hw_fini(adev); + return vega20_ih_hw_fini(ip_block); } -static int vega20_ih_resume(void *handle) +static int vega20_ih_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return vega20_ih_hw_init(adev); + return vega20_ih_hw_init(ip_block); } static bool vega20_ih_is_idle(void *handle) @@ -636,13 +657,13 @@ static bool vega20_ih_is_idle(void *handle) return true; } -static int vega20_ih_wait_for_idle(void *handle) +static int vega20_ih_wait_for_idle(struct amdgpu_ip_block *ip_block) { /* todo */ return -ETIMEDOUT; } -static int vega20_ih_soft_reset(void *handle) +static int vega20_ih_soft_reset(struct amdgpu_ip_block *ip_block) { /* todo */ @@ -696,7 +717,6 @@ static int vega20_ih_set_powergating_state(void *handle, const struct amd_ip_funcs vega20_ih_ip_funcs = { .name = "vega20_ih", .early_init = vega20_ih_early_init, - .late_init = NULL, .sw_init = vega20_ih_sw_init, .sw_fini = vega20_ih_sw_fini, .hw_init = vega20_ih_hw_init, diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c index 792b2eb6bbac..a83505815d39 100644 --- a/drivers/gpu/drm/amd/amdgpu/vi.c +++ b/drivers/gpu/drm/amd/amdgpu/vi.c @@ -1455,9 +1455,9 @@ static const struct amdgpu_asic_funcs vi_asic_funcs = #define CZ_REV_BRISTOL(rev) \ ((rev >= 0xC8 && rev <= 0xCE) || (rev >= 0xE1 && rev <= 0xE6)) -static int vi_common_early_init(void *handle) +static int vi_common_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->flags & AMD_IS_APU) { adev->smc_rreg = &cz_smc_rreg; @@ -1679,9 +1679,9 @@ static int vi_common_early_init(void *handle) return 0; } -static int vi_common_late_init(void *handle) +static int vi_common_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) xgpu_vi_mailbox_get_irq(adev); @@ -1689,9 +1689,9 @@ static int vi_common_late_init(void *handle) return 0; } -static int vi_common_sw_init(void *handle) +static int vi_common_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (amdgpu_sriov_vf(adev)) xgpu_vi_mailbox_add_irq_id(adev); @@ -1699,14 +1699,9 @@ static int vi_common_sw_init(void *handle) return 0; } -static int vi_common_sw_fini(void *handle) +static int vi_common_hw_init(struct amdgpu_ip_block *ip_block) { - return 0; -} - -static int vi_common_hw_init(void *handle) -{ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* move the golden regs per IP block */ vi_init_golden_registers(adev); @@ -1718,9 +1713,9 @@ static int vi_common_hw_init(void *handle) return 0; } -static int vi_common_hw_fini(void *handle) +static int vi_common_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; /* enable the doorbell aperture */ vi_enable_doorbell_aperture(adev, false); @@ -1731,18 +1726,14 @@ static int vi_common_hw_fini(void *handle) return 0; } -static int vi_common_suspend(void *handle) +static int vi_common_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return vi_common_hw_fini(adev); + return vi_common_hw_fini(ip_block); } -static int vi_common_resume(void *handle) +static int vi_common_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - return vi_common_hw_init(adev); + return vi_common_hw_init(ip_block); } static bool vi_common_is_idle(void *handle) @@ -1750,16 +1741,6 @@ static bool vi_common_is_idle(void *handle) return true; } -static int vi_common_wait_for_idle(void *handle) -{ - return 0; -} - -static int vi_common_soft_reset(void *handle) -{ - return 0; -} - static void vi_update_bif_medium_grain_light_sleep(struct amdgpu_device *adev, bool enable) { @@ -2047,19 +2028,14 @@ static const struct amd_ip_funcs vi_common_ip_funcs = { .early_init = vi_common_early_init, .late_init = vi_common_late_init, .sw_init = vi_common_sw_init, - .sw_fini = vi_common_sw_fini, .hw_init = vi_common_hw_init, .hw_fini = vi_common_hw_fini, .suspend = vi_common_suspend, .resume = vi_common_resume, .is_idle = vi_common_is_idle, - .wait_for_idle = vi_common_wait_for_idle, - .soft_reset = vi_common_soft_reset, .set_clockgating_state = vi_common_set_clockgating_state, .set_powergating_state = vi_common_set_powergating_state, .get_clockgating_state = vi_common_get_clockgating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; static const struct amdgpu_ip_block_version vi_common_ip_block = diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index 3e6b4736a7fe..065d87841459 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -365,7 +365,7 @@ static int kfd_ioctl_create_queue(struct file *filep, struct kfd_process *p, p->pasid, dev->id); - err = pqm_create_queue(&p->pqm, dev, filep, &q_properties, &queue_id, + err = pqm_create_queue(&p->pqm, dev, &q_properties, &queue_id, NULL, NULL, NULL, &doorbell_offset_in_process); if (err != 0) goto err_create_queue; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c index 48caecf7e72e..723f1220e1cc 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c @@ -28,6 +28,7 @@ #include "kfd_topology.h" #include "amdgpu.h" #include "amdgpu_amdkfd.h" +#include "amdgpu_xgmi.h" /* GPU Processor ID base for dGPUs for which VCRAT needs to be created. * GPU processor ID are expressed with Bit[31]=1. @@ -2329,6 +2330,8 @@ static int kfd_create_vcrat_image_gpu(void *pcrat_image, continue; if (peer_dev->gpu->kfd->hive_id != kdev->kfd->hive_id) continue; + if (!amdgpu_xgmi_get_is_sharing_enabled(kdev->adev, peer_dev->gpu->adev)) + continue; sub_type_hdr = (typeof(sub_type_hdr))( (char *)sub_type_hdr + sizeof(struct crat_subtype_iolink)); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c index fad1c8f2bc83..956198da7859 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c @@ -534,7 +534,8 @@ static void kfd_cwsr_init(struct kfd_dev *kfd) kfd->cwsr_isa = cwsr_trap_gfx11_hex; kfd->cwsr_isa_size = sizeof(cwsr_trap_gfx11_hex); } else { - BUILD_BUG_ON(sizeof(cwsr_trap_gfx12_hex) > PAGE_SIZE); + BUILD_BUG_ON(sizeof(cwsr_trap_gfx12_hex) + > KFD_CWSR_TMA_OFFSET); kfd->cwsr_isa = cwsr_trap_gfx12_hex; kfd->cwsr_isa_size = sizeof(cwsr_trap_gfx12_hex); } @@ -1392,6 +1393,13 @@ void kfd_dec_compute_active(struct kfd_node *node) WARN_ONCE(count < 0, "Compute profile ref. count error"); } +static bool kfd_compute_active(struct kfd_node *node) +{ + if (atomic_read(&node->kfd->compute_profile)) + return true; + return false; +} + void kgd2kfd_smi_event_throttle(struct kfd_dev *kfd, uint64_t throttle_bitmask) { /* @@ -1485,6 +1493,24 @@ int kgd2kfd_stop_sched(struct kfd_dev *kfd, uint32_t node_id) return node->dqm->ops.halt(node->dqm); } +bool kgd2kfd_compute_active(struct kfd_dev *kfd, uint32_t node_id) +{ + struct kfd_node *node; + + if (!kfd->init_complete) + return false; + + if (node_id >= kfd->num_nodes) { + dev_warn(kfd->adev->dev, "Invalid node ID: %u exceeds %u\n", + node_id, kfd->num_nodes - 1); + return false; + } + + node = kfd->nodes[node_id]; + + return kfd_compute_active(node); +} + #if defined(CONFIG_DEBUG_FS) /* This function will send a package to HIQ to hang the HWS diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c index 648f40091aa3..c79fe9069e22 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c @@ -202,6 +202,8 @@ static int add_queue_mes(struct device_queue_manager *dqm, struct queue *q, int r, queue_type; uint64_t wptr_addr_off; + if (!dqm->sched_running || dqm->sched_halt) + return 0; if (!down_read_trylock(&adev->reset_domain->sem)) return -EIO; @@ -270,6 +272,8 @@ static int remove_queue_mes(struct device_queue_manager *dqm, struct queue *q, int r; struct mes_remove_queue_input queue_input; + if (!dqm->sched_running || dqm->sched_halt) + return 0; if (!down_read_trylock(&adev->reset_domain->sem)) return -EIO; @@ -292,7 +296,7 @@ static int remove_queue_mes(struct device_queue_manager *dqm, struct queue *q, return r; } -static int remove_all_queues_mes(struct device_queue_manager *dqm) +static int remove_all_kfd_queues_mes(struct device_queue_manager *dqm) { struct device_process_node *cur; struct device *dev = dqm->dev->adev->dev; @@ -319,6 +323,33 @@ static int remove_all_queues_mes(struct device_queue_manager *dqm) return retval; } +static int add_all_kfd_queues_mes(struct device_queue_manager *dqm) +{ + struct device_process_node *cur; + struct device *dev = dqm->dev->adev->dev; + struct qcm_process_device *qpd; + struct queue *q; + int retval = 0; + + list_for_each_entry(cur, &dqm->queues, list) { + qpd = cur->qpd; + list_for_each_entry(q, &qpd->queues_list, list) { + if (!q->properties.is_active) + continue; + retval = add_queue_mes(dqm, q, qpd); + if (retval) { + dev_err(dev, "%s: Failed to add queue %d for dev %d", + __func__, + q->properties.queue_id, + dqm->dev->id); + return retval; + } + } + } + + return retval; +} + static int suspend_all_queues_mes(struct device_queue_manager *dqm) { struct amdgpu_device *adev = (struct amdgpu_device *)dqm->dev->adev; @@ -1742,7 +1773,7 @@ static int halt_cpsch(struct device_queue_manager *dqm) KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES, 0, USE_DEFAULT_GRACE_PERIOD, false); else - ret = remove_all_queues_mes(dqm); + ret = remove_all_kfd_queues_mes(dqm); } dqm->sched_halt = true; dqm_unlock(dqm); @@ -1768,6 +1799,9 @@ static int unhalt_cpsch(struct device_queue_manager *dqm) ret = execute_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0, USE_DEFAULT_GRACE_PERIOD); + else + ret = add_all_kfd_queues_mes(dqm); + dqm_unlock(dqm); return ret; @@ -1867,7 +1901,7 @@ static int stop_cpsch(struct device_queue_manager *dqm) if (!dqm->dev->kfd->shared_resources.enable_mes) unmap_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES, 0, USE_DEFAULT_GRACE_PERIOD, false); else - remove_all_queues_mes(dqm); + remove_all_kfd_queues_mes(dqm); dqm->sched_running = false; @@ -2048,7 +2082,7 @@ int amdkfd_fence_wait_timeout(struct device_queue_manager *dqm, { unsigned long end_jiffies = msecs_to_jiffies(timeout_ms) + jiffies; struct device *dev = dqm->dev->adev->dev; - uint64_t *fence_addr = dqm->fence_addr; + uint64_t *fence_addr = dqm->fence_addr; while (*fence_addr != fence_value) { /* Fatal err detected, this response won't come */ @@ -2254,6 +2288,7 @@ static int unmap_queues_cpsch(struct device_queue_manager *dqm, goto out; *dqm->fence_addr = KFD_FENCE_INIT; + mb(); pm_send_query_status(&dqm->packet_mgr, dqm->fence_gpu_addr, KFD_FENCE_COMPLETED); /* should be timed out */ @@ -3173,7 +3208,7 @@ struct copy_context_work_handler_workarea { struct kfd_process *p; }; -static void copy_context_work_handler (struct work_struct *work) +static void copy_context_work_handler(struct work_struct *work) { struct copy_context_work_handler_workarea *workarea; struct mqd_manager *mqd_mgr; @@ -3200,6 +3235,9 @@ static void copy_context_work_handler (struct work_struct *work) struct qcm_process_device *qpd = &pdd->qpd; list_for_each_entry(q, &qpd->queues_list, list) { + if (q->properties.type != KFD_QUEUE_TYPE_COMPUTE) + continue; + mqd_mgr = dqm->mqd_mgrs[KFD_MQD_TYPE_CP]; /* We ignore the return value from get_wave_state diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c index 8ee3d07ffbdf..eacfeb32f35d 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c @@ -445,14 +445,13 @@ svm_migrate_vma_to_vram(struct kfd_node *node, struct svm_range *prange, pr_debug("successful/cpages/npages 0x%lx/0x%lx/0x%lx\n", mpages, cpages, migrate.npages); - kfd_smi_event_migration_end(node, p->lead_thread->pid, - start >> PAGE_SHIFT, end >> PAGE_SHIFT, - 0, node->id, trigger); - svm_range_dma_unmap_dev(adev->dev, scratch, 0, npages); out_free: kvfree(buf); + kfd_smi_event_migration_end(node, p->lead_thread->pid, + start >> PAGE_SHIFT, end >> PAGE_SHIFT, + 0, node->id, trigger, r); out: if (!r && mpages) { pdd = svm_range_get_pdd_by_node(prange, node); @@ -751,14 +750,13 @@ svm_migrate_vma_to_ram(struct kfd_node *node, struct svm_range *prange, svm_migrate_copy_done(adev, mfence); migrate_vma_finalize(&migrate); - kfd_smi_event_migration_end(node, p->lead_thread->pid, - start >> PAGE_SHIFT, end >> PAGE_SHIFT, - node->id, 0, trigger); - svm_range_dma_unmap_dev(adev->dev, scratch, 0, npages); out_free: kvfree(buf); + kfd_smi_event_migration_end(node, p->lead_thread->pid, + start >> PAGE_SHIFT, end >> PAGE_SHIFT, + node->id, 0, trigger, r); out: if (!r && cpages) { mpages = cpages - upages; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h index 26e48fdc8728..9e5ca0b93b2a 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h @@ -1347,7 +1347,6 @@ int pqm_init(struct process_queue_manager *pqm, struct kfd_process *p); void pqm_uninit(struct process_queue_manager *pqm); int pqm_create_queue(struct process_queue_manager *pqm, struct kfd_node *dev, - struct file *f, struct queue_properties *properties, unsigned int *qid, const struct kfd_criu_queue_priv_data *q_data, diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c index d4aa843aacfd..87cd52cf4ee9 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c @@ -271,11 +271,9 @@ static int kfd_get_cu_occupancy(struct attribute *attr, char *buffer) struct kfd_process *proc = NULL; struct kfd_process_device *pdd = NULL; int i; - struct kfd_cu_occupancy cu_occupancy[AMDGPU_MAX_QUEUES]; + struct kfd_cu_occupancy *cu_occupancy; u32 queue_format; - memset(cu_occupancy, 0x0, sizeof(cu_occupancy)); - pdd = container_of(attr, struct kfd_process_device, attr_cu_occupancy); dev = pdd->dev; if (dev->kfd2kgd->get_cu_occupancy == NULL) @@ -293,6 +291,10 @@ static int kfd_get_cu_occupancy(struct attribute *attr, char *buffer) wave_cnt = 0; max_waves_per_cu = 0; + cu_occupancy = kcalloc(AMDGPU_MAX_QUEUES, sizeof(*cu_occupancy), GFP_KERNEL); + if (!cu_occupancy) + return -ENOMEM; + /* * For GFX 9.4.3, fetch the CU occupancy from the first XCC in the partition. * For AQL queues, because of cooperative dispatch we multiply the wave count @@ -318,6 +320,7 @@ static int kfd_get_cu_occupancy(struct attribute *attr, char *buffer) /* Translate wave count to number of compute units */ cu_cnt = (wave_cnt + (max_waves_per_cu - 1)) / max_waves_per_cu; + kfree(cu_occupancy); return snprintf(buffer, PAGE_SIZE, "%d\n", cu_cnt); } @@ -338,8 +341,8 @@ static ssize_t kfd_procfs_show(struct kobject *kobj, struct attribute *attr, attr_sdma); struct kfd_sdma_activity_handler_workarea sdma_activity_work_handler; - INIT_WORK(&sdma_activity_work_handler.sdma_activity_work, - kfd_sdma_activity_worker); + INIT_WORK_ONSTACK(&sdma_activity_work_handler.sdma_activity_work, + kfd_sdma_activity_worker); sdma_activity_work_handler.pdd = pdd; sdma_activity_work_handler.sdma_activity_counter = 0; @@ -347,6 +350,7 @@ static ssize_t kfd_procfs_show(struct kobject *kobj, struct attribute *attr, schedule_work(&sdma_activity_work_handler.sdma_activity_work); flush_work(&sdma_activity_work_handler.sdma_activity_work); + destroy_work_on_stack(&sdma_activity_work_handler.sdma_activity_work); return snprintf(buffer, PAGE_SIZE, "%llu\n", (sdma_activity_work_handler.sdma_activity_counter)/ @@ -850,8 +854,10 @@ struct kfd_process *kfd_create_process(struct task_struct *thread) goto out; } - /* A prior open of /dev/kfd could have already created the process. */ - process = find_process(thread, false); + /* A prior open of /dev/kfd could have already created the process. + * find_process will increase process kref in this case + */ + process = find_process(thread, true); if (process) { pr_debug("Process already found\n"); } else { @@ -899,8 +905,6 @@ struct kfd_process *kfd_create_process(struct task_struct *thread) init_waitqueue_head(&process->wait_irq_drain); } out: - if (!IS_ERR(process)) - kref_get(&process->ref); mutex_unlock(&kfd_processes_mutex); mmput(thread->mm); @@ -1186,10 +1190,8 @@ static void kfd_process_ref_release(struct kref *ref) static struct mmu_notifier *kfd_process_alloc_notifier(struct mm_struct *mm) { - int idx = srcu_read_lock(&kfd_processes_srcu); - struct kfd_process *p = find_process_by_mm(mm); - - srcu_read_unlock(&kfd_processes_srcu, idx); + /* This increments p->ref counter if kfd process p exists */ + struct kfd_process *p = kfd_lookup_process_by_mm(mm); return p ? &p->mmu_notifier : ERR_PTR(-ESRCH); } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c index 01b960b15274..c76db22a1000 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c @@ -235,7 +235,7 @@ void pqm_uninit(struct process_queue_manager *pqm) static int init_user_queue(struct process_queue_manager *pqm, struct kfd_node *dev, struct queue **q, struct queue_properties *q_properties, - struct file *f, unsigned int qid) + unsigned int qid) { int retval; @@ -300,7 +300,6 @@ cleanup: int pqm_create_queue(struct process_queue_manager *pqm, struct kfd_node *dev, - struct file *f, struct queue_properties *properties, unsigned int *qid, const struct kfd_criu_queue_priv_data *q_data, @@ -374,7 +373,7 @@ int pqm_create_queue(struct process_queue_manager *pqm, * allocate_sdma_queue() in create_queue() has the * corresponding check logic. */ - retval = init_user_queue(pqm, dev, &q, properties, f, *qid); + retval = init_user_queue(pqm, dev, &q, properties, *qid); if (retval != 0) goto err_create_queue; pqn->q = q; @@ -395,7 +394,7 @@ int pqm_create_queue(struct process_queue_manager *pqm, goto err_create_queue; } - retval = init_user_queue(pqm, dev, &q, properties, f, *qid); + retval = init_user_queue(pqm, dev, &q, properties, *qid); if (retval != 0) goto err_create_queue; pqn->q = q; @@ -1029,8 +1028,7 @@ int kfd_criu_restore_queue(struct kfd_process *p, print_queue_properties(&qp); - ret = pqm_create_queue(&p->pqm, pdd->dev, NULL, &qp, &queue_id, q_data, mqd, ctl_stack, - NULL); + ret = pqm_create_queue(&p->pqm, pdd->dev, &qp, &queue_id, q_data, mqd, ctl_stack, NULL); if (ret) { pr_err("Failed to create new queue err:%d\n", ret); goto exit; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c index de8b9abf7afc..9b8169761ec5 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c @@ -44,7 +44,7 @@ struct kfd_smi_client { bool suser; }; -#define MAX_KFIFO_SIZE 1024 +#define KFD_MAX_KFIFO_SIZE 8192 static __poll_t kfd_smi_ev_poll(struct file *, struct poll_table_struct *); static ssize_t kfd_smi_ev_read(struct file *, char __user *, size_t, loff_t *); @@ -86,7 +86,7 @@ static ssize_t kfd_smi_ev_read(struct file *filep, char __user *user, struct kfd_smi_client *client = filep->private_data; unsigned char *buf; - size = min_t(size_t, size, MAX_KFIFO_SIZE); + size = min_t(size_t, size, KFD_MAX_KFIFO_SIZE); buf = kmalloc(size, GFP_KERNEL); if (!buf) return -ENOMEM; @@ -292,12 +292,13 @@ void kfd_smi_event_migration_start(struct kfd_node *node, pid_t pid, void kfd_smi_event_migration_end(struct kfd_node *node, pid_t pid, unsigned long start, unsigned long end, - uint32_t from, uint32_t to, uint32_t trigger) + uint32_t from, uint32_t to, uint32_t trigger, + int error_code) { kfd_smi_event_add(pid, node, KFD_SMI_EVENT_MIGRATE_END, KFD_EVENT_FMT_MIGRATE_END( ktime_get_boottime_ns(), pid, start, end - start, - from, to, trigger)); + from, to, trigger, error_code)); } void kfd_smi_event_queue_eviction(struct kfd_node *node, pid_t pid, @@ -354,7 +355,7 @@ int kfd_smi_event_open(struct kfd_node *dev, uint32_t *fd) return -ENOMEM; INIT_LIST_HEAD(&client->list); - ret = kfifo_alloc(&client->fifo, MAX_KFIFO_SIZE, GFP_KERNEL); + ret = kfifo_alloc(&client->fifo, KFD_MAX_KFIFO_SIZE, GFP_KERNEL); if (ret) { kfree(client); return ret; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.h b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.h index 85010b8307f8..503bff13d815 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.h @@ -44,7 +44,8 @@ void kfd_smi_event_migration_start(struct kfd_node *node, pid_t pid, uint32_t trigger); void kfd_smi_event_migration_end(struct kfd_node *node, pid_t pid, unsigned long start, unsigned long end, - uint32_t from, uint32_t to, uint32_t trigger); + uint32_t from, uint32_t to, uint32_t trigger, + int error_code); void kfd_smi_event_queue_eviction(struct kfd_node *node, pid_t pid, uint32_t trigger); void kfd_smi_event_queue_restore(struct kfd_node *node, pid_t pid); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index 1893c27746a5..3e2911895c74 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -3111,8 +3111,6 @@ retry_write_locked: start = max_t(unsigned long, ALIGN_DOWN(addr, size), prange->start); last = min_t(unsigned long, ALIGN(addr + 1, size) - 1, prange->last); if (prange->actual_loc != 0 || best_loc != 0) { - migration = true; - if (best_loc) { r = svm_migrate_to_vram(prange, best_loc, start, last, mm, KFD_MIGRATE_TRIGGER_PAGEFAULT_GPU); @@ -3135,7 +3133,9 @@ retry_write_locked: if (r) { pr_debug("failed %d to migrate svms %p [0x%lx 0x%lx]\n", r, svms, start, last); - goto out_unlock_range; + goto out_migrate_fail; + } else { + migration = true; } } @@ -3145,6 +3145,7 @@ retry_write_locked: pr_debug("failed %d to map svms 0x%p [0x%lx 0x%lx] to gpus\n", r, svms, start, last); +out_migrate_fail: kfd_smi_event_page_fault_end(node, p->lead_thread->pid, addr, migration); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c index 3871591c9aec..9476e30d6baa 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c @@ -1998,6 +1998,8 @@ static void kfd_topology_set_capabilities(struct kfd_topology_device *dev) if (KFD_GC_VERSION(dev->gpu) >= IP_VERSION(9, 4, 2)) dev->node_props.capability |= HSA_CAP_TRAP_DEBUG_PRECISE_MEMORY_OPERATIONS_SUPPORTED; + + dev->node_props.capability |= HSA_CAP_PER_QUEUE_RESET_SUPPORTED; } else { dev->node_props.debug_prop |= HSA_DBG_WATCH_ADDR_MASK_LO_BIT_GFX10 | HSA_DBG_WATCH_ADDR_MASK_HI_BIT; diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index 8d97f17ffe66..f0a6816709ca 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -320,18 +320,18 @@ static bool dm_is_idle(void *handle) return true; } -static int dm_wait_for_idle(void *handle) +static int dm_wait_for_idle(struct amdgpu_ip_block *ip_block) { /* XXX todo */ return 0; } -static bool dm_check_soft_reset(void *handle) +static bool dm_check_soft_reset(struct amdgpu_ip_block *ip_block) { return false; } -static int dm_soft_reset(void *handle) +static int dm_soft_reset(struct amdgpu_ip_block *ip_block) { /* XXX todo */ return 0; @@ -968,7 +968,7 @@ static int dm_set_powergating_state(void *handle, } /* Prototypes of private functions */ -static int dm_early_init(void *handle); +static int dm_early_init(struct amdgpu_ip_block *ip_block); /* Allocate memory for FBC compressed data */ static void amdgpu_dm_fbc_init(struct drm_connector *connector) @@ -1307,6 +1307,29 @@ static int dm_dmub_hw_init(struct amdgpu_device *adev) DRM_INFO("DMUB hardware initialized: version=0x%08X\n", adev->dm.dmcub_fw_version); + /* Keeping sanity checks off if + * DCN31 >= 4.0.59.0 + * DCN314 >= 8.0.16.0 + * Otherwise, turn on sanity checks + */ + switch (amdgpu_ip_version(adev, DCE_HWIP, 0)) { + case IP_VERSION(3, 1, 2): + case IP_VERSION(3, 1, 3): + if (adev->dm.dmcub_fw_version && + adev->dm.dmcub_fw_version >= DMUB_FW_VERSION(4, 0, 0) && + adev->dm.dmcub_fw_version < DMUB_FW_VERSION(4, 0, 59)) + adev->dm.dc->debug.sanity_checks = true; + break; + case IP_VERSION(3, 1, 4): + if (adev->dm.dmcub_fw_version && + adev->dm.dmcub_fw_version >= DMUB_FW_VERSION(4, 0, 0) && + adev->dm.dmcub_fw_version < DMUB_FW_VERSION(8, 0, 16)) + adev->dm.dc->debug.sanity_checks = true; + break; + default: + break; + } + return 0; } @@ -1696,6 +1719,26 @@ dm_allocate_gpu_mem( return da->cpu_ptr; } +void +dm_free_gpu_mem( + struct amdgpu_device *adev, + enum dc_gpu_mem_alloc_type type, + void *pvMem) +{ + struct dal_allocation *da; + + /* walk the da list in DM */ + list_for_each_entry(da, &adev->dm.da_list, list) { + if (pvMem == da->cpu_ptr) { + amdgpu_bo_free_kernel(&da->bo, &da->gpu_addr, &da->cpu_ptr); + list_del(&da->list); + kfree(da); + break; + } + } + +} + static enum dmub_status dm_dmub_send_vbios_gpint_command(struct amdgpu_device *adev, enum dmub_gpint_command command_code, @@ -1762,16 +1805,20 @@ static struct dml2_soc_bb *dm_dmub_get_vbios_bounding_box(struct amdgpu_device * /* Send the chunk */ ret = dm_dmub_send_vbios_gpint_command(adev, send_addrs[i], chunk, 30000); if (ret != DMUB_STATUS_OK) - /* No need to free bb here since it shall be done in dm_sw_fini() */ - return NULL; + goto free_bb; } /* Now ask DMUB to copy the bb */ ret = dm_dmub_send_vbios_gpint_command(adev, DMUB_GPINT__BB_COPY, 1, 200000); if (ret != DMUB_STATUS_OK) - return NULL; + goto free_bb; return bb; + +free_bb: + dm_free_gpu_mem(adev, DC_MEM_ALLOC_TYPE_GART, (void *) bb); + return NULL; + } static enum dmub_ips_disable_type dm_get_default_ips_mode( @@ -1886,7 +1933,11 @@ static int amdgpu_dm_init(struct amdgpu_device *adev) else init_data.flags.gpu_vm_support = (amdgpu_sg_display != 0); } else { - init_data.flags.gpu_vm_support = (amdgpu_sg_display != 0) && (adev->flags & AMD_IS_APU); + if (amdgpu_ip_version(adev, DCE_HWIP, 0) == IP_VERSION(2, 0, 3)) + init_data.flags.gpu_vm_support = (amdgpu_sg_display == 1); + else + init_data.flags.gpu_vm_support = + (amdgpu_sg_display != 0) && (adev->flags & AMD_IS_APU); } adev->mode_info.gpu_vm_support = init_data.flags.gpu_vm_support; @@ -2115,9 +2166,9 @@ error: return -EINVAL; } -static int amdgpu_dm_early_fini(void *handle) +static int amdgpu_dm_early_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_dm_audio_fini(adev); @@ -2509,9 +2560,9 @@ static int dm_dmub_sw_init(struct amdgpu_device *adev) return 0; } -static int dm_sw_init(void *handle) +static int dm_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; adev->dm.cgs_device = amdgpu_cgs_create_device(adev); @@ -2531,9 +2582,9 @@ static int dm_sw_init(void *handle) return load_dmcu_fw(adev); } -static int dm_sw_fini(void *handle) +static int dm_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct dal_allocation *da; list_for_each_entry(da, &adev->dm.da_list, list) { @@ -2541,11 +2592,11 @@ static int dm_sw_fini(void *handle) amdgpu_bo_free_kernel(&da->bo, &da->gpu_addr, &da->cpu_ptr); list_del(&da->list); kfree(da); + adev->dm.bb_from_dmub = NULL; break; } } - adev->dm.bb_from_dmub = NULL; kfree(adev->dm.dmub_fb_info); adev->dm.dmub_fb_info = NULL; @@ -2598,9 +2649,9 @@ static int detect_mst_link_for_all_connectors(struct drm_device *dev) return ret; } -static int dm_late_init(void *handle) +static int dm_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct dmcu_iram_parameters params; unsigned int linear_lut[16]; @@ -2790,7 +2841,7 @@ static int amdgpu_dm_smu_write_watermarks_table(struct amdgpu_device *adev) /** * dm_hw_init() - Initialize DC device - * @handle: The base driver device containing the amdgpu_dm device. + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Initialize the &struct amdgpu_display_manager device. This involves calling * the initializers of each DM component, then populating the struct with them. @@ -2808,9 +2859,9 @@ static int amdgpu_dm_smu_write_watermarks_table(struct amdgpu_device *adev) * - Vblank support * - Debug FS entries, if enabled */ -static int dm_hw_init(void *handle) +static int dm_hw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; int r; /* Create DAL display manager */ @@ -2824,15 +2875,15 @@ static int dm_hw_init(void *handle) /** * dm_hw_fini() - Teardown DC device - * @handle: The base driver device containing the amdgpu_dm device. + * @ip_block: Pointer to the amdgpu_ip_block for this hw instance. * * Teardown components within &struct amdgpu_display_manager that require * cleanup. This involves cleaning up the DRM device, DC, and any modules that * were loaded. Also flush IRQ workqueues and disable them. */ -static int dm_hw_fini(void *handle) +static int dm_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; amdgpu_dm_hpd_fini(adev); @@ -2936,9 +2987,9 @@ static void hpd_rx_irq_work_suspend(struct amdgpu_display_manager *dm) } } -static int dm_suspend(void *handle) +static int dm_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_display_manager *dm = &adev->dm; int ret = 0; @@ -3125,9 +3176,9 @@ cleanup: kfree(bundle); } -static int dm_resume(void *handle) +static int dm_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = handle; + struct amdgpu_device *adev = ip_block->adev; struct drm_device *ddev = adev_to_drm(adev); struct amdgpu_display_manager *dm = &adev->dm; struct amdgpu_dm_connector *aconnector; @@ -3142,8 +3193,7 @@ static int dm_resume(void *handle) struct dm_atomic_state *dm_state = to_dm_atomic_state(dm->atomic_obj.state); enum dc_connection_type new_connection_type = dc_connection_none; struct dc_state *dc_state; - int i, r, j, ret; - bool need_hotplug = false; + int i, r, j; struct dc_commit_streams_params commit_params = {}; if (dm->dc->caps.ips_support) { @@ -3332,23 +3382,16 @@ static int dm_resume(void *handle) aconnector->mst_root) continue; - ret = drm_dp_mst_topology_mgr_resume(&aconnector->mst_mgr, true); - - if (ret < 0) { - dm_helpers_dp_mst_stop_top_mgr(aconnector->dc_link->ctx, - aconnector->dc_link); - need_hotplug = true; - } + drm_dp_mst_topology_queue_probe(&aconnector->mst_mgr); } drm_connector_list_iter_end(&iter); - if (need_hotplug) - drm_kms_helper_hotplug_event(ddev); - amdgpu_dm_irq_resume_late(adev); amdgpu_dm_smu_write_watermarks_table(adev); + drm_kms_helper_hotplug_event(ddev); + return 0; } @@ -3379,8 +3422,6 @@ static const struct amd_ip_funcs amdgpu_dm_funcs = { .soft_reset = dm_soft_reset, .set_clockgating_state = dm_set_clockgating_state, .set_powergating_state = dm_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; const struct amdgpu_ip_block_version dm_ip_block = { @@ -3495,7 +3536,7 @@ void amdgpu_dm_update_connector_after_detect( aconnector->dc_sink = sink; dc_sink_retain(aconnector->dc_sink); amdgpu_dm_update_freesync_caps(connector, - aconnector->edid); + aconnector->drm_edid); } else { amdgpu_dm_update_freesync_caps(connector, NULL); if (!aconnector->dc_sink) { @@ -3554,18 +3595,19 @@ void amdgpu_dm_update_connector_after_detect( aconnector->dc_sink = sink; dc_sink_retain(aconnector->dc_sink); if (sink->dc_edid.length == 0) { - aconnector->edid = NULL; + aconnector->drm_edid = NULL; if (aconnector->dc_link->aux_mode) { - drm_dp_cec_unset_edid( - &aconnector->dm_dp_aux.aux); + drm_dp_cec_unset_edid(&aconnector->dm_dp_aux.aux); } } else { - aconnector->edid = - (struct edid *)sink->dc_edid.raw_edid; + const struct edid *edid = (const struct edid *)sink->dc_edid.raw_edid; + + aconnector->drm_edid = drm_edid_alloc(edid, sink->dc_edid.length); + drm_edid_connector_update(connector, aconnector->drm_edid); if (aconnector->dc_link->aux_mode) - drm_dp_cec_set_edid(&aconnector->dm_dp_aux.aux, - aconnector->edid); + drm_dp_cec_attach(&aconnector->dm_dp_aux.aux, + connector->display_info.source_physical_address); } if (!aconnector->timing_requested) { @@ -3576,17 +3618,16 @@ void amdgpu_dm_update_connector_after_detect( "failed to create aconnector->requested_timing\n"); } - drm_connector_update_edid_property(connector, aconnector->edid); - amdgpu_dm_update_freesync_caps(connector, aconnector->edid); + amdgpu_dm_update_freesync_caps(connector, aconnector->drm_edid); update_connector_ext_caps(aconnector); } else { drm_dp_cec_unset_edid(&aconnector->dm_dp_aux.aux); amdgpu_dm_update_freesync_caps(connector, NULL); - drm_connector_update_edid_property(connector, NULL); aconnector->num_modes = 0; dc_sink_release(aconnector->dc_sink); aconnector->dc_sink = NULL; - aconnector->edid = NULL; + drm_edid_free(aconnector->drm_edid); + aconnector->drm_edid = NULL; kfree(aconnector->timing_requested); aconnector->timing_requested = NULL; /* Set CP to DESIRED if it was ENABLED, so we can re-enable it again on hotplug */ @@ -4622,7 +4663,12 @@ static void amdgpu_dm_backlight_set_level(struct amdgpu_display_manager *dm, if (!rc) DRM_DEBUG("DM: Failed to update backlight via AUX on eDP[%d]\n", bl_idx); } else { - rc = dc_link_set_backlight_level(link, brightness, 0); + struct set_backlight_level_params backlight_level_params = { 0 }; + + backlight_level_params.backlight_pwm_u16_16 = brightness; + backlight_level_params.transition_time_in_ms = 0; + + rc = dc_link_set_backlight_level(link, &backlight_level_params); if (!rc) DRM_DEBUG("DM: Failed to update backlight on eDP[%d]\n", bl_idx); } @@ -5177,15 +5223,20 @@ static ssize_t s3_debug_store(struct device *device, int s3_state; struct drm_device *drm_dev = dev_get_drvdata(device); struct amdgpu_device *adev = drm_to_adev(drm_dev); + struct amdgpu_ip_block *ip_block; + + ip_block = amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_DCE); + if (!ip_block) + return -EINVAL; ret = kstrtoint(buf, 0, &s3_state); if (ret == 0) { if (s3_state) { - dm_resume(adev); + dm_resume(ip_block); drm_kms_helper_hotplug_event(adev_to_drm(adev)); } else - dm_suspend(adev); + dm_suspend(ip_block); } return ret == 0 ? count : 0; @@ -5257,9 +5308,9 @@ static int dm_init_microcode(struct amdgpu_device *adev) return r; } -static int dm_early_init(void *handle) +static int dm_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct amdgpu_mode_info *mode_info = &adev->mode_info; struct atom_context *ctx = mode_info->atom_context; int index = GetIndexIntoMasterTable(DATA, Object_Header); @@ -7122,32 +7173,24 @@ static void amdgpu_dm_connector_funcs_force(struct drm_connector *connector) struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector); struct dc_link *dc_link = aconnector->dc_link; struct dc_sink *dc_em_sink = aconnector->dc_em_sink; - struct edid *edid; - struct i2c_adapter *ddc; + const struct drm_edid *drm_edid; - if (dc_link && dc_link->aux_mode) - ddc = &aconnector->dm_dp_aux.aux.ddc; - else - ddc = &aconnector->i2c->base; - - /* - * Note: drm_get_edid gets edid in the following order: - * 1) override EDID if set via edid_override debugfs, - * 2) firmware EDID if set via edid_firmware module parameter - * 3) regular DDC read. - */ - edid = drm_get_edid(connector, ddc); - if (!edid) { + drm_edid = drm_edid_read(connector); + drm_edid_connector_update(connector, drm_edid); + if (!drm_edid) { DRM_ERROR("No EDID found on connector: %s.\n", connector->name); return; } - aconnector->edid = edid; - + aconnector->drm_edid = drm_edid; /* Update emulated (virtual) sink's EDID */ if (dc_em_sink && dc_link) { + // FIXME: Get rid of drm_edid_raw() + const struct edid *edid = drm_edid_raw(drm_edid); + memset(&dc_em_sink->edid_caps, 0, sizeof(struct dc_edid_caps)); - memmove(dc_em_sink->dc_edid.raw_edid, edid, (edid->extensions + 1) * EDID_LENGTH); + memmove(dc_em_sink->dc_edid.raw_edid, edid, + (edid->extensions + 1) * EDID_LENGTH); dm_helpers_parse_edid_caps( dc_link, &dc_em_sink->dc_edid, @@ -7177,36 +7220,26 @@ static int get_modes(struct drm_connector *connector) static void create_eml_sink(struct amdgpu_dm_connector *aconnector) { struct drm_connector *connector = &aconnector->base; - struct dc_link *dc_link = aconnector->dc_link; struct dc_sink_init_data init_params = { .link = aconnector->dc_link, .sink_signal = SIGNAL_TYPE_VIRTUAL }; - struct edid *edid; - struct i2c_adapter *ddc; + const struct drm_edid *drm_edid; + const struct edid *edid; - if (dc_link->aux_mode) - ddc = &aconnector->dm_dp_aux.aux.ddc; - else - ddc = &aconnector->i2c->base; - - /* - * Note: drm_get_edid gets edid in the following order: - * 1) override EDID if set via edid_override debugfs, - * 2) firmware EDID if set via edid_firmware module parameter - * 3) regular DDC read. - */ - edid = drm_get_edid(connector, ddc); - if (!edid) { + drm_edid = drm_edid_read(connector); + drm_edid_connector_update(connector, drm_edid); + if (!drm_edid) { DRM_ERROR("No EDID found on connector: %s.\n", connector->name); return; } - if (drm_detect_hdmi_monitor(edid)) + if (connector->display_info.is_hdmi) init_params.sink_signal = SIGNAL_TYPE_HDMI_TYPE_A; - aconnector->edid = edid; + aconnector->drm_edid = drm_edid; + edid = drm_edid_raw(drm_edid); // FIXME: Get rid of drm_edid_raw() aconnector->dc_em_sink = dc_link_add_remote_sink( aconnector->dc_link, (uint8_t *)edid, @@ -7313,10 +7346,15 @@ create_validate_stream_for_sink(struct amdgpu_dm_connector *aconnector, const struct drm_connector_state *drm_state = dm_state ? &dm_state->base : NULL; int requested_bpc = drm_state ? drm_state->max_requested_bpc : 8; enum dc_status dc_result = DC_OK; + uint8_t bpc_limit = 6; if (!dm_state) return NULL; + if (aconnector->dc_link->connector_signal == SIGNAL_TYPE_HDMI_TYPE_A || + aconnector->dc_link->dpcd_caps.dongle_type == DISPLAY_DONGLE_DP_HDMI_CONVERTER) + bpc_limit = 8; + do { stream = create_stream_for_sink(connector, drm_mode, dm_state, old_stream, @@ -7337,11 +7375,12 @@ create_validate_stream_for_sink(struct amdgpu_dm_connector *aconnector, dc_result = dm_validate_stream_and_context(adev->dm.dc, stream); if (dc_result != DC_OK) { - DRM_DEBUG_KMS("Mode %dx%d (clk %d) failed DC validation with error %d (%s)\n", + DRM_DEBUG_KMS("Mode %dx%d (clk %d) pixel_encoding:%s color_depth:%s failed validation -- %s\n", drm_mode->hdisplay, drm_mode->vdisplay, drm_mode->clock, - dc_result, + dc_pixel_encoding_to_str(stream->timing.pixel_encoding), + dc_color_depth_to_str(stream->timing.display_color_depth), dc_status_to_str(dc_result)); dc_stream_release(stream); @@ -7349,10 +7388,13 @@ create_validate_stream_for_sink(struct amdgpu_dm_connector *aconnector, requested_bpc -= 2; /* lower bpc to retry validation */ } - } while (stream == NULL && requested_bpc >= 6); + } while (stream == NULL && requested_bpc >= bpc_limit); - if (dc_result == DC_FAIL_ENC_VALIDATE && !aconnector->force_yuv420_output) { - DRM_DEBUG_KMS("Retry forcing YCbCr420 encoding\n"); + if ((dc_result == DC_FAIL_ENC_VALIDATE || + dc_result == DC_EXCEED_DONGLE_CAP) && + !aconnector->force_yuv420_output) { + DRM_DEBUG_KMS("%s:%d Retry forcing yuv420 encoding\n", + __func__, __LINE__); aconnector->force_yuv420_output = true; stream = create_validate_stream_for_sink(aconnector, drm_mode, @@ -7893,16 +7935,16 @@ static void amdgpu_set_panel_orientation(struct drm_connector *connector) } static void amdgpu_dm_connector_ddc_get_modes(struct drm_connector *connector, - struct edid *edid) + const struct drm_edid *drm_edid) { struct amdgpu_dm_connector *amdgpu_dm_connector = to_amdgpu_dm_connector(connector); - if (edid) { + if (drm_edid) { /* empty probed_modes */ INIT_LIST_HEAD(&connector->probed_modes); amdgpu_dm_connector->num_modes = - drm_add_edid_modes(connector, edid); + drm_edid_connector_add_modes(connector); /* sorting the probed modes before calling function * amdgpu_dm_get_native_mode() since EDID can have @@ -7916,10 +7958,10 @@ static void amdgpu_dm_connector_ddc_get_modes(struct drm_connector *connector, amdgpu_dm_get_native_mode(connector); /* Freesync capabilities are reset by calling - * drm_add_edid_modes() and need to be + * drm_edid_connector_add_modes() and need to be * restored here. */ - amdgpu_dm_update_freesync_caps(connector, edid); + amdgpu_dm_update_freesync_caps(connector, drm_edid); } else { amdgpu_dm_connector->num_modes = 0; } @@ -8015,12 +8057,12 @@ static uint add_fs_modes(struct amdgpu_dm_connector *aconnector) } static void amdgpu_dm_connector_add_freesync_modes(struct drm_connector *connector, - struct edid *edid) + const struct drm_edid *drm_edid) { struct amdgpu_dm_connector *amdgpu_dm_connector = to_amdgpu_dm_connector(connector); - if (!(amdgpu_freesync_vid_mode && edid)) + if (!(amdgpu_freesync_vid_mode && drm_edid)) return; if (amdgpu_dm_connector->max_vfreq - amdgpu_dm_connector->min_vfreq > 10) @@ -8033,24 +8075,24 @@ static int amdgpu_dm_connector_get_modes(struct drm_connector *connector) struct amdgpu_dm_connector *amdgpu_dm_connector = to_amdgpu_dm_connector(connector); struct drm_encoder *encoder; - struct edid *edid = amdgpu_dm_connector->edid; + const struct drm_edid *drm_edid = amdgpu_dm_connector->drm_edid; struct dc_link_settings *verified_link_cap = &amdgpu_dm_connector->dc_link->verified_link_cap; const struct dc *dc = amdgpu_dm_connector->dc_link->dc; encoder = amdgpu_dm_connector_to_encoder(connector); - if (!drm_edid_is_valid(edid)) { + if (!drm_edid) { amdgpu_dm_connector->num_modes = drm_add_modes_noedid(connector, 640, 480); if (dc->link_srv->dp_get_encoding_format(verified_link_cap) == DP_128b_132b_ENCODING) amdgpu_dm_connector->num_modes += drm_add_modes_noedid(connector, 1920, 1080); } else { - amdgpu_dm_connector_ddc_get_modes(connector, edid); + amdgpu_dm_connector_ddc_get_modes(connector, drm_edid); if (encoder) amdgpu_dm_connector_add_common_modes(encoder, connector); - amdgpu_dm_connector_add_freesync_modes(connector, edid); + amdgpu_dm_connector_add_freesync_modes(connector, drm_edid); } amdgpu_dm_fbc_init(connector); @@ -9580,7 +9622,7 @@ static void amdgpu_dm_commit_streams(struct drm_atomic_state *state, WARN_ON(!dc_commit_streams(dm->dc, ¶ms)); /* Allow idle optimization when vblank count is 0 for display off */ - if (dm->active_vblank_irq_count == 0) + if ((dm->active_vblank_irq_count == 0) && amdgpu_dm_is_headless(dm->adev)) dc_allow_idle_optimizations(dm->dc, true); mutex_unlock(&dm->dc_lock); @@ -10124,6 +10166,8 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state) for (i = 0; i < crtc_disable_count; i++) pm_runtime_put_autosuspend(dev->dev); pm_runtime_mark_last_busy(dev->dev); + + trace_amdgpu_dm_atomic_commit_tail_finish(state); } static int dm_force_atomic_commit(struct drm_connector *connector) @@ -12024,7 +12068,7 @@ static bool parse_edid_cea(struct amdgpu_dm_connector *aconnector, } static void parse_edid_displayid_vrr(struct drm_connector *connector, - struct edid *edid) + const struct edid *edid) { u8 *edid_ext = NULL; int i; @@ -12067,7 +12111,7 @@ static void parse_edid_displayid_vrr(struct drm_connector *connector, } static int parse_amd_vsdb(struct amdgpu_dm_connector *aconnector, - struct edid *edid, struct amdgpu_hdmi_vsdb_info *vsdb_info) + const struct edid *edid, struct amdgpu_hdmi_vsdb_info *vsdb_info) { u8 *edid_ext = NULL; int i; @@ -12102,7 +12146,8 @@ static int parse_amd_vsdb(struct amdgpu_dm_connector *aconnector, } static int parse_hdmi_amd_vsdb(struct amdgpu_dm_connector *aconnector, - struct edid *edid, struct amdgpu_hdmi_vsdb_info *vsdb_info) + const struct edid *edid, + struct amdgpu_hdmi_vsdb_info *vsdb_info) { u8 *edid_ext = NULL; int i; @@ -12136,7 +12181,7 @@ static int parse_hdmi_amd_vsdb(struct amdgpu_dm_connector *aconnector, * amdgpu_dm_update_freesync_caps - Update Freesync capabilities * * @connector: Connector to query. - * @edid: EDID from monitor + * @drm_edid: DRM EDID from monitor * * Amdgpu supports Freesync in DP and HDMI displays, and it is required to keep * track of some of the display information in the internal data struct used by @@ -12144,19 +12189,16 @@ static int parse_hdmi_amd_vsdb(struct amdgpu_dm_connector *aconnector, * FreeSync parameters. */ void amdgpu_dm_update_freesync_caps(struct drm_connector *connector, - struct edid *edid) + const struct drm_edid *drm_edid) { int i = 0; - struct detailed_timing *timing; - struct detailed_non_pixel *data; - struct detailed_data_monitor_range *range; struct amdgpu_dm_connector *amdgpu_dm_connector = to_amdgpu_dm_connector(connector); struct dm_connector_state *dm_con_state = NULL; struct dc_sink *sink; - struct amdgpu_device *adev = drm_to_adev(connector->dev); struct amdgpu_hdmi_vsdb_info vsdb_info = {0}; + const struct edid *edid; bool freesync_capable = false; enum adaptive_sync_type as_type = ADAPTIVE_SYNC_TYPE_NONE; @@ -12169,13 +12211,13 @@ void amdgpu_dm_update_freesync_caps(struct drm_connector *connector, amdgpu_dm_connector->dc_sink : amdgpu_dm_connector->dc_em_sink; - if (!edid || !sink) { + drm_edid_connector_update(connector, drm_edid); + + if (!drm_edid || !sink) { dm_con_state = to_dm_connector_state(connector->state); amdgpu_dm_connector->min_vfreq = 0; amdgpu_dm_connector->max_vfreq = 0; - connector->display_info.monitor_range.min_vfreq = 0; - connector->display_info.monitor_range.max_vfreq = 0; freesync_capable = false; goto update; @@ -12186,6 +12228,8 @@ void amdgpu_dm_update_freesync_caps(struct drm_connector *connector, if (!adev->dm.freesync_module) goto update; + edid = drm_edid_raw(drm_edid); // FIXME: Get rid of drm_edid_raw() + /* Some eDP panels only have the refresh rate range info in DisplayID */ if ((connector->display_info.monitor_range.min_vfreq == 0 || connector->display_info.monitor_range.max_vfreq == 0)) @@ -12193,67 +12237,10 @@ void amdgpu_dm_update_freesync_caps(struct drm_connector *connector, if (edid && (sink->sink_signal == SIGNAL_TYPE_DISPLAY_PORT || sink->sink_signal == SIGNAL_TYPE_EDP)) { - bool edid_check_required = false; - - if (amdgpu_dm_connector->dc_link && - amdgpu_dm_connector->dc_link->dpcd_caps.allow_invalid_MSA_timing_param) { - if (edid->features & DRM_EDID_FEATURE_CONTINUOUS_FREQ) { - amdgpu_dm_connector->min_vfreq = connector->display_info.monitor_range.min_vfreq; - amdgpu_dm_connector->max_vfreq = connector->display_info.monitor_range.max_vfreq; - if (amdgpu_dm_connector->max_vfreq - - amdgpu_dm_connector->min_vfreq > 10) - freesync_capable = true; - } else { - edid_check_required = edid->version > 1 || - (edid->version == 1 && - edid->revision > 1); - } - } - - if (edid_check_required) { - for (i = 0; i < 4; i++) { - - timing = &edid->detailed_timings[i]; - data = &timing->data.other_data; - range = &data->data.range; - /* - * Check if monitor has continuous frequency mode - */ - if (data->type != EDID_DETAIL_MONITOR_RANGE) - continue; - /* - * Check for flag range limits only. If flag == 1 then - * no additional timing information provided. - * Default GTF, GTF Secondary curve and CVT are not - * supported - */ - if (range->flags != 1) - continue; - - connector->display_info.monitor_range.min_vfreq = range->min_vfreq; - connector->display_info.monitor_range.max_vfreq = range->max_vfreq; - - if (edid->revision >= 4) { - if (data->pad2 & DRM_EDID_RANGE_OFFSET_MIN_VFREQ) - connector->display_info.monitor_range.min_vfreq += 255; - if (data->pad2 & DRM_EDID_RANGE_OFFSET_MAX_VFREQ) - connector->display_info.monitor_range.max_vfreq += 255; - } - - amdgpu_dm_connector->min_vfreq = - connector->display_info.monitor_range.min_vfreq; - amdgpu_dm_connector->max_vfreq = - connector->display_info.monitor_range.max_vfreq; - - break; - } - - if (amdgpu_dm_connector->max_vfreq - - amdgpu_dm_connector->min_vfreq > 10) { - - freesync_capable = true; - } - } + amdgpu_dm_connector->min_vfreq = connector->display_info.monitor_range.min_vfreq; + amdgpu_dm_connector->max_vfreq = connector->display_info.monitor_range.max_vfreq; + if (amdgpu_dm_connector->max_vfreq - amdgpu_dm_connector->min_vfreq > 10) + freesync_capable = true; parse_amd_vsdb(amdgpu_dm_connector, edid, &vsdb_info); if (vsdb_info.replay_mode) { @@ -12262,12 +12249,9 @@ void amdgpu_dm_update_freesync_caps(struct drm_connector *connector, amdgpu_dm_connector->as_type = ADAPTIVE_SYNC_TYPE_EDP; } - } else if (edid && sink->sink_signal == SIGNAL_TYPE_HDMI_TYPE_A) { + } else if (drm_edid && sink->sink_signal == SIGNAL_TYPE_HDMI_TYPE_A) { i = parse_hdmi_amd_vsdb(amdgpu_dm_connector, edid, &vsdb_info); if (i >= 0 && vsdb_info.freesync_supported) { - timing = &edid->detailed_timings[i]; - data = &timing->data.other_data; - amdgpu_dm_connector->min_vfreq = vsdb_info.min_refresh_rate_hz; amdgpu_dm_connector->max_vfreq = vsdb_info.max_refresh_rate_hz; if (amdgpu_dm_connector->max_vfreq - amdgpu_dm_connector->min_vfreq > 10) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h index 90dfffec33cf..6464a8378387 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -673,7 +673,7 @@ struct amdgpu_dm_connector { /* we need to mind the EDID between detect and get modes due to analog/digital/tvencoder */ - struct edid *edid; + const struct drm_edid *drm_edid; /* shared with amdgpu */ struct amdgpu_hpd hpd; @@ -951,7 +951,7 @@ void dm_restore_drm_connector_state(struct drm_device *dev, struct drm_connector *connector); void amdgpu_dm_update_freesync_caps(struct drm_connector *connector, - struct edid *edid); + const struct drm_edid *drm_edid); void amdgpu_dm_trigger_timing_sync(struct drm_device *dev); @@ -1004,6 +1004,9 @@ void *dm_allocate_gpu_mem(struct amdgpu_device *adev, enum dc_gpu_mem_alloc_type type, size_t size, long long *addr); +void dm_free_gpu_mem(struct amdgpu_device *adev, + enum dc_gpu_mem_alloc_type type, + void *addr); bool amdgpu_dm_is_headless(struct amdgpu_device *adev); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c index 288be19db7c1..64a041c2af05 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c @@ -35,8 +35,8 @@ #include "amdgpu_dm_trace.h" #include "amdgpu_dm_debugfs.h" -#define HPD_DETECTION_PERIOD_uS 5000000 -#define HPD_DETECTION_TIME_uS 1000 +#define HPD_DETECTION_PERIOD_uS 2000000 +#define HPD_DETECTION_TIME_uS 100000 void amdgpu_dm_crtc_handle_vblank(struct amdgpu_crtc *acrtc) { @@ -154,6 +154,7 @@ static void amdgpu_dm_crtc_set_panel_sr_feature( amdgpu_dm_psr_enable(vblank_work->stream); if (dm->idle_workqueue && + (dm->dc->config.disable_ips == DMUB_IPS_ENABLE) && dm->dc->idle_optimizations_allowed && dm->idle_workqueue->enable && !dm->idle_workqueue->running) @@ -251,10 +252,8 @@ static void amdgpu_dm_crtc_vblank_control_worker(struct work_struct *work) else if (dm->active_vblank_irq_count) dm->active_vblank_irq_count--; - if (dm->active_vblank_irq_count > 0) { - DRM_DEBUG_KMS("Allow idle optimizations (MALL): false\n"); + if (dm->active_vblank_irq_count > 0) dc_allow_idle_optimizations(dm->dc, false); - } /* * Control PSR based on vblank requirements from OS @@ -272,10 +271,8 @@ static void amdgpu_dm_crtc_vblank_control_worker(struct work_struct *work) vblank_work->acrtc->dm_irq_params.allow_sr_entry); } - if (dm->active_vblank_irq_count == 0) { - DRM_DEBUG_KMS("Allow idle optimizations (MALL): true\n"); + if (dm->active_vblank_irq_count == 0) dc_allow_idle_optimizations(dm->dc, true); - } mutex_unlock(&dm->dc_lock); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c index db56b0aa5454..6a97bb2d9160 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c @@ -1529,7 +1529,6 @@ static ssize_t dp_dsc_clock_en_read(struct file *f, char __user *buf, size_t size, loff_t *pos) { char *rd_buf = NULL; - char *rd_buf_ptr = NULL; struct amdgpu_dm_connector *aconnector = file_inode(f)->i_private; struct display_stream_compressor *dsc; struct dcn_dsc_state dsc_state = {0}; @@ -1543,8 +1542,6 @@ static ssize_t dp_dsc_clock_en_read(struct file *f, char __user *buf, if (!rd_buf) return -ENOMEM; - rd_buf_ptr = rd_buf; - for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && @@ -1558,10 +1555,9 @@ static ssize_t dp_dsc_clock_en_read(struct file *f, char __user *buf, if (dsc) dsc->funcs->dsc_read_state(dsc, &dsc_state); - snprintf(rd_buf_ptr, str_len, + snprintf(rd_buf, str_len, "%d\n", dsc_state.dsc_clock_en); - rd_buf_ptr += str_len; while (size) { if (*pos >= rd_buf_size) @@ -1719,7 +1715,6 @@ static ssize_t dp_dsc_slice_width_read(struct file *f, char __user *buf, size_t size, loff_t *pos) { char *rd_buf = NULL; - char *rd_buf_ptr = NULL; struct amdgpu_dm_connector *aconnector = file_inode(f)->i_private; struct display_stream_compressor *dsc; struct dcn_dsc_state dsc_state = {0}; @@ -1733,8 +1728,6 @@ static ssize_t dp_dsc_slice_width_read(struct file *f, char __user *buf, if (!rd_buf) return -ENOMEM; - rd_buf_ptr = rd_buf; - for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && @@ -1748,10 +1741,9 @@ static ssize_t dp_dsc_slice_width_read(struct file *f, char __user *buf, if (dsc) dsc->funcs->dsc_read_state(dsc, &dsc_state); - snprintf(rd_buf_ptr, str_len, + snprintf(rd_buf, str_len, "%d\n", dsc_state.dsc_slice_width); - rd_buf_ptr += str_len; while (size) { if (*pos >= rd_buf_size) @@ -1907,7 +1899,6 @@ static ssize_t dp_dsc_slice_height_read(struct file *f, char __user *buf, size_t size, loff_t *pos) { char *rd_buf = NULL; - char *rd_buf_ptr = NULL; struct amdgpu_dm_connector *aconnector = file_inode(f)->i_private; struct display_stream_compressor *dsc; struct dcn_dsc_state dsc_state = {0}; @@ -1921,8 +1912,6 @@ static ssize_t dp_dsc_slice_height_read(struct file *f, char __user *buf, if (!rd_buf) return -ENOMEM; - rd_buf_ptr = rd_buf; - for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && @@ -1936,10 +1925,9 @@ static ssize_t dp_dsc_slice_height_read(struct file *f, char __user *buf, if (dsc) dsc->funcs->dsc_read_state(dsc, &dsc_state); - snprintf(rd_buf_ptr, str_len, + snprintf(rd_buf, str_len, "%d\n", dsc_state.dsc_slice_height); - rd_buf_ptr += str_len; while (size) { if (*pos >= rd_buf_size) @@ -2091,7 +2079,6 @@ static ssize_t dp_dsc_bits_per_pixel_read(struct file *f, char __user *buf, size_t size, loff_t *pos) { char *rd_buf = NULL; - char *rd_buf_ptr = NULL; struct amdgpu_dm_connector *aconnector = file_inode(f)->i_private; struct display_stream_compressor *dsc; struct dcn_dsc_state dsc_state = {0}; @@ -2105,8 +2092,6 @@ static ssize_t dp_dsc_bits_per_pixel_read(struct file *f, char __user *buf, if (!rd_buf) return -ENOMEM; - rd_buf_ptr = rd_buf; - for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && @@ -2120,10 +2105,9 @@ static ssize_t dp_dsc_bits_per_pixel_read(struct file *f, char __user *buf, if (dsc) dsc->funcs->dsc_read_state(dsc, &dsc_state); - snprintf(rd_buf_ptr, str_len, + snprintf(rd_buf, str_len, "%d\n", dsc_state.dsc_bits_per_pixel); - rd_buf_ptr += str_len; while (size) { if (*pos >= rd_buf_size) @@ -2270,7 +2254,6 @@ static ssize_t dp_dsc_pic_width_read(struct file *f, char __user *buf, size_t size, loff_t *pos) { char *rd_buf = NULL; - char *rd_buf_ptr = NULL; struct amdgpu_dm_connector *aconnector = file_inode(f)->i_private; struct display_stream_compressor *dsc; struct dcn_dsc_state dsc_state = {0}; @@ -2284,8 +2267,6 @@ static ssize_t dp_dsc_pic_width_read(struct file *f, char __user *buf, if (!rd_buf) return -ENOMEM; - rd_buf_ptr = rd_buf; - for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && @@ -2299,10 +2280,9 @@ static ssize_t dp_dsc_pic_width_read(struct file *f, char __user *buf, if (dsc) dsc->funcs->dsc_read_state(dsc, &dsc_state); - snprintf(rd_buf_ptr, str_len, + snprintf(rd_buf, str_len, "%d\n", dsc_state.dsc_pic_width); - rd_buf_ptr += str_len; while (size) { if (*pos >= rd_buf_size) @@ -2328,7 +2308,6 @@ static ssize_t dp_dsc_pic_height_read(struct file *f, char __user *buf, size_t size, loff_t *pos) { char *rd_buf = NULL; - char *rd_buf_ptr = NULL; struct amdgpu_dm_connector *aconnector = file_inode(f)->i_private; struct display_stream_compressor *dsc; struct dcn_dsc_state dsc_state = {0}; @@ -2342,8 +2321,6 @@ static ssize_t dp_dsc_pic_height_read(struct file *f, char __user *buf, if (!rd_buf) return -ENOMEM; - rd_buf_ptr = rd_buf; - for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && @@ -2357,10 +2334,9 @@ static ssize_t dp_dsc_pic_height_read(struct file *f, char __user *buf, if (dsc) dsc->funcs->dsc_read_state(dsc, &dsc_state); - snprintf(rd_buf_ptr, str_len, + snprintf(rd_buf, str_len, "%d\n", dsc_state.dsc_pic_height); - rd_buf_ptr += str_len; while (size) { if (*pos >= rd_buf_size) @@ -2401,7 +2377,6 @@ static ssize_t dp_dsc_chunk_size_read(struct file *f, char __user *buf, size_t size, loff_t *pos) { char *rd_buf = NULL; - char *rd_buf_ptr = NULL; struct amdgpu_dm_connector *aconnector = file_inode(f)->i_private; struct display_stream_compressor *dsc; struct dcn_dsc_state dsc_state = {0}; @@ -2415,8 +2390,6 @@ static ssize_t dp_dsc_chunk_size_read(struct file *f, char __user *buf, if (!rd_buf) return -ENOMEM; - rd_buf_ptr = rd_buf; - for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && @@ -2430,10 +2403,9 @@ static ssize_t dp_dsc_chunk_size_read(struct file *f, char __user *buf, if (dsc) dsc->funcs->dsc_read_state(dsc, &dsc_state); - snprintf(rd_buf_ptr, str_len, + snprintf(rd_buf, str_len, "%d\n", dsc_state.dsc_chunk_size); - rd_buf_ptr += str_len; while (size) { if (*pos >= rd_buf_size) @@ -2474,7 +2446,6 @@ static ssize_t dp_dsc_slice_bpg_offset_read(struct file *f, char __user *buf, size_t size, loff_t *pos) { char *rd_buf = NULL; - char *rd_buf_ptr = NULL; struct amdgpu_dm_connector *aconnector = file_inode(f)->i_private; struct display_stream_compressor *dsc; struct dcn_dsc_state dsc_state = {0}; @@ -2488,8 +2459,6 @@ static ssize_t dp_dsc_slice_bpg_offset_read(struct file *f, char __user *buf, if (!rd_buf) return -ENOMEM; - rd_buf_ptr = rd_buf; - for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && @@ -2503,10 +2472,9 @@ static ssize_t dp_dsc_slice_bpg_offset_read(struct file *f, char __user *buf, if (dsc) dsc->funcs->dsc_read_state(dsc, &dsc_state); - snprintf(rd_buf_ptr, str_len, + snprintf(rd_buf, str_len, "%d\n", dsc_state.dsc_slice_bpg_offset); - rd_buf_ptr += str_len; while (size) { if (*pos >= rd_buf_size) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c index eea317dcbe8c..b0fea0856866 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c @@ -23,6 +23,8 @@ * */ +#include + #include #include #include @@ -642,6 +644,8 @@ static bool execute_synaptics_rc_command(struct drm_dp_aux *aux, // write rc data memmove(rc_data, data, length); ret = drm_dp_dpcd_write(aux, SYNAPTICS_RC_DATA, rc_data, sizeof(rc_data)); + if (ret < 0) + goto err; } // write rc offset @@ -650,20 +654,21 @@ static bool execute_synaptics_rc_command(struct drm_dp_aux *aux, rc_offset[2] = (unsigned char) (offset >> 16) & 0xFF; rc_offset[3] = (unsigned char) (offset >> 24) & 0xFF; ret = drm_dp_dpcd_write(aux, SYNAPTICS_RC_OFFSET, rc_offset, sizeof(rc_offset)); + if (ret < 0) + goto err; // write rc length rc_length[0] = (unsigned char) length & 0xFF; rc_length[1] = (unsigned char) (length >> 8) & 0xFF; ret = drm_dp_dpcd_write(aux, SYNAPTICS_RC_LENGTH, rc_length, sizeof(rc_length)); + if (ret < 0) + goto err; // write rc cmd rc_cmd = cmd | 0x80; ret = drm_dp_dpcd_write(aux, SYNAPTICS_RC_COMMAND, &rc_cmd, sizeof(rc_cmd)); - - if (ret < 0) { - DRM_ERROR("%s: write cmd ..., err = %d\n", __func__, ret); - return false; - } + if (ret < 0) + goto err; // poll until active is 0 for (i = 0; i < 10; i++) { @@ -686,6 +691,10 @@ static bool execute_synaptics_rc_command(struct drm_dp_aux *aux, drm_dbg_dp(aux->drm_dev, "success = %d\n", success); return success; + +err: + DRM_ERROR("%s: write cmd ..., err = %d\n", __func__, ret); + return false; } static void apply_synaptics_fifo_reset_wa(struct drm_dp_aux *aux) @@ -892,6 +901,60 @@ bool dm_helpers_is_dp_sink_present(struct dc_link *link) return dp_sink_present; } +static int +dm_helpers_probe_acpi_edid(void *data, u8 *buf, unsigned int block, size_t len) +{ + struct drm_connector *connector = data; + struct acpi_device *acpidev = ACPI_COMPANION(connector->dev->dev); + unsigned char start = block * EDID_LENGTH; + void *edid; + int r; + + if (!acpidev) + return -ENODEV; + + /* fetch the entire edid from BIOS */ + r = acpi_video_get_edid(acpidev, ACPI_VIDEO_DISPLAY_LCD, -1, &edid); + if (r < 0) { + drm_dbg(connector->dev, "Failed to get EDID from ACPI: %d\n", r); + return r; + } + if (len > r || start > r || start + len > r) { + r = -EINVAL; + goto cleanup; + } + + memcpy(buf, edid + start, len); + r = 0; + +cleanup: + kfree(edid); + + return r; +} + +static const struct drm_edid * +dm_helpers_read_acpi_edid(struct amdgpu_dm_connector *aconnector) +{ + struct drm_connector *connector = &aconnector->base; + + if (amdgpu_dc_debug_mask & DC_DISABLE_ACPI_EDID) + return NULL; + + switch (connector->connector_type) { + case DRM_MODE_CONNECTOR_LVDS: + case DRM_MODE_CONNECTOR_eDP: + break; + default: + return NULL; + } + + if (connector->force == DRM_FORCE_OFF) + return NULL; + + return drm_edid_read_custom(connector, dm_helpers_probe_acpi_edid, connector); +} + enum dc_edid_status dm_helpers_read_local_edid( struct dc_context *ctx, struct dc_link *link, @@ -902,7 +965,8 @@ enum dc_edid_status dm_helpers_read_local_edid( struct i2c_adapter *ddc; int retry = 3; enum dc_edid_status edid_status; - struct edid *edid; + const struct drm_edid *drm_edid; + const struct edid *edid; if (link->aux_mode) ddc = &aconnector->dm_dp_aux.aux.ddc; @@ -913,26 +977,31 @@ enum dc_edid_status dm_helpers_read_local_edid( * do check sum and retry to make sure read correct edid. */ do { - - edid = drm_get_edid(&aconnector->base, ddc); + drm_edid = dm_helpers_read_acpi_edid(aconnector); + if (drm_edid) + drm_info(connector->dev, "Using ACPI provided EDID for %s\n", connector->name); + else + drm_edid = drm_edid_read_ddc(connector, ddc); + drm_edid_connector_update(connector, drm_edid); /* DP Compliance Test 4.2.2.6 */ if (link->aux_mode && connector->edid_corrupt) drm_dp_send_real_edid_checksum(&aconnector->dm_dp_aux.aux, connector->real_edid_checksum); - if (!edid && connector->edid_corrupt) { + if (!drm_edid && connector->edid_corrupt) { connector->edid_corrupt = false; return EDID_BAD_CHECKSUM; } - if (!edid) + if (!drm_edid) return EDID_NO_RESPONSE; + edid = drm_edid_raw(drm_edid); // FIXME: Get rid of drm_edid_raw() sink->dc_edid.length = EDID_LENGTH * (edid->extensions + 1); memmove(sink->dc_edid.raw_edid, (uint8_t *)edid, sink->dc_edid.length); /* We don't need the original edid anymore */ - kfree(edid); + drm_edid_free(drm_edid); edid_status = dm_helpers_parse_edid_caps( link, @@ -1055,17 +1124,8 @@ void dm_helpers_free_gpu_mem( void *pvMem) { struct amdgpu_device *adev = ctx->driver_context; - struct dal_allocation *da; - /* walk the da list in DM */ - list_for_each_entry(da, &adev->dm.da_list, list) { - if (pvMem == da->cpu_ptr) { - amdgpu_bo_free_kernel(&da->bo, &da->gpu_addr, &da->cpu_ptr); - list_del(&da->list); - kfree(da); - break; - } - } + dm_free_gpu_mem(adev, type, pvMem); } bool dm_helpers_dmub_outbox_interrupt_control(struct dc_context *ctx, bool enable) @@ -1314,4 +1374,4 @@ bool dm_helpers_is_hdr_on(struct dc_context *ctx, struct dc_stream_state *stream { // TODO return false; -} \ No newline at end of file +} diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c index a08e8a0b696c..6e4359490613 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c @@ -129,7 +129,7 @@ dm_dp_mst_connector_destroy(struct drm_connector *connector) dc_sink_release(aconnector->dc_sink); } - kfree(aconnector->edid); + drm_edid_free(aconnector->drm_edid); drm_connector_cleanup(connector); drm_dp_mst_put_port_malloc(aconnector->mst_output_port); @@ -182,7 +182,7 @@ amdgpu_dm_mst_connector_early_unregister(struct drm_connector *connector) dc_sink_release(dc_sink); aconnector->dc_sink = NULL; - aconnector->edid = NULL; + aconnector->drm_edid = NULL; aconnector->dsc_aux = NULL; port->passthrough_aux = NULL; } @@ -302,16 +302,18 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector) if (!aconnector) return drm_add_edid_modes(connector, NULL); - if (!aconnector->edid) { - struct edid *edid; + if (!aconnector->drm_edid) { + const struct drm_edid *drm_edid; - edid = drm_dp_mst_get_edid(connector, &aconnector->mst_root->mst_mgr, aconnector->mst_output_port); + drm_edid = drm_dp_mst_edid_read(connector, + &aconnector->mst_root->mst_mgr, + aconnector->mst_output_port); - if (!edid) { + if (!drm_edid) { amdgpu_dm_set_mst_status(&aconnector->mst_status, MST_REMOTE_EDID, false); - drm_connector_update_edid_property( + drm_edid_connector_update( &aconnector->base, NULL); @@ -345,7 +347,7 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector) return ret; } - aconnector->edid = edid; + aconnector->drm_edid = drm_edid; amdgpu_dm_set_mst_status(&aconnector->mst_status, MST_REMOTE_EDID, true); } @@ -360,10 +362,13 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector) struct dc_sink_init_data init_params = { .link = aconnector->dc_link, .sink_signal = SIGNAL_TYPE_DISPLAY_PORT_MST }; + const struct edid *edid; + + edid = drm_edid_raw(aconnector->drm_edid); // FIXME: Get rid of drm_edid_raw() dc_sink = dc_link_add_remote_sink( aconnector->dc_link, - (uint8_t *)aconnector->edid, - (aconnector->edid->extensions + 1) * EDID_LENGTH, + (uint8_t *)edid, + (edid->extensions + 1) * EDID_LENGTH, &init_params); if (!dc_sink) { @@ -405,7 +410,7 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector) if (aconnector->dc_sink) { amdgpu_dm_update_freesync_caps( - connector, aconnector->edid); + connector, aconnector->drm_edid); #if defined(CONFIG_DRM_AMD_DC_FP) if (!validate_dsc_caps_on_connector(aconnector)) @@ -419,10 +424,9 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector) } } - drm_connector_update_edid_property( - &aconnector->base, aconnector->edid); + drm_edid_connector_update(&aconnector->base, aconnector->drm_edid); - ret = drm_add_edid_modes(connector, aconnector->edid); + ret = drm_edid_connector_add_modes(connector); return ret; } @@ -500,7 +504,7 @@ dm_dp_mst_detect(struct drm_connector *connector, dc_sink_release(aconnector->dc_sink); aconnector->dc_sink = NULL; - aconnector->edid = NULL; + aconnector->drm_edid = NULL; aconnector->dsc_aux = NULL; port->passthrough_aux = NULL; @@ -1120,6 +1124,7 @@ static int compute_mst_dsc_configs_for_link(struct drm_atomic_state *state, int i, k, ret; bool debugfs_overwrite = false; uint16_t fec_overhead_multiplier_x1000 = get_fec_overhead_multiplier(dc_link); + struct drm_connector_state *new_conn_state; memset(params, 0, sizeof(params)); @@ -1127,7 +1132,7 @@ static int compute_mst_dsc_configs_for_link(struct drm_atomic_state *state, return PTR_ERR(mst_state); /* Set up params */ - DRM_DEBUG_DRIVER("%s: MST_DSC Set up params for %d streams\n", __func__, dc_state->stream_count); + DRM_DEBUG_DRIVER("%s: MST_DSC Try to set up params from %d streams\n", __func__, dc_state->stream_count); for (i = 0; i < dc_state->stream_count; i++) { struct dc_dsc_policy dsc_policy = {0}; @@ -1143,6 +1148,14 @@ static int compute_mst_dsc_configs_for_link(struct drm_atomic_state *state, if (!aconnector->mst_output_port) continue; + new_conn_state = drm_atomic_get_new_connector_state(state, &aconnector->base); + + if (!new_conn_state) { + DRM_DEBUG_DRIVER("%s:%d MST_DSC Skip the stream 0x%p with invalid new_conn_state\n", + __func__, __LINE__, stream); + continue; + } + stream->timing.flags.DSC = 0; params[count].timing = &stream->timing; @@ -1175,6 +1188,8 @@ static int compute_mst_dsc_configs_for_link(struct drm_atomic_state *state, count++; } + DRM_DEBUG_DRIVER("%s: MST_DSC Params set up for %d streams\n", __func__, count); + if (count == 0) { ASSERT(0); return 0; @@ -1302,7 +1317,7 @@ static bool is_dsc_need_re_compute( continue; aconnector = (struct amdgpu_dm_connector *) stream->dm_stream_context; - if (!aconnector || !aconnector->dsc_aux) + if (!aconnector) continue; stream_on_link[new_stream_on_link_num] = aconnector; diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.c index adc710fe4a45..8d2cf95ae739 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.c +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.c @@ -78,10 +78,3 @@ void bios_set_scratch_critical_state( uint32_t critial_state = state ? 1 : 0; REG_UPDATE(BIOS_SCRATCH_6, S6_CRITICAL_STATE, critial_state); } - -uint32_t bios_get_vga_enabled_displays( - struct dc_bios *bios) -{ - return REG_READ(BIOS_SCRATCH_3) & 0XFFFF; -} - diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.h b/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.h index e1b4a40a353d..ab162f2fe577 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.h +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser_helper.h @@ -34,7 +34,6 @@ uint8_t *bios_get_image(struct dc_bios *bp, uint32_t offset, bool bios_is_accelerated_mode(struct dc_bios *bios); void bios_set_scratch_acc_mode_change(struct dc_bios *bios, uint32_t state); void bios_set_scratch_critical_state(struct dc_bios *bios, bool state); -uint32_t bios_get_vga_enabled_displays(struct dc_bios *bios); #define GET_IMAGE(type, offset) ((type *) bios_get_image(&bp->base, offset, sizeof(type))) diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c index e93df3d6222e..bc123f1884da 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c @@ -50,12 +50,13 @@ #include "link.h" #include "logger_types.h" + + +#include "yellow_carp_offset.h" #undef DC_LOGGER #define DC_LOGGER \ clk_mgr->base.base.ctx->logger -#include "yellow_carp_offset.h" - #define regCLK1_CLK_PLL_REQ 0x0237 #define regCLK1_CLK_PLL_REQ_BASE_IDX 0 diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_clk_mgr.c index 29eff386505a..91d872d6d392 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_clk_mgr.c @@ -53,9 +53,6 @@ #include "logger_types.h" -#undef DC_LOGGER -#define DC_LOGGER \ - clk_mgr->base.base.ctx->logger #define MAX_INSTANCE 7 @@ -77,6 +74,9 @@ static const struct IP_BASE CLK_BASE = { { { { 0x00016C00, 0x02401800, 0, 0, 0, { { 0x0001B200, 0x0242DC00, 0, 0, 0, 0, 0, 0 } }, { { 0x0001B400, 0x0242E000, 0, 0, 0, 0, 0, 0 } } } }; +#undef DC_LOGGER +#define DC_LOGGER \ + clk_mgr->base.base.ctx->logger #define regCLK1_CLK_PLL_REQ 0x0237 #define regCLK1_CLK_PLL_REQ_BASE_IDX 0 diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c index b46a3afe48ca..b77333817f18 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c @@ -55,6 +55,7 @@ #define DC_LOGGER \ clk_mgr->base.base.ctx->logger + #define regCLK1_CLK_PLL_REQ 0x0237 #define regCLK1_CLK_PLL_REQ_BASE_IDX 0 @@ -132,6 +133,8 @@ static void dcn35_disable_otg_wa(struct clk_mgr *clk_mgr_base, struct dc_state * for (i = 0; i < dc->res_pool->pipe_count; ++i) { struct pipe_ctx *old_pipe = &dc->current_state->res_ctx.pipe_ctx[i]; struct pipe_ctx *new_pipe = &context->res_ctx.pipe_ctx[i]; + struct clk_mgr_internal *clk_mgr_internal = TO_CLK_MGR_INTERNAL(clk_mgr_base); + struct dccg *dccg = clk_mgr_internal->dccg; struct pipe_ctx *pipe = safe_to_lower ? &context->res_ctx.pipe_ctx[i] : &dc->current_state->res_ctx.pipe_ctx[i]; @@ -148,8 +151,21 @@ static void dcn35_disable_otg_wa(struct clk_mgr *clk_mgr_base, struct dc_state * new_pipe->stream_res.stream_enc && new_pipe->stream_res.stream_enc->funcs->is_fifo_enabled && new_pipe->stream_res.stream_enc->funcs->is_fifo_enabled(new_pipe->stream_res.stream_enc); - if (pipe->stream && (pipe->stream->dpms_off || dc_is_virtual_signal(pipe->stream->signal) || - !pipe->stream->link_enc) && !stream_changed_otg_dig_on) { + + bool has_active_hpo = false; + + if (old_pipe->stream && new_pipe->stream && old_pipe->stream == new_pipe->stream) { + has_active_hpo = dccg->ctx->dc->link_srv->dp_is_128b_132b_signal(old_pipe) && + dccg->ctx->dc->link_srv->dp_is_128b_132b_signal(new_pipe); + + } + + + if (!has_active_hpo && !dccg->ctx->dc->link_srv->dp_is_128b_132b_signal(pipe) && + (pipe->stream && (pipe->stream->dpms_off || dc_is_virtual_signal(pipe->stream->signal) || + !pipe->stream->link_enc) && !stream_changed_otg_dig_on)) { + + /* This w/a should not trigger when we have a dig active */ if (disable) { if (pipe->stream_res.tg && pipe->stream_res.tg->funcs->immediate_disable_crtc) @@ -257,11 +273,11 @@ static void dcn35_notify_host_router_bw(struct clk_mgr *clk_mgr_base, struct dc_ struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); uint32_t host_router_bw_kbps[MAX_HOST_ROUTERS_NUM] = { 0 }; int i; - for (i = 0; i < context->stream_count; ++i) { const struct dc_stream_state *stream = context->streams[i]; const struct dc_link *link = stream->link; - uint8_t lowest_dpia_index = 0, hr_index = 0; + uint8_t lowest_dpia_index = 0; + unsigned int hr_index = 0; if (!link) continue; @@ -271,6 +287,8 @@ static void dcn35_notify_host_router_bw(struct clk_mgr *clk_mgr_base, struct dc_ continue; hr_index = (link->link_index - lowest_dpia_index) / 2; + if (hr_index >= MAX_HOST_ROUTERS_NUM) + continue; host_router_bw_kbps[hr_index] += dc_bandwidth_in_kbps_from_timing( &stream->timing, dc_link_get_highest_encoding_format(link)); } @@ -975,11 +993,8 @@ static void dcn35_exit_low_power_state(struct clk_mgr *clk_mgr_base) static bool dcn35_is_ips_supported(struct clk_mgr *clk_mgr_base) { struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); - bool ips_supported = true; - ips_supported = dcn35_smu_get_ips_supported(clk_mgr) ? true : false; - - return ips_supported; + return dcn35_smu_get_ips_supported(clk_mgr) ? true : false; } static void dcn35_init_clocks_fpga(struct clk_mgr *clk_mgr) diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index a88f1b6ea64c..7872c6cabb14 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -621,8 +621,8 @@ dc_stream_forward_crc_window(struct dc_stream_state *stream, * dc_stream_configure_crc() - Configure CRC capture for the given stream. * @dc: DC Object * @stream: The stream to configure CRC on. - * @enable: Enable CRC if true, disable otherwise. * @crc_window: CRC window (x/y start/end) information + * @enable: Enable CRC if true, disable otherwise. * @continuous: Capture CRC on every frame if true. Otherwise, only capture * once. * @@ -1157,6 +1157,8 @@ static void dc_update_visual_confirm_color(struct dc *dc, struct dc_state *conte get_surface_visual_confirm_color(pipe_ctx, &(pipe_ctx->visual_confirm_color)); else if (dc->debug.visual_confirm == VISUAL_CONFIRM_SWIZZLE) get_surface_tile_visual_confirm_color(pipe_ctx, &(pipe_ctx->visual_confirm_color)); + else if (dc->debug.visual_confirm == VISUAL_CONFIRM_HW_CURSOR) + get_cursor_visual_confirm_color(pipe_ctx, &(pipe_ctx->visual_confirm_color)); else { if (dc->ctx->dce_version < DCN_VERSION_2_0) color_space_to_black_color( @@ -1233,16 +1235,8 @@ static void disable_dangling_plane(struct dc *dc, struct dc_state *context) */ if (is_phantom) { if (tg->funcs->enable_crtc) { - int main_pipe_width = 0, main_pipe_height = 0; - struct dc_stream_state *old_paired_stream = dc_state_get_paired_subvp_stream(dc->current_state, old_stream); - - if (old_paired_stream) { - main_pipe_width = old_paired_stream->dst.width; - main_pipe_height = old_paired_stream->dst.height; - } - - if (dc->hwss.blank_phantom) - dc->hwss.blank_phantom(dc, tg, main_pipe_width, main_pipe_height); + if (dc->hwseq->funcs.blank_pixel_data) + dc->hwseq->funcs.blank_pixel_data(dc, pipe, true); tg->funcs->enable_crtc(tg); } } @@ -1437,6 +1431,7 @@ void dc_hardware_init(struct dc *dc) detect_edp_presence(dc); if (dc->ctx->dce_environment != DCE_ENV_VIRTUAL_HW) dc->hwss.init_hw(dc); + dc_dmub_srv_notify_fw_dc_power_state(dc->ctx->dmub_srv, DC_ACPI_CM_POWER_STATE_D0); } void dc_init_callbacks(struct dc *dc, @@ -1876,6 +1871,41 @@ void dc_z10_save_init(struct dc *dc) dc->hwss.z10_save_init(dc); } +/* Set a pipe unlock order based on the change in DET allocation and stores it in dc scratch memory + * Prevents over allocation of DET during unlock process + * e.g. 2 pipe config with different streams with a max of 20 DET segments + * Before: After: + * - Pipe0: 10 DET segments - Pipe0: 12 DET segments + * - Pipe1: 10 DET segments - Pipe1: 8 DET segments + * If Pipe0 gets updated first, 22 DET segments will be allocated + */ +static void determine_pipe_unlock_order(struct dc *dc, struct dc_state *context) +{ + unsigned int i = 0; + struct pipe_ctx *pipe = NULL; + struct timing_generator *tg = NULL; + + if (!dc->config.set_pipe_unlock_order) + return; + + memset(dc->scratch.pipes_to_unlock_first, 0, sizeof(dc->scratch.pipes_to_unlock_first)); + for (i = 0; i < dc->res_pool->pipe_count; i++) { + pipe = &context->res_ctx.pipe_ctx[i]; + tg = pipe->stream_res.tg; + + if (!resource_is_pipe_type(pipe, OTG_MASTER) || + !tg->funcs->is_tg_enabled(tg) || + dc_state_get_pipe_subvp_type(context, pipe) == SUBVP_PHANTOM) { + continue; + } + + if (resource_calculate_det_for_stream(context, pipe) < + resource_calculate_det_for_stream(dc->current_state, &dc->current_state->res_ctx.pipe_ctx[i])) { + dc->scratch.pipes_to_unlock_first[i] = true; + } + } +} + /** * dc_commit_state_no_check - Apply context to the hardware * @@ -1974,6 +2004,7 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c context->streams[i]->update_flags.bits.dsc_changed = prev_dsc_changed; } + determine_pipe_unlock_order(dc, context); /* Program all planes within new context*/ if (dc->res_pool->funcs->prepare_mcache_programming) dc->res_pool->funcs->prepare_mcache_programming(dc, context); @@ -2156,6 +2187,14 @@ enum dc_status dc_commit_streams(struct dc *dc, struct dc_commit_streams_params context->power_source = params->power_source; res = dc_validate_with_context(dc, set, params->stream_count, context, false); + + /* + * Only update link encoder to stream assignment after bandwidth validation passed. + */ + if (res == DC_OK && dc->res_pool->funcs->link_encs_assign) + dc->res_pool->funcs->link_encs_assign( + dc, context, context->streams, context->stream_count); + if (res != DC_OK) { BREAK_TO_DEBUGGER(); goto fail; @@ -2477,41 +2516,35 @@ static enum surface_update_type get_scaling_info_update_type( if (!u->scaling_info) return UPDATE_TYPE_FAST; - if (u->scaling_info->dst_rect.width != u->surface->dst_rect.width + if (u->scaling_info->src_rect.width != u->surface->src_rect.width + || u->scaling_info->src_rect.height != u->surface->src_rect.height + || u->scaling_info->dst_rect.width != u->surface->dst_rect.width || u->scaling_info->dst_rect.height != u->surface->dst_rect.height + || u->scaling_info->clip_rect.width != u->surface->clip_rect.width + || u->scaling_info->clip_rect.height != u->surface->clip_rect.height || u->scaling_info->scaling_quality.integer_scaling != - u->surface->scaling_quality.integer_scaling - ) { + u->surface->scaling_quality.integer_scaling) { update_flags->bits.scaling_change = 1; + if (u->scaling_info->src_rect.width > u->surface->src_rect.width + || u->scaling_info->src_rect.height > u->surface->src_rect.height) + /* Making src rect bigger requires a bandwidth change */ + update_flags->bits.clock_change = 1; + if ((u->scaling_info->dst_rect.width < u->surface->dst_rect.width || u->scaling_info->dst_rect.height < u->surface->dst_rect.height) && (u->scaling_info->dst_rect.width < u->surface->src_rect.width || u->scaling_info->dst_rect.height < u->surface->src_rect.height)) /* Making dst rect smaller requires a bandwidth change */ update_flags->bits.bandwidth_change = 1; + + if (u->scaling_info->src_rect.width > dc->caps.max_optimizable_video_width && + (u->scaling_info->clip_rect.width > u->surface->clip_rect.width || + u->scaling_info->clip_rect.height > u->surface->clip_rect.height)) + /* Changing clip size of a large surface may result in MPC slice count change */ + update_flags->bits.bandwidth_change = 1; } - if (u->scaling_info->src_rect.width != u->surface->src_rect.width - || u->scaling_info->src_rect.height != u->surface->src_rect.height) { - - update_flags->bits.scaling_change = 1; - if (u->scaling_info->src_rect.width > u->surface->src_rect.width - || u->scaling_info->src_rect.height > u->surface->src_rect.height) - /* Making src rect bigger requires a bandwidth change */ - update_flags->bits.clock_change = 1; - } - - if (u->scaling_info->src_rect.width > dc->caps.max_optimizable_video_width && - (u->scaling_info->clip_rect.width > u->surface->clip_rect.width || - u->scaling_info->clip_rect.height > u->surface->clip_rect.height)) - /* Changing clip size of a large surface may result in MPC slice count change */ - update_flags->bits.bandwidth_change = 1; - - if (u->scaling_info->clip_rect.width != u->surface->clip_rect.width || - u->scaling_info->clip_rect.height != u->surface->clip_rect.height) - update_flags->bits.clip_size_change = 1; - if (u->scaling_info->src_rect.x != u->surface->src_rect.x || u->scaling_info->src_rect.y != u->surface->src_rect.y || u->scaling_info->clip_rect.x != u->surface->clip_rect.x @@ -2520,13 +2553,13 @@ static enum surface_update_type get_scaling_info_update_type( || u->scaling_info->dst_rect.y != u->surface->dst_rect.y) update_flags->bits.position_change = 1; + /* process every update flag before returning */ if (update_flags->bits.clock_change || update_flags->bits.bandwidth_change || update_flags->bits.scaling_change) return UPDATE_TYPE_FULL; - if (update_flags->bits.position_change || - update_flags->bits.clip_size_change) + if (update_flags->bits.position_change) return UPDATE_TYPE_MED; return UPDATE_TYPE_FAST; @@ -2617,7 +2650,8 @@ static enum surface_update_type det_surface_update(const struct dc *dc, elevate_update_type(&overall_type, type); } - if (update_flags->bits.lut_3d) { + if (update_flags->bits.lut_3d && + u->surface->mcm_luts.lut3d_data.lut3d_src != DC_CM2_TRANSFER_FUNC_SOURCE_VIDMEM) { type = UPDATE_TYPE_FULL; elevate_update_type(&overall_type, type); } @@ -2637,6 +2671,29 @@ static enum surface_update_type det_surface_update(const struct dc *dc, return overall_type; } +/* May need to flip the desktop plane in cases where MPO plane receives a flip but desktop plane doesn't + * while both planes are flip_immediate + */ +static void force_immediate_gsl_plane_flip(struct dc *dc, struct dc_surface_update *updates, int surface_count) +{ + bool has_flip_immediate_plane = false; + int i; + + for (i = 0; i < surface_count; i++) { + if (updates[i].surface->flip_immediate) { + has_flip_immediate_plane = true; + break; + } + } + + if (has_flip_immediate_plane && surface_count > 1) { + for (i = 0; i < surface_count; i++) { + if (updates[i].surface->flip_immediate) + updates[i].surface->update_flags.bits.addr_update = 1; + } + } +} + static enum surface_update_type check_update_surfaces_for_stream( struct dc *dc, struct dc_surface_update *updates, @@ -2699,6 +2756,9 @@ static enum surface_update_type check_update_surfaces_for_stream( if (stream_update->scaler_sharpener_update) su_flags->bits.scaler_sharpener = 1; + if (stream_update->sharpening_required) + su_flags->bits.sharpening_required = 1; + if (su_flags->raw != 0) overall_type = UPDATE_TYPE_FULL; @@ -2870,10 +2930,20 @@ static void copy_surface_update_to_plane( sizeof(struct dc_transfer_func_distributed_points)); } - if (srf_update->func_shaper) + if (srf_update->cm2_params) { + surface->mcm_shaper_3dlut_setting = srf_update->cm2_params->component_settings.shaper_3dlut_setting; + surface->mcm_lut1d_enable = srf_update->cm2_params->component_settings.lut1d_enable; + surface->mcm_luts = srf_update->cm2_params->cm2_luts; + } + + if (srf_update->func_shaper) { memcpy(&surface->in_shaper_func, srf_update->func_shaper, sizeof(surface->in_shaper_func)); + if (surface->mcm_shaper_3dlut_setting >= DC_CM2_SHAPER_3DLUT_SETTING_ENABLE_SHAPER) + surface->mcm_luts.shaper = &surface->in_shaper_func; + } + if (srf_update->lut3d_func) memcpy(&surface->lut3d_func, srf_update->lut3d_func, sizeof(surface->lut3d_func)); @@ -2886,10 +2956,17 @@ static void copy_surface_update_to_plane( surface->sdr_white_level_nits = srf_update->sdr_white_level_nits; - if (srf_update->blend_tf) + if (srf_update->blend_tf) { memcpy(&surface->blend_tf, srf_update->blend_tf, sizeof(surface->blend_tf)); + if (surface->mcm_lut1d_enable) + surface->mcm_luts.lut1d_func = &surface->blend_tf; + } + + if (srf_update->cm2_params || srf_update->blend_tf) + surface->lut_bank_a = !surface->lut_bank_a; + if (srf_update->input_csc_color_matrix) surface->input_csc_color_matrix = *srf_update->input_csc_color_matrix; @@ -2901,11 +2978,7 @@ static void copy_surface_update_to_plane( if (srf_update->gamut_remap_matrix) surface->gamut_remap_matrix = *srf_update->gamut_remap_matrix; - if (srf_update->cm2_params) { - surface->mcm_shaper_3dlut_setting = srf_update->cm2_params->component_settings.shaper_3dlut_setting; - surface->mcm_lut1d_enable = srf_update->cm2_params->component_settings.lut1d_enable; - surface->mcm_luts = srf_update->cm2_params->cm2_luts; - } + if (srf_update->cursor_csc_color_matrix) surface->cursor_csc_color_matrix = *srf_update->cursor_csc_color_matrix; @@ -3037,6 +3110,8 @@ static void copy_stream_update_to_stream(struct dc *dc, } if (update->scaler_sharpener_update) stream->scaler_sharpener_update = *update->scaler_sharpener_update; + if (update->sharpening_required) + stream->sharpening_required = *update->sharpening_required; } static void backup_planes_and_stream_state( @@ -3153,6 +3228,11 @@ static bool update_planes_and_stream_state(struct dc *dc, context = dc->current_state; update_type = dc_check_update_surfaces_for_stream( dc, srf_updates, surface_count, stream_update, stream_status); + /* It is possible to receive a flip for one plane while there are multiple flip_immediate planes in the same stream. + * E.g. Desktop and MPO plane are flip_immediate but only the MPO plane received a flip + * Force the other flip_immediate planes to flip so GSL doesn't wait for a flip that won't come. + */ + force_immediate_gsl_plane_flip(dc, srf_updates, surface_count); if (update_type == UPDATE_TYPE_FULL) backup_planes_and_stream_state(&dc->scratch.current_state, stream); @@ -3225,8 +3305,7 @@ static bool update_planes_and_stream_state(struct dc *dc, if (update_type != UPDATE_TYPE_MED) continue; - if (surface->update_flags.bits.clip_size_change || - surface->update_flags.bits.position_change) { + if (surface->update_flags.bits.position_change) { for (j = 0; j < dc->res_pool->pipe_count; j++) { struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[j]; @@ -3625,6 +3704,10 @@ static void commit_planes_for_stream_fast(struct dc *dc, struct pipe_ctx *top_pipe_to_program = NULL; struct dc_stream_status *stream_status = NULL; bool should_offload_fams2_flip = false; + bool should_lock_all_pipes = (update_type != UPDATE_TYPE_FAST); + + if (should_lock_all_pipes) + determine_pipe_unlock_order(dc, context); if (dc->debug.fams2_config.bits.enable && dc->debug.fams2_config.bits.enable_offload_flip && @@ -3677,13 +3760,14 @@ static void commit_planes_for_stream_fast(struct dc *dc, if (!pipe_ctx->plane_state) continue; - if (should_update_pipe_for_plane(context, pipe_ctx, plane_state)) + if (!should_update_pipe_for_plane(context, pipe_ctx, plane_state)) continue; + pipe_ctx->plane_state->triplebuffer_flips = false; if (update_type == UPDATE_TYPE_FAST && - dc->hwss.program_triplebuffer != NULL && - !pipe_ctx->plane_state->flip_immediate && dc->debug.enable_tri_buf) { - /*triple buffer for VUpdate only*/ + dc->hwss.program_triplebuffer != NULL && + !pipe_ctx->plane_state->flip_immediate && dc->debug.enable_tri_buf) { + /*triple buffer for VUpdate only*/ pipe_ctx->plane_state->triplebuffer_flips = true; } } @@ -3742,6 +3826,8 @@ static void commit_planes_for_stream(struct dc *dc, bool subvp_curr_use = false; uint8_t current_stream_mask = 0; + if (should_lock_all_pipes) + determine_pipe_unlock_order(dc, context); // Once we apply the new subvp context to hardware it won't be in the // dc->current_state anymore, so we have to cache it before we apply // the new SubVP context @@ -3749,7 +3835,7 @@ static void commit_planes_for_stream(struct dc *dc, dc_exit_ips_for_hw_access(dc); dc_z10_restore(dc); - if (update_type == UPDATE_TYPE_FULL) + if (update_type == UPDATE_TYPE_FULL && dc->optimized_required) hwss_process_outstanding_hw_updates(dc, dc->current_state); for (i = 0; i < dc->res_pool->pipe_count; i++) { @@ -3776,6 +3862,9 @@ static void commit_planes_for_stream(struct dc *dc, context_clock_trace(dc, context); } + if (update_type == UPDATE_TYPE_FULL) + hwss_wait_for_outstanding_hw_updates(dc, dc->current_state); + top_pipe_to_program = resource_get_otg_master_for_stream( &context->res_ctx, stream); @@ -3920,19 +4009,20 @@ static void commit_planes_for_stream(struct dc *dc, struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[j]; if (!pipe_ctx->plane_state) continue; - if (should_update_pipe_for_plane(context, pipe_ctx, plane_state)) + if (!should_update_pipe_for_plane(context, pipe_ctx, plane_state)) continue; pipe_ctx->plane_state->triplebuffer_flips = false; if (update_type == UPDATE_TYPE_FAST && - dc->hwss.program_triplebuffer != NULL && - !pipe_ctx->plane_state->flip_immediate && dc->debug.enable_tri_buf) { - /*triple buffer for VUpdate only*/ - pipe_ctx->plane_state->triplebuffer_flips = true; + dc->hwss.program_triplebuffer != NULL && + !pipe_ctx->plane_state->flip_immediate && dc->debug.enable_tri_buf) { + /*triple buffer for VUpdate only*/ + pipe_ctx->plane_state->triplebuffer_flips = true; } } if (update_type == UPDATE_TYPE_FULL) { /* force vsync flip when reconfiguring pipes to prevent underflow */ plane_state->flip_immediate = false; + plane_state->triplebuffer_flips = false; } } @@ -3953,7 +4043,6 @@ static void commit_planes_for_stream(struct dc *dc, continue; ASSERT(!pipe_ctx->plane_state->triplebuffer_flips); - if (dc->hwss.program_triplebuffer != NULL && dc->debug.enable_tri_buf) { /*turn off triple buffer for full update*/ dc->hwss.program_triplebuffer( @@ -4028,7 +4117,7 @@ static void commit_planes_for_stream(struct dc *dc, /*program triple buffer after lock based on flip type*/ if (dc->hwss.program_triplebuffer != NULL && dc->debug.enable_tri_buf) { - /*only enable triplebuffer for fast_update*/ + /*only enable triplebuffer for fast_update*/ dc->hwss.program_triplebuffer( dc, pipe_ctx, pipe_ctx->plane_state->triplebuffer_flips); } @@ -4777,6 +4866,11 @@ static bool update_planes_and_stream_v1(struct dc *dc, update_type = dc_check_update_surfaces_for_stream( dc, srf_updates, surface_count, stream_update, stream_status); + /* It is possible to receive a flip for one plane while there are multiple flip_immediate planes in the same stream. + * E.g. Desktop and MPO plane are flip_immediate but only the MPO plane received a flip + * Force the other flip_immediate planes to flip so GSL doesn't wait for a flip that won't come. + */ + force_immediate_gsl_plane_flip(dc, srf_updates, surface_count); if (update_type >= UPDATE_TYPE_FULL) { @@ -5338,8 +5432,10 @@ bool dc_set_ips_disable(struct dc *dc, unsigned int disable_ips) void dc_allow_idle_optimizations_internal(struct dc *dc, bool allow, char const *caller_name) { - if (dc->debug.disable_idle_power_optimizations) + if (dc->debug.disable_idle_power_optimizations) { + DC_LOG_DEBUG("%s: disabled\n", __func__); return; + } if (allow != dc->idle_optimizations_allowed) DC_LOG_IPS("%s: allow_idle old=%d new=%d (caller=%s)\n", __func__, @@ -5356,8 +5452,10 @@ void dc_allow_idle_optimizations_internal(struct dc *dc, bool allow, char const return; if (dc->hwss.apply_idle_power_optimizations && dc->clk_mgr != NULL && - dc->hwss.apply_idle_power_optimizations(dc, allow)) + dc->hwss.apply_idle_power_optimizations(dc, allow)) { dc->idle_optimizations_allowed = allow; + DC_LOG_DEBUG("%s: %s\n", __func__, allow ? "enabled" : "disabled"); + } } void dc_exit_ips_for_hw_access_internal(struct dc *dc, const char *caller_name) @@ -5999,7 +6097,12 @@ struct dc_power_profile dc_get_power_profile_for_dc_state(const struct dc_state { struct dc_power_profile profile = { 0 }; - profile.power_level += !context->bw_ctx.bw.dcn.clk.p_state_change_support; + if (!context || !context->clk_mgr || !context->clk_mgr->ctx || !context->clk_mgr->ctx->dc) + return profile; + struct dc *dc = context->clk_mgr->ctx->dc; + + if (dc->res_pool->funcs->get_power_profile) + profile.power_level = dc->res_pool->funcs->get_power_profile(context); return profile; } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c index 801cdbc8117d..af1ea5792560 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c @@ -46,11 +46,6 @@ DC_LOG_IF_TRACE(__VA_ARGS__); \ } while (0) -#define TIMING_TRACE(...) do {\ - if (dc->debug.timing_trace) \ - DC_LOG_SYNC(__VA_ARGS__); \ -} while (0) - #define CLOCK_TRACE(...) do {\ if (dc->debug.clock_trace) \ DC_LOG_BANDWIDTH_CALCS(__VA_ARGS__); \ @@ -306,43 +301,6 @@ void post_surface_trace(struct dc *dc) } -void context_timing_trace( - struct dc *dc, - struct resource_context *res_ctx) -{ - int i; - int h_pos[MAX_PIPES] = {0}, v_pos[MAX_PIPES] = {0}; - struct crtc_position position; - unsigned int underlay_idx = dc->res_pool->underlay_pipe_index; - DC_LOGGER_INIT(dc->ctx->logger); - - - for (i = 0; i < dc->res_pool->pipe_count; i++) { - struct pipe_ctx *pipe_ctx = &res_ctx->pipe_ctx[i]; - /* get_position() returns CRTC vertical/horizontal counter - * hence not applicable for underlay pipe - */ - if (pipe_ctx->stream == NULL || pipe_ctx->pipe_idx == underlay_idx) - continue; - - pipe_ctx->stream_res.tg->funcs->get_position(pipe_ctx->stream_res.tg, &position); - h_pos[i] = position.horizontal_count; - v_pos[i] = position.vertical_count; - } - for (i = 0; i < dc->res_pool->pipe_count; i++) { - struct pipe_ctx *pipe_ctx = &res_ctx->pipe_ctx[i]; - - if (pipe_ctx->stream == NULL || pipe_ctx->pipe_idx == underlay_idx) - continue; - - TIMING_TRACE("OTG_%d H_tot:%d V_tot:%d H_pos:%d V_pos:%d\n", - pipe_ctx->stream_res.tg->inst, - pipe_ctx->stream->timing.h_total, - pipe_ctx->stream->timing.v_total, - h_pos[i], v_pos[i]); - } -} - void context_clock_trace( struct dc *dc, struct dc_state *context) @@ -434,3 +392,43 @@ char *dc_status_to_str(enum dc_status status) return "Unexpected status error"; } + +char *dc_pixel_encoding_to_str(enum dc_pixel_encoding pixel_encoding) +{ + switch (pixel_encoding) { + case PIXEL_ENCODING_RGB: + return "RGB"; + case PIXEL_ENCODING_YCBCR422: + return "YUV422"; + case PIXEL_ENCODING_YCBCR444: + return "YUV444"; + case PIXEL_ENCODING_YCBCR420: + return "YUV420"; + default: + return "Unknown"; + } +} + +char *dc_color_depth_to_str(enum dc_color_depth color_depth) +{ + switch (color_depth) { + case COLOR_DEPTH_666: + return "6-bpc"; + case COLOR_DEPTH_888: + return "8-bpc"; + case COLOR_DEPTH_101010: + return "10-bpc"; + case COLOR_DEPTH_121212: + return "12-bpc"; + case COLOR_DEPTH_141414: + return "14-bpc"; + case COLOR_DEPTH_161616: + return "16-bpc"; + case COLOR_DEPTH_999: + return "9-bpc"; + case COLOR_DEPTH_111111: + return "11-bpc"; + default: + return "Unknown"; + } +} diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c index 7ee2be8f82c4..0419ee7f22a5 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c @@ -312,11 +312,11 @@ void get_mpctree_visual_confirm_color( { const struct tg_color pipe_colors[6] = { {MAX_TG_COLOR_VALUE, 0, 0}, /* red */ - {MAX_TG_COLOR_VALUE, MAX_TG_COLOR_VALUE / 4, 0}, /* orange */ {MAX_TG_COLOR_VALUE, MAX_TG_COLOR_VALUE, 0}, /* yellow */ {0, MAX_TG_COLOR_VALUE, 0}, /* green */ + {0, MAX_TG_COLOR_VALUE, MAX_TG_COLOR_VALUE}, /* cyan */ {0, 0, MAX_TG_COLOR_VALUE}, /* blue */ - {MAX_TG_COLOR_VALUE / 2, 0, MAX_TG_COLOR_VALUE / 2}, /* purple */ + {MAX_TG_COLOR_VALUE, 0, MAX_TG_COLOR_VALUE}, /* magenta */ }; struct pipe_ctx *top_pipe = pipe_ctx; @@ -497,6 +497,23 @@ void get_mclk_switch_visual_confirm_color( } } +void get_cursor_visual_confirm_color( + struct pipe_ctx *pipe_ctx, + struct tg_color *color) +{ + uint32_t color_value = MAX_TG_COLOR_VALUE; + + if (pipe_ctx->stream && pipe_ctx->stream->cursor_position.enable) { + color->color_r_cr = color_value; + color->color_g_y = 0; + color->color_b_cb = 0; + } else { + color->color_r_cr = 0; + color->color_g_y = 0; + color->color_b_cb = color_value; + } +} + void set_p_state_switch_method( struct dc *dc, struct dc_state *context, @@ -1071,8 +1088,13 @@ void hwss_wait_for_outstanding_hw_updates(struct dc *dc, struct dc_state *dc_con if (!pipe_ctx->stream) continue; - if (pipe_ctx->stream_res.tg->funcs->wait_drr_doublebuffer_pending_clear) - pipe_ctx->stream_res.tg->funcs->wait_drr_doublebuffer_pending_clear(pipe_ctx->stream_res.tg); + /* For full update we must wait for all double buffer updates, not just DRR updates. This + * is particularly important for minimal transitions. Only check for OTG_MASTER pipes, + * as non-OTG Master pipes share the same OTG as + */ + if (resource_is_pipe_type(pipe_ctx, OTG_MASTER) && dc->hwss.wait_for_all_pending_updates) { + dc->hwss.wait_for_all_pending_updates(pipe_ctx); + } hubp = pipe_ctx->plane_res.hubp; if (!hubp) diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_exports.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_exports.c index dfdfe22d9e85..457d60eeb486 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_exports.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_exports.c @@ -430,11 +430,10 @@ bool dc_link_get_backlight_level_nits(struct dc_link *link, } bool dc_link_set_backlight_level(const struct dc_link *link, - uint32_t backlight_pwm_u16_16, - uint32_t frame_ramp) + struct set_backlight_level_params *backlight_level_params) { return link->dc->link_srv->edp_set_backlight_level(link, - backlight_pwm_u16_16, frame_ramp); + backlight_level_params); } bool dc_link_set_backlight_level_nits(struct dc_link *link, diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c index c7599c40d4be..33125b95c3a1 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c @@ -765,25 +765,6 @@ static inline void get_vp_scan_direction( *flip_horz_scan_dir = !*flip_horz_scan_dir; } -/* - * This is a preliminary vp size calculation to allow us to check taps support. - * The result is completely overridden afterwards. - */ -static void calculate_viewport_size(struct pipe_ctx *pipe_ctx) -{ - struct scaler_data *data = &pipe_ctx->plane_res.scl_data; - - data->viewport.width = dc_fixpt_ceil(dc_fixpt_mul_int(data->ratios.horz, data->recout.width)); - data->viewport.height = dc_fixpt_ceil(dc_fixpt_mul_int(data->ratios.vert, data->recout.height)); - data->viewport_c.width = dc_fixpt_ceil(dc_fixpt_mul_int(data->ratios.horz_c, data->recout.width)); - data->viewport_c.height = dc_fixpt_ceil(dc_fixpt_mul_int(data->ratios.vert_c, data->recout.height)); - if (pipe_ctx->plane_state->rotation == ROTATION_ANGLE_90 || - pipe_ctx->plane_state->rotation == ROTATION_ANGLE_270) { - swap(data->viewport.width, data->viewport.height); - swap(data->viewport_c.width, data->viewport_c.height); - } -} - static struct rect intersect_rec(const struct rect *r0, const struct rect *r1) { struct rect rec; @@ -1468,6 +1449,7 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx) const struct dc_plane_state *plane_state = pipe_ctx->plane_state; struct dc_crtc_timing *timing = &pipe_ctx->stream->timing; const struct rect odm_slice_src = resource_get_odm_slice_src_rect(pipe_ctx); + struct scaling_taps temp = {0}; bool res = false; DC_LOGGER_INIT(pipe_ctx->stream->ctx->logger); @@ -1525,8 +1507,6 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx) calculate_recout(pipe_ctx); /* depends on pixel format */ calculate_scaling_ratios(pipe_ctx); - /* depends on scaling ratios and recout, does not calculate offset yet */ - calculate_viewport_size(pipe_ctx); /* * LB calculations depend on vp size, h/v_active and scaling ratios @@ -1547,6 +1527,24 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx) pipe_ctx->plane_res.scl_data.lb_params.alpha_en = plane_state->per_pixel_alpha; + // get TAP value with 100x100 dummy data for max scaling qualify, override + // if a new scaling quality required + pipe_ctx->plane_res.scl_data.viewport.width = 100; + pipe_ctx->plane_res.scl_data.viewport.height = 100; + pipe_ctx->plane_res.scl_data.viewport_c.width = 100; + pipe_ctx->plane_res.scl_data.viewport_c.height = 100; + if (pipe_ctx->plane_res.xfm != NULL) + res = pipe_ctx->plane_res.xfm->funcs->transform_get_optimal_number_of_taps( + pipe_ctx->plane_res.xfm, &pipe_ctx->plane_res.scl_data, &plane_state->scaling_quality); + + if (pipe_ctx->plane_res.dpp != NULL) + res = pipe_ctx->plane_res.dpp->funcs->dpp_get_optimal_number_of_taps( + pipe_ctx->plane_res.dpp, &pipe_ctx->plane_res.scl_data, &plane_state->scaling_quality); + + temp = pipe_ctx->plane_res.scl_data.taps; + + calculate_inits_and_viewports(pipe_ctx); + if (pipe_ctx->plane_res.xfm != NULL) res = pipe_ctx->plane_res.xfm->funcs->transform_get_optimal_number_of_taps( pipe_ctx->plane_res.xfm, &pipe_ctx->plane_res.scl_data, &plane_state->scaling_quality); @@ -1573,11 +1571,10 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx) &plane_state->scaling_quality); } - /* - * Depends on recout, scaling ratios, h_active and taps - * May need to re-check lb size after this in some obscure scenario - */ - if (res) + if (res && (pipe_ctx->plane_res.scl_data.taps.v_taps != temp.v_taps || + pipe_ctx->plane_res.scl_data.taps.h_taps != temp.h_taps || + pipe_ctx->plane_res.scl_data.taps.v_taps_c != temp.v_taps_c || + pipe_ctx->plane_res.scl_data.taps.h_taps_c != temp.h_taps_c)) calculate_inits_and_viewports(pipe_ctx); /* @@ -4094,14 +4091,6 @@ enum dc_status dc_validate_global_state( if (!dc->res_pool->funcs->validate_bandwidth(dc, new_ctx, fast_validate)) result = DC_FAIL_BANDWIDTH_VALIDATE; - /* - * Only update link encoder to stream assignment after bandwidth validation passed. - * TODO: Split out assignment and validation. - */ - if (result == DC_OK && dc->res_pool->funcs->link_encs_assign && fast_validate == false) - dc->res_pool->funcs->link_encs_assign( - dc, new_ctx, new_ctx->streams, new_ctx->stream_count); - return result; } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c index 9a406d74c0dd..55dc482d9b36 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c @@ -292,7 +292,9 @@ bool dc_stream_set_cursor_attributes( * 2. If not subvp high refresh, for single display cases, if resolution is >= 5K and refresh rate < 120hz * 3. If not subvp high refresh, for multi display cases, if resolution is >= 4K and refresh rate < 120hz */ - if (dc->debug.allow_sw_cursor_fallback && attributes->height * attributes->width * 4 > 16384) { + if (dc->debug.allow_sw_cursor_fallback && + attributes->height * attributes->width * 4 > 16384 && + !stream->hw_cursor_req) { if (check_subvp_sw_cursor_fallback_req(dc, stream)) return false; } @@ -421,7 +423,6 @@ bool dc_stream_program_cursor_position( /* apply/update visual confirm */ if (dc->debug.visual_confirm == VISUAL_CONFIRM_HW_CURSOR) { /* update software state */ - uint32_t color_value = MAX_TG_COLOR_VALUE; int i; for (i = 0; i < dc->res_pool->pipe_count; i++) { @@ -429,15 +430,7 @@ bool dc_stream_program_cursor_position( /* adjust visual confirm color for all pipes with current stream */ if (stream == pipe_ctx->stream) { - if (stream->cursor_position.enable) { - pipe_ctx->visual_confirm_color.color_r_cr = color_value; - pipe_ctx->visual_confirm_color.color_g_y = 0; - pipe_ctx->visual_confirm_color.color_b_cb = 0; - } else { - pipe_ctx->visual_confirm_color.color_r_cr = 0; - pipe_ctx->visual_confirm_color.color_g_y = 0; - pipe_ctx->visual_confirm_color.color_b_cb = color_value; - } + get_cursor_visual_confirm_color(pipe_ctx, &(pipe_ctx->visual_confirm_color)); /* programming hardware */ if (pipe_ctx->plane_state) @@ -819,12 +812,12 @@ void dc_stream_log(const struct dc *dc, const struct dc_stream_state *stream) stream->dst.height, stream->output_color_space); DC_LOG_DC( - "\tpix_clk_khz: %d, h_total: %d, v_total: %d, pixelencoder:%d, displaycolorDepth:%d\n", + "\tpix_clk_khz: %d, h_total: %d, v_total: %d, pixel_encoding:%s, color_depth:%s\n", stream->timing.pix_clk_100hz / 10, stream->timing.h_total, stream->timing.v_total, - stream->timing.pixel_encoding, - stream->timing.display_color_depth); + dc_pixel_encoding_to_str(stream->timing.pixel_encoding), + dc_color_depth_to_str(stream->timing.display_color_depth)); DC_LOG_DC( "\tlink: %d\n", stream->link->link_index); diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 3992ad73165b..e143fab00a86 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -55,7 +55,7 @@ struct aux_payload; struct set_config_cmd_payload; struct dmub_notification; -#define DC_VER "3.2.301" +#define DC_VER "3.2.309" #define MAX_SURFACES 3 #define MAX_PLANES 6 @@ -225,6 +225,11 @@ struct dc_dmub_caps { bool subvp_psr; bool gecc_enable; uint8_t fams_ver; + bool aux_backlight_support; +}; + +struct dc_scl_caps { + bool sharpener_support; }; struct dc_caps { @@ -292,6 +297,7 @@ struct dc_caps { bool sequential_ono; /* Conservative limit for DCC cases which require ODM4:1 to support*/ uint32_t dcc_plane_width_limit; + struct dc_scl_caps scl_caps; }; struct dc_bug_wa { @@ -463,6 +469,7 @@ struct dc_config { unsigned int enable_fpo_flicker_detection; bool disable_hbr_audio_dp2; bool consolidated_dpia_dp_lt; + bool set_pipe_unlock_order; }; enum visual_confirm { @@ -862,7 +869,6 @@ struct dc_debug_options { bool sanity_checks; bool max_disp_clk; bool surface_trace; - bool timing_trace; bool clock_trace; bool validation_trace; bool bandwidth_calcs_trace; @@ -1061,6 +1067,7 @@ struct dc_debug_options { unsigned int sharpen_policy; unsigned int scale_to_sharpness_policy; bool skip_full_updated_if_possible; + unsigned int enable_oled_edp_power_up_opt; }; @@ -1253,7 +1260,6 @@ union surface_update_flags { uint32_t rotation_change:1; uint32_t swizzle_change:1; uint32_t scaling_change:1; - uint32_t clip_size_change: 1; uint32_t position_change:1; uint32_t in_transfer_func_change:1; uint32_t input_csc_change:1; @@ -1355,6 +1361,7 @@ struct dc_plane_state { enum mpcc_movable_cm_location mcm_location; struct dc_csc_transform cursor_csc_color_matrix; bool adaptive_sharpness_en; + int adaptive_sharpness_policy; int sharpness_level; enum linear_light_scaling linear_light_scaling; unsigned int sdr_white_level_nits; @@ -1461,6 +1468,7 @@ struct dc { struct dc_scratch_space current_state; struct dc_scratch_space new_state; struct dc_stream_state temp_stream; // Used so we don't need to allocate stream on the stack + bool pipes_to_unlock_first[MAX_PIPES]; /* Any of the pipes indicated here should be unlocked first */ } scratch; struct dml2_configuration_options dml2_options; @@ -1513,7 +1521,7 @@ struct dc_surface_update { * change cm2_params.component_settings: Full update * change cm2_params.cm2_luts: Fast update */ - struct dc_cm2_parameters *cm2_params; + const struct dc_cm2_parameters *cm2_params; const struct dc_csc_transform *cursor_csc_color_matrix; unsigned int sdr_white_level_nits; }; @@ -1770,7 +1778,6 @@ struct dc_link { bool dongle_mode_timing_override; bool blank_stream_on_ocs_change; bool read_dpcd204h_on_irq_hpd; - bool disable_assr_for_uhbr; } wa_flags; struct link_mst_stream_allocation_table mst_stream_alloc_table; @@ -1786,6 +1793,7 @@ struct dc_link { // BW ALLOCATON USB4 ONLY struct dc_dpia_bw_alloc dpia_bw_alloc_config; bool skip_implict_edp_power_control; + enum backlight_control_type backlight_control_type; }; /* Return an enumerated dc_link. @@ -2203,8 +2211,7 @@ void dc_link_edp_panel_backlight_power_on(struct dc_link *link, * and 16 bit fractional, where 1.0 is max backlight value. */ bool dc_link_set_backlight_level(const struct dc_link *dc_link, - uint32_t backlight_pwm_u16_16, - uint32_t frame_ramp); + struct set_backlight_level_params *backlight_level_params); /* Set/get nits-based backlight level of an embedded panel (eDP, LVDS). */ bool dc_link_set_backlight_level_nits(struct dc_link *link, diff --git a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c index 1e7de0f03290..f90fc154549a 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c +++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c @@ -519,7 +519,8 @@ void dc_dmub_srv_get_visual_confirm_color_cmd(struct dc *dc, struct pipe_ctx *pi union dmub_rb_cmd cmd = { 0 }; unsigned int panel_inst = 0; - if (!dc_get_edp_link_panel_inst(dc, pipe_ctx->stream->link, &panel_inst)) + if (!dc_get_edp_link_panel_inst(dc, pipe_ctx->stream->link, &panel_inst) && + dc->debug.visual_confirm == VISUAL_CONFIRM_DISABLE) return; memset(&cmd, 0, sizeof(cmd)); @@ -1012,7 +1013,6 @@ static bool dc_can_pipe_disable_cursor(struct pipe_ctx *pipe_ctx) r2 = test_pipe->plane_res.scl_data.recout; r2_r = r2.x + r2.width; r2_b = r2.y + r2.height; - split_pipe = test_pipe; /** * There is another half plane on same layer because of @@ -1294,6 +1294,8 @@ static void dc_dmub_srv_notify_idle(const struct dc *dc, bool allow_idle) memset(&new_signals, 0, sizeof(new_signals)); + new_signals.bits.allow_idle = 1; /* always set */ + if (dc->config.disable_ips == DMUB_IPS_ENABLE || dc->config.disable_ips == DMUB_IPS_DISABLE_DYNAMIC) { new_signals.bits.allow_pg = 1; @@ -1389,7 +1391,7 @@ static void dc_dmub_srv_exit_low_power_state(const struct dc *dc) */ dc_dmub_srv->needs_idle_wake = false; - if (prev_driver_signals.bits.allow_ips2 && + if ((prev_driver_signals.bits.allow_ips2 || prev_driver_signals.all == 0) && (!dc->debug.optimize_ips_handshake || ips_fw->signals.bits.ips2_commit || !ips_fw->signals.bits.in_idle)) { DC_LOG_IPS( @@ -1450,7 +1452,7 @@ static void dc_dmub_srv_exit_low_power_state(const struct dc *dc) } dc_dmub_srv_notify_idle(dc, false); - if (prev_driver_signals.bits.allow_ips1) { + if (prev_driver_signals.bits.allow_ips1 || prev_driver_signals.all == 0) { DC_LOG_IPS( "wait for IPS1 commit clear (ips1_commit=%u ips2_commit=%u)", ips_fw->signals.bits.ips1_commit, @@ -1862,3 +1864,81 @@ void dc_dmub_srv_fams2_passthrough_flip( dm_execute_dmub_cmd_list(dc->ctx, num_cmds, cmds, DM_DMUB_WAIT_TYPE_WAIT); } } + +bool dc_dmub_srv_ips_residency_cntl(struct dc_dmub_srv *dc_dmub_srv, bool start_measurement) +{ + bool result; + + if (!dc_dmub_srv || !dc_dmub_srv->dmub) + return false; + + result = dc_wake_and_execute_gpint(dc_dmub_srv->ctx, DMUB_GPINT__IPS_RESIDENCY, + start_measurement, NULL, DM_DMUB_WAIT_TYPE_WAIT); + + return result; +} + +void dc_dmub_srv_ips_query_residency_info(struct dc_dmub_srv *dc_dmub_srv, struct ips_residency_info *output) +{ + uint32_t i; + enum dmub_gpint_command command_code; + + if (!dc_dmub_srv || !dc_dmub_srv->dmub) + return; + + switch (output->ips_mode) { + case DMUB_IPS_MODE_IPS1_MAX: + command_code = DMUB_GPINT__GET_IPS1_HISTOGRAM_COUNTER; + break; + case DMUB_IPS_MODE_IPS2: + command_code = DMUB_GPINT__GET_IPS2_HISTOGRAM_COUNTER; + break; + case DMUB_IPS_MODE_IPS1_RCG: + command_code = DMUB_GPINT__GET_IPS1_RCG_HISTOGRAM_COUNTER; + break; + case DMUB_IPS_MODE_IPS1_ONO2_ON: + command_code = DMUB_GPINT__GET_IPS1_ONO2_ON_HISTOGRAM_COUNTER; + break; + default: + command_code = DMUB_GPINT__INVALID_COMMAND; + break; + } + + if (command_code == DMUB_GPINT__INVALID_COMMAND) + return; + + // send gpint commands and wait for ack + if (!dc_wake_and_execute_gpint(dc_dmub_srv->ctx, DMUB_GPINT__GET_IPS_RESIDENCY_PERCENT, + (uint16_t)(output->ips_mode), + &output->residency_percent, DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY)) + output->residency_percent = 0; + + if (!dc_wake_and_execute_gpint(dc_dmub_srv->ctx, DMUB_GPINT__GET_IPS_RESIDENCY_ENTRY_COUNTER, + (uint16_t)(output->ips_mode), + &output->entry_counter, DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY)) + output->entry_counter = 0; + + if (!dc_wake_and_execute_gpint(dc_dmub_srv->ctx, DMUB_GPINT__GET_IPS_RESIDENCY_DURATION_US_LO, + (uint16_t)(output->ips_mode), + &output->total_active_time_us[0], DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY)) + output->total_active_time_us[0] = 0; + if (!dc_wake_and_execute_gpint(dc_dmub_srv->ctx, DMUB_GPINT__GET_IPS_RESIDENCY_DURATION_US_HI, + (uint16_t)(output->ips_mode), + &output->total_active_time_us[1], DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY)) + output->total_active_time_us[1] = 0; + + if (!dc_wake_and_execute_gpint(dc_dmub_srv->ctx, DMUB_GPINT__GET_IPS_INACTIVE_RESIDENCY_DURATION_US_LO, + (uint16_t)(output->ips_mode), + &output->total_inactive_time_us[0], DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY)) + output->total_inactive_time_us[0] = 0; + if (!dc_wake_and_execute_gpint(dc_dmub_srv->ctx, DMUB_GPINT__GET_IPS_INACTIVE_RESIDENCY_DURATION_US_HI, + (uint16_t)(output->ips_mode), + &output->total_inactive_time_us[1], DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY)) + output->total_inactive_time_us[1] = 0; + + // NUM_IPS_HISTOGRAM_BUCKETS = 16 + for (i = 0; i < 16; i++) + if (!dc_wake_and_execute_gpint(dc_dmub_srv->ctx, command_code, i, &output->histogram[i], + DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY)) + output->histogram[i] = 0; +} diff --git a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h index 42f0cb672d8b..10b48198b7a6 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h +++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h @@ -209,4 +209,43 @@ void dc_dmub_srv_fams2_passthrough_flip( struct dc_stream_state *stream, struct dc_surface_update *srf_updates, int surface_count); + +/** + * struct ips_residency_info - struct containing info from dmub_ips_residency_stats + * + * @ips_mode: The mode of IPS that the follow stats appertain to + * @residency_percent: The percentage of time spent in given IPS mode in millipercent + * @entry_counter: The number of entries made in to this IPS state + * @total_active_time_us: uint32_t array of length 2 representing time in the given IPS mode + * in microseconds. Index 0 is lower 32 bits, index 1 is upper 32 bits. + * @total_inactive_time_us: uint32_t array of length 2 representing time outside the given IPS mode + * in microseconds. Index 0 is lower 32 bits, index 1 is upper 32 bits. + * @histogram: Histogram of given IPS state durations - bucket definitions in dmub_ips.c + */ +struct ips_residency_info { + enum dmub_ips_mode ips_mode; + unsigned int residency_percent; + unsigned int entry_counter; + unsigned int total_active_time_us[2]; + unsigned int total_inactive_time_us[2]; + unsigned int histogram[16]; +}; + +/** + * bool dc_dmub_srv_ips_residency_cntl() - Controls IPS residency measurement status + * + * @dc_dmub_srv: The DC DMUB service pointer + * @start_measurement: Describes whether to start or stop measurement + * + * Return: true if GPINT was sent successfully, false otherwise + */ +bool dc_dmub_srv_ips_residency_cntl(struct dc_dmub_srv *dc_dmub_srv, bool start_measurement); + +/** + * bool dc_dmub_srv_ips_query_residency_info() - Queries DMCUB for residency info + * + * @dc_dmub_srv: The DC DMUB service pointer + * @output: Output struct to copy the the residency info to + */ +void dc_dmub_srv_ips_query_residency_info(struct dc_dmub_srv *dc_dmub_srv, struct ips_residency_info *output); #endif /* _DMUB_DC_SRV_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h index 41bd95e9177a..8dd6eb044829 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h @@ -1166,6 +1166,7 @@ struct dpcd_caps { int8_t branch_dev_name[6]; int8_t branch_hw_revision; int8_t branch_fw_revision[2]; + int8_t branch_vendor_specific_data[4]; bool allow_invalid_MSA_timing_param; bool panel_mode_edp; @@ -1191,6 +1192,7 @@ struct dpcd_caps { struct edp_psr_info psr_info; struct replay_info pr_info; + uint16_t edp_oled_emission_rate; }; union dpcd_sink_ext_caps { @@ -1204,7 +1206,7 @@ union dpcd_sink_ext_caps { uint8_t oled : 1; uint8_t reserved_2 : 1; uint8_t miniled : 1; - uint8_t reserved : 1; + uint8_t emission_output : 1; } bits; uint8_t raw; }; @@ -1358,6 +1360,9 @@ struct dp_trace { #ifndef DP_TUNNELING_IRQ #define DP_TUNNELING_IRQ (1 << 5) #endif +#ifndef DP_BRANCH_VENDOR_SPECIFIC_START +#define DP_BRANCH_VENDOR_SPECIFIC_START 0x50C +#endif /** USB4 DPCD BW Allocation Registers Chapter 10.7 **/ #ifndef DP_TUNNELING_CAPABILITIES #define DP_TUNNELING_CAPABILITIES 0xE000D /* 1.4a */ diff --git a/drivers/gpu/drm/amd/display/dc/dc_plane.h b/drivers/gpu/drm/amd/display/dc/dc_plane.h index 44afcd989224..bd37ec82b42d 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_plane.h +++ b/drivers/gpu/drm/amd/display/dc/dc_plane.h @@ -26,7 +26,6 @@ #ifndef _DC_PLANE_H_ #define _DC_PLANE_H_ -#include "dc.h" #include "dc_hw_types.h" struct dc_plane_state *dc_create_plane_state(const struct dc *dc); diff --git a/drivers/gpu/drm/amd/display/dc/dc_spl_translate.c b/drivers/gpu/drm/amd/display/dc/dc_spl_translate.c index 603552dbd771..c8d8e335fa37 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_spl_translate.c +++ b/drivers/gpu/drm/amd/display/dc/dc_spl_translate.c @@ -8,13 +8,13 @@ #include "dcn32/dcn32_dpp.h" #include "dcn401/dcn401_dpp.h" -static struct spl_funcs dcn2_spl_funcs = { +static struct spl_callbacks dcn2_spl_callbacks = { .spl_calc_lb_num_partitions = dscl2_spl_calc_lb_num_partitions, }; -static struct spl_funcs dcn32_spl_funcs = { +static struct spl_callbacks dcn32_spl_callbacks = { .spl_calc_lb_num_partitions = dscl32_spl_calc_lb_num_partitions, }; -static struct spl_funcs dcn401_spl_funcs = { +static struct spl_callbacks dcn401_spl_callbacks = { .spl_calc_lb_num_partitions = dscl401_spl_calc_lb_num_partitions, }; static void populate_splrect_from_rect(struct spl_rect *spl_rect, const struct rect *rect) @@ -38,6 +38,7 @@ static void populate_spltaps_from_taps(struct spl_taps *spl_scaling_quality, spl_scaling_quality->h_taps = scaling_quality->h_taps; spl_scaling_quality->v_taps_c = scaling_quality->v_taps_c; spl_scaling_quality->v_taps = scaling_quality->v_taps; + spl_scaling_quality->integer_scaling = scaling_quality->integer_scaling; } static void populate_taps_from_spltaps(struct scaling_taps *scaling_quality, const struct spl_taps *spl_scaling_quality) @@ -76,16 +77,16 @@ void translate_SPL_in_params_from_pipe_ctx(struct pipe_ctx *pipe_ctx, struct spl // This is used to determine the vtap support switch (plane_state->ctx->dce_version) { case DCN_VERSION_2_0: - spl_in->funcs = &dcn2_spl_funcs; + spl_in->callbacks = dcn2_spl_callbacks; break; case DCN_VERSION_3_2: - spl_in->funcs = &dcn32_spl_funcs; + spl_in->callbacks = dcn32_spl_callbacks; break; case DCN_VERSION_4_01: - spl_in->funcs = &dcn401_spl_funcs; + spl_in->callbacks = dcn401_spl_callbacks; break; default: - spl_in->funcs = &dcn2_spl_funcs; + spl_in->callbacks = dcn2_spl_callbacks; } // Make format field from spl_in point to plane_res scl_data format spl_in->basic_in.format = (enum spl_pixel_format)pipe_ctx->plane_res.scl_data.format; @@ -187,14 +188,14 @@ void translate_SPL_in_params_from_pipe_ctx(struct pipe_ctx *pipe_ctx, struct spl spl_in->h_active = pipe_ctx->plane_res.scl_data.h_active; spl_in->v_active = pipe_ctx->plane_res.scl_data.v_active; - spl_in->debug.sharpen_policy = (enum sharpen_policy)pipe_ctx->stream->ctx->dc->debug.sharpen_policy; + spl_in->sharpen_policy = (enum sharpen_policy)plane_state->adaptive_sharpness_policy; spl_in->debug.scale_to_sharpness_policy = (enum scale_to_sharpness_policy)pipe_ctx->stream->ctx->dc->debug.scale_to_sharpness_policy; /* Check if it is stream is in fullscreen and if its HDR. * Use this to determine sharpness levels */ - spl_in->is_fullscreen = dm_helpers_is_fullscreen(pipe_ctx->stream->ctx, pipe_ctx->stream); + spl_in->is_fullscreen = pipe_ctx->stream->sharpening_required; spl_in->is_hdr_on = dm_helpers_is_hdr_on(pipe_ctx->stream->ctx, pipe_ctx->stream); spl_in->sdr_white_level_nits = plane_state->sdr_white_level_nits; } diff --git a/drivers/gpu/drm/amd/display/dc/dc_state.h b/drivers/gpu/drm/amd/display/dc/dc_state.h index caa45db50232..db1e63a7d460 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_state.h +++ b/drivers/gpu/drm/amd/display/dc/dc_state.h @@ -26,7 +26,6 @@ #ifndef _DC_STATE_H_ #define _DC_STATE_H_ -#include "dc.h" #include "inc/core_status.h" struct dc_state *dc_state_create(struct dc *dc, struct dc_state_create_params *params); diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h index 14ea47eda0c8..413970588a26 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_stream.h +++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h @@ -143,6 +143,7 @@ union stream_update_flags { uint32_t crtc_timing_adjust : 1; uint32_t fams_changed : 1; uint32_t scaler_sharpener : 1; + uint32_t sharpening_required : 1; } bits; uint32_t raw; @@ -310,6 +311,7 @@ struct dc_stream_state { struct luminance_data lumin_data; bool scaler_sharpener_update; + bool sharpening_required; }; #define ABM_LEVEL_IMMEDIATE_DISABLE 255 @@ -356,6 +358,7 @@ struct dc_stream_update { struct dc_cursor_position *cursor_position; bool *hw_cursor_req; bool *scaler_sharpener_update; + bool *sharpening_required; }; bool dc_is_stream_unchanged( diff --git a/drivers/gpu/drm/amd/display/dc/dc_types.h b/drivers/gpu/drm/amd/display/dc/dc_types.h index 6d7989b751e2..edf4df1d03b5 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_types.h @@ -179,6 +179,9 @@ struct dc_panel_patch { unsigned int mst_start_top_delay; unsigned int remove_sink_ext_caps; unsigned int disable_colorimetry; + uint8_t blankstream_before_otg_off; + bool oled_optimize_display_on; + unsigned int force_mst_blocked_discovery; }; struct dc_edid_caps { @@ -922,6 +925,12 @@ struct display_endpoint_id { enum display_endpoint_type ep_type; }; +enum backlight_control_type { + BACKLIGHT_CONTROL_PWM = 0, + BACKLIGHT_CONTROL_VESA_AUX = 1, + BACKLIGHT_CONTROL_AMD_AUX = 2, +}; + #if defined(CONFIG_DRM_AMD_SECURE_DISPLAY) struct otg_phy_mux { uint8_t phy_output_num; @@ -1295,4 +1304,31 @@ struct dc_commit_streams_params { enum dc_power_source_type power_source; }; +struct set_backlight_level_params { + /* backlight in pwm */ + uint32_t backlight_pwm_u16_16; + /* brightness ramping */ + uint32_t frame_ramp; + /* backlight control type + * 0: PWM backlight control + * 1: VESA AUX backlight control + * 2: AMD AUX backlight control + */ + enum backlight_control_type control_type; + /* backlight in millinits */ + uint32_t backlight_millinits; + /* transition time in ms */ + uint32_t transition_time_in_ms; + /* minimum luminance in nits */ + uint32_t min_luminance; + /* maximum luminance in nits */ + uint32_t max_luminance; + /* minimum backlight in pwm */ + uint32_t min_backlight_pwm; + /* maximum backlight in pwm */ + uint32_t max_backlight_pwm; + /* AUX HW instance */ + uint8_t aux_inst; +}; + #endif /* DC_TYPES_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/dccg/dcn401/dcn401_dccg.c b/drivers/gpu/drm/amd/display/dc/dccg/dcn401/dcn401_dccg.c index 0b889004509a..d3e46c3cfa57 100644 --- a/drivers/gpu/drm/amd/display/dc/dccg/dcn401/dcn401_dccg.c +++ b/drivers/gpu/drm/amd/display/dc/dccg/dcn401/dcn401_dccg.c @@ -580,9 +580,6 @@ static void dccg401_set_dpstreamclk( int otg_inst, int dp_hpo_inst) { - /* set the dtbclk_p source */ - dccg401_set_dtbclk_p_src(dccg, src, otg_inst); - /* enabled to select one of the DTBCLKs for pipe */ if (src == REFCLK) dccg401_disable_dpstreamclk(dccg, dp_hpo_inst); @@ -805,33 +802,6 @@ static void dccg401_enable_symclk_se(struct dccg *dccg, uint32_t stream_enc_inst { struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg); - switch (link_enc_inst) { - case 0: - REG_UPDATE(SYMCLKA_CLOCK_ENABLE, - SYMCLKA_CLOCK_ENABLE, 1); - if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) - REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKA_ROOT_GATE_DISABLE, 1); - break; - case 1: - REG_UPDATE(SYMCLKB_CLOCK_ENABLE, - SYMCLKB_CLOCK_ENABLE, 1); - if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) - REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKB_ROOT_GATE_DISABLE, 1); - break; - case 2: - REG_UPDATE(SYMCLKC_CLOCK_ENABLE, - SYMCLKC_CLOCK_ENABLE, 1); - if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) - REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKC_ROOT_GATE_DISABLE, 1); - break; - case 3: - REG_UPDATE(SYMCLKD_CLOCK_ENABLE, - SYMCLKD_CLOCK_ENABLE, 1); - if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) - REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKD_ROOT_GATE_DISABLE, 1); - break; - } - switch (stream_enc_inst) { case 0: REG_UPDATE_2(SYMCLKA_CLOCK_ENABLE, @@ -864,37 +834,8 @@ static void dccg401_enable_symclk_se(struct dccg *dccg, uint32_t stream_enc_inst } } -/*get other front end connected to this backend*/ -static uint8_t dccg401_get_number_enabled_symclk_fe_connected_to_be(struct dccg *dccg, uint32_t link_enc_inst) -{ - uint8_t num_enabled_symclk_fe = 0; - uint32_t fe_clk_en[4] = {0}, be_clk_sel[4] = {0}; - struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg); - uint8_t i; - - REG_GET_2(SYMCLKA_CLOCK_ENABLE, SYMCLKA_FE_EN, &fe_clk_en[0], - SYMCLKA_FE_SRC_SEL, &be_clk_sel[0]); - - REG_GET_2(SYMCLKB_CLOCK_ENABLE, SYMCLKB_FE_EN, &fe_clk_en[1], - SYMCLKB_FE_SRC_SEL, &be_clk_sel[1]); - - REG_GET_2(SYMCLKC_CLOCK_ENABLE, SYMCLKC_FE_EN, &fe_clk_en[2], - SYMCLKC_FE_SRC_SEL, &be_clk_sel[2]); - - REG_GET_2(SYMCLKD_CLOCK_ENABLE, SYMCLKD_FE_EN, &fe_clk_en[3], - SYMCLKD_FE_SRC_SEL, &be_clk_sel[3]); - - for (i = 0; i < ARRAY_SIZE(fe_clk_en); i++) { - if (fe_clk_en[i] && be_clk_sel[i] == link_enc_inst) - num_enabled_symclk_fe++; - } - - return num_enabled_symclk_fe; -} - static void dccg401_disable_symclk_se(struct dccg *dccg, uint32_t stream_enc_inst, uint32_t link_enc_inst) { - uint8_t num_enabled_symclk_fe = 0; struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg); switch (stream_enc_inst) { @@ -919,31 +860,6 @@ static void dccg401_disable_symclk_se(struct dccg *dccg, uint32_t stream_enc_ins SYMCLKD_FE_SRC_SEL, 0); break; } - - /*check other enabled symclk fe connected to this be */ - num_enabled_symclk_fe = dccg401_get_number_enabled_symclk_fe_connected_to_be(dccg, link_enc_inst); - /*only turn off backend clk if other front ends attached to this backend are all off, - for mst, only turn off the backend if this is the last front end*/ - if (num_enabled_symclk_fe == 0) { - switch (link_enc_inst) { - case 0: - REG_UPDATE(SYMCLKA_CLOCK_ENABLE, - SYMCLKA_CLOCK_ENABLE, 0); - break; - case 1: - REG_UPDATE(SYMCLKB_CLOCK_ENABLE, - SYMCLKB_CLOCK_ENABLE, 0); - break; - case 2: - REG_UPDATE(SYMCLKC_CLOCK_ENABLE, - SYMCLKC_CLOCK_ENABLE, 0); - break; - case 3: - REG_UPDATE(SYMCLKD_CLOCK_ENABLE, - SYMCLKD_CLOCK_ENABLE, 0); - break; - } - } } static const struct dccg_funcs dccg401_funcs = { diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c index 5c2825bc9a87..d199e4ed2e59 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c @@ -277,7 +277,6 @@ static void dce110_stream_encoder_dp_set_stream_attribute( uint32_t misc1 = 0; uint32_t h_blank; uint32_t h_back_porch; - uint8_t synchronous_clock = 0; /* asynchronous mode */ uint8_t colorimetry_bpc; uint8_t dynamic_range_rgb = 0; /*full range*/ uint8_t dynamic_range_ycbcr = 1; /*bt709*/ @@ -380,7 +379,6 @@ static void dce110_stream_encoder_dp_set_stream_attribute( break; } - misc0 = misc0 | synchronous_clock; misc0 = colorimetry_bpc << 5; if (REG(DP_MSA_TIMING_PARAM1)) { diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c index db7557a1c613..8a3fbf95c48f 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c @@ -76,7 +76,6 @@ UNP_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH_C__GRPH_PRIMARY_SURFACE_ADDRESS_HIGH_C_MAS mmUNP_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH_C, value); - temp = 0; value = 0; temp = address.low_part >> UNP_GRPH_PRIMARY_SURFACE_ADDRESS_C__GRPH_PRIMARY_SURFACE_ADDRESS_C__SHIFT; @@ -112,7 +111,6 @@ UNP_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH_L__GRPH_PRIMARY_SURFACE_ADDRESS_HIGH_L_MAS mmUNP_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH_L, value); - temp = 0; value = 0; temp = address.low_part >> UNP_GRPH_PRIMARY_SURFACE_ADDRESS_L__GRPH_PRIMARY_SURFACE_ADDRESS_L__SHIFT; diff --git a/drivers/gpu/drm/amd/display/dc/dce60/dce60_resource.c b/drivers/gpu/drm/amd/display/dc/dce60/dce60_resource.c index 8db9f7514466..889f314cac65 100644 --- a/drivers/gpu/drm/amd/display/dc/dce60/dce60_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce60/dce60_resource.c @@ -717,7 +717,7 @@ static struct link_encoder *dce60_link_encoder_create( kzalloc(sizeof(struct dce110_link_encoder), GFP_KERNEL); int link_regs_id; - if (!enc110) + if (!enc110 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; link_regs_id = diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c index eaed5d1c398a..dcd2cdfe91eb 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c @@ -365,23 +365,18 @@ bool cm_helper_translate_curve_to_hw_format(struct dc_context *ctx, region_start = -MAX_LOW_POINT; region_end = NUMBER_REGIONS - MAX_LOW_POINT; } else { - /* 11 segments - * segment is from 2^-10 to 2^1 + /* 13 segments + * segment is from 2^-12 to 2^0 * There are less than 256 points, for optimization */ - seg_distr[0] = 3; - seg_distr[1] = 4; - seg_distr[2] = 4; - seg_distr[3] = 4; - seg_distr[4] = 4; - seg_distr[5] = 4; - seg_distr[6] = 4; - seg_distr[7] = 4; - seg_distr[8] = 4; - seg_distr[9] = 4; - seg_distr[10] = 1; + const uint8_t SEG_COUNT = 12; - region_start = -10; + for (i = 0; i < SEG_COUNT; i++) + seg_distr[i] = 4; + + seg_distr[SEG_COUNT] = 1; + + region_start = -SEG_COUNT; region_end = 1; } diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c index 05df502a54f2..88cf47a5ea75 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c @@ -46,7 +46,7 @@ #include "clk_mgr.h" __printf(3, 4) -unsigned int snprintf_count(char *pbuf, unsigned int bufsize, char *fmt, ...) +unsigned int snprintf_count(char *pbuf, unsigned int bufsize, const char *fmt, ...) { int ret_vsnprintf; unsigned int chars_printed; diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_cm_common.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_cm_common.c index f31f0e3abfc0..0690c346f2c5 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_cm_common.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_cm_common.c @@ -140,23 +140,18 @@ bool cm3_helper_translate_curve_to_hw_format( region_start = -MAX_LOW_POINT; region_end = NUMBER_REGIONS - MAX_LOW_POINT; } else { - /* 11 segments - * segment is from 2^-10 to 2^0 + /* 13 segments + * segment is from 2^-12 to 2^0 * There are less than 256 points, for optimization */ - seg_distr[0] = 3; - seg_distr[1] = 4; - seg_distr[2] = 4; - seg_distr[3] = 4; - seg_distr[4] = 4; - seg_distr[5] = 4; - seg_distr[6] = 4; - seg_distr[7] = 4; - seg_distr[8] = 4; - seg_distr[9] = 4; - seg_distr[10] = 1; + const uint8_t SEG_COUNT = 12; - region_start = -10; + for (i = 0; i < SEG_COUNT; i++) + seg_distr[i] = 4; + + seg_distr[SEG_COUNT] = 1; + + region_start = -SEG_COUNT; region_end = 1; } @@ -285,157 +280,6 @@ bool cm3_helper_translate_curve_to_hw_format( return true; } -#define NUM_DEGAMMA_REGIONS 12 - - -bool cm3_helper_translate_curve_to_degamma_hw_format( - const struct dc_transfer_func *output_tf, - struct pwl_params *lut_params) -{ - struct curve_points3 *corner_points; - struct pwl_result_data *rgb_resulted; - struct pwl_result_data *rgb; - struct pwl_result_data *rgb_plus_1; - - int32_t region_start, region_end; - int32_t i; - uint32_t j, k, seg_distr[MAX_REGIONS_NUMBER], increment, start_index, hw_points; - - if (output_tf == NULL || lut_params == NULL || output_tf->type == TF_TYPE_BYPASS) - return false; - - corner_points = lut_params->corner_points; - rgb_resulted = lut_params->rgb_resulted; - hw_points = 0; - - memset(lut_params, 0, sizeof(struct pwl_params)); - memset(seg_distr, 0, sizeof(seg_distr)); - - region_start = -NUM_DEGAMMA_REGIONS; - region_end = 0; - - - for (i = region_end - region_start; i < MAX_REGIONS_NUMBER ; i++) - seg_distr[i] = -1; - /* 12 segments - * segments are from 2^-12 to 0 - */ - for (i = 0; i < NUM_DEGAMMA_REGIONS ; i++) - seg_distr[i] = 4; - - for (k = 0; k < MAX_REGIONS_NUMBER; k++) { - if (seg_distr[k] != -1) - hw_points += (1 << seg_distr[k]); - } - - j = 0; - for (k = 0; k < (region_end - region_start); k++) { - increment = NUMBER_SW_SEGMENTS / (1 << seg_distr[k]); - start_index = (region_start + k + MAX_LOW_POINT) * - NUMBER_SW_SEGMENTS; - for (i = start_index; i < start_index + NUMBER_SW_SEGMENTS; - i += increment) { - if (j == hw_points - 1) - break; - if (i >= TRANSFER_FUNC_POINTS) - return false; - rgb_resulted[j].red = output_tf->tf_pts.red[i]; - rgb_resulted[j].green = output_tf->tf_pts.green[i]; - rgb_resulted[j].blue = output_tf->tf_pts.blue[i]; - j++; - } - } - - /* last point */ - start_index = (region_end + MAX_LOW_POINT) * NUMBER_SW_SEGMENTS; - rgb_resulted[hw_points - 1].red = output_tf->tf_pts.red[start_index]; - rgb_resulted[hw_points - 1].green = output_tf->tf_pts.green[start_index]; - rgb_resulted[hw_points - 1].blue = output_tf->tf_pts.blue[start_index]; - - corner_points[0].red.x = dc_fixpt_pow(dc_fixpt_from_int(2), - dc_fixpt_from_int(region_start)); - corner_points[0].green.x = corner_points[0].red.x; - corner_points[0].blue.x = corner_points[0].red.x; - corner_points[1].red.x = dc_fixpt_pow(dc_fixpt_from_int(2), - dc_fixpt_from_int(region_end)); - corner_points[1].green.x = corner_points[1].red.x; - corner_points[1].blue.x = corner_points[1].red.x; - - corner_points[0].red.y = rgb_resulted[0].red; - corner_points[0].green.y = rgb_resulted[0].green; - corner_points[0].blue.y = rgb_resulted[0].blue; - - /* see comment above, m_arrPoints[1].y should be the Y value for the - * region end (m_numOfHwPoints), not last HW point(m_numOfHwPoints - 1) - */ - corner_points[1].red.y = rgb_resulted[hw_points - 1].red; - corner_points[1].green.y = rgb_resulted[hw_points - 1].green; - corner_points[1].blue.y = rgb_resulted[hw_points - 1].blue; - corner_points[1].red.slope = dc_fixpt_zero; - corner_points[1].green.slope = dc_fixpt_zero; - corner_points[1].blue.slope = dc_fixpt_zero; - - if (output_tf->tf == TRANSFER_FUNCTION_PQ) { - /* for PQ, we want to have a straight line from last HW X point, - * and the slope to be such that we hit 1.0 at 10000 nits. - */ - const struct fixed31_32 end_value = - dc_fixpt_from_int(125); - - corner_points[1].red.slope = dc_fixpt_div( - dc_fixpt_sub(dc_fixpt_one, corner_points[1].red.y), - dc_fixpt_sub(end_value, corner_points[1].red.x)); - corner_points[1].green.slope = dc_fixpt_div( - dc_fixpt_sub(dc_fixpt_one, corner_points[1].green.y), - dc_fixpt_sub(end_value, corner_points[1].green.x)); - corner_points[1].blue.slope = dc_fixpt_div( - dc_fixpt_sub(dc_fixpt_one, corner_points[1].blue.y), - dc_fixpt_sub(end_value, corner_points[1].blue.x)); - } - - lut_params->hw_points_num = hw_points; - - k = 0; - for (i = 1; i < MAX_REGIONS_NUMBER; i++) { - if (seg_distr[k] != -1) { - lut_params->arr_curve_points[k].segments_num = - seg_distr[k]; - lut_params->arr_curve_points[i].offset = - lut_params->arr_curve_points[k].offset + (1 << seg_distr[k]); - } - k++; - } - - if (seg_distr[k] != -1) - lut_params->arr_curve_points[k].segments_num = seg_distr[k]; - - rgb = rgb_resulted; - rgb_plus_1 = rgb_resulted + 1; - - i = 1; - while (i != hw_points + 1) { - if (dc_fixpt_lt(rgb_plus_1->red, rgb->red)) - rgb_plus_1->red = rgb->red; - if (dc_fixpt_lt(rgb_plus_1->green, rgb->green)) - rgb_plus_1->green = rgb->green; - if (dc_fixpt_lt(rgb_plus_1->blue, rgb->blue)) - rgb_plus_1->blue = rgb->blue; - - rgb->delta_red = dc_fixpt_sub(rgb_plus_1->red, rgb->red); - rgb->delta_green = dc_fixpt_sub(rgb_plus_1->green, rgb->green); - rgb->delta_blue = dc_fixpt_sub(rgb_plus_1->blue, rgb->blue); - - ++rgb_plus_1; - ++rgb; - ++i; - } - cm3_helper_convert_to_custom_float(rgb_resulted, - lut_params->corner_points, - hw_points, false); - - return true; -} - bool cm3_helper_convert_to_custom_float( struct pwl_result_data *rgb_resulted, struct curve_points3 *corner_points, diff --git a/drivers/gpu/drm/amd/display/dc/dio/dcn10/dcn10_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dio/dcn10/dcn10_stream_encoder.c index f496e952ceec..d01a8b8f9595 100644 --- a/drivers/gpu/drm/amd/display/dc/dio/dcn10/dcn10_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dio/dcn10/dcn10_stream_encoder.c @@ -255,7 +255,6 @@ void enc1_stream_encoder_dp_set_stream_attribute( uint32_t misc1 = 0; uint32_t h_blank; uint32_t h_back_porch; - uint8_t synchronous_clock = 0; /* asynchronous mode */ uint8_t colorimetry_bpc; uint8_t dp_pixel_encoding = 0; uint8_t dp_component_depth = 0; @@ -362,7 +361,6 @@ void enc1_stream_encoder_dp_set_stream_attribute( break; } - misc0 = misc0 | synchronous_clock; misc0 = colorimetry_bpc << 5; switch (output_color_space) { diff --git a/drivers/gpu/drm/amd/display/dc/dio/dcn314/dcn314_dio_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dio/dcn314/dcn314_dio_stream_encoder.c index 5b343f745cf3..ae81451a3a72 100644 --- a/drivers/gpu/drm/amd/display/dc/dio/dcn314/dcn314_dio_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dio/dcn314/dcn314_dio_stream_encoder.c @@ -83,6 +83,15 @@ void enc314_disable_fifo(struct stream_encoder *enc) REG_UPDATE(DIG_FIFO_CTRL0, DIG_FIFO_ENABLE, 0); } +static bool enc314_is_fifo_enabled(struct stream_encoder *enc) +{ + struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); + uint32_t reset_val; + + REG_GET(DIG_FIFO_CTRL0, DIG_FIFO_ENABLE, &reset_val); + return (reset_val != 0); +} + void enc314_dp_set_odm_combine( struct stream_encoder *enc, bool odm_combine) @@ -468,6 +477,7 @@ static const struct stream_encoder_funcs dcn314_str_enc_funcs = { .enable_fifo = enc314_enable_fifo, .disable_fifo = enc314_disable_fifo, + .is_fifo_enabled = enc314_is_fifo_enabled, .set_input_mode = enc314_set_dig_input_mode, }; diff --git a/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.c index 0a27e0942a12..098c2a01a850 100644 --- a/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.c @@ -447,7 +447,6 @@ void enc401_stream_encoder_dp_set_stream_attribute( uint32_t misc1 = 0; uint32_t h_blank; uint32_t h_back_porch; - uint8_t synchronous_clock = 0; /* asynchronous mode */ uint8_t colorimetry_bpc; uint8_t dp_pixel_encoding = 0; uint8_t dp_component_depth = 0; @@ -603,7 +602,6 @@ void enc401_stream_encoder_dp_set_stream_attribute( break; } - misc0 = misc0 | synchronous_clock; misc0 = colorimetry_bpc << 5; switch (output_color_space) { diff --git a/drivers/gpu/drm/amd/display/dc/dm_services.h b/drivers/gpu/drm/amd/display/dc/dm_services.h index 9405c47ee2a9..f81e5a4e1d6d 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_services.h +++ b/drivers/gpu/drm/amd/display/dc/dm_services.h @@ -143,7 +143,7 @@ void generic_reg_wait(const struct dc_context *ctx, unsigned int delay_between_poll_us, unsigned int time_out_num_tries, const char *func_name, int line); -unsigned int snprintf_count(char *pBuf, unsigned int bufSize, char *fmt, ...); +unsigned int snprintf_count(char *pBuf, unsigned int bufSize, const char *fmt, ...); /* These macros need to be used with soc15 registers in order to retrieve * the actual offset. diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20.c index 565f3c492477..0c8c4a080c50 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20.c @@ -785,12 +785,9 @@ static bool CalculatePrefetchSchedule( if (MyError) { *PrefetchBandwidth = 0; - TimeForFetchingMetaPTE = 0; - TimeForFetchingRowInVBlank = 0; *DestinationLinesToRequestVMInVBlank = 0; *DestinationLinesToRequestRowInVBlank = 0; *DestinationLinesForPrefetch = 0; - LinesToRequestPrefetchPixelData = 0; *VRatioPrefetchY = 0; *VRatioPrefetchC = 0; *RequiredPrefetchPixDataBW = 0; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c index 9d6675ecc5f1..c935903b68e1 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c @@ -845,12 +845,9 @@ static bool CalculatePrefetchSchedule( if (MyError) { *PrefetchBandwidth = 0; - TimeForFetchingMetaPTE = 0; - TimeForFetchingRowInVBlank = 0; *DestinationLinesToRequestVMInVBlank = 0; *DestinationLinesToRequestRowInVBlank = 0; *DestinationLinesForPrefetch = 0; - LinesToRequestPrefetchPixelData = 0; *VRatioPrefetchY = 0; *VRatioPrefetchC = 0; *RequiredPrefetchPixDataBW = 0; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20.c index 4fce64a030b6..390c1a77fda6 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20.c @@ -443,8 +443,6 @@ static void get_meta_and_pte_attr(struct display_mode_lib *mode_lib, blk_bytes = surf_linear ? 256 : get_blk_size_bytes((enum source_macro_tile_size) macro_tile_size); log2_blk_bytes = dml_log2((double) blk_bytes); - log2_blk_height = 0; - log2_blk_width = 0; // remember log rule // "+" in log is multiply @@ -491,8 +489,6 @@ static void get_meta_and_pte_attr(struct display_mode_lib *mode_lib, - log2_meta_req_height; meta_req_width = 1 << log2_meta_req_width; meta_req_height = 1 << log2_meta_req_height; - log2_meta_row_height = 0; - meta_row_width_ub = 0; // the dimensions of a meta row are meta_row_width x meta_row_height in elements. // calculate upper bound of the meta_row_width diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20v2.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20v2.c index 3fa9a5da02f6..843d6004258c 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20v2.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_rq_dlg_calc_20v2.c @@ -443,8 +443,6 @@ static void get_meta_and_pte_attr(struct display_mode_lib *mode_lib, blk_bytes = surf_linear ? 256 : get_blk_size_bytes((enum source_macro_tile_size) macro_tile_size); log2_blk_bytes = dml_log2((double) blk_bytes); - log2_blk_height = 0; - log2_blk_width = 0; // remember log rule // "+" in log is multiply @@ -491,8 +489,6 @@ static void get_meta_and_pte_attr(struct display_mode_lib *mode_lib, - log2_meta_req_height; meta_req_width = 1 << log2_meta_req_width; meta_req_height = 1 << log2_meta_req_height; - log2_meta_row_height = 0; - meta_row_width_ub = 0; // the dimensions of a meta row are meta_row_width x meta_row_height in elements. // calculate upper bound of the meta_row_width diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c b/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c index eb3ed965e48b..cd8cca651419 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c @@ -1049,12 +1049,9 @@ static bool CalculatePrefetchSchedule( if (MyError) { *PrefetchBandwidth = 0; - TimeForFetchingMetaPTE = 0; - TimeForFetchingRowInVBlank = 0; *DestinationLinesToRequestVMInVBlank = 0; *DestinationLinesToRequestRowInVBlank = 0; *DestinationLinesForPrefetch = 0; - LinesToRequestPrefetchPixelData = 0; *VRatioPrefetchY = 0; *VRatioPrefetchC = 0; *RequiredPrefetchPixDataBWLuma = 0; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_rq_dlg_calc_21.c b/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_rq_dlg_calc_21.c index 9e1c18b90805..5718000627b0 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_rq_dlg_calc_21.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_rq_dlg_calc_21.c @@ -435,8 +435,6 @@ static void get_meta_and_pte_attr( blk_bytes = surf_linear ? 256 : get_blk_size_bytes((enum source_macro_tile_size) macro_tile_size); log2_blk_bytes = dml_log2((double) blk_bytes); - log2_blk_height = 0; - log2_blk_width = 0; // remember log rule // "+" in log is multiply @@ -485,8 +483,6 @@ static void get_meta_and_pte_attr( - log2_meta_req_height; meta_req_width = 1 << log2_meta_req_width; meta_req_height = 1 << log2_meta_req_height; - log2_meta_row_height = 0; - meta_row_width_ub = 0; // the dimensions of a meta row are meta_row_width x meta_row_height in elements. // calculate upper bound of the meta_row_width diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c b/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c index 1c10ba4dcdde..cee1b351e105 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c @@ -1280,12 +1280,9 @@ static bool CalculatePrefetchSchedule( if (MyError) { *PrefetchBandwidth = 0; - TimeForFetchingMetaPTE = 0; - TimeForFetchingRowInVBlank = 0; *DestinationLinesToRequestVMInVBlank = 0; *DestinationLinesToRequestRowInVBlank = 0; *DestinationLinesForPrefetch = 0; - LinesToRequestPrefetchPixelData = 0; *VRatioPrefetchY = 0; *VRatioPrefetchC = 0; *RequiredPrefetchPixDataBWLuma = 0; @@ -1775,15 +1772,6 @@ static unsigned int CalculateVMAndRowBytes( *PixelPTEReqWidth = 32768.0 / BytePerPixel; *PTERequestSize = 64; FractionOfPTEReturnDrop = 0; - } else if (MacroTileSizeBytes == 4096) { - PixelPTEReqHeightPTEs = 1; - *PixelPTEReqHeight = MacroTileHeight; - *PixelPTEReqWidth = 8 * *MacroTileWidth; - *PTERequestSize = 64; - if (ScanDirection != dm_vert) - FractionOfPTEReturnDrop = 0; - else - FractionOfPTEReturnDrop = 7.0 / 8; } else if (GPUVMMinPageSize == 4 && MacroTileSizeBytes > 4096) { PixelPTEReqHeightPTEs = 16; *PixelPTEReqHeight = 16 * BlockHeight256Bytes; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_rq_dlg_calc_30.c b/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_rq_dlg_calc_30.c index b28fcc8608ff..76d3bb3c9155 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_rq_dlg_calc_30.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_rq_dlg_calc_30.c @@ -392,8 +392,6 @@ static void get_meta_and_pte_attr(struct display_mode_lib *mode_lib, blk_bytes = surf_linear ? 256 : get_blk_size_bytes((enum source_macro_tile_size) macro_tile_size); log2_blk_bytes = dml_log2((double)blk_bytes); - log2_blk_height = 0; - log2_blk_width = 0; // remember log rule // "+" in log is multiply @@ -464,8 +462,6 @@ static void get_meta_and_pte_attr(struct display_mode_lib *mode_lib, - log2_meta_req_height; meta_req_width = 1 << log2_meta_req_width; meta_req_height = 1 << log2_meta_req_height; - log2_meta_row_height = 0; - meta_row_width_ub = 0; // the dimensions of a meta row are meta_row_width x meta_row_height in elements. // calculate upper bound of the meta_row_width diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c b/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c index 2b275e680379..f567a9023682 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c @@ -1444,12 +1444,9 @@ static bool CalculatePrefetchSchedule( if (MyError) { *PrefetchBandwidth = 0; - TimeForFetchingMetaPTE = 0; - TimeForFetchingRowInVBlank = 0; *DestinationLinesToRequestVMInVBlank = 0; *DestinationLinesToRequestRowInVBlank = 0; *DestinationLinesForPrefetch = 0; - LinesToRequestPrefetchPixelData = 0; *VRatioPrefetchY = 0; *VRatioPrefetchC = 0; *RequiredPrefetchPixDataBWLuma = 0; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_rq_dlg_calc_31.c b/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_rq_dlg_calc_31.c index b57b095cd4a8..c46bda2141ac 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_rq_dlg_calc_31.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_rq_dlg_calc_31.c @@ -413,8 +413,6 @@ static void get_meta_and_pte_attr( log2_blk256_height = dml_log2((double) blk256_height); blk_bytes = surf_linear ? 256 : get_blk_size_bytes((enum source_macro_tile_size) macro_tile_size); log2_blk_bytes = dml_log2((double) blk_bytes); - log2_blk_height = 0; - log2_blk_width = 0; // remember log rule // "+" in log is multiply @@ -481,8 +479,6 @@ static void get_meta_and_pte_attr( log2_meta_req_width = log2_meta_req_bytes + 8 - log2_bytes_per_element - log2_meta_req_height; meta_req_width = 1 << log2_meta_req_width; meta_req_height = 1 << log2_meta_req_height; - log2_meta_row_height = 0; - meta_row_width_ub = 0; // the dimensions of a meta row are meta_row_width x meta_row_height in elements. // calculate upper bound of the meta_row_width diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c b/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c index debfa31583a6..5865e8fa2d8e 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c @@ -1461,12 +1461,9 @@ static bool CalculatePrefetchSchedule( if (MyError) { *PrefetchBandwidth = 0; - TimeForFetchingMetaPTE = 0; - TimeForFetchingRowInVBlank = 0; *DestinationLinesToRequestVMInVBlank = 0; *DestinationLinesToRequestRowInVBlank = 0; *DestinationLinesForPrefetch = 0; - LinesToRequestPrefetchPixelData = 0; *VRatioPrefetchY = 0; *VRatioPrefetchC = 0; *RequiredPrefetchPixDataBWLuma = 0; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_rq_dlg_calc_314.c b/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_rq_dlg_calc_314.c index 61b3bebf24c9..b7d2a0caec11 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_rq_dlg_calc_314.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_rq_dlg_calc_314.c @@ -501,8 +501,6 @@ static void get_meta_and_pte_attr( log2_blk256_height = dml_log2((double) blk256_height); blk_bytes = surf_linear ? 256 : get_blk_size_bytes((enum source_macro_tile_size) macro_tile_size); log2_blk_bytes = dml_log2((double) blk_bytes); - log2_blk_height = 0; - log2_blk_width = 0; // remember log rule // "+" in log is multiply @@ -569,8 +567,6 @@ static void get_meta_and_pte_attr( log2_meta_req_width = log2_meta_req_bytes + 8 - log2_bytes_per_element - log2_meta_req_height; meta_req_width = 1 << log2_meta_req_width; meta_req_height = 1 << log2_meta_req_height; - log2_meta_row_height = 0; - meta_row_width_ub = 0; // the dimensions of a meta row are meta_row_width x meta_row_height in elements. // calculate upper bound of the meta_row_width diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c index d92fb428ee96..86ac7d59fd32 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c @@ -4097,12 +4097,9 @@ bool dml32_CalculatePrefetchSchedule( if (MyError) { *PrefetchBandwidth = 0; - TimeForFetchingMetaPTE = 0; - TimeForFetchingRowInVBlank = 0; *DestinationLinesToRequestVMInVBlank = 0; *DestinationLinesToRequestRowInVBlank = 0; *DestinationLinesForPrefetch = 0; - LinesToRequestPrefetchPixelData = 0; *VRatioPrefetchY = 0; *VRatioPrefetchC = 0; *RequiredPrefetchPixDataBWLuma = 0; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.c index a201dbb743d7..d9e63c4fdd95 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.c @@ -204,8 +204,8 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_51_soc = { .num_states = 8, .sr_exit_time_us = 28.0, .sr_enter_plus_exit_time_us = 30.0, - .sr_exit_z8_time_us = 250.0, - .sr_enter_plus_exit_z8_time_us = 350.0, + .sr_exit_z8_time_us = 263.0, + .sr_enter_plus_exit_z8_time_us = 363.0, .fclk_change_latency_us = 24.0, .usr_retraining_latency_us = 2, .writeback_latency_us = 12.0, diff --git a/drivers/gpu/drm/amd/display/dc/dml/dml1_display_rq_dlg_calc.c b/drivers/gpu/drm/amd/display/dc/dml/dml1_display_rq_dlg_calc.c index d8bfc85e5dcd..88dc2b97e7bf 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dml1_display_rq_dlg_calc.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dml1_display_rq_dlg_calc.c @@ -559,12 +559,11 @@ static void get_surf_rq_param( const struct _vcs_dpi_display_pipe_source_params_st *pipe_src_param, bool is_chroma) { - bool mode_422 = 0; unsigned int vp_width = 0; unsigned int vp_height = 0; unsigned int data_pitch = 0; unsigned int meta_pitch = 0; - unsigned int ppe = mode_422 ? 2 : 1; + unsigned int ppe = 1; bool surf_linear; bool surf_vert; unsigned int bytes_per_element; diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c index 8697eac1e1f7..138b4b1e42ed 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c @@ -859,7 +859,7 @@ static void populate_dml21_plane_config_from_plane_state(struct dml2_context *dm plane->immediate_flip = plane_state->flip_immediate; plane->composition.rect_out_height_spans_vactive = - plane_state->dst_rect.height >= stream->timing.v_addressable && + plane_state->dst_rect.height >= stream->src.height && stream->dst.height >= stream->timing.v_addressable; } @@ -1036,6 +1036,7 @@ void dml21_copy_clocks_to_dc_state(struct dml2_context *in_ctx, struct dc_state context->bw_ctx.bw.dcn.clk.p_state_change_support = in_ctx->v21.mode_programming.programming->uclk_pstate_supported; context->bw_ctx.bw.dcn.clk.dtbclk_en = in_ctx->v21.mode_programming.programming->min_clocks.dcn4x.dtbrefclk_khz > 0; context->bw_ctx.bw.dcn.clk.ref_dtbclk_khz = in_ctx->v21.mode_programming.programming->min_clocks.dcn4x.dtbrefclk_khz; + context->bw_ctx.bw.dcn.clk.socclk_khz = in_ctx->v21.mode_programming.programming->min_clocks.dcn4x.socclk_khz; } void dml21_extract_legacy_watermark_set(const struct dc *in_dc, struct dcn_watermarks *watermark, enum dml2_dchub_watermark_reg_set_index reg_set_idx, struct dml2_context *in_ctx) diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_wrapper.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_wrapper.c index d35dd507cb9f..bbc28b9a15a3 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_wrapper.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_wrapper.c @@ -13,11 +13,11 @@ static bool dml21_allocate_memory(struct dml2_context **dml_ctx) { - *dml_ctx = (struct dml2_context *)kzalloc(sizeof(struct dml2_context), GFP_KERNEL); + *dml_ctx = kzalloc(sizeof(struct dml2_context), GFP_KERNEL); if (!(*dml_ctx)) return false; - (*dml_ctx)->v21.dml_init.dml2_instance = (struct dml2_instance *)kzalloc(sizeof(struct dml2_instance), GFP_KERNEL); + (*dml_ctx)->v21.dml_init.dml2_instance = kzalloc(sizeof(struct dml2_instance), GFP_KERNEL); if (!((*dml_ctx)->v21.dml_init.dml2_instance)) return false; @@ -27,7 +27,7 @@ static bool dml21_allocate_memory(struct dml2_context **dml_ctx) (*dml_ctx)->v21.mode_support.display_config = &(*dml_ctx)->v21.display_config; (*dml_ctx)->v21.mode_programming.display_config = (*dml_ctx)->v21.mode_support.display_config; - (*dml_ctx)->v21.mode_programming.programming = (struct dml2_display_cfg_programming *)kzalloc(sizeof(struct dml2_display_cfg_programming), GFP_KERNEL); + (*dml_ctx)->v21.mode_programming.programming = kzalloc(sizeof(struct dml2_display_cfg_programming), GFP_KERNEL); if (!((*dml_ctx)->v21.mode_programming.programming)) return false; diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_dchub_registers.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_dchub_registers.h index 83fc15bf13cf..25b607e7b726 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_dchub_registers.h +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_dchub_registers.h @@ -88,6 +88,7 @@ struct dml2_display_arb_regs { uint32_t sdpif_request_rate_limit; uint32_t allow_sdpif_rate_limit_when_cstate_req; uint32_t dcfclk_deep_sleep_hysteresis; + uint32_t pstate_stall_threshold; }; struct dml2_cursor_dlg_regs{ diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.c index 0aa4e4d343b0..3d41ffde91c1 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.c @@ -159,6 +159,7 @@ static void create_phantom_stream_from_main_stream(struct dml2_stream_parameters phantom->timing.v_total = meta->v_total; phantom->timing.v_active = meta->v_active; phantom->timing.v_front_porch = meta->v_front_porch; + phantom->timing.v_blank_end = phantom->timing.v_total - phantom->timing.v_front_porch - phantom->timing.v_active; phantom->timing.vblank_nom = phantom->timing.v_total - phantom->timing.v_active; phantom->timing.drr_config.enabled = false; } diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.c index 3ea54fd52e46..92e43a1e4dd4 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.c @@ -12236,6 +12236,8 @@ static void rq_dlg_get_dlg_reg( static void rq_dlg_get_arb_params(const struct dml2_display_cfg *display_cfg, const struct dml2_core_internal_display_mode_lib *mode_lib, struct dml2_display_arb_regs *arb_param) { + double refclk_freq_in_mhz = (display_cfg->overrides.hw.dlg_ref_clk_mhz > 0) ? (double)display_cfg->overrides.hw.dlg_ref_clk_mhz : mode_lib->soc.dchub_refclk_mhz; + arb_param->max_req_outstanding = mode_lib->soc.max_outstanding_reqs; arb_param->min_req_outstanding = mode_lib->soc.max_outstanding_reqs; // turn off the sat level feature if this set to max arb_param->sdpif_request_rate_limit = (3 * mode_lib->ip.words_per_channel * mode_lib->soc.clk_table.dram_config.channel_count) / 4; @@ -12247,6 +12249,7 @@ static void rq_dlg_get_arb_params(const struct dml2_display_cfg *display_cfg, co arb_param->compbuf_size = mode_lib->mp.CompressedBufferSizeInkByte / mode_lib->ip.compressed_buffer_segment_size_in_kbytes; arb_param->allow_sdpif_rate_limit_when_cstate_req = dml_get_hw_debug5(mode_lib); arb_param->dcfclk_deep_sleep_hysteresis = dml_get_dcfclk_deep_sleep_hysteresis(mode_lib); + arb_param->pstate_stall_threshold = (unsigned int)(mode_lib->ip_caps.fams2.max_allow_delay_us * refclk_freq_in_mhz); #ifdef __DML_VBA_DEBUG__ dml2_printf("DML::%s: max_req_outstanding = %d\n", __func__, arb_param->max_req_outstanding); diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_utils.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_utils.c index ab229e1598ae..714b5c39b7e6 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_utils.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_utils.c @@ -425,6 +425,7 @@ static void create_phantom_stream_from_main_stream(struct dml2_stream_parameters phantom->timing.v_total = meta->v_total; phantom->timing.v_active = meta->v_active; phantom->timing.v_front_porch = meta->v_front_porch; + phantom->timing.v_blank_end = phantom->timing.v_total - phantom->timing.v_front_porch - phantom->timing.v_active; phantom->timing.vblank_nom = phantom->timing.v_total - phantom->timing.v_active; phantom->timing.drr_config.enabled = false; } diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.c index dd9971867f74..92269f0e50ed 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.c @@ -1799,6 +1799,7 @@ bool pmo_dcn4_fams2_init_for_pstate_support(struct dml2_pmo_init_for_pstate_supp } if (s->pmo_dcn4.num_pstate_candidates > 0) { + s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.num_pstate_candidates - 1].allow_state_increase = true; s->pmo_dcn4.cur_pstate_candidate = -1; return true; } else { diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.c b/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.c index 866b0abcff1b..9190c1328d5b 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.c @@ -209,8 +209,6 @@ static bool optimize_configuration(struct dml2_context *dml2, struct dml2_wrappe p->cur_display_config->output.OutputEncoder[0], p->cur_mode_support_info->DSCEnabled[0]) - 1; if (odms_needed <= unused_dpps) { - unused_dpps -= odms_needed; - if (odms_needed == 1) { p->new_policy->ODMUse[0] = dml_odm_use_policy_combine_2to1; optimization_done = true; diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp.h b/drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp.h index cd1706d301e7..f09cba8e29cc 100644 --- a/drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp.h +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp.h @@ -690,6 +690,7 @@ struct dcn20_dpp { int lb_memory_size; int lb_bits_per_entry; bool is_write_to_ram_a_safe; + bool dispclk_r_gate_disable; struct scaler_data scl_data; struct pwl_params pwl_data; }; diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.h b/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.h index b110f35ef66b..f236824126e9 100644 --- a/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.h +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.h @@ -572,6 +572,7 @@ struct dcn3_dpp { int lb_memory_size; int lb_bits_per_entry; bool is_write_to_ram_a_safe; + bool dispclk_r_gate_disable; struct scaler_data scl_data; struct pwl_params pwl_data; }; diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn35/dcn35_dpp.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn35/dcn35_dpp.c index 8473c694bfdc..62b7012cda43 100644 --- a/drivers/gpu/drm/amd/display/dc/dpp/dcn35/dcn35_dpp.c +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn35/dcn35_dpp.c @@ -50,13 +50,21 @@ void dpp35_dppclk_control( DPPCLK_RATE_CONTROL, dppclk_div, DPP_CLOCK_ENABLE, 1); else - REG_UPDATE_2(DPP_CONTROL, + if (dpp->dispclk_r_gate_disable) + REG_UPDATE_2(DPP_CONTROL, DPP_CLOCK_ENABLE, 1, DISPCLK_R_GATE_DISABLE, 1); + else + REG_UPDATE(DPP_CONTROL, + DPP_CLOCK_ENABLE, 1); } else - REG_UPDATE_2(DPP_CONTROL, + if (dpp->dispclk_r_gate_disable) + REG_UPDATE_2(DPP_CONTROL, DPP_CLOCK_ENABLE, 0, DISPCLK_R_GATE_DISABLE, 0); + else + REG_UPDATE(DPP_CONTROL, + DPP_CLOCK_ENABLE, 0); } void dpp35_program_bias_and_scale_fcnv( @@ -128,6 +136,10 @@ bool dpp35_construct( (const struct dcn3_dpp_mask *)(tf_mask)); dpp->base.funcs = &dcn35_dpp_funcs; + + // w/a for cursor memory stuck in LS by programming DISPCLK_R_GATE_DISABLE, limit w/a to some ASIC revs + if (dpp->base.ctx->asic_id.hw_internal_rev <= 0x10) + dpp->dispclk_r_gate_disable = true; return ret; } diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_dscl.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_dscl.c index 5105fd580017..2f92e7d4981b 100644 --- a/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_dscl.c +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_dscl.c @@ -1091,7 +1091,8 @@ void dpp401_dscl_set_scaler_manual_scale(struct dpp *dpp_base, /* ISHARP_DELTA_LUT */ dpp401_dscl_set_isharp_filter(dpp, scl_data->dscl_prog_data.isharp_delta); dpp->scl_data.dscl_prog_data.sharpness_level = scl_data->dscl_prog_data.sharpness_level; - dpp->scl_data.dscl_prog_data.isharp_delta = scl_data->dscl_prog_data.isharp_delta; + memcpy(dpp->scl_data.dscl_prog_data.isharp_delta, scl_data->dscl_prog_data.isharp_delta, + sizeof(uint32_t) * ISHARP_LUT_TABLE_SIZE); if (memcmp(&dpp->scl_data, scl_data, sizeof(*scl_data)) == 0) return; diff --git a/drivers/gpu/drm/amd/display/dc/dwb/dcn30/dcn30_cm_common.h b/drivers/gpu/drm/amd/display/dc/dwb/dcn30/dcn30_cm_common.h index bd98b327a6c7..b86347c9b038 100644 --- a/drivers/gpu/drm/amd/display/dc/dwb/dcn30/dcn30_cm_common.h +++ b/drivers/gpu/drm/amd/display/dc/dwb/dcn30/dcn30_cm_common.h @@ -63,10 +63,6 @@ bool cm3_helper_translate_curve_to_hw_format( const struct dc_transfer_func *output_tf, struct pwl_params *lut_params, bool fixpoint); -bool cm3_helper_translate_curve_to_degamma_hw_format( - const struct dc_transfer_func *output_tf, - struct pwl_params *lut_params); - bool cm3_helper_convert_to_custom_float( struct pwl_result_data *rgb_resulted, struct curve_points3 *corner_points, diff --git a/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c b/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c index f344478e9bd4..b099989d9364 100644 --- a/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c +++ b/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c @@ -443,7 +443,6 @@ struct gpio *dal_gpio_create_irq( case GPIO_ID_GPIO_PAD: break; default: - id = GPIO_ID_HPD; ASSERT_CRITICAL(false); return NULL; } diff --git a/drivers/gpu/drm/amd/display/dc/hubbub/dcn10/dcn10_hubbub.h b/drivers/gpu/drm/amd/display/dc/hubbub/dcn10/dcn10_hubbub.h index a1e2cde9c4cc..4bd1dda07719 100644 --- a/drivers/gpu/drm/amd/display/dc/hubbub/dcn10/dcn10_hubbub.h +++ b/drivers/gpu/drm/amd/display/dc/hubbub/dcn10/dcn10_hubbub.h @@ -198,6 +198,8 @@ struct dcn_hubbub_registers { uint32_t DCHUBBUB_ARB_REFCYC_PER_META_TRIP_B; uint32_t DCHUBBUB_ARB_FRAC_URG_BW_MALL_A; uint32_t DCHUBBUB_ARB_FRAC_URG_BW_MALL_B; + uint32_t DCHUBBUB_TIMEOUT_DETECTION_CTRL1; + uint32_t DCHUBBUB_TIMEOUT_DETECTION_CTRL2; }; #define HUBBUB_REG_FIELD_LIST_DCN32(type) \ @@ -313,7 +315,12 @@ struct dcn_hubbub_registers { type DCN_VM_ERROR_VMID;\ type DCN_VM_ERROR_TABLE_LEVEL;\ type DCN_VM_ERROR_PIPE;\ - type DCN_VM_ERROR_INTERRUPT_STATUS + type DCN_VM_ERROR_INTERRUPT_STATUS;\ + type DCHUBBUB_TIMEOUT_ERROR_STATUS;\ + type DCHUBBUB_TIMEOUT_REQ_STALL_THRESHOLD;\ + type DCHUBBUB_TIMEOUT_PSTATE_STALL_THRESHOLD;\ + type DCHUBBUB_TIMEOUT_DETECTION_EN;\ + type DCHUBBUB_TIMEOUT_TIMER_RESET #define HUBBUB_STUTTER_REG_FIELD_LIST(type) \ type DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A;\ diff --git a/drivers/gpu/drm/amd/display/dc/hubbub/dcn401/dcn401_hubbub.c b/drivers/gpu/drm/amd/display/dc/hubbub/dcn401/dcn401_hubbub.c index 37d26fa0b6fb..5d658e9bef64 100644 --- a/drivers/gpu/drm/amd/display/dc/hubbub/dcn401/dcn401_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/hubbub/dcn401/dcn401_hubbub.c @@ -1192,6 +1192,17 @@ static void dcn401_wait_for_det_update(struct hubbub *hubbub, int hubp_inst) } } +static void dcn401_program_timeout_thresholds(struct hubbub *hubbub, struct dml2_display_arb_regs *arb_regs) +{ + struct dcn20_hubbub *hubbub2 = TO_DCN20_HUBBUB(hubbub); + + /* request backpressure and outstanding return threshold (unused)*/ + //REG_UPDATE(DCHUBBUB_TIMEOUT_DETECTION_CTRL1, DCHUBBUB_TIMEOUT_REQ_STALL_THRESHOLD, arb_regs->req_stall_threshold); + + /* P-State stall threshold */ + REG_UPDATE(DCHUBBUB_TIMEOUT_DETECTION_CTRL2, DCHUBBUB_TIMEOUT_PSTATE_STALL_THRESHOLD, arb_regs->pstate_stall_threshold); +} + static const struct hubbub_funcs hubbub4_01_funcs = { .update_dchub = hubbub2_update_dchub, .init_dchub_sys_ctx = hubbub3_init_dchub_sys_ctx, @@ -1215,6 +1226,7 @@ static const struct hubbub_funcs hubbub4_01_funcs = { .program_det_segments = dcn401_program_det_segments, .program_compbuf_segments = dcn401_program_compbuf_segments, .wait_for_det_update = dcn401_wait_for_det_update, + .program_timeout_thresholds = dcn401_program_timeout_thresholds, }; void hubbub401_construct(struct dcn20_hubbub *hubbub2, diff --git a/drivers/gpu/drm/amd/display/dc/hubbub/dcn401/dcn401_hubbub.h b/drivers/gpu/drm/amd/display/dc/hubbub/dcn401/dcn401_hubbub.h index f35f19ba3e18..5f1960722ebd 100644 --- a/drivers/gpu/drm/amd/display/dc/hubbub/dcn401/dcn401_hubbub.h +++ b/drivers/gpu/drm/amd/display/dc/hubbub/dcn401/dcn401_hubbub.h @@ -123,8 +123,12 @@ HUBBUB_SF(DCHUBBUB_CLOCK_CNTL, DCFCLK_R_DCHUBBUB_GATE_DIS, mask_sh),\ HUBBUB_SF(DCHUBBUB_SDPIF_CFG0, SDPIF_PORT_CONTROL, mask_sh),\ HUBBUB_SF(DCHUBBUB_SDPIF_CFG1, SDPIF_MAX_NUM_OUTSTANDING, mask_sh),\ - HUBBUB_SF(DCHUBBUB_MEM_PWR_MODE_CTRL, DET_MEM_PWR_LS_MODE, mask_sh) - + HUBBUB_SF(DCHUBBUB_MEM_PWR_MODE_CTRL, DET_MEM_PWR_LS_MODE, mask_sh),\ + HUBBUB_SF(DCHUBBUB_TIMEOUT_DETECTION_CTRL1, DCHUBBUB_TIMEOUT_ERROR_STATUS, mask_sh),\ + HUBBUB_SF(DCHUBBUB_TIMEOUT_DETECTION_CTRL1, DCHUBBUB_TIMEOUT_REQ_STALL_THRESHOLD, mask_sh),\ + HUBBUB_SF(DCHUBBUB_TIMEOUT_DETECTION_CTRL2, DCHUBBUB_TIMEOUT_PSTATE_STALL_THRESHOLD, mask_sh),\ + HUBBUB_SF(DCHUBBUB_TIMEOUT_DETECTION_CTRL2, DCHUBBUB_TIMEOUT_DETECTION_EN, mask_sh),\ + HUBBUB_SF(DCHUBBUB_TIMEOUT_DETECTION_CTRL2, DCHUBBUB_TIMEOUT_TIMER_RESET, mask_sh) bool hubbub401_program_urgent_watermarks( struct hubbub *hubbub, diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c index 4fbed0298adf..81f4c386c287 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c @@ -1039,7 +1039,8 @@ void dce110_edp_backlight_control( link_transmitter_control(ctx->dc_bios, &cntl); if (enable && link->dpcd_sink_ext_caps.bits.oled && - !link->dc->config.edp_no_power_sequencing) { + !link->dc->config.edp_no_power_sequencing && + !link->local_sink->edid_caps.panel_patch.oled_optimize_display_on) { post_T7_delay += link->panel_config.pps.extra_post_t7_ms; msleep(post_T7_delay); } @@ -3142,9 +3143,10 @@ static void dce110_set_cursor_attribute(struct pipe_ctx *pipe_ctx) } bool dce110_set_backlight_level(struct pipe_ctx *pipe_ctx, - uint32_t backlight_pwm_u16_16, - uint32_t frame_ramp) + struct set_backlight_level_params *backlight_level_params) { + uint32_t backlight_pwm_u16_16 = backlight_level_params->backlight_pwm_u16_16; + uint32_t frame_ramp = backlight_level_params->frame_ramp; struct dc_link *link = pipe_ctx->stream->link; struct dc *dc = link->ctx->dc; struct abm *abm = pipe_ctx->stream_res.abm; @@ -3315,7 +3317,7 @@ void dce110_disable_link_output(struct dc_link *link, * from enable/disable link output and only call edp panel control * in enable_link_dp and disable_link_dp once. */ - if (dmcu != NULL && dmcu->funcs->lock_phy) + if (dmcu != NULL && dmcu->funcs->unlock_phy) dmcu->funcs->unlock_phy(dmcu); dc->link_srv->dp_trace_source_sequence(link, DPCD_SOURCE_SEQ_AFTER_DISABLE_LINK_PHY); } diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.h index ed3cc3648e8e..06789ac3a224 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.h @@ -88,8 +88,7 @@ void dce110_edp_wait_for_hpd_ready( bool power_up); bool dce110_set_backlight_level(struct pipe_ctx *pipe_ctx, - uint32_t backlight_pwm_u16_16, - uint32_t frame_ramp); + struct set_backlight_level_params *params); void dce110_set_abm_immediate_disable(struct pipe_ctx *pipe_ctx); void dce110_set_pipe(struct pipe_ctx *pipe_ctx); void dce110_disable_link_output(struct dc_link *link, diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c index a6a1db5ba8ba..681bb92c6069 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c @@ -3453,7 +3453,6 @@ static bool dcn10_can_pipe_disable_cursor(struct pipe_ctx *pipe_ctx) r2 = test_pipe->plane_res.scl_data.recout; r2_r = r2.x + r2.width; r2_b = r2.y + r2.height; - split_pipe = test_pipe; /** * There is another half plane on same layer because of diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c index a80c08582932..05424a9af58b 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c @@ -1458,8 +1458,12 @@ void dcn20_pipe_control_lock( } else { if (lock) pipe->stream_res.tg->funcs->lock(pipe->stream_res.tg); - else - pipe->stream_res.tg->funcs->unlock(pipe->stream_res.tg); + else { + if (dc->hwseq->funcs.perform_3dlut_wa_unlock) + dc->hwseq->funcs.perform_3dlut_wa_unlock(pipe); + else + pipe->stream_res.tg->funcs->unlock(pipe->stream_res.tg); + } } } @@ -1732,7 +1736,6 @@ static void dcn20_update_dchubp_dpp( if (pipe_ctx->update_flags.bits.scaler || plane_state->update_flags.bits.scaling_change || plane_state->update_flags.bits.position_change || - plane_state->update_flags.bits.clip_size_change || plane_state->update_flags.bits.per_pixel_alpha_change || pipe_ctx->stream->update_flags.bits.scaling) { pipe_ctx->plane_res.scl_data.lb_params.alpha_en = pipe_ctx->plane_state->per_pixel_alpha; @@ -1745,7 +1748,6 @@ static void dcn20_update_dchubp_dpp( if (pipe_ctx->update_flags.bits.viewport || (context == dc->current_state && plane_state->update_flags.bits.position_change) || (context == dc->current_state && plane_state->update_flags.bits.scaling_change) || - (context == dc->current_state && plane_state->update_flags.bits.clip_size_change) || (context == dc->current_state && pipe_ctx->stream->update_flags.bits.scaling)) { hubp->funcs->mem_program_viewport( @@ -2056,22 +2058,15 @@ void dcn20_program_front_end_for_ctx( */ for (i = 0; i < dc->res_pool->pipe_count; i++) { struct dc_stream_state *stream = dc->current_state->res_ctx.pipe_ctx[i].stream; + pipe = &dc->current_state->res_ctx.pipe_ctx[i]; if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable && stream && - dc_state_get_pipe_subvp_type(dc->current_state, &dc->current_state->res_ctx.pipe_ctx[i]) == SUBVP_PHANTOM) { + dc_state_get_pipe_subvp_type(dc->current_state, pipe) == SUBVP_PHANTOM) { struct timing_generator *tg = dc->current_state->res_ctx.pipe_ctx[i].stream_res.tg; if (tg->funcs->enable_crtc) { - if (dc->hwss.blank_phantom) { - int main_pipe_width = 0, main_pipe_height = 0; - struct dc_stream_state *phantom_stream = dc_state_get_paired_subvp_stream(dc->current_state, dc->current_state->res_ctx.pipe_ctx[i].stream); - - if (phantom_stream) { - main_pipe_width = phantom_stream->dst.width; - main_pipe_height = phantom_stream->dst.height; - } - - dc->hwss.blank_phantom(dc, tg, main_pipe_width, main_pipe_height); + if (dc->hwseq->funcs.blank_pixel_data) { + dc->hwseq->funcs.blank_pixel_data(dc, pipe, true); } tg->funcs->enable_crtc(tg); } @@ -2255,9 +2250,9 @@ void dcn20_post_unlock_program_front_end( struct timing_generator *tg = pipe->stream_res.tg; - if (tg->funcs->get_double_buffer_pending) { + if (tg->funcs->get_optc_double_buffer_pending) { for (j = 0; j < TIMEOUT_FOR_PIPE_ENABLE_US / polling_interval_us - && tg->funcs->get_double_buffer_pending(tg); j++) + && tg->funcs->get_optc_double_buffer_pending(tg); j++) udelay(polling_interval_us); } } @@ -2771,7 +2766,6 @@ void dcn20_reset_back_end_for_pipe( struct pipe_ctx *pipe_ctx, struct dc_state *context) { - int i; struct dc_link *link = pipe_ctx->stream->link; const struct link_hwss *link_hwss = get_link_hwss(link, &pipe_ctx->link_res); @@ -2838,19 +2832,16 @@ void dcn20_reset_back_end_for_pipe( } } - for (i = 0; i < dc->res_pool->pipe_count; i++) - if (&dc->current_state->res_ctx.pipe_ctx[i] == pipe_ctx) - break; - - if (i == dc->res_pool->pipe_count) - return; - /* * In case of a dangling plane, setting this to NULL unconditionally * causes failures during reset hw ctx where, if stream is NULL, * it is expected that the pipe_ctx pointers to pipes and plane are NULL. */ pipe_ctx->stream = NULL; + pipe_ctx->top_pipe = NULL; + pipe_ctx->bottom_pipe = NULL; + pipe_ctx->next_odm_pipe = NULL; + pipe_ctx->prev_odm_pipe = NULL; DC_LOG_DEBUG("Reset back end for pipe %d, tg:%d\n", pipe_ctx->pipe_idx, pipe_ctx->stream_res.tg->inst); } diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_hwseq.c index 1ea95f8d4cbc..61efb15572ff 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_hwseq.c @@ -137,7 +137,7 @@ void dcn21_PLAT_58856_wa(struct dc_state *context, struct pipe_ctx *pipe_ctx) pipe_ctx->stream->dpms_off = true; } -static bool dmub_abm_set_pipe(struct abm *abm, uint32_t otg_inst, +bool dcn21_dmub_abm_set_pipe(struct abm *abm, uint32_t otg_inst, uint32_t option, uint32_t panel_inst, uint32_t pwrseq_inst) { union dmub_rb_cmd cmd; @@ -199,7 +199,7 @@ void dcn21_set_abm_immediate_disable(struct pipe_ctx *pipe_ctx) abm->funcs->set_pipe_ex(abm, otg_inst, SET_ABM_PIPE_IMMEDIATELY_DISABLE, panel_cntl->inst, panel_cntl->pwrseq_inst); } else { - dmub_abm_set_pipe(abm, + dcn21_dmub_abm_set_pipe(abm, otg_inst, SET_ABM_PIPE_IMMEDIATELY_DISABLE, panel_cntl->inst, @@ -234,7 +234,7 @@ void dcn21_set_pipe(struct pipe_ctx *pipe_ctx) panel_cntl->inst, panel_cntl->pwrseq_inst); } else { - dmub_abm_set_pipe(abm, otg_inst, + dcn21_dmub_abm_set_pipe(abm, otg_inst, SET_ABM_PIPE_NORMAL, panel_cntl->inst, panel_cntl->pwrseq_inst); @@ -242,14 +242,15 @@ void dcn21_set_pipe(struct pipe_ctx *pipe_ctx) } bool dcn21_set_backlight_level(struct pipe_ctx *pipe_ctx, - uint32_t backlight_pwm_u16_16, - uint32_t frame_ramp) + struct set_backlight_level_params *backlight_level_params) { struct dc_context *dc = pipe_ctx->stream->ctx; struct abm *abm = pipe_ctx->stream_res.abm; struct timing_generator *tg = pipe_ctx->stream_res.tg; struct panel_cntl *panel_cntl = pipe_ctx->stream->link->panel_cntl; uint32_t otg_inst; + uint32_t backlight_pwm_u16_16 = backlight_level_params->backlight_pwm_u16_16; + uint32_t frame_ramp = backlight_level_params->frame_ramp; if (!abm || !tg || !panel_cntl) return false; @@ -257,7 +258,7 @@ bool dcn21_set_backlight_level(struct pipe_ctx *pipe_ctx, otg_inst = tg->inst; if (dc->dc->res_pool->dmcu) { - dce110_set_backlight_level(pipe_ctx, backlight_pwm_u16_16, frame_ramp); + dce110_set_backlight_level(pipe_ctx, backlight_level_params); return true; } @@ -268,7 +269,7 @@ bool dcn21_set_backlight_level(struct pipe_ctx *pipe_ctx, panel_cntl->inst, panel_cntl->pwrseq_inst); } else { - dmub_abm_set_pipe(abm, + dcn21_dmub_abm_set_pipe(abm, otg_inst, SET_ABM_PIPE_NORMAL, panel_cntl->inst, diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_hwseq.h index 9cee9bdb8de9..f72a27ac1bf1 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_hwseq.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_hwseq.h @@ -47,11 +47,12 @@ void dcn21_optimize_pwr_state( void dcn21_PLAT_58856_wa(struct dc_state *context, struct pipe_ctx *pipe_ctx); +bool dcn21_dmub_abm_set_pipe(struct abm *abm, uint32_t otg_inst, + uint32_t option, uint32_t panel_inst, uint32_t pwrseq_inst); void dcn21_set_pipe(struct pipe_ctx *pipe_ctx); void dcn21_set_abm_immediate_disable(struct pipe_ctx *pipe_ctx); bool dcn21_set_backlight_level(struct pipe_ctx *pipe_ctx, - uint32_t backlight_pwm_u16_16, - uint32_t frame_ramp); + struct set_backlight_level_params *params); bool dcn21_is_abm_supported(struct dc *dc, struct dc_state *context, struct dc_stream_state *stream); diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c index bded33575493..e89ebfda4873 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c @@ -245,6 +245,7 @@ static bool dcn30_set_mpc_shaper_3dlut(struct pipe_ctx *pipe_ctx, { struct dpp *dpp_base = pipe_ctx->plane_res.dpp; int mpcc_id = pipe_ctx->plane_res.hubp->inst; + struct dc *dc = pipe_ctx->stream->ctx->dc; struct mpc *mpc = pipe_ctx->stream_res.opp->ctx->dc->res_pool->mpc; bool result = false; int acquired_rmu = 0; @@ -283,8 +284,14 @@ static bool dcn30_set_mpc_shaper_3dlut(struct pipe_ctx *pipe_ctx, result = mpc->funcs->program_3dlut(mpc, &stream->lut3d_func->lut_3d, stream->lut3d_func->state.bits.rmu_mux_num); + if (!result) + DC_LOG_ERROR("%s: program_3dlut failed\n", __func__); + result = mpc->funcs->program_shaper(mpc, shaper_lut, stream->lut3d_func->state.bits.rmu_mux_num); + if (!result) + DC_LOG_ERROR("%s: program_shaper failed\n", __func__); + } else { // loop through the available mux and release the requested mpcc_id mpc->funcs->release_rmu(mpc, mpcc_id); @@ -486,7 +493,6 @@ bool dcn30_mmhubbub_warmup( } /*following is the original: warmup each DWB's mcif buffer*/ for (i = 0; i < num_dwb; i++) { - dwb = dc->res_pool->dwbc[wb_info[i].dwb_pipe_inst]; mcif_wb = dc->res_pool->mcif_wb[wb_info[i].dwb_pipe_inst]; /*warmup is for VM mode only*/ if (wb_info[i].mcif_buf_params.p_vmid == 0) @@ -1185,3 +1191,30 @@ void dcn30_prepare_bandwidth(struct dc *dc, if (!dc->clk_mgr->clks.fw_based_mclk_switching) dc_dmub_srv_p_state_delegate(dc, false, context); } + +void dcn30_wait_for_all_pending_updates(const struct pipe_ctx *pipe_ctx) +{ + struct timing_generator *tg = pipe_ctx->stream_res.tg; + bool pending_updates = false; + unsigned int i; + + if (tg && tg->funcs->is_tg_enabled(tg)) { + // Poll for 100ms maximum + for (i = 0; i < 100000; i++) { + pending_updates = false; + if (tg->funcs->get_optc_double_buffer_pending) + pending_updates |= tg->funcs->get_optc_double_buffer_pending(tg); + + if (tg->funcs->get_otg_double_buffer_pending) + pending_updates |= tg->funcs->get_otg_double_buffer_pending(tg); + + if (tg->funcs->get_pipe_update_pending && pipe_ctx->plane_state) + pending_updates |= tg->funcs->get_pipe_update_pending(tg); + + if (!pending_updates) + break; + + udelay(1); + } + } +} diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.h index 6a153e7ce910..4b90b781c4f2 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.h @@ -96,4 +96,6 @@ void dcn30_set_hubp_blank(const struct dc *dc, void dcn30_prepare_bandwidth(struct dc *dc, struct dc_state *context); +void dcn30_wait_for_all_pending_updates(const struct pipe_ctx *pipe_ctx); + #endif /* __DC_HWSS_DCN30_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_init.c index 2a8dc40d2847..0e8d32e3dbae 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_init.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_init.c @@ -108,7 +108,8 @@ static const struct hw_sequencer_funcs dcn30_funcs = { .set_disp_pattern_generator = dcn30_set_disp_pattern_generator, .get_dcc_en_bits = dcn10_get_dcc_en_bits, .update_visual_confirm_color = dcn10_update_visual_confirm_color, - .is_abm_supported = dcn21_is_abm_supported + .is_abm_supported = dcn21_is_abm_supported, + .wait_for_all_pending_updates = dcn30_wait_for_all_pending_updates, }; static const struct hwseq_private_funcs dcn30_private_funcs = { diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn301/dcn301_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn301/dcn301_init.c index 93e49d87a67c..780ce4c064aa 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn301/dcn301_init.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn301/dcn301_init.c @@ -107,6 +107,7 @@ static const struct hw_sequencer_funcs dcn301_funcs = { .optimize_pwr_state = dcn21_optimize_pwr_state, .exit_optimized_pwr_state = dcn21_exit_optimized_pwr_state, .update_visual_confirm_color = dcn10_update_visual_confirm_color, + .wait_for_all_pending_updates = dcn30_wait_for_all_pending_updates, }; static const struct hwseq_private_funcs dcn301_private_funcs = { diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn301/dcn301_init.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn301/dcn301_init.h index 0bca48ccbfa2..a6e0115a53ee 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn301/dcn301_init.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn301/dcn301_init.h @@ -23,8 +23,8 @@ * */ -#ifndef __DC_DCN30_INIT_H__ -#define __DC_DCN30_INIT_H__ +#ifndef __DC_DCN301_INIT_H__ +#define __DC_DCN301_INIT_H__ struct dc; diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.c index 3d4b31bd9946..03ba01f4ace1 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.c @@ -47,9 +47,11 @@ #include "dce/dmub_outbox.h" #include "link.h" #include "dcn10/dcn10_hwseq.h" +#include "dcn21/dcn21_hwseq.h" #include "inc/link_enc_cfg.h" #include "dcn30/dcn30_vpg.h" #include "dce/dce_i2c_hw.h" +#include "dce/dmub_abm_lcd.h" #define DC_LOGGER_INIT(logger) @@ -517,10 +519,18 @@ static void dcn31_reset_back_end_for_pipe( dc->hwss.set_abm_immediate_disable(pipe_ctx); + link = pipe_ctx->stream->link; + + if ((!pipe_ctx->stream->dpms_off || link->link_status.link_active) && + (link->connector_signal == SIGNAL_TYPE_EDP)) + dc->hwss.blank_stream(pipe_ctx); + pipe_ctx->stream_res.tg->funcs->set_dsc_config( pipe_ctx->stream_res.tg, OPTC_DSC_DISABLED, 0, 0); + pipe_ctx->stream_res.tg->funcs->disable_crtc(pipe_ctx->stream_res.tg); + pipe_ctx->stream_res.tg->funcs->enable_optc_clock(pipe_ctx->stream_res.tg, false); if (pipe_ctx->stream_res.tg->funcs->set_odm_bypass) pipe_ctx->stream_res.tg->funcs->set_odm_bypass( @@ -532,7 +542,6 @@ static void dcn31_reset_back_end_for_pipe( pipe_ctx->stream_res.tg->funcs->set_drr( pipe_ctx->stream_res.tg, NULL); - link = pipe_ctx->stream->link; /* DPMS may already disable or */ /* dpms_off status is incorrect due to fastboot * feature. When system resume from S4 with second @@ -633,3 +642,51 @@ void dcn31_set_static_screen_control(struct pipe_ctx **pipe_ctx, pipe_ctx[i]->stream_res.tg->funcs->set_static_screen_control(pipe_ctx[i]->stream_res.tg, triggers, params->num_frames); } + +static void dmub_abm_set_backlight(struct dc_context *dc, + struct set_backlight_level_params *backlight_level_params, uint32_t panel_inst) +{ + union dmub_rb_cmd cmd; + + memset(&cmd, 0, sizeof(cmd)); + cmd.abm_set_backlight.header.type = DMUB_CMD__ABM; + cmd.abm_set_backlight.header.sub_type = DMUB_CMD__ABM_SET_BACKLIGHT; + cmd.abm_set_backlight.abm_set_backlight_data.frame_ramp = backlight_level_params->frame_ramp; + cmd.abm_set_backlight.abm_set_backlight_data.backlight_user_level = backlight_level_params->backlight_pwm_u16_16; + cmd.abm_set_backlight.abm_set_backlight_data.backlight_control_type = + (enum dmub_backlight_control_type) backlight_level_params->control_type; + cmd.abm_set_backlight.abm_set_backlight_data.min_luminance = backlight_level_params->min_luminance; + cmd.abm_set_backlight.abm_set_backlight_data.max_luminance = backlight_level_params->max_luminance; + cmd.abm_set_backlight.abm_set_backlight_data.min_backlight_pwm = backlight_level_params->min_backlight_pwm; + cmd.abm_set_backlight.abm_set_backlight_data.max_backlight_pwm = backlight_level_params->max_backlight_pwm; + cmd.abm_set_backlight.abm_set_backlight_data.version = DMUB_CMD_ABM_CONTROL_VERSION_1; + cmd.abm_set_backlight.abm_set_backlight_data.panel_mask = (0x01 << panel_inst); + cmd.abm_set_backlight.header.payload_bytes = sizeof(struct dmub_cmd_abm_set_backlight_data); + + dc_wake_and_execute_dmub_cmd(dc, &cmd, DM_DMUB_WAIT_TYPE_WAIT); +} + +bool dcn31_set_backlight_level(struct pipe_ctx *pipe_ctx, + struct set_backlight_level_params *backlight_level_params) +{ + struct dc_context *dc = pipe_ctx->stream->ctx; + struct abm *abm = pipe_ctx->stream_res.abm; + struct timing_generator *tg = pipe_ctx->stream_res.tg; + struct panel_cntl *panel_cntl = pipe_ctx->stream->link->panel_cntl; + uint32_t otg_inst; + + if (!abm || !tg || !panel_cntl) + return false; + + otg_inst = tg->inst; + + dcn21_dmub_abm_set_pipe(abm, + otg_inst, + SET_ABM_PIPE_NORMAL, + panel_cntl->inst, + panel_cntl->pwrseq_inst); + + dmub_abm_set_backlight(dc, backlight_level_params, panel_cntl->inst); + + return true; +} diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.h index b8bc939da155..0d09aa8cfb65 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.h @@ -51,6 +51,8 @@ int dcn31_init_sys_ctx(struct dce_hwseq *hws, struct dc *dc, struct dc_phy_addr_ void dcn31_reset_hw_ctx_wrap( struct dc *dc, struct dc_state *context); +bool dcn31_set_backlight_level(struct pipe_ctx *pipe_ctx, + struct set_backlight_level_params *params); bool dcn31_is_abm_supported(struct dc *dc, struct dc_state *context, struct dc_stream_state *stream); void dcn31_init_pipes(struct dc *dc, struct dc_state *context); @@ -59,5 +61,4 @@ void dcn31_setup_hpo_hw_control(const struct dce_hwseq *hws, bool enable); void dcn31_set_static_screen_control(struct pipe_ctx **pipe_ctx, int num_pipes, const struct dc_static_screen_params *params); - #endif /* __DC_HWSS_DCN31_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_init.c index 56f3c70d4b55..5f8f45b48720 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_init.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_init.c @@ -98,7 +98,7 @@ static const struct hw_sequencer_funcs dcn31_funcs = { .set_flip_control_gsl = dcn20_set_flip_control_gsl, .get_vupdate_offset_from_vsync = dcn10_get_vupdate_offset_from_vsync, .calc_vupdate_position = dcn10_calc_vupdate_position, - .set_backlight_level = dcn21_set_backlight_level, + .set_backlight_level = dcn31_set_backlight_level, .set_abm_immediate_disable = dcn21_set_abm_immediate_disable, .set_pipe = dcn21_set_pipe, .enable_lvds_link_output = dce110_enable_lvds_link_output, diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c index 4e93eeedfc1b..9b88eb72086d 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c @@ -355,6 +355,20 @@ void dcn314_calculate_pix_rate_divider( } } +static bool dcn314_is_pipe_dig_fifo_on(struct pipe_ctx *pipe) +{ + return pipe && pipe->stream + // Check dig's otg instance. + && pipe->stream_res.stream_enc + && pipe->stream_res.stream_enc->funcs->dig_source_otg + && pipe->stream_res.tg->inst == pipe->stream_res.stream_enc->funcs->dig_source_otg(pipe->stream_res.stream_enc) + && pipe->stream->link && pipe->stream->link->link_enc + && pipe->stream->link->link_enc->funcs->is_dig_enabled + && pipe->stream->link->link_enc->funcs->is_dig_enabled(pipe->stream->link->link_enc) + && pipe->stream_res.stream_enc->funcs->is_fifo_enabled + && pipe->stream_res.stream_enc->funcs->is_fifo_enabled(pipe->stream_res.stream_enc); +} + void dcn314_resync_fifo_dccg_dio(struct dce_hwseq *hws, struct dc *dc, struct dc_state *context, unsigned int current_pipe_idx) { unsigned int i; @@ -371,7 +385,11 @@ void dcn314_resync_fifo_dccg_dio(struct dce_hwseq *hws, struct dc *dc, struct dc if (pipe->top_pipe || pipe->prev_odm_pipe) continue; - if (pipe->stream && (pipe->stream->dpms_off || dc_is_virtual_signal(pipe->stream->signal))) { + if (pipe->stream && (pipe->stream->dpms_off || dc_is_virtual_signal(pipe->stream->signal)) && + !pipe->stream->apply_seamless_boot_optimization && + !pipe->stream->apply_edp_fast_boot_optimization) { + if (dcn314_is_pipe_dig_fifo_on(pipe)) + continue; pipe->stream_res.tg->funcs->disable_crtc(pipe->stream_res.tg); reset_sync_context_for_pipe(dc, context, i); otg_disabled[i] = true; @@ -478,7 +496,7 @@ void dcn314_disable_link_output(struct dc_link *link, * from enable/disable link output and only call edp panel control * in enable_link_dp and disable_link_dp once. */ - if (dmcu != NULL && dmcu->funcs->lock_phy) + if (dmcu != NULL && dmcu->funcs->unlock_phy) dmcu->funcs->unlock_phy(dmcu); dc->link_srv->dp_trace_source_sequence(link, DPCD_SOURCE_SEQ_AFTER_DISABLE_LINK_PHY); diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_init.c index 68e6de6b5758..6bdfbf22ce87 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_init.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_init.c @@ -100,7 +100,7 @@ static const struct hw_sequencer_funcs dcn314_funcs = { .set_flip_control_gsl = dcn20_set_flip_control_gsl, .get_vupdate_offset_from_vsync = dcn10_get_vupdate_offset_from_vsync, .calc_vupdate_position = dcn10_calc_vupdate_position, - .set_backlight_level = dcn21_set_backlight_level, + .set_backlight_level = dcn31_set_backlight_level, .set_abm_immediate_disable = dcn21_set_abm_immediate_disable, .set_pipe = dcn21_set_pipe, .enable_lvds_link_output = dce110_enable_lvds_link_output, diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c index 2e8c9f738259..d7f8b2dcaa6b 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c @@ -439,6 +439,7 @@ bool dcn32_set_mpc_shaper_3dlut( { struct dpp *dpp_base = pipe_ctx->plane_res.dpp; int mpcc_id = pipe_ctx->plane_res.hubp->inst; + struct dc *dc = pipe_ctx->stream->ctx->dc; struct mpc *mpc = pipe_ctx->stream_res.opp->ctx->dc->res_pool->mpc; bool result = false; @@ -458,13 +459,13 @@ bool dcn32_set_mpc_shaper_3dlut( if (stream->lut3d_func && stream->lut3d_func->state.bits.initialized == 1) { - result = mpc->funcs->program_3dlut(mpc, - &stream->lut3d_func->lut_3d, - mpcc_id); + result = mpc->funcs->program_3dlut(mpc, &stream->lut3d_func->lut_3d, mpcc_id); + if (!result) + DC_LOG_ERROR("%s: program_3dlut failed\n", __func__); - result = mpc->funcs->program_shaper(mpc, - shaper_lut, - mpcc_id); + result = mpc->funcs->program_shaper(mpc, shaper_lut, mpcc_id); + if (!result) + DC_LOG_ERROR("%s: program_shaper failed\n", __func__); } return result; @@ -1398,10 +1399,10 @@ void dcn32_disable_link_output(struct dc_link *link, link->phy_state.symclk_state = SYMCLK_OFF_TX_OFF; if (signal == SIGNAL_TYPE_EDP && - link->dc->hwss.edp_backlight_control && + link->dc->hwss.edp_power_control && !link->skip_implict_edp_power_control) link->dc->hwss.edp_power_control(link, false); - else if (dmcu != NULL && dmcu->funcs->lock_phy) + else if (dmcu != NULL && dmcu->funcs->unlock_phy) dmcu->funcs->unlock_phy(dmcu); dc->link_srv->dp_trace_source_sequence(link, DPCD_SOURCE_SEQ_AFTER_DISABLE_LINK_PHY); @@ -1698,52 +1699,6 @@ void dcn32_init_blank( hws->funcs.wait_for_blank_complete(opp); } -void dcn32_blank_phantom(struct dc *dc, - struct timing_generator *tg, - int width, - int height) -{ - struct dce_hwseq *hws = dc->hwseq; - enum dc_color_space color_space; - struct tg_color black_color = {0}; - struct output_pixel_processor *opp = NULL; - uint32_t num_opps, opp_id_src0, opp_id_src1; - uint32_t otg_active_width, otg_active_height; - uint32_t i; - - /* program opp dpg blank color */ - color_space = COLOR_SPACE_SRGB; - color_space_to_black_color(dc, color_space, &black_color); - - otg_active_width = width; - otg_active_height = height; - - /* get the OPTC source */ - tg->funcs->get_optc_source(tg, &num_opps, &opp_id_src0, &opp_id_src1); - ASSERT(opp_id_src0 < dc->res_pool->res_cap->num_opp); - - for (i = 0; i < dc->res_pool->res_cap->num_opp; i++) { - if (dc->res_pool->opps[i] != NULL && dc->res_pool->opps[i]->inst == opp_id_src0) { - opp = dc->res_pool->opps[i]; - break; - } - } - - if (opp && opp->funcs->opp_set_disp_pattern_generator) - opp->funcs->opp_set_disp_pattern_generator( - opp, - CONTROLLER_DP_TEST_PATTERN_SOLID_COLOR, - CONTROLLER_DP_COLOR_SPACE_UDEFINED, - COLOR_DEPTH_UNDEFINED, - &black_color, - otg_active_width, - otg_active_height, - 0); - - if (tg->funcs->is_tg_enabled(tg)) - hws->funcs.wait_for_blank_complete(opp); -} - /* phantom stream id's can change often, but can be identical between contexts. * This function checks for the condition the streams are identical to avoid * redundant pipe transitions. diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.h index cac4a08b92a4..0303a5953673 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.h @@ -119,11 +119,6 @@ void dcn32_init_blank( struct dc *dc, struct timing_generator *tg); -void dcn32_blank_phantom(struct dc *dc, - struct timing_generator *tg, - int width, - int height); - bool dcn32_is_pipe_topology_transition_seamless(struct dc *dc, const struct dc_state *cur_ctx, const struct dc_state *new_ctx); diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_init.c index 3422b564ae98..5ecee7e320da 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_init.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_init.c @@ -98,7 +98,7 @@ static const struct hw_sequencer_funcs dcn32_funcs = { .calc_vupdate_position = dcn10_calc_vupdate_position, .apply_idle_power_optimizations = dcn32_apply_idle_power_optimizations, .does_plane_fit_in_mall = NULL, - .set_backlight_level = dcn21_set_backlight_level, + .set_backlight_level = dcn31_set_backlight_level, .set_abm_immediate_disable = dcn21_set_abm_immediate_disable, .hardware_release = dcn30_hardware_release, .set_pipe = dcn21_set_pipe, @@ -117,10 +117,10 @@ static const struct hw_sequencer_funcs dcn32_funcs = { .update_phantom_vp_position = dcn32_update_phantom_vp_position, .update_dsc_pg = dcn32_update_dsc_pg, .apply_update_flags_for_phantom = dcn32_apply_update_flags_for_phantom, - .blank_phantom = dcn32_blank_phantom, .is_pipe_topology_transition_seamless = dcn32_is_pipe_topology_transition_seamless, .calculate_pix_rate_divider = dcn32_calculate_pix_rate_divider, .program_outstanding_updates = dcn32_program_outstanding_updates, + .wait_for_all_pending_updates = dcn30_wait_for_all_pending_updates, }; static const struct hwseq_private_funcs dcn32_private_funcs = { diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c index bd309dbdf7b2..e599cdc465bf 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c @@ -309,6 +309,7 @@ void dcn35_init_hw(struct dc *dc) dc_dmub_srv_query_caps_cmd(dc->ctx->dmub_srv); dc->caps.dmub_caps.psr = dc->ctx->dmub_srv->dmub->feature_caps.psr; dc->caps.dmub_caps.mclk_sw = dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch_ver; + dc->caps.dmub_caps.aux_backlight_support = dc->ctx->dmub_srv->dmub->feature_caps.abm_aux_backlight_support; } if (dc->res_pool->pg_cntl) { @@ -841,6 +842,7 @@ void dcn35_init_pipes(struct dc *dc, struct dc_state *context) uint32_t num_opps = 0; uint32_t opp_id_src0 = OPP_ID_INVALID; uint32_t opp_id_src1 = OPP_ID_INVALID; + uint32_t optc_dsc_state = 0; // Step 1: To find out which OPTC is running & OPTC DSC is ON // We can't use res_pool->res_cap->num_timing_generator to check @@ -849,7 +851,6 @@ void dcn35_init_pipes(struct dc *dc, struct dc_state *context) // Some ASICs would be fused display pipes less than the default setting. // In dcnxx_resource_construct function, driver would obatin real information. for (i = 0; i < dc->res_pool->timing_generator_count; i++) { - uint32_t optc_dsc_state = 0; struct timing_generator *tg = dc->res_pool->timing_generators[i]; if (tg->funcs->is_tg_enabled(tg)) { @@ -864,15 +865,18 @@ void dcn35_init_pipes(struct dc *dc, struct dc_state *context) } } - // Step 2: To power down DSC but skip DSC of running OPTC + // Step 2: To power down DSC but skip DSC of running OPTC for (i = 0; i < dc->res_pool->res_cap->num_dsc; i++) { struct dcn_dsc_state s = {0}; - dc->res_pool->dscs[i]->funcs->dsc_read_state(dc->res_pool->dscs[i], &s); + /* avoid reading DSC state when it is not in use as it may be power gated */ + if (optc_dsc_state) { + dc->res_pool->dscs[i]->funcs->dsc_read_state(dc->res_pool->dscs[i], &s); - if ((s.dsc_opp_source == opp_id_src0 || s.dsc_opp_source == opp_id_src1) && - s.dsc_clock_en && s.dsc_fw_en) - continue; + if ((s.dsc_opp_source == opp_id_src0 || s.dsc_opp_source == opp_id_src1) && + s.dsc_clock_en && s.dsc_fw_en) + continue; + } pg_cntl->funcs->dsc_pg_control(pg_cntl, dc->res_pool->dscs[i]->inst, false); } diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c index 2bbf1fef94fd..fd67779c27a9 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c @@ -101,7 +101,7 @@ static const struct hw_sequencer_funcs dcn35_funcs = { .set_flip_control_gsl = dcn20_set_flip_control_gsl, .get_vupdate_offset_from_vsync = dcn10_get_vupdate_offset_from_vsync, .calc_vupdate_position = dcn10_calc_vupdate_position, - .set_backlight_level = dcn21_set_backlight_level, + .set_backlight_level = dcn31_set_backlight_level, .set_abm_immediate_disable = dcn21_set_abm_immediate_disable, .set_pipe = dcn21_set_pipe, .enable_lvds_link_output = dce110_enable_lvds_link_output, @@ -123,7 +123,6 @@ static const struct hw_sequencer_funcs dcn35_funcs = { .root_clock_control = dcn35_root_clock_control, .set_long_vtotal = dcn35_set_long_vblank, .calculate_pix_rate_divider = dcn32_calculate_pix_rate_divider, - .program_outstanding_updates = dcn32_program_outstanding_updates, }; static const struct hwseq_private_funcs dcn35_private_funcs = { diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_init.c index d00822e8daa5..3c275a1eff58 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_init.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_init.c @@ -100,7 +100,7 @@ static const struct hw_sequencer_funcs dcn351_funcs = { .set_flip_control_gsl = dcn20_set_flip_control_gsl, .get_vupdate_offset_from_vsync = dcn10_get_vupdate_offset_from_vsync, .calc_vupdate_position = dcn10_calc_vupdate_position, - .set_backlight_level = dcn21_set_backlight_level, + .set_backlight_level = dcn31_set_backlight_level, .set_abm_immediate_disable = dcn21_set_abm_immediate_disable, .set_pipe = dcn21_set_pipe, .enable_lvds_link_output = dce110_enable_lvds_link_output, @@ -122,7 +122,6 @@ static const struct hw_sequencer_funcs dcn351_funcs = { .root_clock_control = dcn35_root_clock_control, .set_long_vtotal = dcn35_set_long_vblank, .calculate_pix_rate_divider = dcn32_calculate_pix_rate_divider, - .program_outstanding_updates = dcn32_program_outstanding_updates, .setup_hpo_hw_control = dcn35_setup_hpo_hw_control, }; diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c index 0b743669f23b..e8cc1bfa73f3 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c @@ -506,7 +506,7 @@ void dcn401_populate_mcm_luts(struct dc *dc, dcn401_get_mcm_lut_xable_from_pipe_ctx(dc, pipe_ctx, &shaper_xable, &lut3d_xable, &lut1d_xable); /* 1D LUT */ - if (mcm_luts.lut1d_func && lut3d_xable != MCM_LUT_DISABLE) { + if (mcm_luts.lut1d_func) { memset(&m_lut_params, 0, sizeof(m_lut_params)); if (mcm_luts.lut1d_func->type == TF_TYPE_HWPWL) m_lut_params.pwl = &mcm_luts.lut1d_func->pwl; @@ -521,7 +521,7 @@ void dcn401_populate_mcm_luts(struct dc *dc, mpc->funcs->populate_lut(mpc, MCM_LUT_1DLUT, m_lut_params, lut_bank_a, mpcc_id); } if (mpc->funcs->program_lut_mode) - mpc->funcs->program_lut_mode(mpc, MCM_LUT_1DLUT, lut1d_xable, lut_bank_a, mpcc_id); + mpc->funcs->program_lut_mode(mpc, MCM_LUT_1DLUT, lut1d_xable && m_lut_params.pwl, lut_bank_a, mpcc_id); } /* Shaper */ @@ -669,11 +669,17 @@ bool dcn401_set_mcm_luts(struct pipe_ctx *pipe_ctx, { struct dpp *dpp_base = pipe_ctx->plane_res.dpp; int mpcc_id = pipe_ctx->plane_res.hubp->inst; - struct mpc *mpc = pipe_ctx->stream_res.opp->ctx->dc->res_pool->mpc; + struct dc *dc = pipe_ctx->stream_res.opp->ctx->dc; + struct mpc *mpc = dc->res_pool->mpc; bool result; const struct pwl_params *lut_params = NULL; bool rval; + if (plane_state->mcm_luts.lut3d_data.lut3d_src == DC_CM2_TRANSFER_FUNC_SOURCE_VIDMEM) { + dcn401_populate_mcm_luts(dc, pipe_ctx, plane_state->mcm_luts, plane_state->lut_bank_a); + return true; + } + mpc->funcs->set_movable_cm_location(mpc, MPCC_MOVABLE_CM_LOCATION_BEFORE, mpcc_id); pipe_ctx->plane_state->mcm_location = MPCC_MOVABLE_CM_LOCATION_BEFORE; // 1D LUT @@ -844,6 +850,13 @@ enum dc_status dcn401_enable_stream_timing( odm_slice_width, last_odm_slice_width); } + /* set DTBCLK_P */ + if (dc->res_pool->dccg->funcs->set_dtbclk_p_src) { + if (dc_is_dp_signal(stream->signal) || dc_is_virtual_signal(stream->signal)) { + dc->res_pool->dccg->funcs->set_dtbclk_p_src(dc->res_pool->dccg, DPREFCLK, pipe_ctx->stream_res.tg->inst); + } + } + /* HW program guide assume display already disable * by unplug sequence. OTG assume stop. */ @@ -1004,8 +1017,6 @@ void dcn401_enable_stream(struct pipe_ctx *pipe_ctx) dccg->funcs->enable_symclk32_se(dccg, dp_hpo_inst, phyd32clk); } else { - /* need to set DTBCLK_P source to DPREFCLK for DP8B10B */ - dccg->funcs->set_dtbclk_p_src(dccg, DPREFCLK, tg->inst); dccg->funcs->enable_symclk_se(dccg, stream_enc->stream_enc_inst, link_enc->transmitter - TRANSMITTER_UNIPHY_A); } @@ -1063,7 +1074,6 @@ static bool dcn401_can_pipe_disable_cursor(struct pipe_ctx *pipe_ctx) r2 = test_pipe->plane_res.scl_data.recout; r2_r = r2.x + r2.width; r2_b = r2.y + r2.height; - split_pipe = test_pipe; /** * There is another half plane on same layer because of @@ -1097,6 +1107,58 @@ void adjust_hotspot_between_slices_for_2x_magnify(uint32_t cursor_width, struct } } +static void disable_link_output_symclk_on_tx_off(struct dc_link *link, enum dp_link_encoding link_encoding) +{ + struct dc *dc = link->ctx->dc; + struct pipe_ctx *pipe_ctx = NULL; + uint8_t i; + + for (i = 0; i < MAX_PIPES; i++) { + pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i]; + if (pipe_ctx->stream && pipe_ctx->stream->link == link && pipe_ctx->top_pipe == NULL) { + pipe_ctx->clock_source->funcs->program_pix_clk( + pipe_ctx->clock_source, + &pipe_ctx->stream_res.pix_clk_params, + link_encoding, + &pipe_ctx->pll_settings); + break; + } + } +} + +void dcn401_disable_link_output(struct dc_link *link, + const struct link_resource *link_res, + enum signal_type signal) +{ + struct dc *dc = link->ctx->dc; + const struct link_hwss *link_hwss = get_link_hwss(link, link_res); + struct dmcu *dmcu = dc->res_pool->dmcu; + + if (signal == SIGNAL_TYPE_EDP && + link->dc->hwss.edp_backlight_control && + !link->skip_implict_edp_power_control) + link->dc->hwss.edp_backlight_control(link, false); + else if (dmcu != NULL && dmcu->funcs->lock_phy) + dmcu->funcs->lock_phy(dmcu); + + if (dc_is_tmds_signal(signal) && link->phy_state.symclk_ref_cnts.otg > 0) { + disable_link_output_symclk_on_tx_off(link, DP_UNKNOWN_ENCODING); + link->phy_state.symclk_state = SYMCLK_ON_TX_OFF; + } else { + link_hwss->disable_link_output(link, link_res, signal); + link->phy_state.symclk_state = SYMCLK_OFF_TX_OFF; + } + + if (signal == SIGNAL_TYPE_EDP && + link->dc->hwss.edp_backlight_control && + !link->skip_implict_edp_power_control) + link->dc->hwss.edp_power_control(link, false); + else if (dmcu != NULL && dmcu->funcs->lock_phy) + dmcu->funcs->unlock_phy(dmcu); + + dc->link_srv->dp_trace_source_sequence(link, DPCD_SOURCE_SEQ_AFTER_DISABLE_LINK_PHY); +} + void dcn401_set_cursor_position(struct pipe_ctx *pipe_ctx) { struct dc_cursor_position pos_cpy = pipe_ctx->stream->cursor_position; @@ -1492,6 +1554,11 @@ void dcn401_optimize_bandwidth( pipe_ctx->dlg_regs.min_dst_y_next_start); } } + + /* update timeout thresholds */ + if (hubbub->funcs->program_timeout_thresholds) { + hubbub->funcs->program_timeout_thresholds(hubbub, &context->bw_ctx.bw.dcn.arb_regs); + } } void dcn401_fams2_global_control_lock(struct dc *dc, @@ -1669,7 +1736,7 @@ void dcn401_hardware_release(struct dc *dc) } } -void dcn401_wait_for_det_buffer_update(struct dc *dc, struct dc_state *context, struct pipe_ctx *otg_master) +void dcn401_wait_for_det_buffer_update_under_otg_master(struct dc *dc, struct dc_state *context, struct pipe_ctx *otg_master) { struct pipe_ctx *opp_heads[MAX_PIPES]; struct pipe_ctx *dpp_pipes[MAX_PIPES]; @@ -1695,6 +1762,9 @@ void dcn401_wait_for_det_buffer_update(struct dc *dc, struct dc_state *context, hubbub->funcs->wait_for_det_update) hubbub->funcs->wait_for_det_update(hubbub, dpp_pipe->plane_res.hubp->inst); } + } else { + if (hubbub && opp_heads[slice_idx]->plane_res.hubp && hubbub->funcs->wait_for_det_update) + hubbub->funcs->wait_for_det_update(hubbub, opp_heads[slice_idx]->plane_res.hubp->inst); } } } @@ -1705,7 +1775,6 @@ void dcn401_interdependent_update_lock(struct dc *dc, unsigned int i = 0; struct pipe_ctx *pipe = NULL; struct timing_generator *tg = NULL; - bool pipe_unlocked[MAX_PIPES] = {0}; if (lock) { for (i = 0; i < dc->res_pool->pipe_count; i++) { @@ -1719,48 +1788,91 @@ void dcn401_interdependent_update_lock(struct dc *dc, dc->hwss.pipe_control_lock(dc, pipe, true); } } else { - /* Unlock pipes based on the change in DET allocation instead of pipe index - * Prevents over allocation of DET during unlock process - * e.g. 2 pipe config with different streams with a max of 20 DET segments - * Before: After: - * - Pipe0: 10 DET segments - Pipe0: 12 DET segments - * - Pipe1: 10 DET segments - Pipe1: 8 DET segments - * If Pipe0 gets updated first, 22 DET segments will be allocated - */ + /* Need to free DET being used first and have pipe update, then unlock the remaining pipes*/ for (i = 0; i < dc->res_pool->pipe_count; i++) { pipe = &context->res_ctx.pipe_ctx[i]; tg = pipe->stream_res.tg; - int current_pipe_idx = i; if (!resource_is_pipe_type(pipe, OTG_MASTER) || !tg->funcs->is_tg_enabled(tg) || dc_state_get_pipe_subvp_type(context, pipe) == SUBVP_PHANTOM) { - pipe_unlocked[i] = true; continue; } - // If the same stream exists in old context, ensure the OTG_MASTER pipes for the same stream get compared - struct pipe_ctx *old_otg_master = resource_get_otg_master_for_stream(&dc->current_state->res_ctx, pipe->stream); - - if (old_otg_master) - current_pipe_idx = old_otg_master->pipe_idx; - if (resource_calculate_det_for_stream(context, pipe) < - resource_calculate_det_for_stream(dc->current_state, &dc->current_state->res_ctx.pipe_ctx[current_pipe_idx])) { + if (dc->scratch.pipes_to_unlock_first[i]) { + struct pipe_ctx *old_pipe = &dc->current_state->res_ctx.pipe_ctx[i]; dc->hwss.pipe_control_lock(dc, pipe, false); - pipe_unlocked[i] = true; - dcn401_wait_for_det_buffer_update(dc, context, pipe); + /* Assumes pipe of the same index in current_state is also an OTG_MASTER pipe*/ + dcn401_wait_for_det_buffer_update_under_otg_master(dc, dc->current_state, old_pipe); } } + /* Unlocking the rest of the pipes */ for (i = 0; i < dc->res_pool->pipe_count; i++) { - if (pipe_unlocked[i]) + if (dc->scratch.pipes_to_unlock_first[i]) continue; + pipe = &context->res_ctx.pipe_ctx[i]; + tg = pipe->stream_res.tg; + if (!resource_is_pipe_type(pipe, OTG_MASTER) || + !tg->funcs->is_tg_enabled(tg) || + dc_state_get_pipe_subvp_type(context, pipe) == SUBVP_PHANTOM) { + continue; + } + dc->hwss.pipe_control_lock(dc, pipe, false); } } } +void dcn401_perform_3dlut_wa_unlock(struct pipe_ctx *pipe_ctx) +{ + /* If 3DLUT FL is enabled and 3DLUT is in use, follow the workaround sequence for pipe unlock to make sure that + * HUBP will properly fetch 3DLUT contents after unlock. + * + * This is meant to work around a known HW issue where VREADY will cancel the pending 3DLUT_ENABLE signal regardless + * of whether OTG lock is currently being held or not. + */ + struct pipe_ctx *wa_pipes[MAX_PIPES] = { NULL }; + struct pipe_ctx *odm_pipe, *mpc_pipe; + int i, wa_pipe_ct = 0; + + for (odm_pipe = pipe_ctx; odm_pipe != NULL; odm_pipe = odm_pipe->next_odm_pipe) { + for (mpc_pipe = odm_pipe; mpc_pipe != NULL; mpc_pipe = mpc_pipe->bottom_pipe) { + if (mpc_pipe->plane_state && mpc_pipe->plane_state->mcm_luts.lut3d_data.lut3d_src + == DC_CM2_TRANSFER_FUNC_SOURCE_VIDMEM + && mpc_pipe->plane_state->mcm_shaper_3dlut_setting + == DC_CM2_SHAPER_3DLUT_SETTING_ENABLE_SHAPER_3DLUT) { + wa_pipes[wa_pipe_ct++] = mpc_pipe; + } + } + } + + if (wa_pipe_ct > 0) { + if (pipe_ctx->stream_res.tg->funcs->set_vupdate_keepout) + pipe_ctx->stream_res.tg->funcs->set_vupdate_keepout(pipe_ctx->stream_res.tg, true); + + for (i = 0; i < wa_pipe_ct; ++i) { + if (wa_pipes[i]->plane_res.hubp->funcs->hubp_enable_3dlut_fl) + wa_pipes[i]->plane_res.hubp->funcs->hubp_enable_3dlut_fl(wa_pipes[i]->plane_res.hubp, true); + } + + pipe_ctx->stream_res.tg->funcs->unlock(pipe_ctx->stream_res.tg); + if (pipe_ctx->stream_res.tg->funcs->wait_update_lock_status) + pipe_ctx->stream_res.tg->funcs->wait_update_lock_status(pipe_ctx->stream_res.tg, false); + + for (i = 0; i < wa_pipe_ct; ++i) { + if (wa_pipes[i]->plane_res.hubp->funcs->hubp_enable_3dlut_fl) + wa_pipes[i]->plane_res.hubp->funcs->hubp_enable_3dlut_fl(wa_pipes[i]->plane_res.hubp, true); + } + + if (pipe_ctx->stream_res.tg->funcs->set_vupdate_keepout) + pipe_ctx->stream_res.tg->funcs->set_vupdate_keepout(pipe_ctx->stream_res.tg, false); + } else { + pipe_ctx->stream_res.tg->funcs->unlock(pipe_ctx->stream_res.tg); + } +} + void dcn401_program_outstanding_updates(struct dc *dc, struct dc_state *context) { @@ -1770,3 +1882,125 @@ void dcn401_program_outstanding_updates(struct dc *dc, if (hubbub->funcs->program_compbuf_segments) hubbub->funcs->program_compbuf_segments(hubbub, context->bw_ctx.bw.dcn.arb_regs.compbuf_size, true); } + +void dcn401_reset_back_end_for_pipe( + struct dc *dc, + struct pipe_ctx *pipe_ctx, + struct dc_state *context) +{ + struct dc_link *link = pipe_ctx->stream->link; + const struct link_hwss *link_hwss = get_link_hwss(link, &pipe_ctx->link_res); + + DC_LOGGER_INIT(dc->ctx->logger); + if (pipe_ctx->stream_res.stream_enc == NULL) { + pipe_ctx->stream = NULL; + return; + } + + /* DPMS may already disable or */ + /* dpms_off status is incorrect due to fastboot + * feature. When system resume from S4 with second + * screen only, the dpms_off would be true but + * VBIOS lit up eDP, so check link status too. + */ + if (!pipe_ctx->stream->dpms_off || link->link_status.link_active) + dc->link_srv->set_dpms_off(pipe_ctx); + else if (pipe_ctx->stream_res.audio) + dc->hwss.disable_audio_stream(pipe_ctx); + + /* free acquired resources */ + if (pipe_ctx->stream_res.audio) { + /*disable az_endpoint*/ + pipe_ctx->stream_res.audio->funcs->az_disable(pipe_ctx->stream_res.audio); + + /*free audio*/ + if (dc->caps.dynamic_audio == true) { + /*we have to dynamic arbitrate the audio endpoints*/ + /*we free the resource, need reset is_audio_acquired*/ + update_audio_usage(&dc->current_state->res_ctx, dc->res_pool, + pipe_ctx->stream_res.audio, false); + pipe_ctx->stream_res.audio = NULL; + } + } + + /* by upper caller loop, parent pipe: pipe0, will be reset last. + * back end share by all pipes and will be disable only when disable + * parent pipe. + */ + if (pipe_ctx->top_pipe == NULL) { + + dc->hwss.set_abm_immediate_disable(pipe_ctx); + + pipe_ctx->stream_res.tg->funcs->disable_crtc(pipe_ctx->stream_res.tg); + + pipe_ctx->stream_res.tg->funcs->enable_optc_clock(pipe_ctx->stream_res.tg, false); + if (pipe_ctx->stream_res.tg->funcs->set_odm_bypass) + pipe_ctx->stream_res.tg->funcs->set_odm_bypass( + pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing); + + if (pipe_ctx->stream_res.tg->funcs->set_drr) + pipe_ctx->stream_res.tg->funcs->set_drr( + pipe_ctx->stream_res.tg, NULL); + /* TODO - convert symclk_ref_cnts for otg to a bit map to solve + * the case where the same symclk is shared across multiple otg + * instances + */ + if (dc_is_hdmi_tmds_signal(pipe_ctx->stream->signal)) + link->phy_state.symclk_ref_cnts.otg = 0; + if (link->phy_state.symclk_state == SYMCLK_ON_TX_OFF) { + link_hwss->disable_link_output(link, + &pipe_ctx->link_res, pipe_ctx->stream->signal); + link->phy_state.symclk_state = SYMCLK_OFF_TX_OFF; + } + + /* reset DTBCLK_P */ + if (dc->res_pool->dccg->funcs->set_dtbclk_p_src) + dc->res_pool->dccg->funcs->set_dtbclk_p_src(dc->res_pool->dccg, REFCLK, pipe_ctx->stream_res.tg->inst); + } + +/* + * In case of a dangling plane, setting this to NULL unconditionally + * causes failures during reset hw ctx where, if stream is NULL, + * it is expected that the pipe_ctx pointers to pipes and plane are NULL. + */ + pipe_ctx->stream = NULL; + pipe_ctx->top_pipe = NULL; + pipe_ctx->bottom_pipe = NULL; + pipe_ctx->next_odm_pipe = NULL; + pipe_ctx->prev_odm_pipe = NULL; + DC_LOG_DEBUG("Reset back end for pipe %d, tg:%d\n", + pipe_ctx->pipe_idx, pipe_ctx->stream_res.tg->inst); +} + +void dcn401_reset_hw_ctx_wrap( + struct dc *dc, + struct dc_state *context) +{ + int i; + struct dce_hwseq *hws = dc->hwseq; + + /* Reset Back End*/ + for (i = dc->res_pool->pipe_count - 1; i >= 0 ; i--) { + struct pipe_ctx *pipe_ctx_old = + &dc->current_state->res_ctx.pipe_ctx[i]; + struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; + + if (!pipe_ctx_old->stream) + continue; + + if (pipe_ctx_old->top_pipe || pipe_ctx_old->prev_odm_pipe) + continue; + + if (!pipe_ctx->stream || + pipe_need_reprogram(pipe_ctx_old, pipe_ctx)) { + struct clock_source *old_clk = pipe_ctx_old->clock_source; + + if (hws->funcs.reset_back_end_for_pipe) + hws->funcs.reset_back_end_for_pipe(dc, pipe_ctx_old, dc->current_state); + if (hws->funcs.enable_stream_gating) + hws->funcs.enable_stream_gating(dc, pipe_ctx_old); + if (old_clk) + old_clk->funcs->cs_power_down(old_clk); + } + } +} diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.h index a27e62081685..28a513dfc005 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.h @@ -55,6 +55,10 @@ void dcn401_populate_mcm_luts(struct dc *dc, bool lut_bank_a); void dcn401_setup_hpo_hw_control(const struct dce_hwseq *hws, bool enable); +void dcn401_disable_link_output(struct dc_link *link, + const struct link_resource *link_res, + enum signal_type signal); + void dcn401_set_cursor_position(struct pipe_ctx *pipe_ctx); bool dcn401_apply_idle_power_optimizations(struct dc *dc, bool enable); @@ -81,7 +85,16 @@ void dcn401_hardware_release(struct dc *dc); void dcn401_update_odm(struct dc *dc, struct dc_state *context, struct pipe_ctx *otg_master); void adjust_hotspot_between_slices_for_2x_magnify(uint32_t cursor_width, struct dc_cursor_position *pos_cpy); -void dcn401_wait_for_det_buffer_update(struct dc *dc, struct dc_state *context, struct pipe_ctx *otg_master); +void dcn401_wait_for_det_buffer_update_under_otg_master(struct dc *dc, struct dc_state *context, struct pipe_ctx *otg_master); void dcn401_interdependent_update_lock(struct dc *dc, struct dc_state *context, bool lock); void dcn401_program_outstanding_updates(struct dc *dc, struct dc_state *context); +void dcn401_reset_back_end_for_pipe( + struct dc *dc, + struct pipe_ctx *pipe_ctx, + struct dc_state *context); +void dcn401_reset_hw_ctx_wrap( + struct dc *dc, + struct dc_state *context); +void dcn401_perform_3dlut_wa_unlock(struct pipe_ctx *pipe_ctx); + #endif /* __DC_HWSS_DCN401_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_init.c index a2ca07235c83..23e4f208152e 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_init.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_init.c @@ -77,14 +77,14 @@ static const struct hw_sequencer_funcs dcn401_funcs = { .calc_vupdate_position = dcn10_calc_vupdate_position, .apply_idle_power_optimizations = dcn401_apply_idle_power_optimizations, .does_plane_fit_in_mall = NULL, - .set_backlight_level = dcn21_set_backlight_level, + .set_backlight_level = dcn31_set_backlight_level, .set_abm_immediate_disable = dcn21_set_abm_immediate_disable, .hardware_release = dcn401_hardware_release, .set_pipe = dcn21_set_pipe, .enable_lvds_link_output = dce110_enable_lvds_link_output, .enable_tmds_link_output = dce110_enable_tmds_link_output, .enable_dp_link_output = dce110_enable_dp_link_output, - .disable_link_output = dcn32_disable_link_output, + .disable_link_output = dcn401_disable_link_output, .set_disp_pattern_generator = dcn30_set_disp_pattern_generator, .get_dcc_en_bits = dcn10_get_dcc_en_bits, .enable_phantom_streams = dcn32_enable_phantom_streams, @@ -93,13 +93,13 @@ static const struct hw_sequencer_funcs dcn401_funcs = { .update_phantom_vp_position = dcn32_update_phantom_vp_position, .update_dsc_pg = dcn32_update_dsc_pg, .apply_update_flags_for_phantom = dcn32_apply_update_flags_for_phantom, - .blank_phantom = dcn32_blank_phantom, .wait_for_dcc_meta_propagation = dcn401_wait_for_dcc_meta_propagation, .is_pipe_topology_transition_seamless = dcn32_is_pipe_topology_transition_seamless, .fams2_global_control_lock = dcn401_fams2_global_control_lock, .fams2_update_config = dcn401_fams2_update_config, .fams2_global_control_lock_fast = dcn401_fams2_global_control_lock_fast, .program_outstanding_updates = dcn401_program_outstanding_updates, + .wait_for_all_pending_updates = dcn30_wait_for_all_pending_updates, }; static const struct hwseq_private_funcs dcn401_private_funcs = { @@ -111,7 +111,7 @@ static const struct hwseq_private_funcs dcn401_private_funcs = { .power_down = dce110_power_down, .enable_display_power_gating = dcn10_dummy_display_power_gating, .blank_pixel_data = dcn20_blank_pixel_data, - .reset_hw_ctx_wrap = dcn20_reset_hw_ctx_wrap, + .reset_hw_ctx_wrap = dcn401_reset_hw_ctx_wrap, .enable_stream_timing = dcn401_enable_stream_timing, .edp_backlight_control = dce110_edp_backlight_control, .setup_vupdate_interrupt = dcn20_setup_vupdate_interrupt, @@ -136,8 +136,9 @@ static const struct hwseq_private_funcs dcn401_private_funcs = { .update_mall_sel = dcn32_update_mall_sel, .calculate_dccg_k1_k2_values = NULL, .apply_single_controller_ctx_to_hw = dce110_apply_single_controller_ctx_to_hw, - .reset_back_end_for_pipe = dcn20_reset_back_end_for_pipe, + .reset_back_end_for_pipe = dcn401_reset_back_end_for_pipe, .populate_mcm_luts = NULL, + .perform_3dlut_wa_unlock = dcn401_perform_3dlut_wa_unlock, }; void dcn401_hw_sequencer_init_functions(struct dc *dc) diff --git a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h index ac9205625623..66fdc5805d0a 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h @@ -365,8 +365,7 @@ struct hw_sequencer_funcs { void (*clear_status_bits)(struct dc *dc, unsigned int mask); bool (*set_backlight_level)(struct pipe_ctx *pipe_ctx, - uint32_t backlight_pwm_u16_16, - uint32_t frame_ramp); + struct set_backlight_level_params *params); void (*set_abm_immediate_disable)(struct pipe_ctx *pipe_ctx); @@ -462,6 +461,7 @@ struct hw_sequencer_funcs { void (*program_outstanding_updates)(struct dc *dc, struct dc_state *context); void (*setup_hpo_hw_control)(const struct dce_hwseq *hws, bool enable); + void (*wait_for_all_pending_updates)(const struct pipe_ctx *pipe_ctx); }; void color_space_to_black_color( @@ -504,6 +504,10 @@ void get_mclk_switch_visual_confirm_color( struct pipe_ctx *pipe_ctx, struct tg_color *color); +void get_cursor_visual_confirm_color( + struct pipe_ctx *pipe_ctx, + struct tg_color *color); + void set_p_state_switch_method( struct dc *dc, struct dc_state *context, diff --git a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer_private.h b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer_private.h index 0ac675456979..22a5d4a03c98 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer_private.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer_private.h @@ -182,6 +182,7 @@ struct hwseq_private_funcs { struct pipe_ctx *pipe_ctx, struct dc_cm2_func_luts mcm_luts, bool lut_bank_a); + void (*perform_3dlut_wa_unlock)(struct pipe_ctx *pipe_ctx); }; struct dce_hwseq { diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_status.h b/drivers/gpu/drm/amd/display/dc/inc/core_status.h index fa5edd03d004..b5afd8c3103d 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/core_status.h +++ b/drivers/gpu/drm/amd/display/dc/inc/core_status.h @@ -60,5 +60,7 @@ enum dc_status { }; char *dc_status_to_str(enum dc_status status); +char *dc_pixel_encoding_to_str(enum dc_pixel_encoding pixel_encoding); +char *dc_color_depth_to_str(enum dc_color_depth color_depth); #endif /* _CORE_STATUS_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_types.h b/drivers/gpu/drm/amd/display/dc/inc/core_types.h index bfb8b8502d20..8597e866bfe6 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/core_types.h +++ b/drivers/gpu/drm/amd/display/dc/inc/core_types.h @@ -215,6 +215,10 @@ struct resource_funcs { void (*get_panel_config_defaults)(struct dc_panel_config *panel_config); void (*build_pipe_pix_clk_params)(struct pipe_ctx *pipe_ctx); + /* + * Get indicator of power from a context that went through full validation + */ + int (*get_power_profile)(const struct dc_state *context); }; struct audio_support{ diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h index 67c32401893e..6c1d41c0f099 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h @@ -228,6 +228,7 @@ struct hubbub_funcs { void (*program_det_segments)(struct hubbub *hubbub, int hubp_inst, unsigned det_buffer_size_seg); void (*program_compbuf_segments)(struct hubbub *hubbub, unsigned compbuf_size_seg, bool safe_to_increase); void (*wait_for_det_update)(struct hubbub *hubbub, int hubp_inst); + void (*program_timeout_thresholds)(struct hubbub *hubbub, struct dml2_display_arb_regs *arb_regs); }; struct hubbub { diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h index 3d4c8bd42b49..b74e18cc1e66 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h @@ -342,7 +342,11 @@ struct timing_generator_funcs { void (*wait_drr_doublebuffer_pending_clear)(struct timing_generator *tg); void (*set_long_vtotal)(struct timing_generator *optc, const struct long_vtotal_params *params); void (*wait_odm_doublebuffer_pending_clear)(struct timing_generator *tg); - bool (*get_double_buffer_pending)(struct timing_generator *tg); + bool (*get_optc_double_buffer_pending)(struct timing_generator *tg); + bool (*get_otg_double_buffer_pending)(struct timing_generator *tg); + bool (*get_pipe_update_pending)(struct timing_generator *tg); + void (*set_vupdate_keepout)(struct timing_generator *tg, bool enable); + bool (*wait_update_lock_status)(struct timing_generator *tg, bool locked); }; #endif diff --git a/drivers/gpu/drm/amd/display/dc/inc/link.h b/drivers/gpu/drm/amd/display/dc/inc/link.h index 72a8479e1f2d..f04292086c08 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/link.h +++ b/drivers/gpu/drm/amd/display/dc/inc/link.h @@ -248,8 +248,7 @@ struct link_service { uint32_t *backlight_millinits_avg, uint32_t *backlight_millinits_peak); bool (*edp_set_backlight_level)(const struct dc_link *link, - uint32_t backlight_pwm_u16_16, - uint32_t frame_ramp); + struct set_backlight_level_params *backlight_level_params); bool (*edp_set_backlight_level_nits)(struct dc_link *link, bool isHDR, uint32_t backlight_millinits, diff --git a/drivers/gpu/drm/amd/display/dc/link/link_detection.c b/drivers/gpu/drm/amd/display/dc/link/link_detection.c index d21ee9d12d26..e026c728042a 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_detection.c +++ b/drivers/gpu/drm/amd/display/dc/link/link_detection.c @@ -48,6 +48,9 @@ #include "dm_helpers.h" #include "clk_mgr.h" + // Offset DPCD 050Eh == 0x5A +#define MST_HUB_ID_0x5A 0x5A + #define DC_LOGGER \ link->ctx->logger #define DC_LOGGER_INIT(logger) @@ -692,6 +695,15 @@ static void apply_dpia_mst_dsc_always_on_wa(struct dc_link *link) link->dpcd_caps.dsc_caps.dsc_basic_caps.fields.dsc_support.DSC_SUPPORT && !link->dc->debug.dpia_debug.bits.disable_mst_dsc_work_around) link->wa_flags.dpia_mst_dsc_always_on = true; + + if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA && + link->type == dc_connection_mst_branch && + link->dpcd_caps.branch_dev_id == DP_BRANCH_DEVICE_ID_90CC24 && + link->dpcd_caps.branch_vendor_specific_data[2] == MST_HUB_ID_0x5A && + link->dpcd_caps.dsc_caps.dsc_basic_caps.fields.dsc_support.DSC_SUPPORT && + !link->dc->debug.dpia_debug.bits.disable_mst_dsc_work_around) { + link->wa_flags.dpia_mst_dsc_always_on = true; + } } static void revert_dpia_mst_dsc_always_on_wa(struct dc_link *link) diff --git a/drivers/gpu/drm/amd/display/dc/link/link_dpms.c b/drivers/gpu/drm/amd/display/dc/link/link_dpms.c index c4e03482ba9a..41cab9ad6885 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_dpms.c +++ b/drivers/gpu/drm/amd/display/dc/link/link_dpms.c @@ -2082,6 +2082,9 @@ static enum dc_status enable_link_dp(struct dc_state *state, if (link_settings->link_rate == LINK_RATE_LOW) skip_video_pattern = false; + if (stream->sink_patches.oled_optimize_display_on) + set_default_brightness_aux(link); + if (perform_link_training_with_retries(link_settings, skip_video_pattern, lt_attempts, @@ -2105,10 +2108,14 @@ static enum dc_status enable_link_dp(struct dc_state *state, if (link->dpcd_sink_ext_caps.bits.oled == 1 || link->dpcd_sink_ext_caps.bits.sdr_aux_backlight_control == 1 || link->dpcd_sink_ext_caps.bits.hdr_aux_backlight_control == 1) { - set_default_brightness_aux(link); - if (link->dpcd_sink_ext_caps.bits.oled == 1) - msleep(bl_oled_enable_delay); - edp_backlight_enable_aux(link, true); + if (!stream->sink_patches.oled_optimize_display_on) { + set_default_brightness_aux(link); + if (link->dpcd_sink_ext_caps.bits.oled == 1) + msleep(bl_oled_enable_delay); + edp_backlight_enable_aux(link, true); + } else { + edp_backlight_enable_aux(link, true); + } } return status; diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c index d78c8ec4de79..9dabaf682171 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c @@ -51,9 +51,10 @@ #include "dc_dmub_srv.h" #include "gpio_service_interface.h" +#define DC_TRACE_LEVEL_MESSAGE(...) /* do nothing */ + #define DC_LOGGER \ link->ctx->logger -#define DC_TRACE_LEVEL_MESSAGE(...) /* do nothing */ #ifndef MAX #define MAX(X, Y) ((X) > (Y) ? (X) : (Y)) @@ -1207,6 +1208,13 @@ static void get_active_converter_info( dp_hw_fw_revision.ieee_fw_rev, sizeof(dp_hw_fw_revision.ieee_fw_rev)); } + + core_link_read_dpcd( + link, + DP_BRANCH_VENDOR_SPECIFIC_START, + (uint8_t *)link->dpcd_caps.branch_vendor_specific_data, + sizeof(link->dpcd_caps.branch_vendor_specific_data)); + if (link->dpcd_caps.dpcd_rev.raw >= DPCD_REV_14 && link->dpcd_caps.dongle_type != DISPLAY_DONGLE_NONE) { union dp_dfp_cap_ext dfp_cap_ext; @@ -1625,7 +1633,11 @@ static bool retrieve_link_cap(struct dc_link *link) } /* Read DP tunneling information. */ - status = dpcd_get_tunneling_device_data(link); + if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA) { + status = dpcd_get_tunneling_device_data(link); + if (status != DC_OK) + dm_error("%s: Read tunneling device data failed.\n", __func__); + } dpcd_set_source_specific_data(link); /* Sink may need to configure internals based on vendor, so allow some @@ -1842,6 +1854,9 @@ static bool retrieve_link_cap(struct dc_link *link) DP_FEC_CAPABILITY, &link->dpcd_caps.fec_cap.raw, sizeof(link->dpcd_caps.fec_cap.raw)); + if (status != DC_OK) + DC_LOG_ERROR("%s:%d: core_link_read_dpcd (DP_FEC_CAPABILITY) failed\n", __func__, __LINE__); + status = core_link_read_dpcd( link, DP_DSC_SUPPORT, @@ -1864,6 +1879,9 @@ static bool retrieve_link_cap(struct dc_link *link) DP_DSC_BRANCH_OVERALL_THROUGHPUT_0, link->dpcd_caps.dsc_caps.dsc_branch_decoder_caps.raw, sizeof(link->dpcd_caps.dsc_caps.dsc_branch_decoder_caps.raw)); + if (status != DC_OK) + DC_LOG_ERROR("%s:%d: core_link_read_dpcd (DP_DSC_BRANCH_OVERALL_THROUGHPUT_0) failed\n", __func__, __LINE__); + DC_LOG_DSC("DSC branch decoder capability is read at link %d", link->link_index); DC_LOG_DSC("\tBRANCH_OVERALL_THROUGHPUT_0 = 0x%02x", link->dpcd_caps.dsc_caps.dsc_branch_decoder_caps.fields.BRANCH_OVERALL_THROUGHPUT_0); @@ -2055,6 +2073,14 @@ void detect_edp_sink_caps(struct dc_link *link) core_link_read_dpcd(link, DP_SINK_PR_MAX_NUMBER_OF_DEVIATION_LINE, &link->dpcd_caps.pr_info.max_deviation_line, sizeof(link->dpcd_caps.pr_info.max_deviation_line)); + + /* + * OLED Emission Rate info + */ + if (link->dpcd_sink_ext_caps.bits.emission_output) + core_link_read_dpcd(link, DP_SINK_EMISSION_RATE, + (uint8_t *)&link->dpcd_caps.edp_oled_emission_rate, + sizeof(link->dpcd_caps.edp_oled_emission_rate)); } bool dp_get_max_link_enc_cap(const struct dc_link *link, struct dc_link_settings *max_link_enc_cap) diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia.c index 6af42ba9885c..0d123e647652 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia.c @@ -59,12 +59,18 @@ enum dc_status dpcd_get_tunneling_device_data(struct dc_link *link) dpcd_dp_tun_data, sizeof(dpcd_dp_tun_data)); + if (status != DC_OK) + goto err; + status = core_link_read_dpcd( link, DP_USB4_ROUTER_TOPOLOGY_ID, dpcd_topology_data, sizeof(dpcd_topology_data)); + if (status != DC_OK) + goto err; + link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.raw = dpcd_dp_tun_data[DP_TUNNELING_CAPABILITIES_SUPPORT - DP_TUNNELING_CAPABILITIES_SUPPORT]; link->dpcd_caps.usb4_dp_tun_info.dpia_info.raw = @@ -75,6 +81,7 @@ enum dc_status dpcd_get_tunneling_device_data(struct dc_link *link) for (i = 0; i < DPCD_USB4_TOPOLOGY_ID_LEN; i++) link->dpcd_caps.usb4_dp_tun_info.usb4_topology_id[i] = dpcd_topology_data[i]; +err: return status; } diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_irq_handler.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_irq_handler.c index 96bf135b6f05..48abeaa88678 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_irq_handler.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_irq_handler.c @@ -221,21 +221,11 @@ static void handle_hpd_irq_replay_sink(struct dc_link *link) &replay_error_status.raw, sizeof(replay_error_status.raw)); - link->replay_settings.config.replay_error_status.bits.LINK_CRC_ERROR = - replay_error_status.bits.LINK_CRC_ERROR; - link->replay_settings.config.replay_error_status.bits.DESYNC_ERROR = - replay_configuration.bits.DESYNC_ERROR_STATUS; - link->replay_settings.config.replay_error_status.bits.STATE_TRANSITION_ERROR = - replay_configuration.bits.STATE_TRANSITION_ERROR_STATUS; - - if (link->replay_settings.config.replay_error_status.bits.LINK_CRC_ERROR || - link->replay_settings.config.replay_error_status.bits.DESYNC_ERROR || - link->replay_settings.config.replay_error_status.bits.STATE_TRANSITION_ERROR) { + if (replay_error_status.bits.LINK_CRC_ERROR || + replay_configuration.bits.DESYNC_ERROR_STATUS || + replay_configuration.bits.STATE_TRANSITION_ERROR_STATUS) { bool allow_active; - if (link->replay_settings.config.replay_error_status.bits.DESYNC_ERROR) - link->replay_settings.config.received_desync_error_hpd = 1; - if (link->replay_settings.config.force_disable_desync_error_check) return; diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c index 27b881f947e8..754c895e1bfb 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c @@ -272,7 +272,7 @@ void dp_wait_for_training_aux_rd_interval( struct dc_link *link, uint32_t wait_in_micro_secs) { - fsleep(wait_in_micro_secs); + usleep_range_state(wait_in_micro_secs, wait_in_micro_secs, TASK_UNINTERRUPTIBLE); DC_LOG_HW_LINK_TRAINING("%s:\n wait = %d\n", __func__, @@ -1107,9 +1107,13 @@ enum dc_status dpcd_set_link_settings( status = core_link_write_dpcd(link, DP_DOWNSPREAD_CTRL, &downspread.raw, sizeof(downspread)); + if (status != DC_OK) + DC_LOG_ERROR("%s:%d: core_link_write_dpcd (DP_DOWNSPREAD_CTRL) failed\n", __func__, __LINE__); status = core_link_write_dpcd(link, DP_LANE_COUNT_SET, &lane_count_set.raw, 1); + if (status != DC_OK) + DC_LOG_ERROR("%s:%d: core_link_write_dpcd (DP_LANE_COUNT_SET) failed\n", __func__, __LINE__); if (link->dpcd_caps.dpcd_rev.raw >= DPCD_REV_13 && lt_settings->link_settings.use_link_rate_set == true) { @@ -1125,12 +1129,19 @@ enum dc_status dpcd_set_link_settings( supported_link_rates, sizeof(supported_link_rates)); } status = core_link_write_dpcd(link, DP_LINK_BW_SET, &rate, 1); + if (status != DC_OK) + DC_LOG_ERROR("%s:%d: core_link_write_dpcd (DP_LINK_BW_SET) failed\n", __func__, __LINE__); + status = core_link_write_dpcd(link, DP_LINK_RATE_SET, <_settings->link_settings.link_rate_set, 1); + if (status != DC_OK) + DC_LOG_ERROR("%s:%d: core_link_write_dpcd (DP_LINK_RATE_SET) failed\n", __func__, __LINE__); } else { rate = get_dpcd_link_rate(<_settings->link_settings); status = core_link_write_dpcd(link, DP_LINK_BW_SET, &rate, 1); + if (status != DC_OK) + DC_LOG_ERROR("%s:%d: core_link_write_dpcd (DP_LINK_BW_SET) failed\n", __func__, __LINE__); } if (rate) { diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_fixed_vs_pe_retimer.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_fixed_vs_pe_retimer.c index b5cf75975fff..ccf8096dde29 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_fixed_vs_pe_retimer.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_fixed_vs_pe_retimer.c @@ -412,7 +412,6 @@ enum link_training_result dp_perform_fixed_vs_pe_training_sequence( /* 5. check CR done*/ if (dp_is_cr_done(lane_count, dpcd_lane_status)) { - status = LINK_TRAINING_SUCCESS; break; } diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c index 3aa05a2be6c0..e0e3bb865359 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c @@ -157,31 +157,13 @@ bool edp_set_backlight_level_nits(struct dc_link *link, uint32_t backlight_millinits, uint32_t transition_time_in_ms) { - struct dpcd_source_backlight_set dpcd_backlight_set; - uint8_t backlight_control = isHDR ? 1 : 0; - if (!link || (link->connector_signal != SIGNAL_TYPE_EDP && link->connector_signal != SIGNAL_TYPE_DISPLAY_PORT)) return false; - // OLEDs have no PWM, they can only use AUX - if (link->dpcd_sink_ext_caps.bits.oled == 1) - backlight_control = 1; - - *(uint32_t *)&dpcd_backlight_set.backlight_level_millinits = backlight_millinits; - *(uint16_t *)&dpcd_backlight_set.backlight_transition_time_ms = (uint16_t)transition_time_in_ms; - - - if (!link->dpcd_caps.panel_luminance_control) { - if (core_link_write_dpcd(link, DP_SOURCE_BACKLIGHT_LEVEL, - (uint8_t *)(&dpcd_backlight_set), - sizeof(dpcd_backlight_set)) != DC_OK) - return false; - - if (core_link_write_dpcd(link, DP_SOURCE_BACKLIGHT_CONTROL, - &backlight_control, 1) != DC_OK) - return false; - } else { + // use internal backlight control if dmub capabilities are not present + if (link->backlight_control_type == BACKLIGHT_CONTROL_VESA_AUX && + !link->dc->caps.dmub_caps.aux_backlight_support) { uint8_t backlight_enable = 0; struct target_luminance_value *target_luminance = NULL; @@ -205,6 +187,24 @@ bool edp_set_backlight_level_nits(struct dc_link *link, (uint8_t *)(target_luminance), sizeof(struct target_luminance_value)) != DC_OK) return false; + } else if (link->backlight_control_type == BACKLIGHT_CONTROL_AMD_AUX) { + struct dpcd_source_backlight_set dpcd_backlight_set; + *(uint32_t *)&dpcd_backlight_set.backlight_level_millinits = backlight_millinits; + *(uint16_t *)&dpcd_backlight_set.backlight_transition_time_ms = (uint16_t)transition_time_in_ms; + + uint8_t backlight_control = isHDR ? 1 : 0; + // OLEDs have no PWM, they can only use AUX + if (link->dpcd_sink_ext_caps.bits.oled == 1) + backlight_control = 1; + + if (core_link_write_dpcd(link, DP_SOURCE_BACKLIGHT_LEVEL, + (uint8_t *)(&dpcd_backlight_set), + sizeof(dpcd_backlight_set)) != DC_OK) + return false; + + if (core_link_write_dpcd(link, DP_SOURCE_BACKLIGHT_CONTROL, + &backlight_control, 1) != DC_OK) + return false; } return true; @@ -519,11 +519,11 @@ static struct pipe_ctx *get_pipe_from_link(const struct dc_link *link) } bool edp_set_backlight_level(const struct dc_link *link, - uint32_t backlight_pwm_u16_16, - uint32_t frame_ramp) + struct set_backlight_level_params *backlight_level_params) { struct dc *dc = link->ctx->dc; - + uint32_t backlight_pwm_u16_16 = backlight_level_params->backlight_pwm_u16_16; + uint32_t frame_ramp = backlight_level_params->frame_ramp; DC_LOGGER_INIT(link->ctx->logger); DC_LOG_BACKLIGHT("New Backlight level: %d (0x%X)\n", backlight_pwm_u16_16, backlight_pwm_u16_16); @@ -544,10 +544,11 @@ bool edp_set_backlight_level(const struct dc_link *link, return false; } + backlight_level_params->frame_ramp = frame_ramp; + dc->hwss.set_backlight_level( pipe_ctx, - backlight_pwm_u16_16, - frame_ramp); + backlight_level_params); } return true; } @@ -940,8 +941,7 @@ bool edp_setup_replay(struct dc_link *link, const struct dc_stream_state *stream struct replay_context replay_context = { 0 }; unsigned int lineTimeInNs = 0; - - union replay_enable_and_configuration replay_config; + union replay_enable_and_configuration replay_config = { 0 }; union dpcd_alpm_configuration alpm_config; @@ -1168,9 +1168,6 @@ static void edp_set_assr_enable(const struct dc *pDC, struct dc_link *link, link_enc_index = link->link_enc->transmitter - TRANSMITTER_UNIPHY_A; if (link_res->hpo_dp_link_enc) { - if (link->wa_flags.disable_assr_for_uhbr) - return; - link_enc_index = link_res->hpo_dp_link_enc->inst; use_hpo_dp_link_enc = true; } diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h index 30dc8c24c008..bcfa6ac5d4e7 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h @@ -36,8 +36,7 @@ bool edp_get_backlight_level_nits(struct dc_link *link, uint32_t *backlight_millinits_avg, uint32_t *backlight_millinits_peak); bool edp_set_backlight_level(const struct dc_link *link, - uint32_t backlight_pwm_u16_16, - uint32_t frame_ramp); + struct set_backlight_level_params *backlight_level_params); bool edp_set_backlight_level_nits(struct dc_link *link, bool isHDR, uint32_t backlight_millinits, diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.h index b7a57f98553d..40757f20d73f 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.h +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.h @@ -202,6 +202,7 @@ struct dcn_optc_registers { uint32_t OPTC_CLOCK_CONTROL; uint32_t OPTC_WIDTH_CONTROL2; uint32_t OTG_PSTATE_REGISTER; + uint32_t OTG_PIPE_UPDATE_STATUS; }; #define TG_COMMON_MASK_SH_LIST_DCN(mask_sh)\ @@ -566,6 +567,12 @@ struct dcn_optc_registers { type OTG_H_TIMING_DIV_MODE_DB_UPDATE_PENDING;\ type OPTC_DOUBLE_BUFFER_PENDING;\ +#define TG_REG_FIELD_LIST_DCN2_0(type) \ + type OTG_FLIP_PENDING;\ + type OTG_DC_REG_UPDATE_PENDING;\ + type OTG_CURSOR_UPDATE_PENDING;\ + type OTG_VUPDATE_KEEPOUT_STATUS;\ + #define TG_REG_FIELD_LIST_DCN3_2(type) \ type OTG_H_TIMING_DIV_MODE_MANUAL; @@ -600,6 +607,7 @@ struct dcn_optc_registers { struct dcn_optc_shift { TG_REG_FIELD_LIST(uint8_t) + TG_REG_FIELD_LIST_DCN2_0(uint8_t) TG_REG_FIELD_LIST_DCN3_2(uint8_t) TG_REG_FIELD_LIST_DCN3_5(uint8_t) TG_REG_FIELD_LIST_DCN401(uint8_t) @@ -607,6 +615,7 @@ struct dcn_optc_shift { struct dcn_optc_mask { TG_REG_FIELD_LIST(uint32_t) + TG_REG_FIELD_LIST_DCN2_0(uint32_t) TG_REG_FIELD_LIST_DCN3_2(uint32_t) TG_REG_FIELD_LIST_DCN3_5(uint32_t) TG_REG_FIELD_LIST_DCN401(uint32_t) diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn20/dcn20_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn20/dcn20_optc.h index 364034b19028..928e110b95fb 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn20/dcn20_optc.h +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn20/dcn20_optc.h @@ -43,7 +43,8 @@ SRI(OPTC_MEMORY_CONFIG, ODM, inst),\ SR(DWB_SOURCE_SELECT),\ SRI(OTG_MANUAL_FLOW_CONTROL, OTG, inst), \ - SRI(OTG_DRR_CONTROL, OTG, inst) + SRI(OTG_DRR_CONTROL, OTG, inst),\ + SRI(OTG_PIPE_UPDATE_STATUS, OTG, inst) #define TG_COMMON_MASK_SH_LIST_DCN2_0(mask_sh)\ TG_COMMON_MASK_SH_LIST_DCN(mask_sh),\ @@ -53,6 +54,10 @@ SF(OTG0_OTG_GLOBAL_CONTROL2, GLOBAL_UPDATE_LOCK_EN, mask_sh),\ SF(OTG0_OTG_GLOBAL_CONTROL2, DIG_UPDATE_LOCATION, mask_sh),\ SF(OTG0_OTG_DOUBLE_BUFFER_CONTROL, OTG_RANGE_TIMING_DBUF_UPDATE_MODE, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_FLIP_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_DC_REG_UPDATE_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_CURSOR_UPDATE_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh),\ SF(OTG0_OTG_GSL_WINDOW_X, OTG_GSL_WINDOW_START_X, mask_sh),\ SF(OTG0_OTG_GSL_WINDOW_X, OTG_GSL_WINDOW_END_X, mask_sh), \ SF(OTG0_OTG_GSL_WINDOW_Y, OTG_GSL_WINDOW_START_Y, mask_sh),\ diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn30/dcn30_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn30/dcn30_optc.c index abcd03d78668..4c95c0958612 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn30/dcn30_optc.c +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn30/dcn30_optc.c @@ -271,6 +271,48 @@ void optc3_set_odm_combine(struct timing_generator *optc, int *opp_id, int opp_c optc1->opp_count = opp_cnt; } +/* OTG status register that indicates OPTC update is pending */ +bool optc3_get_optc_double_buffer_pending(struct timing_generator *optc) +{ + struct optc *optc1 = DCN10TG_FROM_TG(optc); + uint32_t update_pending = 0; + + REG_GET(OPTC_INPUT_GLOBAL_CONTROL, + OPTC_DOUBLE_BUFFER_PENDING, + &update_pending); + + return (update_pending == 1); +} + +/* OTG status register that indicates OTG update is pending */ +bool optc3_get_otg_update_pending(struct timing_generator *optc) +{ + struct optc *optc1 = DCN10TG_FROM_TG(optc); + uint32_t update_pending = 0; + + REG_GET(OTG_DOUBLE_BUFFER_CONTROL, + OTG_UPDATE_PENDING, + &update_pending); + + return (update_pending == 1); +} + +/* OTG status register that indicates surface update is pending */ +bool optc3_get_pipe_update_pending(struct timing_generator *optc) +{ + struct optc *optc1 = DCN10TG_FROM_TG(optc); + uint32_t flip_pending = 0; + uint32_t dc_update_pending = 0; + + REG_GET_2(OTG_PIPE_UPDATE_STATUS, + OTG_FLIP_PENDING, + &flip_pending, + OTG_DC_REG_UPDATE_PENDING, + &dc_update_pending); + + return (flip_pending == 1 || dc_update_pending == 1); +} + /** * optc3_set_timing_double_buffer() - DRR double buffering control * @@ -375,6 +417,9 @@ static struct timing_generator_funcs dcn30_tg_funcs = { .get_hw_timing = optc1_get_hw_timing, .wait_drr_doublebuffer_pending_clear = optc3_wait_drr_doublebuffer_pending_clear, .is_two_pixels_per_container = optc1_is_two_pixels_per_container, + .get_optc_double_buffer_pending = optc3_get_optc_double_buffer_pending, + .get_otg_double_buffer_pending = optc3_get_otg_update_pending, + .get_pipe_update_pending = optc3_get_pipe_update_pending, }; void dcn30_timing_generator_init(struct optc *optc1) diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn30/dcn30_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn30/dcn30_optc.h index bda974d432ea..e2303f9eaf13 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn30/dcn30_optc.h +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn30/dcn30_optc.h @@ -109,7 +109,8 @@ SRI(OPTC_BYTES_PER_PIXEL, ODM, inst),\ SRI(OPTC_WIDTH_CONTROL, ODM, inst),\ SRI(OPTC_MEMORY_CONFIG, ODM, inst),\ - SR(DWB_SOURCE_SELECT) + SR(DWB_SOURCE_SELECT),\ + SRI(OTG_PIPE_UPDATE_STATUS, OTG, inst) #define DCN30_VTOTAL_REGS_SF(mask_sh) @@ -209,6 +210,7 @@ SF(ODM0_OPTC_INPUT_CLOCK_CONTROL, OPTC_INPUT_CLK_GATE_DIS, mask_sh),\ SF(ODM0_OPTC_INPUT_GLOBAL_CONTROL, OPTC_UNDERFLOW_OCCURRED_STATUS, mask_sh),\ SF(ODM0_OPTC_INPUT_GLOBAL_CONTROL, OPTC_UNDERFLOW_CLEAR, mask_sh),\ + SF(ODM0_OPTC_INPUT_GLOBAL_CONTROL, OPTC_DOUBLE_BUFFER_PENDING, mask_sh),\ SF(VTG0_CONTROL, VTG0_ENABLE, mask_sh),\ SF(VTG0_CONTROL, VTG0_FP2, mask_sh),\ SF(VTG0_CONTROL, VTG0_VCOUNT_INIT, mask_sh),\ @@ -319,7 +321,11 @@ SF(OTG0_OTG_DRR_V_TOTAL_CHANGE, OTG_DRR_V_TOTAL_CHANGE_LIMIT, mask_sh),\ SF(OTG0_OTG_H_TIMING_CNTL, OTG_H_TIMING_DIV_MODE, mask_sh),\ SF(OTG0_OTG_DOUBLE_BUFFER_CONTROL, OTG_DRR_TIMING_DBUF_UPDATE_PENDING, mask_sh),\ - SF(OTG0_OTG_DOUBLE_BUFFER_CONTROL, OTG_DRR_TIMING_DBUF_UPDATE_MODE, mask_sh) + SF(OTG0_OTG_DOUBLE_BUFFER_CONTROL, OTG_DRR_TIMING_DBUF_UPDATE_MODE, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_FLIP_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_DC_REG_UPDATE_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_CURSOR_UPDATE_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh),\ void dcn30_timing_generator_init(struct optc *optc1); @@ -356,4 +362,7 @@ void optc3_set_odm_combine(struct timing_generator *optc, int *opp_id, int opp_c void optc3_wait_drr_doublebuffer_pending_clear(struct timing_generator *optc); void optc3_tg_init(struct timing_generator *optc); void optc3_set_vtotal_min_max(struct timing_generator *optc, int vtotal_min, int vtotal_max); +bool optc3_get_optc_double_buffer_pending(struct timing_generator *optc); +bool optc3_get_otg_update_pending(struct timing_generator *optc); +bool optc3_get_pipe_update_pending(struct timing_generator *optc); #endif /* __DC_OPTC_DCN30_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn301/dcn301_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn301/dcn301_optc.c index 1a22ae89fb55..d7a45ef2d01b 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn301/dcn301_optc.c +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn301/dcn301_optc.c @@ -169,6 +169,9 @@ static struct timing_generator_funcs dcn30_tg_funcs = { .get_hw_timing = optc1_get_hw_timing, .wait_drr_doublebuffer_pending_clear = optc3_wait_drr_doublebuffer_pending_clear, .is_two_pixels_per_container = optc1_is_two_pixels_per_container, + .get_optc_double_buffer_pending = optc3_get_optc_double_buffer_pending, + .get_otg_double_buffer_pending = optc3_get_otg_update_pending, + .get_pipe_update_pending = optc3_get_pipe_update_pending, }; void dcn301_timing_generator_init(struct optc *optc1) diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn31/dcn31_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn31/dcn31_optc.h index 30b81a448ce2..fbbe86d00c2e 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn31/dcn31_optc.h +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn31/dcn31_optc.h @@ -99,7 +99,8 @@ SRI(OPTC_MEMORY_CONFIG, ODM, inst),\ SRI(OTG_CRC_CNTL2, OTG, inst),\ SR(DWB_SOURCE_SELECT),\ - SRI(OTG_DRR_CONTROL, OTG, inst) + SRI(OTG_DRR_CONTROL, OTG, inst),\ + SRI(OTG_PIPE_UPDATE_STATUS, OTG, inst) #define OPTC_COMMON_MASK_SH_LIST_DCN3_1(mask_sh)\ SF(OTG0_OTG_VSTARTUP_PARAM, VSTARTUP_START, mask_sh),\ @@ -254,7 +255,11 @@ SF(OTG0_OTG_CRC_CNTL2, OTG_CRC_DATA_STREAM_COMBINE_MODE, mask_sh),\ SF(OTG0_OTG_CRC_CNTL2, OTG_CRC_DATA_STREAM_SPLIT_MODE, mask_sh),\ SF(OTG0_OTG_CRC_CNTL2, OTG_CRC_DATA_FORMAT, mask_sh),\ - SF(OTG0_OTG_DRR_CONTROL, OTG_V_TOTAL_LAST_USED_BY_DRR, mask_sh) + SF(OTG0_OTG_DRR_CONTROL, OTG_V_TOTAL_LAST_USED_BY_DRR, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_FLIP_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_DC_REG_UPDATE_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_CURSOR_UPDATE_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh),\ void dcn31_timing_generator_init(struct optc *optc1); diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn314/dcn314_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn314/dcn314_optc.h index 99c098e76116..0ff72b97b465 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn314/dcn314_optc.h +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn314/dcn314_optc.h @@ -98,7 +98,8 @@ SRI(OPTC_BYTES_PER_PIXEL, ODM, inst),\ SRI(OPTC_WIDTH_CONTROL, ODM, inst),\ SRI(OPTC_MEMORY_CONFIG, ODM, inst),\ - SRI(OTG_DRR_CONTROL, OTG, inst) + SRI(OTG_DRR_CONTROL, OTG, inst),\ + SRI(OTG_PIPE_UPDATE_STATUS, OTG, inst) #define OPTC_COMMON_MASK_SH_LIST_DCN3_14(mask_sh)\ SF(OTG0_OTG_VSTARTUP_PARAM, VSTARTUP_START, mask_sh),\ @@ -248,7 +249,11 @@ SF(OTG0_OTG_H_TIMING_CNTL, OTG_H_TIMING_DIV_MODE, mask_sh),\ SF(OTG0_OTG_H_TIMING_CNTL, OTG_H_TIMING_DIV_MODE_MANUAL, mask_sh),\ SF(OTG0_OTG_DOUBLE_BUFFER_CONTROL, OTG_DRR_TIMING_DBUF_UPDATE_MODE, mask_sh),\ - SF(OTG0_OTG_DRR_CONTROL, OTG_V_TOTAL_LAST_USED_BY_DRR, mask_sh) + SF(OTG0_OTG_DRR_CONTROL, OTG_V_TOTAL_LAST_USED_BY_DRR, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_FLIP_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_DC_REG_UPDATE_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_CURSOR_UPDATE_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh),\ void dcn314_timing_generator_init(struct optc *optc1); diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.c index 00094f0e8470..c217f653b3c8 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.c +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.c @@ -297,18 +297,6 @@ static void optc32_set_drr( optc32_setup_manual_trigger(optc); } -bool optc32_get_double_buffer_pending(struct timing_generator *optc) -{ - struct optc *optc1 = DCN10TG_FROM_TG(optc); - uint32_t update_pending = 0; - - REG_GET(OPTC_INPUT_GLOBAL_CONTROL, - OPTC_DOUBLE_BUFFER_PENDING, - &update_pending); - - return (update_pending == 1); -} - static struct timing_generator_funcs dcn32_tg_funcs = { .validate_timing = optc1_validate_timing, .program_timing = optc1_program_timing, @@ -373,7 +361,9 @@ static struct timing_generator_funcs dcn32_tg_funcs = { .setup_manual_trigger = optc2_setup_manual_trigger, .get_hw_timing = optc1_get_hw_timing, .is_two_pixels_per_container = optc1_is_two_pixels_per_container, - .get_double_buffer_pending = optc32_get_double_buffer_pending, + .get_optc_double_buffer_pending = optc3_get_optc_double_buffer_pending, + .get_otg_double_buffer_pending = optc3_get_otg_update_pending, + .get_pipe_update_pending = optc3_get_pipe_update_pending, }; void dcn32_timing_generator_init(struct optc *optc1) diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.h index 665d7c52f67c..0b0964a9da74 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.h +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.h @@ -177,7 +177,11 @@ SF(OTG0_OTG_H_TIMING_CNTL, OTG_H_TIMING_DIV_MODE, mask_sh),\ SF(OTG0_OTG_H_TIMING_CNTL, OTG_H_TIMING_DIV_MODE_MANUAL, mask_sh),\ SF(OTG0_OTG_DOUBLE_BUFFER_CONTROL, OTG_DRR_TIMING_DBUF_UPDATE_MODE, mask_sh),\ - SF(OTG0_OTG_DRR_CONTROL, OTG_V_TOTAL_LAST_USED_BY_DRR, mask_sh) + SF(OTG0_OTG_DRR_CONTROL, OTG_V_TOTAL_LAST_USED_BY_DRR, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_FLIP_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_DC_REG_UPDATE_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_CURSOR_UPDATE_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh) void dcn32_timing_generator_init(struct optc *optc1); void optc32_set_h_timing_div_manual_mode(struct timing_generator *optc, bool manual_mode); @@ -185,6 +189,5 @@ void optc32_get_odm_combine_segments(struct timing_generator *tg, int *odm_combi void optc32_set_odm_bypass(struct timing_generator *optc, const struct dc_crtc_timing *dc_crtc_timing); void optc32_wait_odm_doublebuffer_pending_clear(struct timing_generator *tg); -bool optc32_get_double_buffer_pending(struct timing_generator *optc); #endif /* __DC_OPTC_DCN32_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.h index d077e2392379..be749ab41dce 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.h +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.h @@ -67,7 +67,11 @@ SF(OTG0_OTG_CRC1_WINDOWB_Y_CONTROL_READBACK, OTG_CRC1_WINDOWB_Y_END_READBACK, mask_sh),\ SF(OPTC_CLOCK_CONTROL, OPTC_FGCG_REP_DIS, mask_sh),\ SF(OTG0_OTG_V_COUNT_STOP_CONTROL, OTG_V_COUNT_STOP, mask_sh),\ - SF(OTG0_OTG_V_COUNT_STOP_CONTROL2, OTG_V_COUNT_STOP_TIMER, mask_sh) + SF(OTG0_OTG_V_COUNT_STOP_CONTROL2, OTG_V_COUNT_STOP_TIMER, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_FLIP_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_DC_REG_UPDATE_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_CURSOR_UPDATE_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh) void dcn35_timing_generator_init(struct optc *optc1); diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.c index a5d6a7dca554..783ca9acc762 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.c +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.c @@ -430,6 +430,35 @@ static void optc401_program_global_sync( REG_UPDATE(OTG_PSTATE_REGISTER, OTG_PSTATE_KEEPOUT_START, pstate_keepout); } +static void optc401_set_vupdate_keepout(struct timing_generator *tg, bool enable) +{ + struct optc *optc1 = DCN10TG_FROM_TG(tg); + + REG_SET_3(OTG_VUPDATE_KEEPOUT, 0, + MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_START_OFFSET, 0, + MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_END_OFFSET, optc1->vready_offset + 10, + OTG_MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_EN, enable); + + return; +} + +static bool optc401_wait_update_lock_status(struct timing_generator *tg, bool locked) +{ + struct optc *optc1 = DCN10TG_FROM_TG(tg); + uint32_t lock_status = 0; + + REG_WAIT(OTG_MASTER_UPDATE_LOCK, + UPDATE_LOCK_STATUS, locked, + 1, 150000); + + REG_GET(OTG_MASTER_UPDATE_LOCK, UPDATE_LOCK_STATUS, &lock_status); + + if (lock_status != locked) + return false; + + return true; +} + static struct timing_generator_funcs dcn401_tg_funcs = { .validate_timing = optc1_validate_timing, .program_timing = optc1_program_timing, @@ -493,7 +522,11 @@ static struct timing_generator_funcs dcn401_tg_funcs = { .setup_manual_trigger = optc2_setup_manual_trigger, .get_hw_timing = optc1_get_hw_timing, .is_two_pixels_per_container = optc1_is_two_pixels_per_container, - .get_double_buffer_pending = optc32_get_double_buffer_pending, + .get_optc_double_buffer_pending = optc3_get_optc_double_buffer_pending, + .get_otg_double_buffer_pending = optc3_get_otg_update_pending, + .get_pipe_update_pending = optc3_get_pipe_update_pending, + .set_vupdate_keepout = optc401_set_vupdate_keepout, + .wait_update_lock_status = optc401_wait_update_lock_status, }; void dcn401_timing_generator_init(struct optc *optc1) diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.h index bb13a645802d..1be89571986f 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.h +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.h @@ -159,7 +159,11 @@ SF(OTG0_OTG_PSTATE_REGISTER, OTG_PSTATE_KEEPOUT_START, mask_sh),\ SF(OTG0_OTG_PSTATE_REGISTER, OTG_PSTATE_EXTEND, mask_sh),\ SF(OTG0_OTG_PSTATE_REGISTER, OTG_UNBLANK, mask_sh),\ - SF(OTG0_OTG_PSTATE_REGISTER, OTG_PSTATE_ALLOW_WIDTH_MIN, mask_sh) + SF(OTG0_OTG_PSTATE_REGISTER, OTG_PSTATE_ALLOW_WIDTH_MIN, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_FLIP_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_DC_REG_UPDATE_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_CURSOR_UPDATE_PENDING, mask_sh),\ + SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh) void dcn401_timing_generator_init(struct optc *optc1); diff --git a/drivers/gpu/drm/amd/display/dc/resource/dce100/dce100_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dce100/dce100_resource.c index 53a5f4cb648c..e698543ec937 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dce100/dce100_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dce100/dce100_resource.c @@ -623,7 +623,7 @@ static struct link_encoder *dce100_link_encoder_create( kzalloc(sizeof(struct dce110_link_encoder), GFP_KERNEL); int link_regs_id; - if (!enc110) + if (!enc110 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; link_regs_id = diff --git a/drivers/gpu/drm/amd/display/dc/resource/dce110/dce110_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dce110/dce110_resource.c index 91da5cf85b69..035c6cfdaee5 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dce110/dce110_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dce110/dce110_resource.c @@ -668,7 +668,7 @@ static struct link_encoder *dce110_link_encoder_create( kzalloc(sizeof(struct dce110_link_encoder), GFP_KERNEL); int link_regs_id; - if (!enc110) + if (!enc110 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; link_regs_id = diff --git a/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.c index 162856c523e4..480a50967385 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.c @@ -629,7 +629,7 @@ static struct link_encoder *dce112_link_encoder_create( kzalloc(sizeof(struct dce110_link_encoder), GFP_KERNEL); int link_regs_id; - if (!enc110) + if (!enc110 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; link_regs_id = diff --git a/drivers/gpu/drm/amd/display/dc/resource/dce120/dce120_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dce120/dce120_resource.c index 621825a51f46..c63c59623433 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dce120/dce120_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dce120/dce120_resource.c @@ -706,7 +706,7 @@ static struct link_encoder *dce120_link_encoder_create( kzalloc(sizeof(struct dce110_link_encoder), GFP_KERNEL); int link_regs_id; - if (!enc110) + if (!enc110 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; link_regs_id = diff --git a/drivers/gpu/drm/amd/display/dc/resource/dce80/dce80_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dce80/dce80_resource.c index a73d3c6ef425..3d5113f010bb 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dce80/dce80_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dce80/dce80_resource.c @@ -723,7 +723,7 @@ static struct link_encoder *dce80_link_encoder_create( kzalloc(sizeof(struct dce110_link_encoder), GFP_KERNEL); int link_regs_id; - if (!enc110) + if (!enc110 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; link_regs_id = diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.c index 563c5eec83ff..770a380cc03d 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.c @@ -533,7 +533,6 @@ static const struct dc_debug_options debug_defaults_drv = { .sanity_checks = true, .disable_dmcu = false, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, /* raven smu dones't allow 0 disp clk, @@ -560,18 +559,6 @@ static const struct dc_debug_options debug_defaults_drv = { .using_dml2 = false, }; -static const struct dc_debug_options debug_defaults_diags = { - .disable_dmcu = false, - .force_abm_enable = false, - .timing_trace = true, - .clock_trace = true, - .disable_stutter = true, - .disable_pplib_clock_request = true, - .disable_pplib_wm_range = true, - .underflow_assert_delay_us = 0xFFFFFFFF, - .enable_legacy_fast_update = true, -}; - static void dcn10_dpp_destroy(struct dpp **dpp) { kfree(TO_DCN10_DPP(*dpp)); @@ -751,7 +738,7 @@ static struct link_encoder *dcn10_link_encoder_create( kzalloc(sizeof(struct dcn10_link_encoder), GFP_KERNEL); int link_regs_id; - if (!enc10) + if (!enc10 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; link_regs_id = @@ -1400,8 +1387,6 @@ static bool dcn10_resource_construct( if (dc->ctx->dce_environment == DCE_ENV_PRODUCTION_DRV) dc->debug = debug_defaults_drv; - else - dc->debug = debug_defaults_diags; /************************************************* * Create resources * diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c index eea2b3b307cd..189d0c85872e 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c @@ -706,7 +706,6 @@ static const struct resource_caps res_cap_nv14 = { static const struct dc_debug_options debug_defaults_drv = { .disable_dmcu = false, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_pplib_clock_request = true, .pipe_split_policy = MPC_SPLIT_AVOID_MULT_DISP, @@ -920,7 +919,7 @@ struct link_encoder *dcn20_link_encoder_create( kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); int link_regs_id; - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; link_regs_id = diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn201/dcn201_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn201/dcn201_resource.c index fc54483b9104..d3d67d366523 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn201/dcn201_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn201/dcn201_resource.c @@ -600,7 +600,6 @@ static const struct dc_plane_cap plane_cap = { static const struct dc_debug_options debug_defaults_drv = { .disable_dmcu = true, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_pplib_clock_request = true, .pipe_split_policy = MPC_SPLIT_DYNAMIC, @@ -797,7 +796,7 @@ static struct link_encoder *dcn201_link_encoder_create( kzalloc(sizeof(struct dcn20_link_encoder), GFP_ATOMIC); struct dcn10_link_encoder *enc10; - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; enc10 = &enc20->enc10; diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.c index 347e6aaea582..021ba8ac5c8c 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.c @@ -610,7 +610,6 @@ static const struct dc_plane_cap plane_cap = { static const struct dc_debug_options debug_defaults_drv = { .disable_dmcu = false, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_pplib_clock_request = true, .min_disp_clk_khz = 100000, @@ -1298,7 +1297,7 @@ static struct link_encoder *dcn21_link_encoder_create( kzalloc(sizeof(struct dcn21_link_encoder), GFP_KERNEL); int link_regs_id; - if (!enc21) + if (!enc21 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; link_regs_id = diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c index 5040a4c6ed18..cd31e4f16c14 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c @@ -711,7 +711,6 @@ static const struct dc_plane_cap plane_cap = { static const struct dc_debug_options debug_defaults_drv = { .disable_dmcu = true, //No DMCU on DCN30 .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_pplib_clock_request = true, .pipe_split_policy = MPC_SPLIT_DYNAMIC, @@ -927,7 +926,7 @@ static struct link_encoder *dcn30_link_encoder_create( struct dcn20_link_encoder *enc20 = kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; dcn30_link_encoder_construct(enc20, diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn301/dcn301_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn301/dcn301_resource.c index 7d04739c3ba1..a9816affd312 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn301/dcn301_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn301/dcn301_resource.c @@ -682,7 +682,6 @@ static const struct dc_plane_cap plane_cap = { static const struct dc_debug_options debug_defaults_drv = { .disable_dmcu = true, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_dpp_power_gate = false, .disable_hubp_power_gate = false, @@ -883,7 +882,7 @@ static struct link_encoder *dcn301_link_encoder_create( struct dcn20_link_encoder *enc20 = kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; dcn301_link_encoder_construct(enc20, diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn302/dcn302_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn302/dcn302_resource.c index 5791b5cc2875..02af8b8f4d27 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn302/dcn302_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn302/dcn302_resource.c @@ -81,7 +81,6 @@ static const struct dc_debug_options debug_defaults_drv = { .disable_dmcu = true, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_pplib_clock_request = true, .pipe_split_policy = MPC_SPLIT_DYNAMIC, @@ -893,7 +892,7 @@ static struct link_encoder *dcn302_link_encoder_create( { struct dcn20_link_encoder *enc20 = kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; dcn30_link_encoder_construct(enc20, enc_init_data, &link_enc_feature, diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c index 63f0f882c861..7002a8dd358a 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c @@ -82,7 +82,6 @@ static const struct dc_debug_options debug_defaults_drv = { .disable_dmcu = true, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_pplib_clock_request = true, .pipe_split_policy = MPC_SPLIT_AVOID, @@ -839,7 +838,7 @@ static struct link_encoder *dcn303_link_encoder_create( { struct dcn20_link_encoder *enc20 = kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; dcn30_link_encoder_construct(enc20, enc_init_data, &link_enc_feature, diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c index ac8cb20e2e3b..c16cf1c8f7f9 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c @@ -858,7 +858,6 @@ static const struct dc_plane_cap plane_cap = { static const struct dc_debug_options debug_defaults_drv = { .disable_dmcu = true, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_pplib_clock_request = false, .pipe_split_policy = MPC_SPLIT_DYNAMIC, @@ -869,7 +868,7 @@ static const struct dc_debug_options debug_defaults_drv = { .max_downscale_src_width = 4096,/*upto true 4K*/ .disable_pplib_wm_range = false, .scl_reset_length10 = true, - .sanity_checks = true, + .sanity_checks = false, .underflow_assert_delay_us = 0xFFFFFFFF, .dwb_fi_phase = -1, // -1 = disable, .dmub_command_table = true, @@ -1093,7 +1092,7 @@ static struct link_encoder *dcn31_link_encoder_create( struct dcn20_link_encoder *enc20 = kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; dcn31_link_encoder_construct(enc20, diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c index 169924d0a839..c0f48c78e968 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c @@ -876,7 +876,6 @@ static const struct dc_debug_options debug_defaults_drv = { .replay_skip_crtc_disabled = true, .disable_dmcu = true, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_dpp_power_gate = false, .disable_hubp_power_gate = false, @@ -889,7 +888,7 @@ static const struct dc_debug_options debug_defaults_drv = { .max_downscale_src_width = 4096,/*upto true 4k*/ .disable_pplib_wm_range = false, .scl_reset_length10 = true, - .sanity_checks = true, + .sanity_checks = false, .underflow_assert_delay_us = 0xFFFFFFFF, .dwb_fi_phase = -1, // -1 = disable, .dmub_command_table = true, @@ -1149,7 +1148,7 @@ static struct link_encoder *dcn31_link_encoder_create( struct dcn20_link_encoder *enc20 = kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; dcn31_link_encoder_construct(enc20, diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c index 3f4b9dba4112..6c3295259a81 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c @@ -858,7 +858,6 @@ static const struct dc_debug_options debug_defaults_drv = { .disable_z10 = true, /*hw not support it*/ .disable_dmcu = true, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_pplib_clock_request = false, .pipe_split_policy = MPC_SPLIT_DYNAMIC, @@ -1091,7 +1090,7 @@ static struct link_encoder *dcn31_link_encoder_create( struct dcn20_link_encoder *enc20 = kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; dcn31_link_encoder_construct(enc20, @@ -1812,6 +1811,11 @@ static void dcn315_get_panel_config_defaults(struct dc_panel_config *panel_confi *panel_config = panel_config_defaults; } +static int dcn315_get_power_profile(const struct dc_state *context) +{ + return !context->bw_ctx.bw.dcn.clk.p_state_change_support; +} + static struct dc_cap_funcs cap_funcs = { .get_dcc_compression_cap = dcn20_get_dcc_compression_cap }; @@ -1840,6 +1844,7 @@ static struct resource_funcs dcn315_res_pool_funcs = { .update_bw_bounding_box = dcn315_update_bw_bounding_box, .patch_unknown_plane_state = dcn20_patch_unknown_plane_state, .get_panel_config_defaults = dcn315_get_panel_config_defaults, + .get_power_profile = dcn315_get_power_profile, }; static bool dcn315_resource_construct( diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c index 5fd52c5fcee4..6edaaadcb173 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c @@ -853,7 +853,6 @@ static const struct dc_debug_options debug_defaults_drv = { .disable_z10 = true, /*hw not support it*/ .disable_dmcu = true, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_pplib_clock_request = false, .pipe_split_policy = MPC_SPLIT_DYNAMIC, @@ -1085,7 +1084,7 @@ static struct link_encoder *dcn31_link_encoder_create( struct dcn20_link_encoder *enc20 = kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; dcn31_link_encoder_construct(enc20, diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c index a124ad9bd108..01d1a11d5545 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c @@ -689,7 +689,6 @@ static const struct dc_plane_cap plane_cap = { static const struct dc_debug_options debug_defaults_drv = { .disable_dmcu = true, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_pplib_clock_request = false, .pipe_split_policy = MPC_SPLIT_AVOID, // Due to CRB, no need to MPC split anymore @@ -1039,7 +1038,7 @@ static struct link_encoder *dcn32_link_encoder_create( struct dcn20_link_encoder *enc20 = kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; #undef REG_STRUCT @@ -1990,6 +1989,10 @@ unsigned int dcn32_calculate_mall_ways_from_bytes(const struct dc *dc, unsigned return 0; } + if (dc->caps.max_cab_allocation_bytes == 0) { + return 0xffffffff; + } + /* add 2 lines for worst case alignment */ cache_lines_used = total_size_in_mall_bytes / dc->caps.cache_line_size + 2; diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h index 7901792afb7b..86c6e5e8c42e 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h @@ -1054,7 +1054,8 @@ unsigned int dcn32_calculate_mall_ways_from_bytes(const struct dc *dc, unsigned SRI_ARR(OPTC_BYTES_PER_PIXEL, ODM, inst), \ SRI_ARR(OPTC_WIDTH_CONTROL, ODM, inst), \ SRI_ARR(OPTC_MEMORY_CONFIG, ODM, inst), \ - SRI_ARR(OTG_DRR_CONTROL, OTG, inst) + SRI_ARR(OTG_DRR_CONTROL, OTG, inst), \ + SRI_ARR(OTG_PIPE_UPDATE_STATUS, OTG, inst) /* HUBP */ diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c index 827a94f84f10..5cb74fd9cb7d 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c @@ -686,7 +686,6 @@ static const struct dc_plane_cap plane_cap = { static const struct dc_debug_options debug_defaults_drv = { .disable_dmcu = true, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_pplib_clock_request = false, .pipe_split_policy = MPC_SPLIT_AVOID, @@ -1035,7 +1034,7 @@ static struct link_encoder *dcn321_link_encoder_create( struct dcn20_link_encoder *enc20 = kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; #undef REG_STRUCT diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c index 893a9d9ee870..6cc2960b6104 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c @@ -712,7 +712,6 @@ static const struct dc_plane_cap plane_cap = { static const struct dc_debug_options debug_defaults_drv = { .disable_dmcu = true, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_pplib_clock_request = false, .pipe_split_policy = MPC_SPLIT_AVOID, @@ -1074,7 +1073,7 @@ static struct link_encoder *dcn35_link_encoder_create( struct dcn20_link_encoder *enc20 = kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; #undef REG_STRUCT diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c index 70abd32ce2ad..d87e2641cda1 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c @@ -692,7 +692,6 @@ static const struct dc_plane_cap plane_cap = { static const struct dc_debug_options debug_defaults_drv = { .disable_dmcu = true, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_pplib_clock_request = false, .pipe_split_policy = MPC_SPLIT_AVOID, @@ -1054,7 +1053,7 @@ static struct link_encoder *dcn35_link_encoder_create( struct dcn20_link_encoder *enc20 = kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; #undef REG_STRUCT diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.c index 9d56fbdcd06a..db93bac247c0 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.c @@ -685,7 +685,6 @@ static const struct dc_plane_cap plane_cap = { static const struct dc_debug_options debug_defaults_drv = { .disable_dmcu = true, .force_abm_enable = false, - .timing_trace = false, .clock_trace = true, .disable_pplib_clock_request = false, .pipe_split_policy = MPC_SPLIT_AVOID, @@ -1032,7 +1031,7 @@ static struct link_encoder *dcn401_link_encoder_create( struct dcn20_link_encoder *enc20 = kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL); - if (!enc20) + if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs)) return NULL; #undef REG_STRUCT @@ -1579,7 +1578,8 @@ static void dcn401_destroy_resource_pool(struct resource_pool **pool) } static struct dc_cap_funcs cap_funcs = { - .get_dcc_compression_cap = dcn20_get_dcc_compression_cap + .get_dcc_compression_cap = dcn20_get_dcc_compression_cap, + .get_subvp_en = dcn32_subvp_in_use, }; static void dcn401_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params) @@ -1688,6 +1688,45 @@ static void dcn401_build_pipe_pix_clk_params(struct pipe_ctx *pipe_ctx) } } +static int dcn401_get_power_profile(const struct dc_state *context) +{ + int uclk_mhz = context->bw_ctx.bw.dcn.clk.dramclk_khz / 1000; + int dpm_level = 0; + + for (int i = 0; i < context->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_memclk_levels; i++) { + if (context->clk_mgr->bw_params->clk_table.entries[i].memclk_mhz == 0 || + uclk_mhz < context->clk_mgr->bw_params->clk_table.entries[i].memclk_mhz) + break; + if (uclk_mhz > context->clk_mgr->bw_params->clk_table.entries[i].memclk_mhz) + dpm_level++; + } + + return dpm_level; +} + +static unsigned int dcn401_calc_num_avail_chans_for_mall(struct dc *dc, unsigned int num_chans) +{ + unsigned int num_available_chans = 1; + + /* channels for MALL must be a power of 2 */ + while (num_chans > 1) { + num_available_chans = (num_available_chans << 1); + num_chans = (num_chans >> 1); + } + + /* cannot be odd */ + num_available_chans &= ~1; + + /* clamp to max available channels for MALL per ASIC */ + if (ASICREV_IS_GC_12_0_0_A0(dc->ctx->asic_id.hw_internal_rev)) { + num_available_chans = num_available_chans > 16 ? 16 : num_available_chans; + } else if (ASICREV_IS_GC_12_0_1_A0(dc->ctx->asic_id.hw_internal_rev)) { + num_available_chans = num_available_chans > 8 ? 8 : num_available_chans; + } + + return num_available_chans; +} + static struct resource_funcs dcn401_res_pool_funcs = { .destroy = dcn401_destroy_resource_pool, .link_enc_create = dcn401_link_encoder_create, @@ -1714,6 +1753,7 @@ static struct resource_funcs dcn401_res_pool_funcs = { .prepare_mcache_programming = dcn401_prepare_mcache_programming, .build_pipe_pix_clk_params = dcn401_build_pipe_pix_clk_params, .calculate_mall_ways_from_bytes = dcn32_calculate_mall_ways_from_bytes, + .get_power_profile = dcn401_get_power_profile, }; static uint32_t read_pipe_fuses(struct dc_context *ctx) @@ -1795,14 +1835,12 @@ static bool dcn401_resource_construct( dc->caps.min_horizontal_blanking_period = 80; dc->caps.dmdata_alloc_size = 2048; dc->caps.mall_size_per_mem_channel = 4; - /* total size = mall per channel * num channels * 1024 * 1024 */ - dc->caps.mall_size_total = dc->caps.mall_size_per_mem_channel * dc->ctx->dc_bios->vram_info.num_chans * 1048576; dc->caps.cursor_cache_size = dc->caps.max_cursor_size * dc->caps.max_cursor_size * 8; dc->caps.cache_line_size = 64; dc->caps.cache_num_ways = 16; /* Calculate the available MALL space */ - dc->caps.max_cab_allocation_bytes = dcn32_calc_num_avail_chans_for_mall( + dc->caps.max_cab_allocation_bytes = dcn401_calc_num_avail_chans_for_mall( dc, dc->ctx->dc_bios->vram_info.num_chans) * dc->caps.mall_size_per_mem_channel * 1024 * 1024; dc->caps.mall_size_total = dc->caps.max_cab_allocation_bytes; @@ -1867,6 +1905,7 @@ static bool dcn401_resource_construct( dc->config.prefer_easf = true; dc->config.dc_mode_clk_limit_support = true; dc->config.enable_windowed_mpo_odm = true; + dc->config.set_pipe_unlock_order = true; /* Need to ensure DET gets freed before allocating */ /* read VBIOS LTTPR caps */ { if (ctx->dc_bios->funcs->get_lttpr_caps) { @@ -2132,6 +2171,7 @@ static bool dcn401_resource_construct( /* SPL */ spl_init_easf_filter_coeffs(); spl_init_blur_scale_coeffs(); + dc->caps.scl_caps.sharpener_support = true; return true; diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.h index 514d1ce20df9..7c8d61db153d 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.h +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.h @@ -536,8 +536,9 @@ void dcn401_prepare_mcache_programming(struct dc *dc, struct dc_state *context); SRI_ARR(OPTC_WIDTH_CONTROL, ODM, inst), \ SRI_ARR(OPTC_WIDTH_CONTROL2, ODM, inst), \ SRI_ARR(OPTC_MEMORY_CONFIG, ODM, inst), \ - SRI_ARR(OTG_DRR_CONTROL, OTG, inst), \ - SRI_ARR(OTG_PSTATE_REGISTER, OTG, inst) + SRI_ARR(OTG_DRR_CONTROL, OTG, inst), \ + SRI_ARR(OTG_PSTATE_REGISTER, OTG, inst), \ + SRI_ARR(OTG_PIPE_UPDATE_STATUS, OTG, inst) /* HUBBUB */ #define HUBBUB_REG_LIST_DCN4_01_RI(id) \ @@ -609,7 +610,9 @@ void dcn401_prepare_mcache_programming(struct dc *dc, struct dc_state *context); SR(DCHUBBUB_CLOCK_CNTL), \ SR(DCHUBBUB_SDPIF_CFG0), \ SR(DCHUBBUB_SDPIF_CFG1), \ - SR(DCHUBBUB_MEM_PWR_MODE_CTRL) + SR(DCHUBBUB_MEM_PWR_MODE_CTRL), \ + SR(DCHUBBUB_TIMEOUT_DETECTION_CTRL1), \ + SR(DCHUBBUB_TIMEOUT_DETECTION_CTRL2) /* DCCG */ diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl.c b/drivers/gpu/drm/amd/display/dc/spl/dc_spl.c index 014e8a296f0c..614276200aa0 100644 --- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl.c +++ b/drivers/gpu/drm/amd/display/dc/spl/dc_spl.c @@ -848,13 +848,13 @@ static bool spl_get_isharp_en(struct spl_in *spl_in, * surfaces based on policy setting */ if (!spl_is_yuv420(spl_in->basic_in.format) && - (spl_in->debug.sharpen_policy == SHARPEN_YUV)) + (spl_in->sharpen_policy == SHARPEN_YUV)) return enable_isharp; else if ((spl_is_yuv420(spl_in->basic_in.format) && !fullscreen) && - (spl_in->debug.sharpen_policy == SHARPEN_RGB_FULLSCREEN_YUV)) + (spl_in->sharpen_policy == SHARPEN_RGB_FULLSCREEN_YUV)) return enable_isharp; else if (!spl_in->is_fullscreen && - spl_in->debug.sharpen_policy == SHARPEN_FULLSCREEN_ALL) + spl_in->sharpen_policy == SHARPEN_FULLSCREEN_ALL) return enable_isharp; /* @@ -868,6 +868,60 @@ static bool spl_get_isharp_en(struct spl_in *spl_in, return enable_isharp; } +/* Calculate number of tap with adaptive scaling off */ +static void spl_get_taps_non_adaptive_scaler( + struct spl_scratch *spl_scratch, const struct spl_taps *in_taps) +{ + if (in_taps->h_taps == 0) { + if (spl_fixpt_ceil(spl_scratch->scl_data.ratios.horz) > 1) + spl_scratch->scl_data.taps.h_taps = spl_min(2 * spl_fixpt_ceil( + spl_scratch->scl_data.ratios.horz), 8); + else + spl_scratch->scl_data.taps.h_taps = 4; + } else + spl_scratch->scl_data.taps.h_taps = in_taps->h_taps; + + if (in_taps->v_taps == 0) { + if (spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert) > 1) + spl_scratch->scl_data.taps.v_taps = spl_min(spl_fixpt_ceil(spl_fixpt_mul_int( + spl_scratch->scl_data.ratios.vert, 2)), 8); + else + spl_scratch->scl_data.taps.v_taps = 4; + } else + spl_scratch->scl_data.taps.v_taps = in_taps->v_taps; + + if (in_taps->v_taps_c == 0) { + if (spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert_c) > 1) + spl_scratch->scl_data.taps.v_taps_c = spl_min(spl_fixpt_ceil(spl_fixpt_mul_int( + spl_scratch->scl_data.ratios.vert_c, 2)), 8); + else + spl_scratch->scl_data.taps.v_taps_c = 4; + } else + spl_scratch->scl_data.taps.v_taps_c = in_taps->v_taps_c; + + if (in_taps->h_taps_c == 0) { + if (spl_fixpt_ceil(spl_scratch->scl_data.ratios.horz_c) > 1) + spl_scratch->scl_data.taps.h_taps_c = spl_min(2 * spl_fixpt_ceil( + spl_scratch->scl_data.ratios.horz_c), 8); + else + spl_scratch->scl_data.taps.h_taps_c = 4; + } else if ((in_taps->h_taps_c % 2) != 0 && in_taps->h_taps_c != 1) + /* Only 1 and even h_taps_c are supported by hw */ + spl_scratch->scl_data.taps.h_taps_c = in_taps->h_taps_c - 1; + else + spl_scratch->scl_data.taps.h_taps_c = in_taps->h_taps_c; + + if (IDENTITY_RATIO(spl_scratch->scl_data.ratios.horz)) + spl_scratch->scl_data.taps.h_taps = 1; + if (IDENTITY_RATIO(spl_scratch->scl_data.ratios.vert)) + spl_scratch->scl_data.taps.v_taps = 1; + if (IDENTITY_RATIO(spl_scratch->scl_data.ratios.horz_c)) + spl_scratch->scl_data.taps.h_taps_c = 1; + if (IDENTITY_RATIO(spl_scratch->scl_data.ratios.vert_c)) + spl_scratch->scl_data.taps.v_taps_c = 1; + +} + /* Calculate optimal number of taps */ static bool spl_get_optimal_number_of_taps( int max_downscale_src_width, struct spl_in *spl_in, struct spl_scratch *spl_scratch, @@ -882,8 +936,22 @@ static bool spl_get_optimal_number_of_taps( if (spl_scratch->scl_data.viewport.width > spl_scratch->scl_data.h_active && max_downscale_src_width != 0 && - spl_scratch->scl_data.viewport.width > max_downscale_src_width) + spl_scratch->scl_data.viewport.width > max_downscale_src_width) { + spl_get_taps_non_adaptive_scaler(spl_scratch, in_taps); + *enable_easf_v = false; + *enable_easf_h = false; + *enable_isharp = false; return false; + } + + /* Disable adaptive scaler and sharpener when integer scaling is enabled */ + if (spl_in->scaling_quality.integer_scaling) { + spl_get_taps_non_adaptive_scaler(spl_scratch, in_taps); + *enable_easf_v = false; + *enable_easf_h = false; + *enable_isharp = false; + return true; + } /* Check if we are using EASF or not */ skip_easf = enable_easf(spl_in, spl_scratch); @@ -893,43 +961,9 @@ static bool spl_get_optimal_number_of_taps( * From programming guide: taps = min{ ceil(2*H_RATIO,1), 8} for downscaling * taps = 4 for upscaling */ - if (skip_easf) { - if (in_taps->h_taps == 0) { - if (spl_fixpt_ceil(spl_scratch->scl_data.ratios.horz) > 1) - spl_scratch->scl_data.taps.h_taps = spl_min(2 * spl_fixpt_ceil( - spl_scratch->scl_data.ratios.horz), 8); - else - spl_scratch->scl_data.taps.h_taps = 4; - } else - spl_scratch->scl_data.taps.h_taps = in_taps->h_taps; - if (in_taps->v_taps == 0) { - if (spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert) > 1) - spl_scratch->scl_data.taps.v_taps = spl_min(spl_fixpt_ceil(spl_fixpt_mul_int( - spl_scratch->scl_data.ratios.vert, 2)), 8); - else - spl_scratch->scl_data.taps.v_taps = 4; - } else - spl_scratch->scl_data.taps.v_taps = in_taps->v_taps; - if (in_taps->v_taps_c == 0) { - if (spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert_c) > 1) - spl_scratch->scl_data.taps.v_taps_c = spl_min(spl_fixpt_ceil(spl_fixpt_mul_int( - spl_scratch->scl_data.ratios.vert_c, 2)), 8); - else - spl_scratch->scl_data.taps.v_taps_c = 4; - } else - spl_scratch->scl_data.taps.v_taps_c = in_taps->v_taps_c; - if (in_taps->h_taps_c == 0) { - if (spl_fixpt_ceil(spl_scratch->scl_data.ratios.horz_c) > 1) - spl_scratch->scl_data.taps.h_taps_c = spl_min(2 * spl_fixpt_ceil( - spl_scratch->scl_data.ratios.horz_c), 8); - else - spl_scratch->scl_data.taps.h_taps_c = 4; - } else if ((in_taps->h_taps_c % 2) != 0 && in_taps->h_taps_c != 1) - /* Only 1 and even h_taps_c are supported by hw */ - spl_scratch->scl_data.taps.h_taps_c = in_taps->h_taps_c - 1; - else - spl_scratch->scl_data.taps.h_taps_c = in_taps->h_taps_c; - } else { + if (skip_easf) + spl_get_taps_non_adaptive_scaler(spl_scratch, in_taps); + else { if (spl_is_yuv420(spl_in->basic_in.format)) { spl_scratch->scl_data.taps.h_taps = 6; spl_scratch->scl_data.taps.v_taps = 6; @@ -954,7 +988,7 @@ static bool spl_get_optimal_number_of_taps( else lb_config = LB_MEMORY_CONFIG_0; // Determine max vtap support by calculating how much line buffer can fit - spl_in->funcs->spl_calc_lb_num_partitions(spl_in->basic_out.alpha_en, &spl_scratch->scl_data, + spl_in->callbacks.spl_calc_lb_num_partitions(spl_in->basic_out.alpha_en, &spl_scratch->scl_data, lb_config, &num_part_y, &num_part_c); /* MAX_V_TAPS = MIN (NUM_LINES - MAX(CEILING(V_RATIO,1)-2, 0), 8) */ if (spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert) > 2) @@ -1590,7 +1624,8 @@ static void spl_set_isharp_data(struct dscl_prog_data *dscl_prog_data, spl_build_isharp_1dlut_from_reference_curve(ratio, setup, adp_sharpness, scale_to_sharpness_policy); - dscl_prog_data->isharp_delta = spl_get_pregen_filter_isharp_1D_lut(setup); + memcpy(dscl_prog_data->isharp_delta, spl_get_pregen_filter_isharp_1D_lut(setup), + sizeof(uint32_t) * ISHARP_LUT_TABLE_SIZE); dscl_prog_data->sharpness_level = adp_sharpness.sharpness_level; dscl_prog_data->isharp_en = 1; // ISHARP_EN @@ -1753,12 +1788,12 @@ bool spl_calculate_scaler_params(struct spl_in *spl_in, struct spl_out *spl_out) // Clamp spl_clamp_viewport(&spl_scratch.scl_data.viewport); - if (!res) - return res; - // Save all calculated parameters in dscl_prog_data structure to program hw registers spl_set_dscl_prog_data(spl_in, &spl_scratch, spl_out, enable_easf_v, enable_easf_h, enable_isharp); + if (!res) + return res; + if (spl_in->lls_pref == LLS_PREF_YES) { if (spl_in->is_hdr_on) setup = HDR_L; diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.h b/drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.h index afcc66206ca2..89af91e19b6c 100644 --- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.h +++ b/drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.h @@ -7,7 +7,6 @@ #include "dc_spl_types.h" -#define ISHARP_LUT_TABLE_SIZE 32 const uint32_t *spl_get_filter_isharp_1D_lut_0(void); const uint32_t *spl_get_filter_isharp_1D_lut_0p5x(void); const uint32_t *spl_get_filter_isharp_1D_lut_1p0x(void); diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_types.h b/drivers/gpu/drm/amd/display/dc/spl/dc_spl_types.h index 2a74ff5fdfdb..55d557df4aa5 100644 --- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_types.h +++ b/drivers/gpu/drm/amd/display/dc/spl/dc_spl_types.h @@ -5,10 +5,8 @@ #ifndef __DC_SPL_TYPES_H__ #define __DC_SPL_TYPES_H__ +#include "spl_debug.h" #include "spl_os_types.h" // swap -#ifndef SPL_ASSERT -#define SPL_ASSERT(_bool) ((void *)0) -#endif #include "spl_fixpt31_32.h" // fixed31_32 and related functions #include "spl_custom_float.h" // custom float and related functions @@ -252,6 +250,7 @@ enum isharp_en { ISHARP_DISABLE, ISHARP_ENABLE }; +#define ISHARP_LUT_TABLE_SIZE 32 // Below struct holds values that can be directly used to program // hardware registers. No conversion/clamping is required struct dscl_prog_data { @@ -402,7 +401,7 @@ struct dscl_prog_data { uint32_t isharp_nl_en; // ISHARP_NL_EN ? TODO:check this struct isharp_lba isharp_lba; // ISHARP_LBA struct isharp_fmt isharp_fmt; // ISHARP_FMT - const uint32_t *isharp_delta; + uint32_t isharp_delta[ISHARP_LUT_TABLE_SIZE]; struct isharp_nldelta_sclip isharp_nldelta_sclip; // ISHARP_NLDELTA_SCLIP /* blur and scale filter */ const uint16_t *filter_blur_scale_v; @@ -498,7 +497,7 @@ enum scale_to_sharpness_policy { SCALE_TO_SHARPNESS_ADJ_YUV = 1, SCALE_TO_SHARPNESS_ADJ_ALL = 2 }; -struct spl_funcs { +struct spl_callbacks { void (*spl_calc_lb_num_partitions) (bool alpha_en, const struct spl_scaler_data *scl_data, @@ -510,7 +509,6 @@ struct spl_funcs { struct spl_debug { int visual_confirm_base_offset; int visual_confirm_dpp_offset; - enum sharpen_policy sharpen_policy; enum scale_to_sharpness_policy scale_to_sharpness_policy; }; @@ -520,7 +518,7 @@ struct spl_in { // Basic slice information int odm_slice_index; // ODM Slice Index using get_odm_split_index struct spl_taps scaling_quality; // Explicit Scaling Quality - struct spl_funcs *funcs; + struct spl_callbacks callbacks; // Inputs for isharp and EASF struct adaptive_sharpness adaptive_sharpness; // Adaptive Sharpness enum linear_light_scaling lls_pref; // Linear Light Scaling @@ -532,6 +530,7 @@ struct spl_in { int h_active; int v_active; int sdr_white_level_nits; + enum sharpen_policy sharpen_policy; }; // end of SPL inputs diff --git a/drivers/gpu/drm/amd/display/dc/spl/spl_debug.h b/drivers/gpu/drm/amd/display/dc/spl/spl_debug.h index 5696dafd0894..a6f6132df241 100644 --- a/drivers/gpu/drm/amd/display/dc/spl/spl_debug.h +++ b/drivers/gpu/drm/amd/display/dc/spl/spl_debug.h @@ -5,21 +5,26 @@ #ifndef SPL_DEBUG_H #define SPL_DEBUG_H -#ifdef SPL_ASSERT -#undef SPL_ASSERT -#endif -#define SPL_ASSERT(b) +#if defined(CONFIG_HAVE_KGDB) || defined(CONFIG_KGDB) +#define SPL_ASSERT_CRITICAL(expr) do { \ + if (WARN_ON(!(expr))) { \ + kgdb_breakpoint(); \ + } \ +} while (0) +#else +#define SPL_ASSERT_CRITICAL(expr) do { \ + if (WARN_ON(!(expr))) { \ + ; \ + } \ +} while (0) +#endif /* CONFIG_HAVE_KGDB || CONFIG_KGDB */ -#define SPL_ASSERT_CRITICAL(expr) do {if (expr)/* Do nothing */; } while (0) +#if defined(CONFIG_DEBUG_KERNEL_DC) +#define SPL_ASSERT(expr) SPL_ASSERT_CRITICAL(expr) +#else +#define SPL_ASSERT(expr) WARN_ON(!(expr)) +#endif /* CONFIG_DEBUG_KERNEL_DC */ -#ifdef SPL_DALMSG -#undef SPL_DALMSG -#endif -#define SPL_DALMSG(b) - -#ifdef SPL_DAL_ASSERT_MSG -#undef SPL_DAL_ASSERT_MSG -#endif -#define SPL_DAL_ASSERT_MSG(b, m) +#define SPL_BREAK_TO_DEBUGGER() SPL_ASSERT(0) #endif // SPL_DEBUG_H diff --git a/drivers/gpu/drm/amd/display/dc/spl/spl_fixpt31_32.c b/drivers/gpu/drm/amd/display/dc/spl/spl_fixpt31_32.c index a95565df5487..131f1e3949d3 100644 --- a/drivers/gpu/drm/amd/display/dc/spl/spl_fixpt31_32.c +++ b/drivers/gpu/drm/amd/display/dc/spl/spl_fixpt31_32.c @@ -22,14 +22,14 @@ static inline unsigned long long abs_i64( * result = dividend / divisor * *remainder = dividend % divisor */ -static inline unsigned long long complete_integer_division_u64( +static inline unsigned long long spl_complete_integer_division_u64( unsigned long long dividend, unsigned long long divisor, unsigned long long *remainder) { unsigned long long result; - ASSERT(divisor); + SPL_ASSERT(divisor); result = spl_div64_u64_rem(dividend, divisor, remainder); @@ -60,10 +60,10 @@ struct spl_fixed31_32 spl_fixpt_from_fraction(long long numerator, long long den /* determine integer part */ - unsigned long long res_value = complete_integer_division_u64( + unsigned long long res_value = spl_complete_integer_division_u64( arg1_value, arg2_value, &remainder); - ASSERT(res_value <= LONG_MAX); + SPL_ASSERT(res_value <= (unsigned long long)LONG_MAX); /* determine fractional part */ { @@ -85,7 +85,7 @@ struct spl_fixed31_32 spl_fixpt_from_fraction(long long numerator, long long den { unsigned long long summand = (remainder << 1) >= arg2_value; - ASSERT(res_value <= LLONG_MAX - summand); + SPL_ASSERT(res_value <= (unsigned long long)LLONG_MAX - summand); res_value += summand; } @@ -118,19 +118,19 @@ struct spl_fixed31_32 spl_fixpt_mul(struct spl_fixed31_32 arg1, struct spl_fixed res.value = arg1_int * arg2_int; - ASSERT(res.value <= (long long)LONG_MAX); + SPL_ASSERT(res.value <= (long long)LONG_MAX); res.value <<= FIXED31_32_BITS_PER_FRACTIONAL_PART; tmp = arg1_int * arg2_fra; - ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value)); + SPL_ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value)); res.value += tmp; tmp = arg2_int * arg1_fra; - ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value)); + SPL_ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value)); res.value += tmp; @@ -139,7 +139,7 @@ struct spl_fixed31_32 spl_fixpt_mul(struct spl_fixed31_32 arg1, struct spl_fixed tmp = (tmp >> FIXED31_32_BITS_PER_FRACTIONAL_PART) + (tmp >= (unsigned long long)spl_fixpt_half.value); - ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value)); + SPL_ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value)); res.value += tmp; @@ -163,17 +163,17 @@ struct spl_fixed31_32 spl_fixpt_sqr(struct spl_fixed31_32 arg) res.value = arg_int * arg_int; - ASSERT(res.value <= (long long)LONG_MAX); + SPL_ASSERT(res.value <= (long long)LONG_MAX); res.value <<= FIXED31_32_BITS_PER_FRACTIONAL_PART; tmp = arg_int * arg_fra; - ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value)); + SPL_ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value)); res.value += tmp; - ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value)); + SPL_ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value)); res.value += tmp; @@ -182,7 +182,7 @@ struct spl_fixed31_32 spl_fixpt_sqr(struct spl_fixed31_32 arg) tmp = (tmp >> FIXED31_32_BITS_PER_FRACTIONAL_PART) + (tmp >= (unsigned long long)spl_fixpt_half.value); - ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value)); + SPL_ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value)); res.value += tmp; @@ -196,7 +196,7 @@ struct spl_fixed31_32 spl_fixpt_recip(struct spl_fixed31_32 arg) * Good idea to use Newton's method */ - ASSERT(arg.value); + SPL_ASSERT(arg.value); return spl_fixpt_from_fraction( spl_fixpt_one.value, @@ -286,7 +286,7 @@ struct spl_fixed31_32 spl_fixpt_cos(struct spl_fixed31_32 arg) * * Calculated as Taylor series. */ -static struct spl_fixed31_32 fixed31_32_exp_from_taylor_series(struct spl_fixed31_32 arg) +static struct spl_fixed31_32 spl_fixed31_32_exp_from_taylor_series(struct spl_fixed31_32 arg) { unsigned int n = 9; @@ -295,7 +295,7 @@ static struct spl_fixed31_32 fixed31_32_exp_from_taylor_series(struct spl_fixed3 n + 1); /* TODO find correct res */ - ASSERT(spl_fixpt_lt(arg, spl_fixpt_one)); + SPL_ASSERT(spl_fixpt_lt(arg, spl_fixpt_one)); do res = spl_fixpt_add( @@ -337,22 +337,22 @@ struct spl_fixed31_32 spl_fixpt_exp(struct spl_fixed31_32 arg) spl_fixpt_ln2, m)); - ASSERT(m != 0); + SPL_ASSERT(m != 0); - ASSERT(spl_fixpt_lt( + SPL_ASSERT(spl_fixpt_lt( spl_fixpt_abs(r), spl_fixpt_one)); if (m > 0) return spl_fixpt_shl( - fixed31_32_exp_from_taylor_series(r), + spl_fixed31_32_exp_from_taylor_series(r), (unsigned char)m); else return spl_fixpt_div_int( - fixed31_32_exp_from_taylor_series(r), + spl_fixed31_32_exp_from_taylor_series(r), 1LL << -m); } else if (arg.value != 0) - return fixed31_32_exp_from_taylor_series(arg); + return spl_fixed31_32_exp_from_taylor_series(arg); else return spl_fixpt_one; } @@ -364,7 +364,7 @@ struct spl_fixed31_32 spl_fixpt_log(struct spl_fixed31_32 arg) struct spl_fixed31_32 error; - ASSERT(arg.value > 0); + SPL_ASSERT(arg.value > 0); /* TODO if arg is negative, return NaN */ /* TODO if arg is zero, return -INF */ @@ -396,7 +396,7 @@ struct spl_fixed31_32 spl_fixpt_log(struct spl_fixed31_32 arg) * part in 32 bits. It is used in hw programming (scaler) */ -static inline unsigned int ux_dy( +static inline unsigned int spl_ux_dy( long long value, unsigned int integer_bits, unsigned int fractional_bits) @@ -415,13 +415,13 @@ static inline unsigned int ux_dy( return result | fractional_part; } -static inline unsigned int clamp_ux_dy( +static inline unsigned int spl_clamp_ux_dy( long long value, unsigned int integer_bits, unsigned int fractional_bits, unsigned int min_clamp) { - unsigned int truncated_val = ux_dy(value, integer_bits, fractional_bits); + unsigned int truncated_val = spl_ux_dy(value, integer_bits, fractional_bits); if (value >= (1LL << (integer_bits + FIXED31_32_BITS_PER_FRACTIONAL_PART))) return (1 << (integer_bits + fractional_bits)) - 1; @@ -433,40 +433,40 @@ static inline unsigned int clamp_ux_dy( unsigned int spl_fixpt_u4d19(struct spl_fixed31_32 arg) { - return ux_dy(arg.value, 4, 19); + return spl_ux_dy(arg.value, 4, 19); } unsigned int spl_fixpt_u3d19(struct spl_fixed31_32 arg) { - return ux_dy(arg.value, 3, 19); + return spl_ux_dy(arg.value, 3, 19); } unsigned int spl_fixpt_u2d19(struct spl_fixed31_32 arg) { - return ux_dy(arg.value, 2, 19); + return spl_ux_dy(arg.value, 2, 19); } unsigned int spl_fixpt_u0d19(struct spl_fixed31_32 arg) { - return ux_dy(arg.value, 0, 19); + return spl_ux_dy(arg.value, 0, 19); } unsigned int spl_fixpt_clamp_u0d14(struct spl_fixed31_32 arg) { - return clamp_ux_dy(arg.value, 0, 14, 1); + return spl_clamp_ux_dy(arg.value, 0, 14, 1); } unsigned int spl_fixpt_clamp_u0d10(struct spl_fixed31_32 arg) { - return clamp_ux_dy(arg.value, 0, 10, 1); + return spl_clamp_ux_dy(arg.value, 0, 10, 1); } int spl_fixpt_s4d19(struct spl_fixed31_32 arg) { if (arg.value < 0) - return -(int)ux_dy(spl_fixpt_abs(arg).value, 4, 19); + return -(int)spl_ux_dy(spl_fixpt_abs(arg).value, 4, 19); else - return ux_dy(arg.value, 4, 19); + return spl_ux_dy(arg.value, 4, 19); } struct spl_fixed31_32 spl_fixpt_from_ux_dy(unsigned int value, diff --git a/drivers/gpu/drm/amd/display/dc/spl/spl_fixpt31_32.h b/drivers/gpu/drm/amd/display/dc/spl/spl_fixpt31_32.h index 8a045e2f8699..ed2647f9a099 100644 --- a/drivers/gpu/drm/amd/display/dc/spl/spl_fixpt31_32.h +++ b/drivers/gpu/drm/amd/display/dc/spl/spl_fixpt31_32.h @@ -5,11 +5,8 @@ #ifndef __SPL_FIXED31_32_H__ #define __SPL_FIXED31_32_H__ -#include "os_types.h" +#include "spl_debug.h" #include "spl_os_types.h" // swap -#ifndef ASSERT -#define ASSERT(_bool) ((void *)0) -#endif #ifndef LLONG_MAX #define LLONG_MAX 9223372036854775807ll @@ -194,7 +191,7 @@ static inline struct spl_fixed31_32 spl_fixpt_clamp( */ static inline struct spl_fixed31_32 spl_fixpt_shl(struct spl_fixed31_32 arg, unsigned char shift) { - ASSERT(((arg.value >= 0) && (arg.value <= LLONG_MAX >> shift)) || + SPL_ASSERT(((arg.value >= 0) && (arg.value <= LLONG_MAX >> shift)) || ((arg.value < 0) && (arg.value >= ~(LLONG_MAX >> shift)))); arg.value = arg.value << shift; @@ -231,7 +228,7 @@ static inline struct spl_fixed31_32 spl_fixpt_add(struct spl_fixed31_32 arg1, st { struct spl_fixed31_32 res; - ASSERT(((arg1.value >= 0) && (LLONG_MAX - arg1.value >= arg2.value)) || + SPL_ASSERT(((arg1.value >= 0) && (LLONG_MAX - arg1.value >= arg2.value)) || ((arg1.value < 0) && (LLONG_MIN - arg1.value <= arg2.value))); res.value = arg1.value + arg2.value; @@ -256,7 +253,7 @@ static inline struct spl_fixed31_32 spl_fixpt_sub(struct spl_fixed31_32 arg1, st { struct spl_fixed31_32 res; - ASSERT(((arg2.value >= 0) && (LLONG_MIN + arg2.value <= arg1.value)) || + SPL_ASSERT(((arg2.value >= 0) && (LLONG_MIN + arg2.value <= arg1.value)) || ((arg2.value < 0) && (LLONG_MAX + arg2.value >= arg1.value))); res.value = arg1.value - arg2.value; @@ -448,7 +445,7 @@ static inline int spl_fixpt_round(struct spl_fixed31_32 arg) const long long summand = spl_fixpt_half.value; - ASSERT(LLONG_MAX - (long long)arg_value >= summand); + SPL_ASSERT(LLONG_MAX - (long long)arg_value >= summand); arg_value += summand; @@ -469,7 +466,7 @@ static inline int spl_fixpt_ceil(struct spl_fixed31_32 arg) const long long summand = spl_fixpt_one.value - spl_fixpt_epsilon.value; - ASSERT(LLONG_MAX - (long long)arg_value >= summand); + SPL_ASSERT(LLONG_MAX - (long long)arg_value >= summand); arg_value += summand; @@ -504,7 +501,7 @@ static inline struct spl_fixed31_32 spl_fixpt_truncate(struct spl_fixed31_32 arg bool negative = arg.value < 0; if (frac_bits >= FIXED31_32_BITS_PER_FRACTIONAL_PART) { - ASSERT(frac_bits == FIXED31_32_BITS_PER_FRACTIONAL_PART); + SPL_ASSERT(frac_bits == FIXED31_32_BITS_PER_FRACTIONAL_PART); return arg; } diff --git a/drivers/gpu/drm/amd/display/dc/spl/spl_os_types.h b/drivers/gpu/drm/amd/display/dc/spl/spl_os_types.h index 709706ed4f2c..2e6ba71960ac 100644 --- a/drivers/gpu/drm/amd/display/dc/spl/spl_os_types.h +++ b/drivers/gpu/drm/amd/display/dc/spl/spl_os_types.h @@ -6,6 +6,8 @@ #ifndef _SPL_OS_TYPES_H_ #define _SPL_OS_TYPES_H_ +#include "spl_debug.h" + #include #include #include @@ -18,7 +20,6 @@ * general debug capabilities * */ -#define SPL_BREAK_TO_DEBUGGER() ASSERT(0) static inline uint64_t spl_div_u64_rem(uint64_t dividend, uint32_t divisor, uint32_t *remainder) { diff --git a/drivers/gpu/drm/amd/display/dmub/dmub_srv.h b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h index fe5b6f7a3eb1..b353c4ceb60d 100644 --- a/drivers/gpu/drm/amd/display/dmub/dmub_srv.h +++ b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h @@ -301,6 +301,7 @@ struct dmub_srv_hw_params { bool disallow_phy_access; bool disable_sldo_opt; bool enable_non_transparent_setconfig; + bool lower_hbr3_phy_ssc; }; /** @@ -570,6 +571,14 @@ struct dmub_notification { }; }; +/* enum dmub_ips_mode - IPS mode identifier */ +enum dmub_ips_mode { + DMUB_IPS_MODE_IPS1_MAX = 0, + DMUB_IPS_MODE_IPS2, + DMUB_IPS_MODE_IPS1_RCG, + DMUB_IPS_MODE_IPS1_ONO2_ON +}; + /** * DMUB firmware version helper macro - useful for checking if the version * of a firmware to know if feature or functionality is supported or present. diff --git a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h index ebcf68bfae2b..b800a507d1e0 100644 --- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h @@ -170,6 +170,11 @@ #pragma pack(push, 1) #define ABM_NUM_OF_ACE_SEGMENTS 5 +/** + * Debug FW state offset + */ +#define DMUB_DEBUG_FW_STATE_OFFSET 0x300 + union abm_flags { struct { /** @@ -490,6 +495,7 @@ struct dmub_feature_caps { uint8_t gecc_enable; uint8_t replay_supported; uint8_t replay_reserved[3]; + uint8_t abm_aux_backlight_support; }; struct dmub_visual_confirm_color { @@ -689,7 +695,8 @@ union dmub_fw_boot_options { uint32_t ips_disable: 3; /* options to disable ips support*/ uint32_t ips_sequential_ono: 1; /**< 1 to enable sequential ONO IPS sequence */ uint32_t disable_sldo_opt: 1; /**< 1 to disable SLDO optimizations */ - uint32_t reserved : 7; /**< reserved */ + uint32_t lower_hbr3_phy_ssc: 1; /**< 1 to lower hbr3 phy ssc to 0.125 percent */ + uint32_t reserved : 6; /**< reserved */ } bits; /**< boot bits */ uint32_t all; /**< 32-bit access to bits */ }; @@ -721,6 +728,7 @@ enum dmub_shared_state_feature_id { DMUB_SHARED_SHARE_FEATURE__INVALID = 0, DMUB_SHARED_SHARE_FEATURE__IPS_FW = 1, DMUB_SHARED_SHARE_FEATURE__IPS_DRIVER = 2, + DMUB_SHARED_SHARE_FEATURE__DEBUG_SETUP = 3, DMUB_SHARED_STATE_FEATURE__LAST, /* Total number of features. */ }; @@ -747,7 +755,8 @@ union dmub_shared_state_ips_driver_signals { uint32_t allow_ips1 : 1; /**< 1 is IPS1 is allowed */ uint32_t allow_ips2 : 1; /**< 1 is IPS1 is allowed */ uint32_t allow_z10 : 1; /**< 1 if Z10 is allowed */ - uint32_t reserved_bits : 28; /**< Reversed bits */ + uint32_t allow_idle: 1; /**< 1 if driver is allowing idle */ + uint32_t reserved_bits : 27; /**< Reversed bits */ } bits; uint32_t all; }; @@ -757,6 +766,14 @@ union dmub_shared_state_ips_driver_signals { */ #define DMUB_SHARED_STATE__IPS_FW_VERSION 1 +struct dmub_shared_state_debug_setup { + union { + struct { + uint32_t exclude_points[62]; + } profile_mode; + }; +}; + /** * struct dmub_shared_state_ips_fw - Firmware state for IPS. */ @@ -809,6 +826,7 @@ struct dmub_shared_state_feature_block { struct dmub_shared_state_feature_common common; /**< Generic data */ struct dmub_shared_state_ips_fw ips_fw; /**< IPS firmware state */ struct dmub_shared_state_ips_driver ips_driver; /**< IPS driver state */ + struct dmub_shared_state_debug_setup debug_setup; /**< Debug setup */ } data; /**< Shared state data. */ }; /* 256-bytes, fixed */ @@ -1050,12 +1068,111 @@ enum dmub_gpint_command { */ DMUB_GPINT__GET_TRACE_BUFFER_MASK_WORD3 = 119, + /** + * DESC: Set IPS residency measurement + * ARGS: 0 - Disable ips measurement + * 1 - Enable ips measurement + */ + DMUB_GPINT__IPS_RESIDENCY = 121, + /** * DESC: Enable measurements for various task duration * ARGS: 0 - Disable measurement * 1 - Enable measurement */ DMUB_GPINT__TRACE_DMUB_WAKE_ACTIVITY = 123, + + /** + * DESC: Gets IPS residency in microseconds + * ARGS: 0 - Return IPS1 residency + * 1 - Return IPS2 residency + * 2 - Return IPS1_RCG residency + * 3 - Return IPS1_ONO2_ON residency + * RETURN: Total residency in microseconds - lower 32 bits + */ + DMUB_GPINT__GET_IPS_RESIDENCY_DURATION_US_LO = 124, + + /** + * DESC: Gets IPS1 histogram counts + * ARGS: Bucket index + * RETURN: Total count for the bucket + */ + DMUB_GPINT__GET_IPS1_HISTOGRAM_COUNTER = 125, + + /** + * DESC: Gets IPS2 histogram counts + * ARGS: Bucket index + * RETURN: Total count for the bucket + */ + DMUB_GPINT__GET_IPS2_HISTOGRAM_COUNTER = 126, + + /** + * DESC: Gets IPS residency + * ARGS: 0 - Return IPS1 residency + * 1 - Return IPS2 residency + * 2 - Return IPS1_RCG residency + * 3 - Return IPS1_ONO2_ON residency + * RETURN: Total residency in milli-percent. + */ + DMUB_GPINT__GET_IPS_RESIDENCY_PERCENT = 127, + + /** + * DESC: Gets IPS1_RCG histogram counts + * ARGS: Bucket index + * RETURN: Total count for the bucket + */ + DMUB_GPINT__GET_IPS1_RCG_HISTOGRAM_COUNTER = 128, + + /** + * DESC: Gets IPS1_ONO2_ON histogram counts + * ARGS: Bucket index + * RETURN: Total count for the bucket + */ + DMUB_GPINT__GET_IPS1_ONO2_ON_HISTOGRAM_COUNTER = 129, + + /** + * DESC: Gets IPS entry counter during residency measurement + * ARGS: 0 - Return IPS1 entry counts + * 1 - Return IPS2 entry counts + * 2 - Return IPS1_RCG entry counts + * 3 - Return IPS2_ONO2_ON entry counts + * RETURN: Entry counter for selected IPS mode + */ + DMUB_GPINT__GET_IPS_RESIDENCY_ENTRY_COUNTER = 130, + + /** + * DESC: Gets IPS inactive residency in microseconds + * ARGS: 0 - Return IPS1_MAX residency + * 1 - Return IPS2 residency + * 2 - Return IPS1_RCG residency + * 3 - Return IPS1_ONO2_ON residency + * RETURN: Total inactive residency in microseconds - lower 32 bits + */ + DMUB_GPINT__GET_IPS_INACTIVE_RESIDENCY_DURATION_US_LO = 131, + + /** + * DESC: Gets IPS inactive residency in microseconds + * ARGS: 0 - Return IPS1_MAX residency + * 1 - Return IPS2 residency + * 2 - Return IPS1_RCG residency + * 3 - Return IPS1_ONO2_ON residency + * RETURN: Total inactive residency in microseconds - upper 32 bits + */ + DMUB_GPINT__GET_IPS_INACTIVE_RESIDENCY_DURATION_US_HI = 132, + + /** + * DESC: Gets IPS residency in microseconds + * ARGS: 0 - Return IPS1 residency + * 1 - Return IPS2 residency + * 2 - Return IPS1_RCG residency + * 3 - Return IPS1_ONO2_ON residency + * RETURN: Total residency in microseconds - upper 32 bits + */ + DMUB_GPINT__GET_IPS_RESIDENCY_DURATION_US_HI = 133, + /** + * DESC: Setup debug configs. + */ + DMUB_GPINT__SETUP_DEBUG_MODE = 136, }; /** @@ -1306,9 +1423,10 @@ enum dmub_out_cmd_type { /* DMUB_CMD__DPIA command sub-types. */ enum dmub_cmd_dpia_type { DMUB_CMD__DPIA_DIG1_DPIA_CONTROL = 0, - DMUB_CMD__DPIA_SET_CONFIG_ACCESS = 1, + DMUB_CMD__DPIA_SET_CONFIG_ACCESS = 1, // will be replaced by DPIA_SET_CONFIG_REQUEST DMUB_CMD__DPIA_MST_ALLOC_SLOTS = 2, DMUB_CMD__DPIA_SET_TPS_NOTIFICATION = 3, + DMUB_CMD__DPIA_SET_CONFIG_REQUEST = 4, }; /* DMUB_OUT_CMD__DPIA_NOTIFICATION command types. */ @@ -2097,7 +2215,7 @@ struct dmub_rb_cmd_dig1_dpia_control { }; /** - * SET_CONFIG Command Payload + * SET_CONFIG Command Payload (deprecated) */ struct set_config_cmd_payload { uint8_t msg_type; /* set config message type */ @@ -2105,7 +2223,7 @@ struct set_config_cmd_payload { }; /** - * Data passed from driver to FW in a DMUB_CMD__DPIA_SET_CONFIG_ACCESS command. + * Data passed from driver to FW in a DMUB_CMD__DPIA_SET_CONFIG_ACCESS command. (deprecated) */ struct dmub_cmd_set_config_control_data { struct set_config_cmd_payload cmd_pkt; @@ -2113,6 +2231,17 @@ struct dmub_cmd_set_config_control_data { uint8_t immed_status; /* Immediate status returned in case of error */ }; +/** + * SET_CONFIG Request Command Payload + */ +struct set_config_request_cmd_payload { + uint8_t instance; /* DPIA instance */ + uint8_t immed_status; /* Immediate status returned in case of error */ + uint8_t msg_type; /* set config message type */ + uint8_t reserved; + uint32_t msg_data; /* set config message data */ +}; + /** * DMUB command structure for SET_CONFIG command. */ @@ -2121,6 +2250,14 @@ struct dmub_rb_cmd_set_config_access { struct dmub_cmd_set_config_control_data set_config_control; /* set config data */ }; +/** + * DMUB command structure for SET_CONFIG request command. + */ +struct dmub_rb_cmd_set_config_request { + struct dmub_cmd_header header; /* header */ + struct set_config_request_cmd_payload payload; /* set config request payload */ +}; + /** * Data passed from driver to FW in a DMUB_CMD__DPIA_MST_ALLOC_SLOTS command. */ @@ -4289,6 +4426,24 @@ struct dmub_rb_cmd_abm_set_pipe { struct dmub_cmd_abm_set_pipe_data abm_set_pipe_data; }; +/** + * Type of backlight control method to be used by ABM module + */ +enum dmub_backlight_control_type { + /** + * PWM Backlight control + */ + DMU_BACKLIGHT_CONTROL_PWM = 0, + /** + * VESA Aux-based backlight control + */ + DMU_BACKLIGHT_CONTROL_VESA_AUX = 1, + /** + * AMD DPCD Aux-based backlight control + */ + DMU_BACKLIGHT_CONTROL_AMD_AUX = 2, +}; + /** * Data passed from driver to FW in a DMUB_CMD__ABM_SET_BACKLIGHT command. */ @@ -4315,10 +4470,43 @@ struct dmub_cmd_abm_set_backlight_data { */ uint8_t panel_mask; + /** + * AUX HW Instance. + */ + uint8_t aux_inst; + /** * Explicit padding to 4 byte boundary. */ - uint8_t pad[2]; + uint8_t pad[1]; + + /** + * Backlight control type. + * Value 0 is PWM backlight control. + * Value 1 is VAUX backlight control. + * Value 2 is AMD DPCD AUX backlight control. + */ + enum dmub_backlight_control_type backlight_control_type; + + /** + * Minimum luminance in nits. + */ + uint32_t min_luminance; + + /** + * Maximum luminance in nits. + */ + uint32_t max_luminance; + + /** + * Minimum backlight in pwm. + */ + uint32_t min_backlight_pwm; + + /** + * Maximum backlight in pwm. + */ + uint32_t max_backlight_pwm; }; /** @@ -5022,7 +5210,34 @@ struct dmub_rb_cmd_get_usbc_cable_id { enum dmub_cmd_secure_display_type { DMUB_CMD__SECURE_DISPLAY_TEST_CMD = 0, /* test command to only check if inbox message works */ DMUB_CMD__SECURE_DISPLAY_CRC_STOP_UPDATE, - DMUB_CMD__SECURE_DISPLAY_CRC_WIN_NOTIFY + DMUB_CMD__SECURE_DISPLAY_CRC_WIN_NOTIFY, + DMUB_CMD__SECURE_DISPLAY_MULTIPLE_CRC_STOP_UPDATE, + DMUB_CMD__SECURE_DISPLAY_MULTIPLE_CRC_WIN_NOTIFY +}; + +#define MAX_ROI_NUM 2 + +struct dmub_cmd_roi_info { + uint16_t x_start; + uint16_t x_end; + uint16_t y_start; + uint16_t y_end; + uint8_t otg_id; + uint8_t phy_id; +}; + +struct dmub_cmd_roi_window_ctl { + uint16_t x_start; + uint16_t x_end; + uint16_t y_start; + uint16_t y_end; + bool enable; +}; + +struct dmub_cmd_roi_ctl_info { + uint8_t otg_id; + uint8_t phy_id; + struct dmub_cmd_roi_window_ctl roi_ctl[MAX_ROI_NUM]; }; /** @@ -5033,14 +5248,8 @@ struct dmub_rb_cmd_secure_display { /** * Data passed from driver to dmub firmware. */ - struct dmub_cmd_roi_info { - uint16_t x_start; - uint16_t x_end; - uint16_t y_start; - uint16_t y_end; - uint8_t otg_id; - uint8_t phy_id; - } roi_info; + struct dmub_cmd_roi_info roi_info; + struct dmub_cmd_roi_ctl_info mul_roi_ctl; }; /** @@ -5318,7 +5527,11 @@ union dmub_rb_cmd { /** * Definition of a DMUB_CMD__DPIA_SET_CONFIG_ACCESS command. */ - struct dmub_rb_cmd_set_config_access set_config_access; + struct dmub_rb_cmd_set_config_access set_config_access; // (deprecated) + /** + * Definition of a DMUB_CMD__DPIA_SET_CONFIG_ACCESS command. + */ + struct dmub_rb_cmd_set_config_request set_config_request; /** * Definition of a DMUB_CMD__DPIA_MST_ALLOC_SLOTS command. */ diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.c index 2ccad79053c5..e5e77bd3c31e 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.c +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.c @@ -426,6 +426,7 @@ void dmub_dcn35_enable_dmub_boot_options(struct dmub_srv *dmub, const struct dmu boot_options.bits.ips_sequential_ono = params->ips_sequential_ono; boot_options.bits.disable_sldo_opt = params->disable_sldo_opt; boot_options.bits.enable_non_transparent_setconfig = params->enable_non_transparent_setconfig; + boot_options.bits.lower_hbr3_phy_ssc = params->lower_hbr3_phy_ssc; REG_WRITE(DMCUB_SCRATCH14, boot_options.all); } @@ -463,7 +464,7 @@ uint32_t dmub_dcn35_get_current_time(struct dmub_srv *dmub) void dmub_dcn35_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnostic_data *diag_data) { - uint32_t is_dmub_enabled, is_soft_reset, is_sec_reset; + uint32_t is_dmub_enabled, is_soft_reset; uint32_t is_traceport_enabled, is_cw6_enabled; if (!dmub || !diag_data) @@ -513,9 +514,6 @@ void dmub_dcn35_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnosti REG_GET(DMCUB_CNTL2, DMCUB_SOFT_RESET, &is_soft_reset); diag_data->is_dmcub_soft_reset = is_soft_reset; - REG_GET(DMCUB_SEC_CNTL, DMCUB_SEC_RESET_STATUS, &is_sec_reset); - diag_data->is_dmcub_secure_reset = is_sec_reset; - REG_GET(DMCUB_CNTL, DMCUB_TRACEPORT_EN, &is_traceport_enabled); diag_data->is_traceport_en = is_traceport_enabled; diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c index db16066bc893..a3f3ff5d49ac 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c @@ -497,6 +497,7 @@ enum dmub_status const struct dmub_fw_meta_info *fw_info; uint32_t fw_state_size = DMUB_FW_STATE_SIZE; uint32_t trace_buffer_size = DMUB_TRACE_BUFFER_SIZE; + uint32_t shared_state_size = DMUB_FW_HEADER_SHARED_STATE_SIZE; uint32_t window_sizes[DMUB_WINDOW_TOTAL] = { 0 }; if (!dmub->sw_init) @@ -514,6 +515,7 @@ enum dmub_status fw_state_size = fw_info->fw_region_size; trace_buffer_size = fw_info->trace_buffer_size; + shared_state_size = fw_info->shared_state_size; /** * If DM didn't fill in a version, then fill it in based on @@ -534,7 +536,7 @@ enum dmub_status window_sizes[DMUB_WINDOW_5_TRACEBUFF] = trace_buffer_size; window_sizes[DMUB_WINDOW_6_FW_STATE] = fw_state_size; window_sizes[DMUB_WINDOW_7_SCRATCH_MEM] = DMUB_SCRATCH_MEM_SIZE; - window_sizes[DMUB_WINDOW_SHARED_STATE] = DMUB_FW_HEADER_SHARED_STATE_SIZE; + window_sizes[DMUB_WINDOW_SHARED_STATE] = max(DMUB_FW_HEADER_SHARED_STATE_SIZE, shared_state_size); out->fb_size = dmub_srv_calc_regions_for_memory_type(params, out, window_sizes, DMUB_WINDOW_MEMORY_TYPE_FB); diff --git a/drivers/gpu/drm/amd/display/include/dpcd_defs.h b/drivers/gpu/drm/amd/display/include/dpcd_defs.h index aee5170f5fb2..de8f3cfed6c8 100644 --- a/drivers/gpu/drm/amd/display/include/dpcd_defs.h +++ b/drivers/gpu/drm/amd/display/include/dpcd_defs.h @@ -164,18 +164,19 @@ enum dpcd_psr_sink_states { PSR_SINK_STATE_SINK_INTERNAL_ERROR = 7, }; -#define DP_SOURCE_SEQUENCE 0x30c -#define DP_SOURCE_TABLE_REVISION 0x310 -#define DP_SOURCE_PAYLOAD_SIZE 0x311 -#define DP_SOURCE_SINK_CAP 0x317 -#define DP_SOURCE_BACKLIGHT_LEVEL 0x320 -#define DP_SOURCE_BACKLIGHT_CURRENT_PEAK 0x326 -#define DP_SOURCE_BACKLIGHT_CONTROL 0x32E -#define DP_SOURCE_BACKLIGHT_ENABLE 0x32F -#define DP_SOURCE_MINIMUM_HBLANK_SUPPORTED 0x340 +#define DP_SOURCE_SEQUENCE 0x30C +#define DP_SOURCE_TABLE_REVISION 0x310 +#define DP_SOURCE_PAYLOAD_SIZE 0x311 +#define DP_SOURCE_SINK_CAP 0x317 +#define DP_SOURCE_BACKLIGHT_LEVEL 0x320 +#define DP_SOURCE_BACKLIGHT_CURRENT_PEAK 0x326 +#define DP_SOURCE_BACKLIGHT_CONTROL 0x32E +#define DP_SOURCE_BACKLIGHT_ENABLE 0x32F +#define DP_SOURCE_MINIMUM_HBLANK_SUPPORTED 0x340 #define DP_SINK_PR_REPLAY_STATUS 0x378 #define DP_SINK_PR_PIXEL_DEVIATION_PER_LINE 0x379 #define DP_SINK_PR_MAX_NUMBER_OF_DEVIATION_LINE 0x37A +#define DP_SINK_EMISSION_RATE 0x37E /* Remove once drm_dp_helper.h is updated upstream */ #ifndef DP_TOTAL_LTTPR_CNT diff --git a/drivers/gpu/drm/amd/display/include/logger_interface.h b/drivers/gpu/drm/amd/display/include/logger_interface.h index 02c23b04d34b..058f882d5bdd 100644 --- a/drivers/gpu/drm/amd/display/include/logger_interface.h +++ b/drivers/gpu/drm/amd/display/include/logger_interface.h @@ -52,10 +52,6 @@ void update_surface_trace( void post_surface_trace(struct dc *dc); -void context_timing_trace( - struct dc *dc, - struct resource_context *res_ctx); - void context_clock_trace( struct dc *dc, struct dc_state *context); diff --git a/drivers/gpu/drm/amd/display/include/logger_types.h b/drivers/gpu/drm/amd/display/include/logger_types.h index a48d564d1660..4d68c1c6e210 100644 --- a/drivers/gpu/drm/amd/display/include/logger_types.h +++ b/drivers/gpu/drm/amd/display/include/logger_types.h @@ -61,11 +61,13 @@ #define DC_LOG_ALL_TF_CHANNELS(...) pr_debug("[GAMMA]:"__VA_ARGS__) #define DC_LOG_DSC(...) drm_dbg_dp((DC_LOGGER)->dev, __VA_ARGS__) #define DC_LOG_SMU(...) pr_debug("[SMU_MSG]:"__VA_ARGS__) -#define DC_LOG_MALL(...) pr_debug("[MALL]:"__VA_ARGS__) #define DC_LOG_DWB(...) drm_dbg((DC_LOGGER)->dev, __VA_ARGS__) #define DC_LOG_DP2(...) drm_dbg_dp((DC_LOGGER)->dev, __VA_ARGS__) #define DC_LOG_AUTO_DPM_TEST(...) pr_debug("[AutoDPMTest]: "__VA_ARGS__) #define DC_LOG_IPS(...) pr_debug("[IPS]: "__VA_ARGS__) +#define DC_LOG_MALL(...) pr_debug("[MALL]:"__VA_ARGS__) +#define DC_LOG_REGISTER_READ(...) pr_debug("[REGISTER_READ]: "__VA_ARGS__) +#define DC_LOG_REGISTER_WRITE(...) pr_debug("[REGISTER_WRITE]: "__VA_ARGS__) struct dc_log_buffer_ctx { char *buf; diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c index 3699e633801d..a71df052cf25 100644 --- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c +++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c @@ -1399,71 +1399,6 @@ static void scale_gamma_dx(struct pwl_float_data *pwl_rgb, pwl_rgb[i-1].b, 2), pwl_rgb[i-2].b); } -/* todo: all these scale_gamma functions are inherently the same but - * take different structures as params or different format for ramp - * values. We could probably implement it in a more generic fashion - */ -static void scale_user_regamma_ramp(struct pwl_float_data *pwl_rgb, - const struct regamma_ramp *ramp, - struct dividers dividers) -{ - unsigned short max_driver = 0xFFFF; - unsigned short max_os = 0xFF00; - unsigned short scaler = max_os; - uint32_t i; - struct pwl_float_data *rgb = pwl_rgb; - struct pwl_float_data *rgb_last = rgb + GAMMA_RGB_256_ENTRIES - 1; - - i = 0; - do { - if (ramp->gamma[i] > max_os || - ramp->gamma[i + 256] > max_os || - ramp->gamma[i + 512] > max_os) { - scaler = max_driver; - break; - } - i++; - } while (i != GAMMA_RGB_256_ENTRIES); - - i = 0; - do { - rgb->r = dc_fixpt_from_fraction( - ramp->gamma[i], scaler); - rgb->g = dc_fixpt_from_fraction( - ramp->gamma[i + 256], scaler); - rgb->b = dc_fixpt_from_fraction( - ramp->gamma[i + 512], scaler); - - ++rgb; - ++i; - } while (i != GAMMA_RGB_256_ENTRIES); - - rgb->r = dc_fixpt_mul(rgb_last->r, - dividers.divider1); - rgb->g = dc_fixpt_mul(rgb_last->g, - dividers.divider1); - rgb->b = dc_fixpt_mul(rgb_last->b, - dividers.divider1); - - ++rgb; - - rgb->r = dc_fixpt_mul(rgb_last->r, - dividers.divider2); - rgb->g = dc_fixpt_mul(rgb_last->g, - dividers.divider2); - rgb->b = dc_fixpt_mul(rgb_last->b, - dividers.divider2); - - ++rgb; - - rgb->r = dc_fixpt_mul(rgb_last->r, - dividers.divider3); - rgb->g = dc_fixpt_mul(rgb_last->g, - dividers.divider3); - rgb->b = dc_fixpt_mul(rgb_last->b, - dividers.divider3); -} - /* * RS3+ color transform DDI - 1D LUT adjustment is composed with regamma here * Input is evenly distributed in the output color space as specified in @@ -1663,106 +1598,6 @@ static bool calculate_interpolated_hardware_curve( return true; } -/* The "old" interpolation uses a complicated scheme to build an array of - * coefficients while also using an array of 0-255 normalized to 0-1 - * Then there's another loop using both of the above + new scaled user ramp - * and we concatenate them. It also searches for points of interpolation and - * uses enums for positions. - * - * This function uses a different approach: - * user ramp is always applied on X with 0/255, 1/255, 2/255, ..., 255/255 - * To find index for hwX , we notice the following: - * i/255 <= hwX < (i+1)/255 <=> i <= 255*hwX < i+1 - * See apply_lut_1d which is the same principle, but on 4K entry 1D LUT - * - * Once the index is known, combined Y is simply: - * user_ramp(index) + (hwX-index/255)*(user_ramp(index+1) - user_ramp(index) - * - * We should switch to this method in all cases, it's simpler and faster - * ToDo one day - for now this only applies to ADL regamma to avoid regression - * for regular use cases (sRGB and PQ) - */ -static void interpolate_user_regamma(uint32_t hw_points_num, - struct pwl_float_data *rgb_user, - bool apply_degamma, - struct dc_transfer_func_distributed_points *tf_pts) -{ - uint32_t i; - uint32_t color = 0; - int32_t index; - int32_t index_next; - struct fixed31_32 *tf_point; - struct fixed31_32 hw_x; - struct fixed31_32 norm_factor = - dc_fixpt_from_int(255); - struct fixed31_32 norm_x; - struct fixed31_32 index_f; - struct fixed31_32 lut1; - struct fixed31_32 lut2; - struct fixed31_32 delta_lut; - struct fixed31_32 delta_index; - const struct fixed31_32 one = dc_fixpt_from_int(1); - - i = 0; - /* fixed_pt library has problems handling too small values */ - while (i != 32) { - tf_pts->red[i] = dc_fixpt_zero; - tf_pts->green[i] = dc_fixpt_zero; - tf_pts->blue[i] = dc_fixpt_zero; - ++i; - } - while (i <= hw_points_num + 1) { - for (color = 0; color < 3; color++) { - if (color == 0) - tf_point = &tf_pts->red[i]; - else if (color == 1) - tf_point = &tf_pts->green[i]; - else - tf_point = &tf_pts->blue[i]; - - if (apply_degamma) { - if (color == 0) - hw_x = coordinates_x[i].regamma_y_red; - else if (color == 1) - hw_x = coordinates_x[i].regamma_y_green; - else - hw_x = coordinates_x[i].regamma_y_blue; - } else - hw_x = coordinates_x[i].x; - - if (dc_fixpt_le(one, hw_x)) - hw_x = one; - - norm_x = dc_fixpt_mul(norm_factor, hw_x); - index = dc_fixpt_floor(norm_x); - if (index < 0 || index > 255) - continue; - - index_f = dc_fixpt_from_int(index); - index_next = (index == 255) ? index : index + 1; - - if (color == 0) { - lut1 = rgb_user[index].r; - lut2 = rgb_user[index_next].r; - } else if (color == 1) { - lut1 = rgb_user[index].g; - lut2 = rgb_user[index_next].g; - } else { - lut1 = rgb_user[index].b; - lut2 = rgb_user[index_next].b; - } - - // we have everything now, so interpolate - delta_lut = dc_fixpt_sub(lut2, lut1); - delta_index = dc_fixpt_sub(norm_x, index_f); - - *tf_point = dc_fixpt_add(lut1, - dc_fixpt_mul(delta_index, delta_lut)); - } - ++i; - } -} - static void build_new_custom_resulted_curve( uint32_t hw_points_num, struct dc_transfer_func_distributed_points *tf_pts) @@ -1784,29 +1619,6 @@ static void build_new_custom_resulted_curve( } } -static void apply_degamma_for_user_regamma(struct pwl_float_data_ex *rgb_regamma, - uint32_t hw_points_num, struct calculate_buffer *cal_buffer) -{ - uint32_t i; - - struct gamma_coefficients coeff; - struct pwl_float_data_ex *rgb = rgb_regamma; - const struct hw_x_point *coord_x = coordinates_x; - - build_coefficients(&coeff, TRANSFER_FUNCTION_SRGB); - - i = 0; - while (i != hw_points_num + 1) { - rgb->r = translate_from_linear_space_ex( - coord_x->x, &coeff, 0, cal_buffer); - rgb->g = rgb->r; - rgb->b = rgb->r; - ++coord_x; - ++rgb; - ++i; - } -} - static bool map_regamma_hw_to_x_user( const struct dc_gamma *ramp, struct pixel_gamma_point *coeff128, @@ -1855,125 +1667,6 @@ static bool map_regamma_hw_to_x_user( #define _EXTRA_POINTS 3 -bool calculate_user_regamma_coeff(struct dc_transfer_func *output_tf, - const struct regamma_lut *regamma, - struct calculate_buffer *cal_buffer, - const struct dc_gamma *ramp) -{ - struct gamma_coefficients coeff; - const struct hw_x_point *coord_x = coordinates_x; - uint32_t i = 0; - - do { - coeff.a0[i] = dc_fixpt_from_fraction( - regamma->coeff.A0[i], 10000000); - coeff.a1[i] = dc_fixpt_from_fraction( - regamma->coeff.A1[i], 1000); - coeff.a2[i] = dc_fixpt_from_fraction( - regamma->coeff.A2[i], 1000); - coeff.a3[i] = dc_fixpt_from_fraction( - regamma->coeff.A3[i], 1000); - coeff.user_gamma[i] = dc_fixpt_from_fraction( - regamma->coeff.gamma[i], 1000); - - ++i; - } while (i != 3); - - i = 0; - /* fixed_pt library has problems handling too small values */ - while (i != 32) { - output_tf->tf_pts.red[i] = dc_fixpt_zero; - output_tf->tf_pts.green[i] = dc_fixpt_zero; - output_tf->tf_pts.blue[i] = dc_fixpt_zero; - ++coord_x; - ++i; - } - while (i != MAX_HW_POINTS + 1) { - output_tf->tf_pts.red[i] = translate_from_linear_space_ex( - coord_x->x, &coeff, 0, cal_buffer); - output_tf->tf_pts.green[i] = translate_from_linear_space_ex( - coord_x->x, &coeff, 1, cal_buffer); - output_tf->tf_pts.blue[i] = translate_from_linear_space_ex( - coord_x->x, &coeff, 2, cal_buffer); - ++coord_x; - ++i; - } - - if (ramp && ramp->type == GAMMA_CS_TFM_1D) - apply_lut_1d(ramp, MAX_HW_POINTS, &output_tf->tf_pts); - - // this function just clamps output to 0-1 - build_new_custom_resulted_curve(MAX_HW_POINTS, &output_tf->tf_pts); - output_tf->type = TF_TYPE_DISTRIBUTED_POINTS; - - return true; -} - -bool calculate_user_regamma_ramp(struct dc_transfer_func *output_tf, - const struct regamma_lut *regamma, - struct calculate_buffer *cal_buffer, - const struct dc_gamma *ramp) -{ - struct dc_transfer_func_distributed_points *tf_pts = &output_tf->tf_pts; - struct dividers dividers; - - struct pwl_float_data *rgb_user = NULL; - struct pwl_float_data_ex *rgb_regamma = NULL; - bool ret = false; - - if (regamma == NULL) - return false; - - output_tf->type = TF_TYPE_DISTRIBUTED_POINTS; - - rgb_user = kcalloc(GAMMA_RGB_256_ENTRIES + _EXTRA_POINTS, - sizeof(*rgb_user), - GFP_KERNEL); - if (!rgb_user) - goto rgb_user_alloc_fail; - - rgb_regamma = kcalloc(MAX_HW_POINTS + _EXTRA_POINTS, - sizeof(*rgb_regamma), - GFP_KERNEL); - if (!rgb_regamma) - goto rgb_regamma_alloc_fail; - - dividers.divider1 = dc_fixpt_from_fraction(3, 2); - dividers.divider2 = dc_fixpt_from_int(2); - dividers.divider3 = dc_fixpt_from_fraction(5, 2); - - scale_user_regamma_ramp(rgb_user, ®amma->ramp, dividers); - - if (regamma->flags.bits.applyDegamma == 1) { - apply_degamma_for_user_regamma(rgb_regamma, MAX_HW_POINTS, cal_buffer); - copy_rgb_regamma_to_coordinates_x(coordinates_x, - MAX_HW_POINTS, rgb_regamma); - } - - interpolate_user_regamma(MAX_HW_POINTS, rgb_user, - regamma->flags.bits.applyDegamma, tf_pts); - - // no custom HDR curves! - tf_pts->end_exponent = 0; - tf_pts->x_point_at_y1_red = 1; - tf_pts->x_point_at_y1_green = 1; - tf_pts->x_point_at_y1_blue = 1; - - if (ramp && ramp->type == GAMMA_CS_TFM_1D) - apply_lut_1d(ramp, MAX_HW_POINTS, &output_tf->tf_pts); - - // this function just clamps output to 0-1 - build_new_custom_resulted_curve(MAX_HW_POINTS, tf_pts); - - ret = true; - - kfree(rgb_regamma); -rgb_regamma_alloc_fail: - kfree(rgb_user); -rgb_user_alloc_fail: - return ret; -} - bool mod_color_calculate_degamma_params(struct dc_color_caps *dc_caps, struct dc_transfer_func *input_tf, const struct dc_gamma *ramp, bool map_user_ramp) diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.h b/drivers/gpu/drm/amd/display/modules/color/color_gamma.h index ee5c466613de..97e55278940e 100644 --- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.h +++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.h @@ -115,15 +115,4 @@ bool mod_color_calculate_degamma_params(struct dc_color_caps *dc_caps, struct dc_transfer_func *output_tf, const struct dc_gamma *ramp, bool mapUserRamp); -bool calculate_user_regamma_coeff(struct dc_transfer_func *output_tf, - const struct regamma_lut *regamma, - struct calculate_buffer *cal_buffer, - const struct dc_gamma *ramp); - -bool calculate_user_regamma_ramp(struct dc_transfer_func *output_tf, - const struct regamma_lut *regamma, - struct calculate_buffer *cal_buffer, - const struct dc_gamma *ramp); - - #endif /* COLOR_MOD_COLOR_GAMMA_H_ */ diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c index bbd259cea4f4..f980a84dceef 100644 --- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c +++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c @@ -48,6 +48,7 @@ #define VSYNCS_BETWEEN_FLIP_THRESHOLD 2 #define FREESYNC_CONSEC_FLIP_AFTER_VSYNC 5 #define FREESYNC_VSYNC_TO_FLIP_DELTA_IN_US 500 +#define MICRO_HZ_TO_HZ(x) (x / 1000000) struct core_freesync { struct mod_freesync public; @@ -128,13 +129,26 @@ unsigned int mod_freesync_calc_v_total_from_refresh( unsigned int v_total; unsigned int frame_duration_in_ns; + if (refresh_in_uhz == 0) + return stream->timing.v_total; + frame_duration_in_ns = ((unsigned int)(div64_u64((1000000000ULL * 1000000), refresh_in_uhz))); - v_total = div64_u64(div64_u64(((unsigned long long)( - frame_duration_in_ns) * (stream->timing.pix_clk_100hz / 10)), - stream->timing.h_total) + 500000, 1000000); + if (MICRO_HZ_TO_HZ(refresh_in_uhz) <= stream->timing.min_refresh_in_uhz) { + /* When the target refresh rate is the minimum panel refresh rate, + * round down the vtotal value to avoid stretching vblank over + * panel's vtotal boundary. + */ + v_total = div64_u64(div64_u64(((unsigned long long)( + frame_duration_in_ns) * (stream->timing.pix_clk_100hz / 10)), + stream->timing.h_total), 1000000); + } else { + v_total = div64_u64(div64_u64(((unsigned long long)( + frame_duration_in_ns) * (stream->timing.pix_clk_100hz / 10)), + stream->timing.h_total) + 500000, 1000000); + } /* v_total cannot be less than nominal */ if (v_total < stream->timing.v_total) { diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c index c996365e84b0..1d41dd58f6bc 100644 --- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c +++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c @@ -27,6 +27,11 @@ #include "hdcp.h" +static inline uint16_t get_hdmi_rxstatus_msg_size(const uint8_t rxstatus[2]) +{ + return HDCP_2_2_HDMI_RXSTATUS_MSG_SZ_HI(rxstatus[1]) << 8 | rxstatus[0]; +} + static inline enum mod_hdcp_status check_receiver_id_list_ready(struct mod_hdcp *hdcp) { uint8_t is_ready = 0; @@ -35,8 +40,7 @@ static inline enum mod_hdcp_status check_receiver_id_list_ready(struct mod_hdcp is_ready = HDCP_2_2_DP_RXSTATUS_READY(hdcp->auth.msg.hdcp2.rxstatus_dp) ? 1 : 0; else is_ready = (HDCP_2_2_HDMI_RXSTATUS_READY(hdcp->auth.msg.hdcp2.rxstatus[1]) && - (HDCP_2_2_HDMI_RXSTATUS_MSG_SZ_HI(hdcp->auth.msg.hdcp2.rxstatus[1]) << 8 | - hdcp->auth.msg.hdcp2.rxstatus[0])) ? 1 : 0; + get_hdmi_rxstatus_msg_size(hdcp->auth.msg.hdcp2.rxstatus) != 0) ? 1 : 0; return is_ready ? MOD_HDCP_STATUS_SUCCESS : MOD_HDCP_STATUS_HDCP2_RX_ID_LIST_NOT_READY; } @@ -84,15 +88,13 @@ static inline enum mod_hdcp_status check_link_integrity_failure_dp( static enum mod_hdcp_status check_ake_cert_available(struct mod_hdcp *hdcp) { enum mod_hdcp_status status; - uint16_t size; if (is_dp_hdcp(hdcp)) { status = MOD_HDCP_STATUS_SUCCESS; } else { status = mod_hdcp_read_rxstatus(hdcp); if (status == MOD_HDCP_STATUS_SUCCESS) { - size = HDCP_2_2_HDMI_RXSTATUS_MSG_SZ_HI(hdcp->auth.msg.hdcp2.rxstatus[1]) << 8 | - hdcp->auth.msg.hdcp2.rxstatus[0]; + const uint16_t size = get_hdmi_rxstatus_msg_size(hdcp->auth.msg.hdcp2.rxstatus); status = (size == sizeof(hdcp->auth.msg.hdcp2.ake_cert)) ? MOD_HDCP_STATUS_SUCCESS : MOD_HDCP_STATUS_HDCP2_AKE_CERT_PENDING; @@ -104,7 +106,6 @@ static enum mod_hdcp_status check_ake_cert_available(struct mod_hdcp *hdcp) static enum mod_hdcp_status check_h_prime_available(struct mod_hdcp *hdcp) { enum mod_hdcp_status status; - uint8_t size; status = mod_hdcp_read_rxstatus(hdcp); if (status != MOD_HDCP_STATUS_SUCCESS) @@ -115,8 +116,7 @@ static enum mod_hdcp_status check_h_prime_available(struct mod_hdcp *hdcp) MOD_HDCP_STATUS_SUCCESS : MOD_HDCP_STATUS_HDCP2_H_PRIME_PENDING; } else { - size = HDCP_2_2_HDMI_RXSTATUS_MSG_SZ_HI(hdcp->auth.msg.hdcp2.rxstatus[1]) << 8 | - hdcp->auth.msg.hdcp2.rxstatus[0]; + const uint16_t size = get_hdmi_rxstatus_msg_size(hdcp->auth.msg.hdcp2.rxstatus); status = (size == sizeof(hdcp->auth.msg.hdcp2.ake_h_prime)) ? MOD_HDCP_STATUS_SUCCESS : MOD_HDCP_STATUS_HDCP2_H_PRIME_PENDING; @@ -128,7 +128,6 @@ out: static enum mod_hdcp_status check_pairing_info_available(struct mod_hdcp *hdcp) { enum mod_hdcp_status status; - uint8_t size; status = mod_hdcp_read_rxstatus(hdcp); if (status != MOD_HDCP_STATUS_SUCCESS) @@ -139,8 +138,7 @@ static enum mod_hdcp_status check_pairing_info_available(struct mod_hdcp *hdcp) MOD_HDCP_STATUS_SUCCESS : MOD_HDCP_STATUS_HDCP2_PAIRING_INFO_PENDING; } else { - size = HDCP_2_2_HDMI_RXSTATUS_MSG_SZ_HI(hdcp->auth.msg.hdcp2.rxstatus[1]) << 8 | - hdcp->auth.msg.hdcp2.rxstatus[0]; + const uint16_t size = get_hdmi_rxstatus_msg_size(hdcp->auth.msg.hdcp2.rxstatus); status = (size == sizeof(hdcp->auth.msg.hdcp2.ake_pairing_info)) ? MOD_HDCP_STATUS_SUCCESS : MOD_HDCP_STATUS_HDCP2_PAIRING_INFO_PENDING; @@ -152,7 +150,6 @@ out: static enum mod_hdcp_status poll_l_prime_available(struct mod_hdcp *hdcp) { enum mod_hdcp_status status = MOD_HDCP_STATUS_FAILURE; - uint8_t size; uint16_t max_wait = 20; // units of ms uint16_t num_polls = 5; uint16_t wait_time = max_wait / num_polls; @@ -167,8 +164,7 @@ static enum mod_hdcp_status poll_l_prime_available(struct mod_hdcp *hdcp) if (status != MOD_HDCP_STATUS_SUCCESS) break; - size = HDCP_2_2_HDMI_RXSTATUS_MSG_SZ_HI(hdcp->auth.msg.hdcp2.rxstatus[1]) << 8 | - hdcp->auth.msg.hdcp2.rxstatus[0]; + const uint16_t size = get_hdmi_rxstatus_msg_size(hdcp->auth.msg.hdcp2.rxstatus); status = (size == sizeof(hdcp->auth.msg.hdcp2.lc_l_prime)) ? MOD_HDCP_STATUS_SUCCESS : MOD_HDCP_STATUS_HDCP2_L_PRIME_PENDING; @@ -181,7 +177,6 @@ static enum mod_hdcp_status poll_l_prime_available(struct mod_hdcp *hdcp) static enum mod_hdcp_status check_stream_ready_available(struct mod_hdcp *hdcp) { enum mod_hdcp_status status; - uint8_t size; if (is_dp_hdcp(hdcp)) { status = MOD_HDCP_STATUS_INVALID_OPERATION; @@ -189,8 +184,7 @@ static enum mod_hdcp_status check_stream_ready_available(struct mod_hdcp *hdcp) status = mod_hdcp_read_rxstatus(hdcp); if (status != MOD_HDCP_STATUS_SUCCESS) goto out; - size = HDCP_2_2_HDMI_RXSTATUS_MSG_SZ_HI(hdcp->auth.msg.hdcp2.rxstatus[1]) << 8 | - hdcp->auth.msg.hdcp2.rxstatus[0]; + const uint16_t size = get_hdmi_rxstatus_msg_size(hdcp->auth.msg.hdcp2.rxstatus); status = (size == sizeof(hdcp->auth.msg.hdcp2.repeater_auth_stream_ready)) ? MOD_HDCP_STATUS_SUCCESS : MOD_HDCP_STATUS_HDCP2_STREAM_READY_PENDING; @@ -249,8 +243,7 @@ static uint8_t process_rxstatus(struct mod_hdcp *hdcp, sizeof(hdcp->auth.msg.hdcp2.rx_id_list); else hdcp->auth.msg.hdcp2.rx_id_list_size = - HDCP_2_2_HDMI_RXSTATUS_MSG_SZ_HI(hdcp->auth.msg.hdcp2.rxstatus[1]) << 8 | - hdcp->auth.msg.hdcp2.rxstatus[0]; + get_hdmi_rxstatus_msg_size(hdcp->auth.msg.hdcp2.rxstatus); } out: return (*status == MOD_HDCP_STATUS_SUCCESS); diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h index 3f91926a50e9..7eefcb0f5070 100644 --- a/drivers/gpu/drm/amd/include/amd_shared.h +++ b/drivers/gpu/drm/amd/include/amd_shared.h @@ -28,6 +28,8 @@ #define AMD_MAX_USEC_TIMEOUT 1000000 /* 1000 ms */ +struct amdgpu_ip_block; + /* * Chip flags @@ -337,6 +339,11 @@ enum DC_DEBUG_MASK { * @DC_FORCE_IPS_ENABLE: If set, force enable all IPS, all the time. */ DC_FORCE_IPS_ENABLE = 0x4000, + /** + * @DC_DISABLE_ACPI_EDID: If set, don't attempt to fetch EDID for + * eDP display from ACPI _DDC method. + */ + DC_DISABLE_ACPI_EDID = 0x8000, }; enum amd_dpm_forced_level; @@ -377,30 +384,30 @@ enum amd_dpm_forced_level; */ struct amd_ip_funcs { char *name; - int (*early_init)(void *handle); - int (*late_init)(void *handle); - int (*sw_init)(void *handle); - int (*sw_fini)(void *handle); - int (*early_fini)(void *handle); - int (*hw_init)(void *handle); - int (*hw_fini)(void *handle); - void (*late_fini)(void *handle); - int (*prepare_suspend)(void *handle); - int (*suspend)(void *handle); - int (*resume)(void *handle); + int (*early_init)(struct amdgpu_ip_block *ip_block); + int (*late_init)(struct amdgpu_ip_block *ip_block); + int (*sw_init)(struct amdgpu_ip_block *ip_block); + int (*sw_fini)(struct amdgpu_ip_block *ip_block); + int (*early_fini)(struct amdgpu_ip_block *ip_block); + int (*hw_init)(struct amdgpu_ip_block *ip_block); + int (*hw_fini)(struct amdgpu_ip_block *ip_block); + void (*late_fini)(struct amdgpu_ip_block *ip_block); + int (*prepare_suspend)(struct amdgpu_ip_block *ip_block); + int (*suspend)(struct amdgpu_ip_block *ip_block); + int (*resume)(struct amdgpu_ip_block *ip_block); bool (*is_idle)(void *handle); - int (*wait_for_idle)(void *handle); - bool (*check_soft_reset)(void *handle); - int (*pre_soft_reset)(void *handle); - int (*soft_reset)(void *handle); - int (*post_soft_reset)(void *handle); + int (*wait_for_idle)(struct amdgpu_ip_block *ip_block); + bool (*check_soft_reset)(struct amdgpu_ip_block *ip_block); + int (*pre_soft_reset)(struct amdgpu_ip_block *ip_block); + int (*soft_reset)(struct amdgpu_ip_block *ip_block); + int (*post_soft_reset)(struct amdgpu_ip_block *ip_block); int (*set_clockgating_state)(void *handle, enum amd_clockgating_state state); int (*set_powergating_state)(void *handle, enum amd_powergating_state state); void (*get_clockgating_state)(void *handle, u64 *flags); - void (*dump_ip_state)(void *handle); - void (*print_ip_state)(void *handle, struct drm_printer *p); + void (*dump_ip_state)(struct amdgpu_ip_block *ip_block); + void (*print_ip_state)(struct amdgpu_ip_block *ip_block, struct drm_printer *p); }; diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_4_1_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_4_1_0_sh_mask.h index f42a276499cd..5d9d5fea6e06 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_4_1_0_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_4_1_0_sh_mask.h @@ -6199,10 +6199,12 @@ #define DCHUBBUB_CTRL_STATUS__ROB_UNDERFLOW_STATUS__SHIFT 0x1 #define DCHUBBUB_CTRL_STATUS__ROB_OVERFLOW_STATUS__SHIFT 0x2 #define DCHUBBUB_CTRL_STATUS__ROB_OVERFLOW_CLEAR__SHIFT 0x3 +#define DCHUBBUB_CTRL_STATUS__DCHUBBUB_HW_DEBUG__SHIFT 0x4 #define DCHUBBUB_CTRL_STATUS__CSTATE_SWATH_CHK_GOOD_MODE__SHIFT 0x1f #define DCHUBBUB_CTRL_STATUS__ROB_UNDERFLOW_STATUS_MASK 0x00000002L #define DCHUBBUB_CTRL_STATUS__ROB_OVERFLOW_STATUS_MASK 0x00000004L #define DCHUBBUB_CTRL_STATUS__ROB_OVERFLOW_CLEAR_MASK 0x00000008L +#define DCHUBBUB_CTRL_STATUS__DCHUBBUB_HW_DEBUG_MASK 0x3FFFFFF0L #define DCHUBBUB_CTRL_STATUS__CSTATE_SWATH_CHK_GOOD_MODE_MASK 0x80000000L //DCHUBBUB_TIMEOUT_DETECTION_CTRL1 #define DCHUBBUB_TIMEOUT_DETECTION_CTRL1__DCHUBBUB_TIMEOUT_ERROR_STATUS__SHIFT 0x0 diff --git a/drivers/gpu/drm/amd/include/asic_reg/mmhub/mmhub_1_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/mmhub/mmhub_1_0_offset.h index 2c3ce243861a..380e44230bda 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/mmhub/mmhub_1_0_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/mmhub/mmhub_1_0_offset.h @@ -1232,6 +1232,29 @@ #define mmMC_VM_MX_L1_PERFCOUNTER_HI 0x059d #define mmMC_VM_MX_L1_PERFCOUNTER_HI_BASE_IDX 0 +// Stand Alone Walker Registers +#define VMC_TAP_PDE_REQUEST_SNOOP_OFFSET 8 +#define VMC_TAP_PTE_REQUEST_SNOOP_OFFSET 11 +#define mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32 0x0606 +#define mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32_BASE_IDX 0 +#define mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32 0x0607 +#define mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32_BASE_IDX 0 +#define mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_START_ADDR_LO32 0x0608 +#define mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_START_ADDR_LO32_BASE_IDX 0 +#define mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_START_ADDR_HI32 0x0609 +#define mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_START_ADDR_HI32_BASE_IDX 0 +#define mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_END_ADDR_LO32 0x060a +#define mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_END_ADDR_LO32_BASE_IDX 0 +#define mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_END_ADDR_HI32 0x060b +#define mmVM_L2_SAW_CONTEXT0_PAGE_TABLE_END_ADDR_HI32_BASE_IDX 0 +#define mmVM_L2_SAW_CONTEXT0_CNTL 0x0604 +#define mmVM_L2_SAW_CONTEXT0_CNTL_BASE_IDX 0 +#define CONTEXT0_CNTL_ENABLE_OFFSET 0 +#define CONTEXT0_CNTL_PAGE_TABLE_DEPTH_OFFSET 1 +#define mmVM_L2_SAW_CONTEXTS_DISABLE 0x060c +#define mmVM_L2_SAW_CONTEXTS_DISABLE_BASE_IDX 0 +#define mmVM_L2_SAW_CNTL4 0x0603 +#define mmVM_L2_SAW_CNTL4_BASE_IDX 0 // addressBlock: mmhub_utcl2_atcl2dec // base address: 0x69900 diff --git a/drivers/gpu/drm/amd/include/kgd_pp_interface.h b/drivers/gpu/drm/amd/include/kgd_pp_interface.h index 19a48d98830a..bb27c0d2a9ae 100644 --- a/drivers/gpu/drm/amd/include/kgd_pp_interface.h +++ b/drivers/gpu/drm/amd/include/kgd_pp_interface.h @@ -119,6 +119,8 @@ enum pp_clock_type { OD_ACOUSTIC_TARGET, OD_FAN_TARGET_TEMPERATURE, OD_FAN_MINIMUM_PWM, + OD_FAN_ZERO_RPM_ENABLE, + OD_FAN_ZERO_RPM_STOP_TEMP, }; enum amd_pp_sensors { @@ -199,6 +201,8 @@ enum PP_OD_DPM_TABLE_COMMAND { PP_OD_EDIT_ACOUSTIC_TARGET, PP_OD_EDIT_FAN_TARGET_TEMPERATURE, PP_OD_EDIT_FAN_MINIMUM_PWM, + PP_OD_EDIT_FAN_ZERO_RPM_ENABLE, + PP_OD_EDIT_FAN_ZERO_RPM_STOP_TEMP, }; struct pp_states_info { @@ -336,7 +340,8 @@ enum pp_policy_soc_pstate { #define MAX_CLKS 4 #define NUM_VCN 4 #define NUM_JPEG_ENG 32 - +#define MAX_XCC 8 +#define NUM_XCP 8 struct seq_file; enum amd_pp_clock_type; struct amd_pp_simple_clock_info; @@ -350,6 +355,15 @@ struct pp_smu_wm_range_sets; struct pp_smu_nv_clock_table; struct dpm_clocks; +struct amdgpu_xcp_metrics { + /* Utilization Instantaneous (%) */ + u32 gfx_busy_inst[MAX_XCC]; + u16 jpeg_busy[NUM_JPEG_ENG]; + u16 vcn_busy[NUM_VCN]; + /* Utilization Accumulated (%) */ + u64 gfx_busy_acc[MAX_XCC]; +}; + struct amd_pm_funcs { /* export for dpm on ci and si */ int (*pre_set_power_state)(void *handle); @@ -872,6 +886,97 @@ struct gpu_metrics_v1_5 { uint16_t padding; }; +struct gpu_metrics_v1_6 { + struct metrics_table_header common_header; + + /* Temperature (Celsius) */ + uint16_t temperature_hotspot; + uint16_t temperature_mem; + uint16_t temperature_vrsoc; + + /* Power (Watts) */ + uint16_t curr_socket_power; + + /* Utilization (%) */ + uint16_t average_gfx_activity; + uint16_t average_umc_activity; // memory controller + + /* Energy (15.259uJ (2^-16) units) */ + uint64_t energy_accumulator; + + /* Driver attached timestamp (in ns) */ + uint64_t system_clock_counter; + + /* Accumulation cycle counter */ + uint32_t accumulation_counter; + + /* Accumulated throttler residencies */ + uint32_t prochot_residency_acc; + uint32_t ppt_residency_acc; + uint32_t socket_thm_residency_acc; + uint32_t vr_thm_residency_acc; + uint32_t hbm_thm_residency_acc; + + /* Clock Lock Status. Each bit corresponds to clock instance */ + uint32_t gfxclk_lock_status; + + /* Link width (number of lanes) and speed (in 0.1 GT/s) */ + uint16_t pcie_link_width; + uint16_t pcie_link_speed; + + /* XGMI bus width and bitrate (in Gbps) */ + uint16_t xgmi_link_width; + uint16_t xgmi_link_speed; + + /* Utilization Accumulated (%) */ + uint32_t gfx_activity_acc; + uint32_t mem_activity_acc; + + /*PCIE accumulated bandwidth (GB/sec) */ + uint64_t pcie_bandwidth_acc; + + /*PCIE instantaneous bandwidth (GB/sec) */ + uint64_t pcie_bandwidth_inst; + + /* PCIE L0 to recovery state transition accumulated count */ + uint64_t pcie_l0_to_recov_count_acc; + + /* PCIE replay accumulated count */ + uint64_t pcie_replay_count_acc; + + /* PCIE replay rollover accumulated count */ + uint64_t pcie_replay_rover_count_acc; + + /* PCIE NAK sent accumulated count */ + uint32_t pcie_nak_sent_count_acc; + + /* PCIE NAK received accumulated count */ + uint32_t pcie_nak_rcvd_count_acc; + + /* XGMI accumulated data transfer size(KiloBytes) */ + uint64_t xgmi_read_data_acc[NUM_XGMI_LINKS]; + uint64_t xgmi_write_data_acc[NUM_XGMI_LINKS]; + + /* PMFW attached timestamp (10ns resolution) */ + uint64_t firmware_timestamp; + + /* Current clocks (Mhz) */ + uint16_t current_gfxclk[MAX_GFX_CLKS]; + uint16_t current_socclk[MAX_CLKS]; + uint16_t current_vclk0[MAX_CLKS]; + uint16_t current_dclk0[MAX_CLKS]; + uint16_t current_uclk; + + /* Number of current partition */ + uint16_t num_partition; + + /* XCP metrics stats */ + struct amdgpu_xcp_metrics xcp_stats[NUM_XCP]; + + /* PCIE other end recovery counter */ + uint32_t pcie_lc_perf_other_end_recovery; +}; + /* * gpu_metrics_v2_0 is not recommended as it's not naturally aligned. * Use gpu_metrics_v2_1 or later instead. diff --git a/drivers/gpu/drm/amd/include/mes_v11_api_def.h b/drivers/gpu/drm/amd/include/mes_v11_api_def.h index 21ceafce1f9b..eb46cb10c24d 100644 --- a/drivers/gpu/drm/amd/include/mes_v11_api_def.h +++ b/drivers/gpu/drm/amd/include/mes_v11_api_def.h @@ -230,13 +230,23 @@ union MESAPI_SET_HW_RESOURCES { uint32_t disable_add_queue_wptr_mc_addr : 1; uint32_t enable_mes_event_int_logging : 1; uint32_t enable_reg_active_poll : 1; - uint32_t reserved : 21; + uint32_t use_disable_queue_in_legacy_uq_preemption : 1; + uint32_t send_write_data : 1; + uint32_t os_tdr_timeout_override : 1; + uint32_t use_rs64mem_for_proc_gang_ctx : 1; + uint32_t use_add_queue_unmap_flag_addr : 1; + uint32_t enable_mes_sch_stb_log : 1; + uint32_t limit_single_process : 1; + uint32_t is_strix_tmz_wa_enabled :1; + uint32_t reserved : 13; }; uint32_t uint32_t_all; }; uint32_t oversubscription_timer; uint64_t doorbell_info; uint64_t event_intr_history_gpu_mc_ptr; + uint64_t timestamp; + uint32_t os_tdr_timeout_in_sec; }; uint32_t max_dwords_in_api[API_FRAME_SIZE_IN_DWORDS]; @@ -563,6 +573,11 @@ enum MESAPI_MISC_OPCODE { MESAPI_MISC__READ_REG, MESAPI_MISC__WAIT_REG_MEM, MESAPI_MISC__SET_SHADER_DEBUGGER, + MESAPI_MISC__NOTIFY_WORK_ON_UNMAPPED_QUEUE, + MESAPI_MISC__NOTIFY_TO_UNMAP_PROCESSES, + MESAPI_MISC__CHANGE_CONFIG, + MESAPI_MISC__LAUNCH_CLEANER_SHADER, + MESAPI_MISC__MAX, }; @@ -617,6 +632,31 @@ struct SET_SHADER_DEBUGGER { uint32_t trap_en; }; +enum MESAPI_MISC__CHANGE_CONFIG_OPTION { + MESAPI_MISC__CHANGE_CONFIG_OPTION_LIMIT_SINGLE_PROCESS = 0, + MESAPI_MISC__CHANGE_CONFIG_OPTION_ENABLE_HWS_LOGGING_BUFFER = 1, + MESAPI_MISC__CHANGE_CONFIG_OPTION_CHANGE_TDR_CONFIG = 2, + + MESAPI_MISC__CHANGE_CONFIG_OPTION_MAX = 0x1F +}; + +struct CHANGE_CONFIG { + enum MESAPI_MISC__CHANGE_CONFIG_OPTION opcode; + union { + struct { + uint32_t limit_single_process : 1; + uint32_t enable_hws_logging_buffer : 1; + uint32_t reserved : 31; + } bits; + uint32_t all; + } option; + + struct { + uint32_t tdr_level; + uint32_t tdr_delay; + } tdr_config; +}; + union MESAPI__MISC { struct { union MES_API_HEADER header; @@ -631,6 +671,7 @@ union MESAPI__MISC { struct WAIT_REG_MEM wait_reg_mem; struct SET_SHADER_DEBUGGER set_shader_debugger; enum MES_AMD_PRIORITY_LEVEL queue_sch_level; + struct CHANGE_CONFIG change_config; uint32_t data[MISC_DATA_MAX_SIZE_IN_DWORDS]; }; diff --git a/drivers/gpu/drm/amd/include/mes_v12_api_def.h b/drivers/gpu/drm/amd/include/mes_v12_api_def.h index 101e2fe962c6..c9b2ca5cf75f 100644 --- a/drivers/gpu/drm/amd/include/mes_v12_api_def.h +++ b/drivers/gpu/drm/amd/include/mes_v12_api_def.h @@ -643,6 +643,10 @@ enum MESAPI_MISC_OPCODE { MESAPI_MISC__SET_SHADER_DEBUGGER, MESAPI_MISC__NOTIFY_WORK_ON_UNMAPPED_QUEUE, MESAPI_MISC__NOTIFY_TO_UNMAP_PROCESSES, + MESAPI_MISC__QUERY_HUNG_ENGINE_ID, + MESAPI_MISC__CHANGE_CONFIG, + MESAPI_MISC__LAUNCH_CLEANER_SHADER, + MESAPI_MISC__SETUP_MES_DBGEXT, MESAPI_MISC__MAX, }; @@ -713,6 +717,31 @@ struct SET_GANG_SUBMIT { uint32_t slave_gang_context_array_index; }; +enum MESAPI_MISC__CHANGE_CONFIG_OPTION { + MESAPI_MISC__CHANGE_CONFIG_OPTION_LIMIT_SINGLE_PROCESS = 0, + MESAPI_MISC__CHANGE_CONFIG_OPTION_ENABLE_HWS_LOGGING_BUFFER = 1, + MESAPI_MISC__CHANGE_CONFIG_OPTION_CHANGE_TDR_CONFIG = 2, + + MESAPI_MISC__CHANGE_CONFIG_OPTION_MAX = 0x1F +}; + +struct CHANGE_CONFIG { + enum MESAPI_MISC__CHANGE_CONFIG_OPTION opcode; + union { + struct { + uint32_t limit_single_process : 1; + uint32_t enable_hws_logging_buffer : 1; + uint32_t reserved : 30; + } bits; + uint32_t all; + } option; + + struct { + uint32_t tdr_level; + uint32_t tdr_delay; + } tdr_config; +}; + union MESAPI__MISC { struct { union MES_API_HEADER header; @@ -726,7 +755,7 @@ union MESAPI__MISC { struct WAIT_REG_MEM wait_reg_mem; struct SET_SHADER_DEBUGGER set_shader_debugger; enum MES_AMD_PRIORITY_LEVEL queue_sch_level; - + struct CHANGE_CONFIG change_config; uint32_t data[MISC_DATA_MAX_SIZE_IN_DWORDS]; }; uint64_t timestamp; diff --git a/drivers/gpu/drm/amd/pm/amdgpu_pm.c b/drivers/gpu/drm/amd/pm/amdgpu_pm.c index d5d6ab484e5a..136e8193867c 100644 --- a/drivers/gpu/drm/amd/pm/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/pm/amdgpu_pm.c @@ -145,15 +145,12 @@ static ssize_t amdgpu_get_power_dpm_state(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; amdgpu_dpm_get_current_power_state(adev, &pm); - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); return sysfs_emit(buf, "%s\n", @@ -185,11 +182,9 @@ static ssize_t amdgpu_set_power_dpm_state(struct device *dev, else return -EINVAL; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); + ret = pm_runtime_resume_and_get(ddev->dev); + if (ret < 0) return ret; - } amdgpu_dpm_set_power_state(adev, state); @@ -273,15 +268,12 @@ static ssize_t amdgpu_get_power_dpm_force_performance_level(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; level = amdgpu_dpm_get_performance_level(adev); - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); return sysfs_emit(buf, "%s\n", @@ -336,11 +328,9 @@ static ssize_t amdgpu_set_power_dpm_force_performance_level(struct device *dev, return -EINVAL; } - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); + ret = pm_runtime_resume_and_get(ddev->dev); + if (ret < 0) return ret; - } mutex_lock(&adev->pm.stable_pstate_ctx_lock); if (amdgpu_dpm_force_performance_level(adev, level)) { @@ -374,16 +364,13 @@ static ssize_t amdgpu_get_pp_num_states(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; if (amdgpu_dpm_get_pp_num_states(adev, &data)) memset(&data, 0, sizeof(data)); - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); buf_len = sysfs_emit(buf, "states: %d\n", data.nums); @@ -412,17 +399,14 @@ static ssize_t amdgpu_get_pp_cur_state(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; amdgpu_dpm_get_current_power_state(adev, &pm); ret = amdgpu_dpm_get_pp_num_states(adev, &data); - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); if (ret) @@ -485,11 +469,9 @@ static ssize_t amdgpu_set_pp_force_state(struct device *dev, idx = array_index_nospec(idx, ARRAY_SIZE(data.states)); - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); + ret = pm_runtime_resume_and_get(ddev->dev); + if (ret < 0) return ret; - } ret = amdgpu_dpm_get_pp_num_states(adev, &data); if (ret) @@ -544,15 +526,12 @@ static ssize_t amdgpu_get_pp_table(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; size = amdgpu_dpm_get_pp_table(adev, &table); - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); if (size <= 0) @@ -580,11 +559,9 @@ static ssize_t amdgpu_set_pp_table(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); + ret = pm_runtime_resume_and_get(ddev->dev); + if (ret < 0) return ret; - } ret = amdgpu_dpm_set_pp_table(adev, buf, count); @@ -808,11 +785,9 @@ static ssize_t amdgpu_set_pp_od_clk_voltage(struct device *dev, tmp_str++; } - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); + ret = pm_runtime_resume_and_get(ddev->dev); + if (ret < 0) return ret; - } if (amdgpu_dpm_set_fine_grain_clk_vol(adev, type, @@ -865,11 +840,9 @@ static ssize_t amdgpu_get_pp_od_clk_voltage(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; for (clk_index = 0 ; clk_index < 6 ; clk_index++) { ret = amdgpu_dpm_emit_clock_levels(adev, od_clocks[clk_index], buf, &size); @@ -888,7 +861,6 @@ static ssize_t amdgpu_get_pp_od_clk_voltage(struct device *dev, if (size == 0) size = sysfs_emit(buf, "\n"); - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); return size; @@ -929,11 +901,9 @@ static ssize_t amdgpu_set_pp_features(struct device *dev, if (ret) return -EINVAL; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); + ret = pm_runtime_resume_and_get(ddev->dev); + if (ret < 0) return ret; - } ret = amdgpu_dpm_set_ppfeature_status(adev, featuremask); @@ -960,17 +930,14 @@ static ssize_t amdgpu_get_pp_features(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; size = amdgpu_dpm_get_ppfeature_status(adev, buf); if (size <= 0) size = sysfs_emit(buf, "\n"); - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); return size; @@ -1029,11 +996,9 @@ static ssize_t amdgpu_get_pp_dpm_clock(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; ret = amdgpu_dpm_emit_clock_levels(adev, type, buf, &size); if (ret == -ENOENT) @@ -1042,7 +1007,6 @@ static ssize_t amdgpu_get_pp_dpm_clock(struct device *dev, if (size == 0) size = sysfs_emit(buf, "\n"); - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); return size; @@ -1102,11 +1066,9 @@ static ssize_t amdgpu_set_pp_dpm_clock(struct device *dev, if (ret) return ret; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); + ret = pm_runtime_resume_and_get(ddev->dev); + if (ret < 0) return ret; - } ret = amdgpu_dpm_force_clock_level(adev, type, mask); @@ -1283,15 +1245,12 @@ static ssize_t amdgpu_get_pp_sclk_od(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; value = amdgpu_dpm_get_sclk_od(adev); - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); return sysfs_emit(buf, "%d\n", value); @@ -1317,11 +1276,9 @@ static ssize_t amdgpu_set_pp_sclk_od(struct device *dev, if (ret) return -EINVAL; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); + ret = pm_runtime_resume_and_get(ddev->dev); + if (ret < 0) return ret; - } amdgpu_dpm_set_sclk_od(adev, (uint32_t)value); @@ -1345,15 +1302,12 @@ static ssize_t amdgpu_get_pp_mclk_od(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; value = amdgpu_dpm_get_mclk_od(adev); - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); return sysfs_emit(buf, "%d\n", value); @@ -1379,11 +1333,9 @@ static ssize_t amdgpu_set_pp_mclk_od(struct device *dev, if (ret) return -EINVAL; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); + ret = pm_runtime_resume_and_get(ddev->dev); + if (ret < 0) return ret; - } amdgpu_dpm_set_mclk_od(adev, (uint32_t)value); @@ -1427,17 +1379,14 @@ static ssize_t amdgpu_get_pp_power_profile_mode(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; size = amdgpu_dpm_get_power_profile_mode(adev, buf); if (size <= 0) size = sysfs_emit(buf, "\n"); - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); return size; @@ -1492,11 +1441,9 @@ static ssize_t amdgpu_set_pp_power_profile_mode(struct device *dev, } parameter[parameter_size] = profile_mode; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); + ret = pm_runtime_resume_and_get(ddev->dev); + if (ret < 0) return ret; - } ret = amdgpu_dpm_set_power_profile_mode(adev, parameter, parameter_size); @@ -1520,16 +1467,13 @@ static int amdgpu_hwmon_get_sensor_generic(struct amdgpu_device *adev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - r = pm_runtime_get_sync(adev_to_drm(adev)->dev); - if (r < 0) { - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); - return r; - } + r = pm_runtime_get_if_active(adev->dev); + if (r <= 0) + return r ?: -EPERM; /* get the sensor value */ r = amdgpu_dpm_read_sensor(adev, sensor, query, &size); - pm_runtime_mark_last_busy(adev_to_drm(adev)->dev); pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); return r; @@ -1639,15 +1583,12 @@ static ssize_t amdgpu_get_pcie_bw(struct device *dev, if (!adev->asic_funcs->get_pcie_usage) return -ENODATA; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; amdgpu_asic_get_pcie_usage(adev, &count0, &count1); - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); return sysfs_emit(buf, "%llu %llu %i\n", @@ -1770,11 +1711,9 @@ static ssize_t amdgpu_get_apu_thermal_cap(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = drm_to_adev(ddev); - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; ret = amdgpu_dpm_get_apu_thermal_limit(adev, &limit); if (!ret) @@ -1782,7 +1721,6 @@ static ssize_t amdgpu_get_apu_thermal_cap(struct device *dev, else size = sysfs_emit(buf, "failed to get thermal limit\n"); - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); return size; @@ -1807,14 +1745,14 @@ static ssize_t amdgpu_set_apu_thermal_cap(struct device *dev, return -EINVAL; } - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); + ret = pm_runtime_resume_and_get(ddev->dev); + if (ret < 0) return ret; - } ret = amdgpu_dpm_set_apu_thermal_limit(adev, value); if (ret) { + pm_runtime_mark_last_busy(ddev->dev); + pm_runtime_put_autosuspend(ddev->dev); dev_err(dev, "failed to update thermal limit\n"); return ret; } @@ -1849,15 +1787,12 @@ static ssize_t amdgpu_get_pm_metrics(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; size = amdgpu_dpm_get_pm_metrics(adev, buf, PAGE_SIZE); - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); return size; @@ -1890,11 +1825,9 @@ static ssize_t amdgpu_get_gpu_metrics(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); - return ret; - } + ret = pm_runtime_get_if_active(ddev->dev); + if (ret <= 0) + return ret ?: -EPERM; size = amdgpu_dpm_get_gpu_metrics(adev, &gpu_metrics); if (size <= 0) @@ -1906,7 +1839,6 @@ static ssize_t amdgpu_get_gpu_metrics(struct device *dev, memcpy(buf, gpu_metrics, size); out: - pm_runtime_mark_last_busy(ddev->dev); pm_runtime_put_autosuspend(ddev->dev); return size; @@ -2008,11 +1940,9 @@ static ssize_t amdgpu_set_smartshift_bias(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - r = pm_runtime_get_sync(ddev->dev); - if (r < 0) { - pm_runtime_put_autosuspend(ddev->dev); + r = pm_runtime_resume_and_get(ddev->dev); + if (r < 0) return r; - } r = kstrtoint(buf, 10, &bias); if (r) @@ -2335,11 +2265,9 @@ static ssize_t amdgpu_set_pm_policy_attr(struct device *dev, policy_attr = container_of(attr, struct amdgpu_pm_policy_attr, dev_attr); - ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(ddev->dev); + ret = pm_runtime_resume_and_get(ddev->dev); + if (ret < 0) return ret; - } ret = amdgpu_dpm_set_pm_policy(adev, policy_attr->id, val); @@ -2772,15 +2700,12 @@ static ssize_t amdgpu_hwmon_get_pwm1_enable(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(adev_to_drm(adev)->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); - return ret; - } + ret = pm_runtime_get_if_active(adev->dev); + if (ret <= 0) + return ret ?: -EPERM; ret = amdgpu_dpm_get_fan_control_mode(adev, &pwm_mode); - pm_runtime_mark_last_busy(adev_to_drm(adev)->dev); pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); if (ret) @@ -2817,11 +2742,9 @@ static ssize_t amdgpu_hwmon_set_pwm1_enable(struct device *dev, else return -EINVAL; - ret = pm_runtime_get_sync(adev_to_drm(adev)->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); + ret = pm_runtime_resume_and_get(adev->dev); + if (ret < 0) return ret; - } ret = amdgpu_dpm_set_fan_control_mode(adev, pwm_mode); @@ -2866,11 +2789,9 @@ static ssize_t amdgpu_hwmon_set_pwm1(struct device *dev, if (err) return err; - err = pm_runtime_get_sync(adev_to_drm(adev)->dev); - if (err < 0) { - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); + err = pm_runtime_resume_and_get(adev->dev); + if (err < 0) return err; - } err = amdgpu_dpm_get_fan_control_mode(adev, &pwm_mode); if (err) @@ -2907,15 +2828,12 @@ static ssize_t amdgpu_hwmon_get_pwm1(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - err = pm_runtime_get_sync(adev_to_drm(adev)->dev); - if (err < 0) { - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); - return err; - } + err = pm_runtime_get_if_active(adev->dev); + if (err <= 0) + return err ?: -EPERM; err = amdgpu_dpm_get_fan_speed_pwm(adev, &speed); - pm_runtime_mark_last_busy(adev_to_drm(adev)->dev); pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); if (err) @@ -2937,15 +2855,12 @@ static ssize_t amdgpu_hwmon_get_fan1_input(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - err = pm_runtime_get_sync(adev_to_drm(adev)->dev); - if (err < 0) { - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); - return err; - } + err = pm_runtime_get_if_active(adev->dev); + if (err <= 0) + return err ?: -EPERM; err = amdgpu_dpm_get_fan_speed_rpm(adev, &speed); - pm_runtime_mark_last_busy(adev_to_drm(adev)->dev); pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); if (err) @@ -3001,15 +2916,12 @@ static ssize_t amdgpu_hwmon_get_fan1_target(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - err = pm_runtime_get_sync(adev_to_drm(adev)->dev); - if (err < 0) { - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); - return err; - } + err = pm_runtime_get_if_active(adev->dev); + if (err <= 0) + return err ?: -EPERM; err = amdgpu_dpm_get_fan_speed_rpm(adev, &rpm); - pm_runtime_mark_last_busy(adev_to_drm(adev)->dev); pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); if (err) @@ -3036,11 +2948,9 @@ static ssize_t amdgpu_hwmon_set_fan1_target(struct device *dev, if (err) return err; - err = pm_runtime_get_sync(adev_to_drm(adev)->dev); - if (err < 0) { - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); + err = pm_runtime_resume_and_get(adev->dev); + if (err < 0) return err; - } err = amdgpu_dpm_get_fan_control_mode(adev, &pwm_mode); if (err) @@ -3076,15 +2986,12 @@ static ssize_t amdgpu_hwmon_get_fan1_enable(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(adev_to_drm(adev)->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); - return ret; - } + ret = pm_runtime_get_if_active(adev->dev); + if (ret <= 0) + return ret ?: -EPERM; ret = amdgpu_dpm_get_fan_control_mode(adev, &pwm_mode); - pm_runtime_mark_last_busy(adev_to_drm(adev)->dev); pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); if (ret) @@ -3119,11 +3026,9 @@ static ssize_t amdgpu_hwmon_set_fan1_enable(struct device *dev, else return -EINVAL; - err = pm_runtime_get_sync(adev_to_drm(adev)->dev); - if (err < 0) { - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); + err = pm_runtime_resume_and_get(adev->dev); + if (err < 0) return err; - } err = amdgpu_dpm_set_fan_control_mode(adev, pwm_mode); @@ -3248,11 +3153,9 @@ static ssize_t amdgpu_hwmon_show_power_cap_generic(struct device *dev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - r = pm_runtime_get_sync(adev_to_drm(adev)->dev); - if (r < 0) { - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); - return r; - } + r = pm_runtime_get_if_active(adev->dev); + if (r <= 0) + return r ?: -EPERM; r = amdgpu_dpm_get_power_limit(adev, &limit, pp_limit_level, power_type); @@ -3262,7 +3165,6 @@ static ssize_t amdgpu_hwmon_show_power_cap_generic(struct device *dev, else size = sysfs_emit(buf, "\n"); - pm_runtime_mark_last_busy(adev_to_drm(adev)->dev); pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); return size; @@ -3339,11 +3241,9 @@ static ssize_t amdgpu_hwmon_set_power_cap(struct device *dev, value = value / 1000000; /* convert to Watt */ value |= limit_type << 24; - err = pm_runtime_get_sync(adev_to_drm(adev)->dev); - if (err < 0) { - pm_runtime_put_autosuspend(adev_to_drm(adev)->dev); + err = pm_runtime_resume_and_get(adev->dev); + if (err < 0) return err; - } err = amdgpu_dpm_set_power_limit(adev, value); @@ -3787,17 +3687,14 @@ static int amdgpu_retrieve_od_settings(struct amdgpu_device *adev, if (adev->in_suspend && !adev->in_runpm) return -EPERM; - ret = pm_runtime_get_sync(adev->dev); - if (ret < 0) { - pm_runtime_put_autosuspend(adev->dev); - return ret; - } + ret = pm_runtime_get_if_active(adev->dev); + if (ret <= 0) + return ret ?: -EPERM; size = amdgpu_dpm_print_clock_levels(adev, od_type, buf); if (size == 0) size = sysfs_emit(buf, "\n"); - pm_runtime_mark_last_busy(adev->dev); pm_runtime_put_autosuspend(adev->dev); return size; @@ -3879,23 +3776,23 @@ amdgpu_distribute_custom_od_settings(struct amdgpu_device *adev, if (ret) return ret; - ret = pm_runtime_get_sync(adev->dev); + ret = pm_runtime_resume_and_get(adev->dev); if (ret < 0) - goto err_out0; + return ret; ret = amdgpu_dpm_odn_edit_dpm_table(adev, cmd_type, parameter, parameter_size); if (ret) - goto err_out1; + goto err_out; if (cmd_type == PP_OD_COMMIT_DPM_TABLE) { ret = amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL); if (ret) - goto err_out1; + goto err_out; } pm_runtime_mark_last_busy(adev->dev); @@ -3903,9 +3800,8 @@ amdgpu_distribute_custom_od_settings(struct amdgpu_device *adev, return count; -err_out1: +err_out: pm_runtime_mark_last_busy(adev->dev); -err_out0: pm_runtime_put_autosuspend(adev->dev); return ret; @@ -4213,6 +4109,117 @@ static umode_t fan_minimum_pwm_visible(struct amdgpu_device *adev) return umode; } +/** + * DOC: fan_zero_rpm_enable + * + * The amdgpu driver provides a sysfs API for checking and adjusting the + * zero RPM feature. + * + * Reading back the file shows you the current setting and the permitted + * ranges if changable. + * + * Writing an integer to the file, change the setting accordingly. + * + * When you have finished the editing, write "c" (commit) to the file to commit + * your changes. + * + * If you want to reset to the default value, write "r" (reset) to the file to + * reset them. + */ +static ssize_t fan_zero_rpm_enable_show(struct kobject *kobj, + struct kobj_attribute *attr, + char *buf) +{ + struct od_kobj *container = container_of(kobj, struct od_kobj, kobj); + struct amdgpu_device *adev = (struct amdgpu_device *)container->priv; + + return (ssize_t)amdgpu_retrieve_od_settings(adev, OD_FAN_ZERO_RPM_ENABLE, buf); +} + +static ssize_t fan_zero_rpm_enable_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, + size_t count) +{ + struct od_kobj *container = container_of(kobj, struct od_kobj, kobj); + struct amdgpu_device *adev = (struct amdgpu_device *)container->priv; + + return (ssize_t)amdgpu_distribute_custom_od_settings(adev, + PP_OD_EDIT_FAN_ZERO_RPM_ENABLE, + buf, + count); +} + +static umode_t fan_zero_rpm_enable_visible(struct amdgpu_device *adev) +{ + umode_t umode = 0000; + + if (adev->pm.od_feature_mask & OD_OPS_SUPPORT_FAN_ZERO_RPM_ENABLE_RETRIEVE) + umode |= S_IRUSR | S_IRGRP | S_IROTH; + + if (adev->pm.od_feature_mask & OD_OPS_SUPPORT_FAN_ZERO_RPM_ENABLE_SET) + umode |= S_IWUSR; + + return umode; +} + +/** + * DOC: fan_zero_rpm_stop_temperature + * + * The amdgpu driver provides a sysfs API for checking and adjusting the + * zero RPM stop temperature feature. + * + * Reading back the file shows you the current setting and the permitted + * ranges if changable. + * + * Writing an integer to the file, change the setting accordingly. + * + * When you have finished the editing, write "c" (commit) to the file to commit + * your changes. + * + * If you want to reset to the default value, write "r" (reset) to the file to + * reset them. + * + * This setting works only if the Zero RPM setting is enabled. It adjusts the + * temperature below which the fan can stop. + */ +static ssize_t fan_zero_rpm_stop_temp_show(struct kobject *kobj, + struct kobj_attribute *attr, + char *buf) +{ + struct od_kobj *container = container_of(kobj, struct od_kobj, kobj); + struct amdgpu_device *adev = (struct amdgpu_device *)container->priv; + + return (ssize_t)amdgpu_retrieve_od_settings(adev, OD_FAN_ZERO_RPM_STOP_TEMP, buf); +} + +static ssize_t fan_zero_rpm_stop_temp_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, + size_t count) +{ + struct od_kobj *container = container_of(kobj, struct od_kobj, kobj); + struct amdgpu_device *adev = (struct amdgpu_device *)container->priv; + + return (ssize_t)amdgpu_distribute_custom_od_settings(adev, + PP_OD_EDIT_FAN_ZERO_RPM_STOP_TEMP, + buf, + count); +} + +static umode_t fan_zero_rpm_stop_temp_visible(struct amdgpu_device *adev) +{ + umode_t umode = 0000; + + if (adev->pm.od_feature_mask & OD_OPS_SUPPORT_FAN_ZERO_RPM_STOP_TEMP_RETRIEVE) + umode |= S_IRUSR | S_IRGRP | S_IROTH; + + if (adev->pm.od_feature_mask & OD_OPS_SUPPORT_FAN_ZERO_RPM_STOP_TEMP_SET) + umode |= S_IWUSR; + + return umode; +} + static struct od_feature_set amdgpu_od_set = { .containers = { [0] = { @@ -4258,6 +4265,22 @@ static struct od_feature_set amdgpu_od_set = { .store = fan_minimum_pwm_store, }, }, + [5] = { + .name = "fan_zero_rpm_enable", + .ops = { + .is_visible = fan_zero_rpm_enable_visible, + .show = fan_zero_rpm_enable_show, + .store = fan_zero_rpm_enable_store, + }, + }, + [6] = { + .name = "fan_zero_rpm_stop_temperature", + .ops = { + .is_visible = fan_zero_rpm_stop_temp_visible, + .show = fan_zero_rpm_stop_temp_show, + .store = fan_zero_rpm_stop_temp_store, + }, + }, }, }, }, @@ -4758,11 +4781,9 @@ static int amdgpu_debugfs_pm_info_show(struct seq_file *m, void *unused) if (adev->in_suspend && !adev->in_runpm) return -EPERM; - r = pm_runtime_get_sync(dev->dev); - if (r < 0) { - pm_runtime_put_autosuspend(dev->dev); + r = pm_runtime_resume_and_get(dev->dev); + if (r < 0) return r; - } if (amdgpu_dpm_debugfs_print_current_performance_level(adev, m)) { r = amdgpu_debugfs_pm_info_pp(m, adev); @@ -4777,7 +4798,6 @@ static int amdgpu_debugfs_pm_info_show(struct seq_file *m, void *unused) seq_printf(m, "\n"); out: - pm_runtime_mark_last_busy(dev->dev); pm_runtime_put_autosuspend(dev->dev); return r; diff --git a/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h b/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h index f5bf41f21c41..363af8990aa2 100644 --- a/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h +++ b/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h @@ -328,6 +328,10 @@ struct config_table_setting #define OD_OPS_SUPPORT_FAN_TARGET_TEMPERATURE_SET BIT(7) #define OD_OPS_SUPPORT_FAN_MINIMUM_PWM_RETRIEVE BIT(8) #define OD_OPS_SUPPORT_FAN_MINIMUM_PWM_SET BIT(9) +#define OD_OPS_SUPPORT_FAN_ZERO_RPM_ENABLE_RETRIEVE BIT(10) +#define OD_OPS_SUPPORT_FAN_ZERO_RPM_ENABLE_SET BIT(11) +#define OD_OPS_SUPPORT_FAN_ZERO_RPM_STOP_TEMP_RETRIEVE BIT(12) +#define OD_OPS_SUPPORT_FAN_ZERO_RPM_STOP_TEMP_SET BIT(13) struct amdgpu_pm { struct mutex mutex; diff --git a/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c b/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c index e8b6989a40f3..8908646ad620 100644 --- a/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c +++ b/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c @@ -2954,9 +2954,9 @@ static int kv_dpm_get_temp(void *handle) return actual_temp; } -static int kv_dpm_early_init(void *handle) +static int kv_dpm_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->powerplay.pp_funcs = &kv_dpm_funcs; adev->powerplay.pp_handle = adev; @@ -2965,10 +2965,10 @@ static int kv_dpm_early_init(void *handle) return 0; } -static int kv_dpm_late_init(void *handle) +static int kv_dpm_late_init(struct amdgpu_ip_block *ip_block) { /* powerdown unused blocks for now */ - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (!adev->pm.dpm_enabled) return 0; @@ -2979,11 +2979,10 @@ static int kv_dpm_late_init(void *handle) return 0; } -static int kv_dpm_sw_init(void *handle) +static int kv_dpm_sw_init(struct amdgpu_ip_block *ip_block) { int ret; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - + struct amdgpu_device *adev = ip_block->adev; ret = amdgpu_irq_add_id(adev, AMDGPU_IRQ_CLIENTID_LEGACY, 230, &adev->pm.dpm.thermal.irq); if (ret) @@ -3024,9 +3023,9 @@ dpm_failed: return ret; } -static int kv_dpm_sw_fini(void *handle) +static int kv_dpm_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; flush_work(&adev->pm.dpm.thermal.work); @@ -3035,10 +3034,10 @@ static int kv_dpm_sw_fini(void *handle) return 0; } -static int kv_dpm_hw_init(void *handle) +static int kv_dpm_hw_init(struct amdgpu_ip_block *ip_block) { int ret; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (!amdgpu_dpm) return 0; @@ -3053,9 +3052,9 @@ static int kv_dpm_hw_init(void *handle) return ret; } -static int kv_dpm_hw_fini(void *handle) +static int kv_dpm_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->pm.dpm_enabled) kv_dpm_disable(adev); @@ -3063,9 +3062,9 @@ static int kv_dpm_hw_fini(void *handle) return 0; } -static int kv_dpm_suspend(void *handle) +static int kv_dpm_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->pm.dpm_enabled) { /* disable dpm */ @@ -3076,10 +3075,10 @@ static int kv_dpm_suspend(void *handle) return 0; } -static int kv_dpm_resume(void *handle) +static int kv_dpm_resume(struct amdgpu_ip_block *ip_block) { int ret; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->pm.dpm_enabled) { /* asic init will reset to the boot state */ @@ -3100,17 +3099,6 @@ static bool kv_dpm_is_idle(void *handle) return true; } -static int kv_dpm_wait_for_idle(void *handle) -{ - return 0; -} - - -static int kv_dpm_soft_reset(void *handle) -{ - return 0; -} - static int kv_dpm_set_interrupt_state(struct amdgpu_device *adev, struct amdgpu_irq_src *src, unsigned type, @@ -3314,12 +3302,8 @@ static const struct amd_ip_funcs kv_dpm_ip_funcs = { .suspend = kv_dpm_suspend, .resume = kv_dpm_resume, .is_idle = kv_dpm_is_idle, - .wait_for_idle = kv_dpm_wait_for_idle, - .soft_reset = kv_dpm_soft_reset, .set_clockgating_state = kv_dpm_set_clockgating_state, .set_powergating_state = kv_dpm_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; const struct amdgpu_ip_block_version kv_smu_ip_block = { diff --git a/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c b/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c index a1baa13ab2c2..ee23a0f897c5 100644 --- a/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c +++ b/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c @@ -4755,13 +4755,15 @@ static int si_populate_memory_timing_parameters(struct amdgpu_device *adev, u32 dram_timing; u32 dram_timing2; u32 burst_time; + int ret; arb_regs->mc_arb_rfsh_rate = (u8)si_calculate_memory_refresh_rate(adev, pl->sclk); - amdgpu_atombios_set_engine_dram_timings(adev, - pl->sclk, - pl->mclk); + ret = amdgpu_atombios_set_engine_dram_timings(adev, pl->sclk, + pl->mclk); + if (ret) + return ret; dram_timing = RREG32(MC_ARB_DRAM_TIMING); dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2); @@ -7619,10 +7621,10 @@ static int si_dpm_process_interrupt(struct amdgpu_device *adev, return 0; } -static int si_dpm_late_init(void *handle) +static int si_dpm_late_init(struct amdgpu_ip_block *ip_block) { int ret; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (!adev->pm.dpm_enabled) return 0; @@ -7716,10 +7718,10 @@ static int si_dpm_init_microcode(struct amdgpu_device *adev) return err; } -static int si_dpm_sw_init(void *handle) +static int si_dpm_sw_init(struct amdgpu_ip_block *ip_block) { int ret; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; ret = amdgpu_irq_add_id(adev, AMDGPU_IRQ_CLIENTID_LEGACY, 230, &adev->pm.dpm.thermal.irq); if (ret) @@ -7763,9 +7765,9 @@ dpm_failed: return ret; } -static int si_dpm_sw_fini(void *handle) +static int si_dpm_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; flush_work(&adev->pm.dpm.thermal.work); @@ -7774,11 +7776,11 @@ static int si_dpm_sw_fini(void *handle) return 0; } -static int si_dpm_hw_init(void *handle) +static int si_dpm_hw_init(struct amdgpu_ip_block *ip_block) { int ret; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (!amdgpu_dpm) return 0; @@ -7793,9 +7795,9 @@ static int si_dpm_hw_init(void *handle) return ret; } -static int si_dpm_hw_fini(void *handle) +static int si_dpm_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->pm.dpm_enabled) si_dpm_disable(adev); @@ -7803,9 +7805,9 @@ static int si_dpm_hw_fini(void *handle) return 0; } -static int si_dpm_suspend(void *handle) +static int si_dpm_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->pm.dpm_enabled) { /* disable dpm */ @@ -7816,10 +7818,10 @@ static int si_dpm_suspend(void *handle) return 0; } -static int si_dpm_resume(void *handle) +static int si_dpm_resume(struct amdgpu_ip_block *ip_block) { int ret; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->pm.dpm_enabled) { /* asic init will reset to the boot state */ @@ -7841,17 +7843,12 @@ static bool si_dpm_is_idle(void *handle) return true; } -static int si_dpm_wait_for_idle(void *handle) +static int si_dpm_wait_for_idle(struct amdgpu_ip_block *ip_block) { /* XXX */ return 0; } -static int si_dpm_soft_reset(void *handle) -{ - return 0; -} - static int si_dpm_set_clockgating_state(void *handle, enum amd_clockgating_state state) { @@ -7928,10 +7925,10 @@ static void si_dpm_print_power_state(void *handle, amdgpu_dpm_print_ps_status(adev, rps); } -static int si_dpm_early_init(void *handle) +static int si_dpm_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; adev->powerplay.pp_funcs = &si_dpm_funcs; adev->powerplay.pp_handle = adev; @@ -8047,11 +8044,8 @@ static const struct amd_ip_funcs si_dpm_ip_funcs = { .resume = si_dpm_resume, .is_idle = si_dpm_is_idle, .wait_for_idle = si_dpm_wait_for_idle, - .soft_reset = si_dpm_soft_reset, .set_clockgating_state = si_dpm_set_clockgating_state, .set_powergating_state = si_dpm_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; const struct amdgpu_ip_block_version si_smu_ip_block = diff --git a/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c index a71c6117d7e5..26624a716fc6 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c +++ b/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c @@ -75,11 +75,10 @@ static void amd_powerplay_destroy(struct amdgpu_device *adev) hwmgr = NULL; } -static int pp_early_init(void *handle) +static int pp_early_init(struct amdgpu_ip_block *ip_block) { int ret; - struct amdgpu_device *adev = handle; - + struct amdgpu_device *adev = ip_block->adev; ret = amd_powerplay_create(adev); if (ret != 0) @@ -131,9 +130,9 @@ static void pp_swctf_delayed_work_handler(struct work_struct *work) orderly_poweroff(true); } -static int pp_sw_init(void *handle) +static int pp_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = handle; + struct amdgpu_device *adev = ip_block->adev; struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle; int ret = 0; @@ -148,9 +147,9 @@ static int pp_sw_init(void *handle) return ret; } -static int pp_sw_fini(void *handle) +static int pp_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = handle; + struct amdgpu_device *adev = ip_block->adev; struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle; hwmgr_sw_fini(hwmgr); @@ -160,10 +159,10 @@ static int pp_sw_fini(void *handle) return 0; } -static int pp_hw_init(void *handle) +static int pp_hw_init(struct amdgpu_ip_block *ip_block) { int ret = 0; - struct amdgpu_device *adev = handle; + struct amdgpu_device *adev = ip_block->adev; struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle; ret = hwmgr_hw_init(hwmgr); @@ -174,10 +173,9 @@ static int pp_hw_init(void *handle) return ret; } -static int pp_hw_fini(void *handle) +static int pp_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = handle; - struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle; + struct pp_hwmgr *hwmgr = ip_block->adev->powerplay.pp_handle; cancel_delayed_work_sync(&hwmgr->swctf_delayed_work); @@ -217,9 +215,9 @@ static void pp_reserve_vram_for_smu(struct amdgpu_device *adev) } } -static int pp_late_init(void *handle) +static int pp_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = handle; + struct amdgpu_device *adev = ip_block->adev; struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle; if (hwmgr && hwmgr->pm_en) @@ -231,9 +229,9 @@ static int pp_late_init(void *handle) return 0; } -static void pp_late_fini(void *handle) +static void pp_late_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = handle; + struct amdgpu_device *adev = ip_block->adev; if (adev->pm.smu_prv_buffer) amdgpu_bo_free_kernel(&adev->pm.smu_prv_buffer, NULL, NULL); @@ -246,25 +244,15 @@ static bool pp_is_idle(void *handle) return false; } -static int pp_wait_for_idle(void *handle) -{ - return 0; -} - -static int pp_sw_reset(void *handle) -{ - return 0; -} - static int pp_set_powergating_state(void *handle, enum amd_powergating_state state) { return 0; } -static int pp_suspend(void *handle) +static int pp_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = handle; + struct amdgpu_device *adev = ip_block->adev; struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle; cancel_delayed_work_sync(&hwmgr->swctf_delayed_work); @@ -272,10 +260,9 @@ static int pp_suspend(void *handle) return hwmgr_suspend(hwmgr); } -static int pp_resume(void *handle) +static int pp_resume(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = handle; - struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle; + struct pp_hwmgr *hwmgr = ip_block->adev->powerplay.pp_handle; return hwmgr_resume(hwmgr); } @@ -298,12 +285,8 @@ static const struct amd_ip_funcs pp_ip_funcs = { .suspend = pp_suspend, .resume = pp_resume, .is_idle = pp_is_idle, - .wait_for_idle = pp_wait_for_idle, - .soft_reset = pp_sw_reset, .set_clockgating_state = pp_set_clockgating_state, .set_powergating_state = pp_set_powergating_state, - .dump_ip_state = NULL, - .print_ip_state = NULL, }; const struct amdgpu_ip_block_version pp_smu_ip_block = diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/ppatomctrl.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/ppatomctrl.c index b56298d9da98..fe24219c3bf4 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/ppatomctrl.c +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/ppatomctrl.c @@ -28,7 +28,6 @@ #include "ppatomctrl.h" #include "atombios.h" #include "cgs_common.h" -#include "ppevvmath.h" #define MEM_ID_MASK 0xff000000 #define MEM_ID_SHIFT 24 @@ -677,433 +676,6 @@ bool atomctrl_get_pp_assign_pin( return bRet; } -int atomctrl_calculate_voltage_evv_on_sclk( - struct pp_hwmgr *hwmgr, - uint8_t voltage_type, - uint32_t sclk, - uint16_t virtual_voltage_Id, - uint16_t *voltage, - uint16_t dpm_level, - bool debug) -{ - ATOM_ASIC_PROFILING_INFO_V3_4 *getASICProfilingInfo; - struct amdgpu_device *adev = hwmgr->adev; - EFUSE_LINEAR_FUNC_PARAM sRO_fuse; - EFUSE_LINEAR_FUNC_PARAM sCACm_fuse; - EFUSE_LINEAR_FUNC_PARAM sCACb_fuse; - EFUSE_LOGISTIC_FUNC_PARAM sKt_Beta_fuse; - EFUSE_LOGISTIC_FUNC_PARAM sKv_m_fuse; - EFUSE_LOGISTIC_FUNC_PARAM sKv_b_fuse; - EFUSE_INPUT_PARAMETER sInput_FuseValues; - READ_EFUSE_VALUE_PARAMETER sOutput_FuseValues; - - uint32_t ul_RO_fused, ul_CACb_fused, ul_CACm_fused, ul_Kt_Beta_fused, ul_Kv_m_fused, ul_Kv_b_fused; - fInt fSM_A0, fSM_A1, fSM_A2, fSM_A3, fSM_A4, fSM_A5, fSM_A6, fSM_A7; - fInt fMargin_RO_a, fMargin_RO_b, fMargin_RO_c, fMargin_fixed, fMargin_FMAX_mean, fMargin_Plat_mean, fMargin_FMAX_sigma, fMargin_Plat_sigma, fMargin_DC_sigma; - fInt fLkg_FT, repeat; - fInt fMicro_FMAX, fMicro_CR, fSigma_FMAX, fSigma_CR, fSigma_DC, fDC_SCLK, fSquared_Sigma_DC, fSquared_Sigma_CR, fSquared_Sigma_FMAX; - fInt fRLL_LoadLine, fDerateTDP, fVDDC_base, fA_Term, fC_Term, fB_Term, fRO_DC_margin; - fInt fRO_fused, fCACm_fused, fCACb_fused, fKv_m_fused, fKv_b_fused, fKt_Beta_fused, fFT_Lkg_V0NORM; - fInt fSclk_margin, fSclk, fEVV_V; - fInt fV_min, fV_max, fT_prod, fLKG_Factor, fT_FT, fV_FT, fV_x, fTDP_Power, fTDP_Power_right, fTDP_Power_left, fTDP_Current, fV_NL; - uint32_t ul_FT_Lkg_V0NORM; - fInt fLn_MaxDivMin, fMin, fAverage, fRange; - fInt fRoots[2]; - fInt fStepSize = GetScaledFraction(625, 100000); - - int result; - - getASICProfilingInfo = (ATOM_ASIC_PROFILING_INFO_V3_4 *) - smu_atom_get_data_table(hwmgr->adev, - GetIndexIntoMasterTable(DATA, ASIC_ProfilingInfo), - NULL, NULL, NULL); - - if (!getASICProfilingInfo) - return -1; - - if (getASICProfilingInfo->asHeader.ucTableFormatRevision < 3 || - (getASICProfilingInfo->asHeader.ucTableFormatRevision == 3 && - getASICProfilingInfo->asHeader.ucTableContentRevision < 4)) - return -1; - - /*----------------------------------------------------------- - *GETTING MULTI-STEP PARAMETERS RELATED TO CURRENT DPM LEVEL - *----------------------------------------------------------- - */ - fRLL_LoadLine = Divide(getASICProfilingInfo->ulLoadLineSlop, 1000); - - switch (dpm_level) { - case 1: - fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM1), 1000); - break; - case 2: - fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM2), 1000); - break; - case 3: - fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM3), 1000); - break; - case 4: - fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM4), 1000); - break; - case 5: - fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM5), 1000); - break; - case 6: - fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM6), 1000); - break; - case 7: - fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM7), 1000); - break; - default: - pr_err("DPM Level not supported\n"); - fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM0), 1000); - } - - /*------------------------- - * DECODING FUSE VALUES - * ------------------------ - */ - /*Decode RO_Fused*/ - sRO_fuse = getASICProfilingInfo->sRoFuse; - - sInput_FuseValues.usEfuseIndex = sRO_fuse.usEfuseIndex; - sInput_FuseValues.ucBitShift = sRO_fuse.ucEfuseBitLSB; - sInput_FuseValues.ucBitLength = sRO_fuse.ucEfuseLength; - - sOutput_FuseValues.sEfuse = sInput_FuseValues; - - result = amdgpu_atom_execute_table(adev->mode_info.atom_context, - GetIndexIntoMasterTable(COMMAND, ReadEfuseValue), - (uint32_t *)&sOutput_FuseValues, sizeof(sOutput_FuseValues)); - - if (result) - return result; - - /* Finally, the actual fuse value */ - ul_RO_fused = le32_to_cpu(sOutput_FuseValues.ulEfuseValue); - fMin = GetScaledFraction(le32_to_cpu(sRO_fuse.ulEfuseMin), 1); - fRange = GetScaledFraction(le32_to_cpu(sRO_fuse.ulEfuseEncodeRange), 1); - fRO_fused = fDecodeLinearFuse(ul_RO_fused, fMin, fRange, sRO_fuse.ucEfuseLength); - - sCACm_fuse = getASICProfilingInfo->sCACm; - - sInput_FuseValues.usEfuseIndex = sCACm_fuse.usEfuseIndex; - sInput_FuseValues.ucBitShift = sCACm_fuse.ucEfuseBitLSB; - sInput_FuseValues.ucBitLength = sCACm_fuse.ucEfuseLength; - - sOutput_FuseValues.sEfuse = sInput_FuseValues; - - result = amdgpu_atom_execute_table(adev->mode_info.atom_context, - GetIndexIntoMasterTable(COMMAND, ReadEfuseValue), - (uint32_t *)&sOutput_FuseValues, sizeof(sOutput_FuseValues)); - - if (result) - return result; - - ul_CACm_fused = le32_to_cpu(sOutput_FuseValues.ulEfuseValue); - fMin = GetScaledFraction(le32_to_cpu(sCACm_fuse.ulEfuseMin), 1000); - fRange = GetScaledFraction(le32_to_cpu(sCACm_fuse.ulEfuseEncodeRange), 1000); - - fCACm_fused = fDecodeLinearFuse(ul_CACm_fused, fMin, fRange, sCACm_fuse.ucEfuseLength); - - sCACb_fuse = getASICProfilingInfo->sCACb; - - sInput_FuseValues.usEfuseIndex = sCACb_fuse.usEfuseIndex; - sInput_FuseValues.ucBitShift = sCACb_fuse.ucEfuseBitLSB; - sInput_FuseValues.ucBitLength = sCACb_fuse.ucEfuseLength; - sOutput_FuseValues.sEfuse = sInput_FuseValues; - - result = amdgpu_atom_execute_table(adev->mode_info.atom_context, - GetIndexIntoMasterTable(COMMAND, ReadEfuseValue), - (uint32_t *)&sOutput_FuseValues, sizeof(sOutput_FuseValues)); - - if (result) - return result; - - ul_CACb_fused = le32_to_cpu(sOutput_FuseValues.ulEfuseValue); - fMin = GetScaledFraction(le32_to_cpu(sCACb_fuse.ulEfuseMin), 1000); - fRange = GetScaledFraction(le32_to_cpu(sCACb_fuse.ulEfuseEncodeRange), 1000); - - fCACb_fused = fDecodeLinearFuse(ul_CACb_fused, fMin, fRange, sCACb_fuse.ucEfuseLength); - - sKt_Beta_fuse = getASICProfilingInfo->sKt_b; - - sInput_FuseValues.usEfuseIndex = sKt_Beta_fuse.usEfuseIndex; - sInput_FuseValues.ucBitShift = sKt_Beta_fuse.ucEfuseBitLSB; - sInput_FuseValues.ucBitLength = sKt_Beta_fuse.ucEfuseLength; - - sOutput_FuseValues.sEfuse = sInput_FuseValues; - - result = amdgpu_atom_execute_table(adev->mode_info.atom_context, - GetIndexIntoMasterTable(COMMAND, ReadEfuseValue), - (uint32_t *)&sOutput_FuseValues, sizeof(sOutput_FuseValues)); - - if (result) - return result; - - ul_Kt_Beta_fused = le32_to_cpu(sOutput_FuseValues.ulEfuseValue); - fAverage = GetScaledFraction(le32_to_cpu(sKt_Beta_fuse.ulEfuseEncodeAverage), 1000); - fRange = GetScaledFraction(le32_to_cpu(sKt_Beta_fuse.ulEfuseEncodeRange), 1000); - - fKt_Beta_fused = fDecodeLogisticFuse(ul_Kt_Beta_fused, - fAverage, fRange, sKt_Beta_fuse.ucEfuseLength); - - sKv_m_fuse = getASICProfilingInfo->sKv_m; - - sInput_FuseValues.usEfuseIndex = sKv_m_fuse.usEfuseIndex; - sInput_FuseValues.ucBitShift = sKv_m_fuse.ucEfuseBitLSB; - sInput_FuseValues.ucBitLength = sKv_m_fuse.ucEfuseLength; - - sOutput_FuseValues.sEfuse = sInput_FuseValues; - - result = amdgpu_atom_execute_table(adev->mode_info.atom_context, - GetIndexIntoMasterTable(COMMAND, ReadEfuseValue), - (uint32_t *)&sOutput_FuseValues, sizeof(sOutput_FuseValues)); - if (result) - return result; - - ul_Kv_m_fused = le32_to_cpu(sOutput_FuseValues.ulEfuseValue); - fAverage = GetScaledFraction(le32_to_cpu(sKv_m_fuse.ulEfuseEncodeAverage), 1000); - fRange = GetScaledFraction((le32_to_cpu(sKv_m_fuse.ulEfuseEncodeRange) & 0x7fffffff), 1000); - fRange = fMultiply(fRange, ConvertToFraction(-1)); - - fKv_m_fused = fDecodeLogisticFuse(ul_Kv_m_fused, - fAverage, fRange, sKv_m_fuse.ucEfuseLength); - - sKv_b_fuse = getASICProfilingInfo->sKv_b; - - sInput_FuseValues.usEfuseIndex = sKv_b_fuse.usEfuseIndex; - sInput_FuseValues.ucBitShift = sKv_b_fuse.ucEfuseBitLSB; - sInput_FuseValues.ucBitLength = sKv_b_fuse.ucEfuseLength; - sOutput_FuseValues.sEfuse = sInput_FuseValues; - - result = amdgpu_atom_execute_table(adev->mode_info.atom_context, - GetIndexIntoMasterTable(COMMAND, ReadEfuseValue), - (uint32_t *)&sOutput_FuseValues, sizeof(sOutput_FuseValues)); - - if (result) - return result; - - ul_Kv_b_fused = le32_to_cpu(sOutput_FuseValues.ulEfuseValue); - fAverage = GetScaledFraction(le32_to_cpu(sKv_b_fuse.ulEfuseEncodeAverage), 1000); - fRange = GetScaledFraction(le32_to_cpu(sKv_b_fuse.ulEfuseEncodeRange), 1000); - - fKv_b_fused = fDecodeLogisticFuse(ul_Kv_b_fused, - fAverage, fRange, sKv_b_fuse.ucEfuseLength); - - /* Decoding the Leakage - No special struct container */ - /* - * usLkgEuseIndex=56 - * ucLkgEfuseBitLSB=6 - * ucLkgEfuseLength=10 - * ulLkgEncodeLn_MaxDivMin=69077 - * ulLkgEncodeMax=1000000 - * ulLkgEncodeMin=1000 - * ulEfuseLogisticAlpha=13 - */ - - sInput_FuseValues.usEfuseIndex = getASICProfilingInfo->usLkgEuseIndex; - sInput_FuseValues.ucBitShift = getASICProfilingInfo->ucLkgEfuseBitLSB; - sInput_FuseValues.ucBitLength = getASICProfilingInfo->ucLkgEfuseLength; - - sOutput_FuseValues.sEfuse = sInput_FuseValues; - - result = amdgpu_atom_execute_table(adev->mode_info.atom_context, - GetIndexIntoMasterTable(COMMAND, ReadEfuseValue), - (uint32_t *)&sOutput_FuseValues, sizeof(sOutput_FuseValues)); - - if (result) - return result; - - ul_FT_Lkg_V0NORM = le32_to_cpu(sOutput_FuseValues.ulEfuseValue); - fLn_MaxDivMin = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulLkgEncodeLn_MaxDivMin), 10000); - fMin = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulLkgEncodeMin), 10000); - - fFT_Lkg_V0NORM = fDecodeLeakageID(ul_FT_Lkg_V0NORM, - fLn_MaxDivMin, fMin, getASICProfilingInfo->ucLkgEfuseLength); - fLkg_FT = fFT_Lkg_V0NORM; - - /*------------------------------------------- - * PART 2 - Grabbing all required values - *------------------------------------------- - */ - fSM_A0 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A0), 1000000), - ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A0_sign))); - fSM_A1 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A1), 1000000), - ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A1_sign))); - fSM_A2 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A2), 100000), - ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A2_sign))); - fSM_A3 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A3), 1000000), - ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A3_sign))); - fSM_A4 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A4), 1000000), - ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A4_sign))); - fSM_A5 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A5), 1000), - ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A5_sign))); - fSM_A6 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A6), 1000), - ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A6_sign))); - fSM_A7 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A7), 1000), - ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A7_sign))); - - fMargin_RO_a = ConvertToFraction(le32_to_cpu(getASICProfilingInfo->ulMargin_RO_a)); - fMargin_RO_b = ConvertToFraction(le32_to_cpu(getASICProfilingInfo->ulMargin_RO_b)); - fMargin_RO_c = ConvertToFraction(le32_to_cpu(getASICProfilingInfo->ulMargin_RO_c)); - - fMargin_fixed = ConvertToFraction(le32_to_cpu(getASICProfilingInfo->ulMargin_fixed)); - - fMargin_FMAX_mean = GetScaledFraction( - le32_to_cpu(getASICProfilingInfo->ulMargin_Fmax_mean), 10000); - fMargin_Plat_mean = GetScaledFraction( - le32_to_cpu(getASICProfilingInfo->ulMargin_plat_mean), 10000); - fMargin_FMAX_sigma = GetScaledFraction( - le32_to_cpu(getASICProfilingInfo->ulMargin_Fmax_sigma), 10000); - fMargin_Plat_sigma = GetScaledFraction( - le32_to_cpu(getASICProfilingInfo->ulMargin_plat_sigma), 10000); - - fMargin_DC_sigma = GetScaledFraction( - le32_to_cpu(getASICProfilingInfo->ulMargin_DC_sigma), 100); - fMargin_DC_sigma = fDivide(fMargin_DC_sigma, ConvertToFraction(1000)); - - fCACm_fused = fDivide(fCACm_fused, ConvertToFraction(100)); - fCACb_fused = fDivide(fCACb_fused, ConvertToFraction(100)); - fKt_Beta_fused = fDivide(fKt_Beta_fused, ConvertToFraction(100)); - fKv_m_fused = fNegate(fDivide(fKv_m_fused, ConvertToFraction(100))); - fKv_b_fused = fDivide(fKv_b_fused, ConvertToFraction(10)); - - fSclk = GetScaledFraction(sclk, 100); - - fV_max = fDivide(GetScaledFraction( - le32_to_cpu(getASICProfilingInfo->ulMaxVddc), 1000), ConvertToFraction(4)); - fT_prod = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulBoardCoreTemp), 10); - fLKG_Factor = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulEvvLkgFactor), 100); - fT_FT = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulLeakageTemp), 10); - fV_FT = fDivide(GetScaledFraction( - le32_to_cpu(getASICProfilingInfo->ulLeakageVoltage), 1000), ConvertToFraction(4)); - fV_min = fDivide(GetScaledFraction( - le32_to_cpu(getASICProfilingInfo->ulMinVddc), 1000), ConvertToFraction(4)); - - /*----------------------- - * PART 3 - *----------------------- - */ - - fA_Term = fAdd(fMargin_RO_a, fAdd(fMultiply(fSM_A4, fSclk), fSM_A5)); - fB_Term = fAdd(fAdd(fMultiply(fSM_A2, fSclk), fSM_A6), fMargin_RO_b); - fC_Term = fAdd(fMargin_RO_c, - fAdd(fMultiply(fSM_A0, fLkg_FT), - fAdd(fMultiply(fSM_A1, fMultiply(fLkg_FT, fSclk)), - fAdd(fMultiply(fSM_A3, fSclk), - fSubtract(fSM_A7, fRO_fused))))); - - fVDDC_base = fSubtract(fRO_fused, - fSubtract(fMargin_RO_c, - fSubtract(fSM_A3, fMultiply(fSM_A1, fSclk)))); - fVDDC_base = fDivide(fVDDC_base, fAdd(fMultiply(fSM_A0, fSclk), fSM_A2)); - - repeat = fSubtract(fVDDC_base, - fDivide(fMargin_DC_sigma, ConvertToFraction(1000))); - - fRO_DC_margin = fAdd(fMultiply(fMargin_RO_a, - fGetSquare(repeat)), - fAdd(fMultiply(fMargin_RO_b, repeat), - fMargin_RO_c)); - - fDC_SCLK = fSubtract(fRO_fused, - fSubtract(fRO_DC_margin, - fSubtract(fSM_A3, - fMultiply(fSM_A2, repeat)))); - fDC_SCLK = fDivide(fDC_SCLK, fAdd(fMultiply(fSM_A0, repeat), fSM_A1)); - - fSigma_DC = fSubtract(fSclk, fDC_SCLK); - - fMicro_FMAX = fMultiply(fSclk, fMargin_FMAX_mean); - fMicro_CR = fMultiply(fSclk, fMargin_Plat_mean); - fSigma_FMAX = fMultiply(fSclk, fMargin_FMAX_sigma); - fSigma_CR = fMultiply(fSclk, fMargin_Plat_sigma); - - fSquared_Sigma_DC = fGetSquare(fSigma_DC); - fSquared_Sigma_CR = fGetSquare(fSigma_CR); - fSquared_Sigma_FMAX = fGetSquare(fSigma_FMAX); - - fSclk_margin = fAdd(fMicro_FMAX, - fAdd(fMicro_CR, - fAdd(fMargin_fixed, - fSqrt(fAdd(fSquared_Sigma_FMAX, - fAdd(fSquared_Sigma_DC, fSquared_Sigma_CR)))))); - /* - fA_Term = fSM_A4 * (fSclk + fSclk_margin) + fSM_A5; - fB_Term = fSM_A2 * (fSclk + fSclk_margin) + fSM_A6; - fC_Term = fRO_DC_margin + fSM_A0 * fLkg_FT + fSM_A1 * fLkg_FT * (fSclk + fSclk_margin) + fSM_A3 * (fSclk + fSclk_margin) + fSM_A7 - fRO_fused; - */ - - fA_Term = fAdd(fMultiply(fSM_A4, fAdd(fSclk, fSclk_margin)), fSM_A5); - fB_Term = fAdd(fMultiply(fSM_A2, fAdd(fSclk, fSclk_margin)), fSM_A6); - fC_Term = fAdd(fRO_DC_margin, - fAdd(fMultiply(fSM_A0, fLkg_FT), - fAdd(fMultiply(fMultiply(fSM_A1, fLkg_FT), - fAdd(fSclk, fSclk_margin)), - fAdd(fMultiply(fSM_A3, - fAdd(fSclk, fSclk_margin)), - fSubtract(fSM_A7, fRO_fused))))); - - SolveQuadracticEqn(fA_Term, fB_Term, fC_Term, fRoots); - - if (GreaterThan(fRoots[0], fRoots[1])) - fEVV_V = fRoots[1]; - else - fEVV_V = fRoots[0]; - - if (GreaterThan(fV_min, fEVV_V)) - fEVV_V = fV_min; - else if (GreaterThan(fEVV_V, fV_max)) - fEVV_V = fSubtract(fV_max, fStepSize); - - fEVV_V = fRoundUpByStepSize(fEVV_V, fStepSize, 0); - - /*----------------- - * PART 4 - *----------------- - */ - - fV_x = fV_min; - - while (GreaterThan(fAdd(fV_max, fStepSize), fV_x)) { - fTDP_Power_left = fMultiply(fMultiply(fMultiply(fAdd( - fMultiply(fCACm_fused, fV_x), fCACb_fused), fSclk), - fGetSquare(fV_x)), fDerateTDP); - - fTDP_Power_right = fMultiply(fFT_Lkg_V0NORM, fMultiply(fLKG_Factor, - fMultiply(fExponential(fMultiply(fAdd(fMultiply(fKv_m_fused, - fT_prod), fKv_b_fused), fV_x)), fV_x))); - fTDP_Power_right = fMultiply(fTDP_Power_right, fExponential(fMultiply( - fKt_Beta_fused, fT_prod))); - fTDP_Power_right = fDivide(fTDP_Power_right, fExponential(fMultiply( - fAdd(fMultiply(fKv_m_fused, fT_prod), fKv_b_fused), fV_FT))); - fTDP_Power_right = fDivide(fTDP_Power_right, fExponential(fMultiply( - fKt_Beta_fused, fT_FT))); - - fTDP_Power = fAdd(fTDP_Power_left, fTDP_Power_right); - - fTDP_Current = fDivide(fTDP_Power, fV_x); - - fV_NL = fAdd(fV_x, fDivide(fMultiply(fTDP_Current, fRLL_LoadLine), - ConvertToFraction(10))); - - fV_NL = fRoundUpByStepSize(fV_NL, fStepSize, 0); - - if (GreaterThan(fV_max, fV_NL) && - (GreaterThan(fV_NL, fEVV_V) || - Equal(fV_NL, fEVV_V))) { - fV_NL = fMultiply(fV_NL, ConvertToFraction(1000)); - - *voltage = (uint16_t)fV_NL.partial.real; - break; - } else - fV_x = fAdd(fV_x, fStepSize); - } - - return result; -} - /** * atomctrl_get_voltage_evv_on_sclk: gets voltage via call to ATOM COMMAND table. * @hwmgr: input: pointer to hwManager diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/ppatomctrl.h b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/ppatomctrl.h index 1f987e846628..22b0ac12df97 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/ppatomctrl.h +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/ppatomctrl.h @@ -316,8 +316,6 @@ extern int atomctrl_get_engine_pll_dividers_kong(struct pp_hwmgr *hwmgr, pp_atomctrl_clock_dividers_kong *dividers); extern int atomctrl_read_efuse(struct pp_hwmgr *hwmgr, uint16_t start_index, uint16_t end_index, uint32_t *efuse); -extern int atomctrl_calculate_voltage_evv_on_sclk(struct pp_hwmgr *hwmgr, uint8_t voltage_type, - uint32_t sclk, uint16_t virtual_voltage_Id, uint16_t *voltage, uint16_t dpm_level, bool debug); extern int atomctrl_get_engine_pll_dividers_ai(struct pp_hwmgr *hwmgr, uint32_t clock_value, pp_atomctrl_clock_dividers_ai *dividers); extern int atomctrl_set_ac_timing_ai(struct pp_hwmgr *hwmgr, uint32_t memory_clock, uint8_t level); diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/ppevvmath.h b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/ppevvmath.h deleted file mode 100644 index 409aeec6baa9..000000000000 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/ppevvmath.h +++ /dev/null @@ -1,561 +0,0 @@ -/* - * Copyright 2015 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR - * OTHER DEALINGS IN THE SOFTWARE. - * - */ -#include - -enum ppevvmath_constants { - /* We multiply all original integers with 2^SHIFT_AMOUNT to get the fInt representation */ - SHIFT_AMOUNT = 16, - - /* Change this value to change the number of decimal places in the final output - 5 is a good default */ - PRECISION = 5, - - SHIFTED_2 = (2 << SHIFT_AMOUNT), - - /* 32767 - Might change in the future */ - MAX = (1 << (SHIFT_AMOUNT - 1)) - 1, -}; - -/* ------------------------------------------------------------------------------- - * NEW TYPE - fINT - * ------------------------------------------------------------------------------- - * A variable of type fInt can be accessed in 3 ways using the dot (.) operator - * fInt A; - * A.full => The full number as it is. Generally not easy to read - * A.partial.real => Only the integer portion - * A.partial.decimal => Only the fractional portion - */ -typedef union _fInt { - int full; - struct _partial { - unsigned int decimal: SHIFT_AMOUNT; /*Needs to always be unsigned*/ - int real: 32 - SHIFT_AMOUNT; - } partial; -} fInt; - -/* ------------------------------------------------------------------------------- - * Function Declarations - * ------------------------------------------------------------------------------- - */ -static fInt ConvertToFraction(int); /* Use this to convert an INT to a FINT */ -static fInt Convert_ULONG_ToFraction(uint32_t); /* Use this to convert an uint32_t to a FINT */ -static fInt GetScaledFraction(int, int); /* Use this to convert an INT to a FINT after scaling it by a factor */ -static int ConvertBackToInteger(fInt); /* Convert a FINT back to an INT that is scaled by 1000 (i.e. last 3 digits are the decimal digits) */ - -static fInt fNegate(fInt); /* Returns -1 * input fInt value */ -static fInt fAdd (fInt, fInt); /* Returns the sum of two fInt numbers */ -static fInt fSubtract (fInt A, fInt B); /* Returns A-B - Sometimes easier than Adding negative numbers */ -static fInt fMultiply (fInt, fInt); /* Returns the product of two fInt numbers */ -static fInt fDivide (fInt A, fInt B); /* Returns A/B */ -static fInt fGetSquare(fInt); /* Returns the square of a fInt number */ -static fInt fSqrt(fInt); /* Returns the Square Root of a fInt number */ - -static int uAbs(int); /* Returns the Absolute value of the Int */ -static int uPow(int base, int exponent); /* Returns base^exponent an INT */ - -static void SolveQuadracticEqn(fInt, fInt, fInt, fInt[]); /* Returns the 2 roots via the array */ -static bool Equal(fInt, fInt); /* Returns true if two fInts are equal to each other */ -static bool GreaterThan(fInt A, fInt B); /* Returns true if A > B */ - -static fInt fExponential(fInt exponent); /* Can be used to calculate e^exponent */ -static fInt fNaturalLog(fInt value); /* Can be used to calculate ln(value) */ - -/* Fuse decoding functions - * ------------------------------------------------------------------------------------- - */ -static fInt fDecodeLinearFuse(uint32_t fuse_value, fInt f_min, fInt f_range, uint32_t bitlength); -static fInt fDecodeLogisticFuse(uint32_t fuse_value, fInt f_average, fInt f_range, uint32_t bitlength); -static fInt fDecodeLeakageID (uint32_t leakageID_fuse, fInt ln_max_div_min, fInt f_min, uint32_t bitlength); - -/* Internal Support Functions - Use these ONLY for testing or adding to internal functions - * ------------------------------------------------------------------------------------- - * Some of the following functions take two INTs as their input - This is unsafe for a variety of reasons. - */ -static fInt Divide (int, int); /* Divide two INTs and return result as FINT */ -static fInt fNegate(fInt); - -static int uGetScaledDecimal (fInt); /* Internal function */ -static int GetReal (fInt A); /* Internal function */ - -/* ------------------------------------------------------------------------------------- - * TROUBLESHOOTING INFORMATION - * ------------------------------------------------------------------------------------- - * 1) ConvertToFraction - InputOutOfRangeException: Only accepts numbers smaller than MAX (default: 32767) - * 2) fAdd - OutputOutOfRangeException: Output bigger than MAX (default: 32767) - * 3) fMultiply - OutputOutOfRangeException: - * 4) fGetSquare - OutputOutOfRangeException: - * 5) fDivide - DivideByZeroException - * 6) fSqrt - NegativeSquareRootException: Input cannot be a negative number - */ - -/* ------------------------------------------------------------------------------------- - * START OF CODE - * ------------------------------------------------------------------------------------- - */ -static fInt fExponential(fInt exponent) /*Can be used to calculate e^exponent*/ -{ - uint32_t i; - bool bNegated = false; - - fInt fPositiveOne = ConvertToFraction(1); - fInt fZERO = ConvertToFraction(0); - - fInt lower_bound = Divide(78, 10000); - fInt solution = fPositiveOne; /*Starting off with baseline of 1 */ - fInt error_term; - - static const uint32_t k_array[11] = {55452, 27726, 13863, 6931, 4055, 2231, 1178, 606, 308, 155, 78}; - static const uint32_t expk_array[11] = {2560000, 160000, 40000, 20000, 15000, 12500, 11250, 10625, 10313, 10156, 10078}; - - if (GreaterThan(fZERO, exponent)) { - exponent = fNegate(exponent); - bNegated = true; - } - - while (GreaterThan(exponent, lower_bound)) { - for (i = 0; i < 11; i++) { - if (GreaterThan(exponent, GetScaledFraction(k_array[i], 10000))) { - exponent = fSubtract(exponent, GetScaledFraction(k_array[i], 10000)); - solution = fMultiply(solution, GetScaledFraction(expk_array[i], 10000)); - } - } - } - - error_term = fAdd(fPositiveOne, exponent); - - solution = fMultiply(solution, error_term); - - if (bNegated) - solution = fDivide(fPositiveOne, solution); - - return solution; -} - -static fInt fNaturalLog(fInt value) -{ - uint32_t i; - fInt upper_bound = Divide(8, 1000); - fInt fNegativeOne = ConvertToFraction(-1); - fInt solution = ConvertToFraction(0); /*Starting off with baseline of 0 */ - fInt error_term; - - static const uint32_t k_array[10] = {160000, 40000, 20000, 15000, 12500, 11250, 10625, 10313, 10156, 10078}; - static const uint32_t logk_array[10] = {27726, 13863, 6931, 4055, 2231, 1178, 606, 308, 155, 78}; - - while (GreaterThan(fAdd(value, fNegativeOne), upper_bound)) { - for (i = 0; i < 10; i++) { - if (GreaterThan(value, GetScaledFraction(k_array[i], 10000))) { - value = fDivide(value, GetScaledFraction(k_array[i], 10000)); - solution = fAdd(solution, GetScaledFraction(logk_array[i], 10000)); - } - } - } - - error_term = fAdd(fNegativeOne, value); - - return fAdd(solution, error_term); -} - -static fInt fDecodeLinearFuse(uint32_t fuse_value, fInt f_min, fInt f_range, uint32_t bitlength) -{ - fInt f_fuse_value = Convert_ULONG_ToFraction(fuse_value); - fInt f_bit_max_value = Convert_ULONG_ToFraction((uPow(2, bitlength)) - 1); - - fInt f_decoded_value; - - f_decoded_value = fDivide(f_fuse_value, f_bit_max_value); - f_decoded_value = fMultiply(f_decoded_value, f_range); - f_decoded_value = fAdd(f_decoded_value, f_min); - - return f_decoded_value; -} - - -static fInt fDecodeLogisticFuse(uint32_t fuse_value, fInt f_average, fInt f_range, uint32_t bitlength) -{ - fInt f_fuse_value = Convert_ULONG_ToFraction(fuse_value); - fInt f_bit_max_value = Convert_ULONG_ToFraction((uPow(2, bitlength)) - 1); - - fInt f_CONSTANT_NEG13 = ConvertToFraction(-13); - fInt f_CONSTANT1 = ConvertToFraction(1); - - fInt f_decoded_value; - - f_decoded_value = fSubtract(fDivide(f_bit_max_value, f_fuse_value), f_CONSTANT1); - f_decoded_value = fNaturalLog(f_decoded_value); - f_decoded_value = fMultiply(f_decoded_value, fDivide(f_range, f_CONSTANT_NEG13)); - f_decoded_value = fAdd(f_decoded_value, f_average); - - return f_decoded_value; -} - -static fInt fDecodeLeakageID (uint32_t leakageID_fuse, fInt ln_max_div_min, fInt f_min, uint32_t bitlength) -{ - fInt fLeakage; - fInt f_bit_max_value = Convert_ULONG_ToFraction((uPow(2, bitlength)) - 1); - - fLeakage = fMultiply(ln_max_div_min, Convert_ULONG_ToFraction(leakageID_fuse)); - fLeakage = fDivide(fLeakage, f_bit_max_value); - fLeakage = fExponential(fLeakage); - fLeakage = fMultiply(fLeakage, f_min); - - return fLeakage; -} - -static fInt ConvertToFraction(int X) /*Add all range checking here. Is it possible to make fInt a private declaration? */ -{ - fInt temp; - - if (X <= MAX) - temp.full = (X << SHIFT_AMOUNT); - else - temp.full = 0; - - return temp; -} - -static fInt fNegate(fInt X) -{ - fInt CONSTANT_NEGONE = ConvertToFraction(-1); - return fMultiply(X, CONSTANT_NEGONE); -} - -static fInt Convert_ULONG_ToFraction(uint32_t X) -{ - fInt temp; - - if (X <= MAX) - temp.full = (X << SHIFT_AMOUNT); - else - temp.full = 0; - - return temp; -} - -static fInt GetScaledFraction(int X, int factor) -{ - int times_shifted, factor_shifted; - bool bNEGATED; - fInt fValue; - - times_shifted = 0; - factor_shifted = 0; - bNEGATED = false; - - if (X < 0) { - X = -1*X; - bNEGATED = true; - } - - if (factor < 0) { - factor = -1*factor; - bNEGATED = !bNEGATED; /*If bNEGATED = true due to X < 0, this will cover the case of negative cancelling negative */ - } - - if ((X > MAX) || factor > MAX) { - if ((X/factor) <= MAX) { - while (X > MAX) { - X = X >> 1; - times_shifted++; - } - - while (factor > MAX) { - factor = factor >> 1; - factor_shifted++; - } - } else { - fValue.full = 0; - return fValue; - } - } - - if (factor == 1) - return ConvertToFraction(X); - - fValue = fDivide(ConvertToFraction(X * uPow(-1, bNEGATED)), ConvertToFraction(factor)); - - fValue.full = fValue.full << times_shifted; - fValue.full = fValue.full >> factor_shifted; - - return fValue; -} - -/* Addition using two fInts */ -static fInt fAdd (fInt X, fInt Y) -{ - fInt Sum; - - Sum.full = X.full + Y.full; - - return Sum; -} - -/* Addition using two fInts */ -static fInt fSubtract (fInt X, fInt Y) -{ - fInt Difference; - - Difference.full = X.full - Y.full; - - return Difference; -} - -static bool Equal(fInt A, fInt B) -{ - if (A.full == B.full) - return true; - else - return false; -} - -static bool GreaterThan(fInt A, fInt B) -{ - if (A.full > B.full) - return true; - else - return false; -} - -static fInt fMultiply (fInt X, fInt Y) /* Uses 64-bit integers (int64_t) */ -{ - fInt Product; - int64_t tempProduct; - - /*The following is for a very specific common case: Non-zero number with ONLY fractional portion*/ - /* TEMPORARILY DISABLED - CAN BE USED TO IMPROVE PRECISION - bool X_LessThanOne, Y_LessThanOne; - - X_LessThanOne = (X.partial.real == 0 && X.partial.decimal != 0 && X.full >= 0); - Y_LessThanOne = (Y.partial.real == 0 && Y.partial.decimal != 0 && Y.full >= 0); - - if (X_LessThanOne && Y_LessThanOne) { - Product.full = X.full * Y.full; - return Product - }*/ - - tempProduct = ((int64_t)X.full) * ((int64_t)Y.full); /*Q(16,16)*Q(16,16) = Q(32, 32) - Might become a negative number! */ - tempProduct = tempProduct >> 16; /*Remove lagging 16 bits - Will lose some precision from decimal; */ - Product.full = (int)tempProduct; /*The int64_t will lose the leading 16 bits that were part of the integer portion */ - - return Product; -} - -static fInt fDivide (fInt X, fInt Y) -{ - fInt fZERO, fQuotient; - int64_t longlongX, longlongY; - - fZERO = ConvertToFraction(0); - - if (Equal(Y, fZERO)) - return fZERO; - - longlongX = (int64_t)X.full; - longlongY = (int64_t)Y.full; - - longlongX = longlongX << 16; /*Q(16,16) -> Q(32,32) */ - - div64_s64(longlongX, longlongY); /*Q(32,32) divided by Q(16,16) = Q(16,16) Back to original format */ - - fQuotient.full = (int)longlongX; - return fQuotient; -} - -static int ConvertBackToInteger (fInt A) /*THIS is the function that will be used to check with the Golden settings table*/ -{ - fInt fullNumber, scaledDecimal, scaledReal; - - scaledReal.full = GetReal(A) * uPow(10, PRECISION-1); /* DOUBLE CHECK THISSSS!!! */ - - scaledDecimal.full = uGetScaledDecimal(A); - - fullNumber = fAdd(scaledDecimal, scaledReal); - - return fullNumber.full; -} - -static fInt fGetSquare(fInt A) -{ - return fMultiply(A, A); -} - -/* x_new = x_old - (x_old^2 - C) / (2 * x_old) */ -static fInt fSqrt(fInt num) -{ - fInt F_divide_Fprime, Fprime; - fInt test; - fInt twoShifted; - int seed, counter, error; - fInt x_new, x_old, C, y; - - fInt fZERO = ConvertToFraction(0); - - /* (0 > num) is the same as (num < 0), i.e., num is negative */ - - if (GreaterThan(fZERO, num) || Equal(fZERO, num)) - return fZERO; - - C = num; - - if (num.partial.real > 3000) - seed = 60; - else if (num.partial.real > 1000) - seed = 30; - else if (num.partial.real > 100) - seed = 10; - else - seed = 2; - - counter = 0; - - if (Equal(num, fZERO)) /*Square Root of Zero is zero */ - return fZERO; - - twoShifted = ConvertToFraction(2); - x_new = ConvertToFraction(seed); - - do { - counter++; - - x_old.full = x_new.full; - - test = fGetSquare(x_old); /*1.75*1.75 is reverting back to 1 when shifted down */ - y = fSubtract(test, C); /*y = f(x) = x^2 - C; */ - - Fprime = fMultiply(twoShifted, x_old); - F_divide_Fprime = fDivide(y, Fprime); - - x_new = fSubtract(x_old, F_divide_Fprime); - - error = ConvertBackToInteger(x_new) - ConvertBackToInteger(x_old); - - if (counter > 20) /*20 is already way too many iterations. If we dont have an answer by then, we never will*/ - return x_new; - - } while (uAbs(error) > 0); - - return x_new; -} - -static void SolveQuadracticEqn(fInt A, fInt B, fInt C, fInt Roots[]) -{ - fInt *pRoots = &Roots[0]; - fInt temp, root_first, root_second; - fInt f_CONSTANT10, f_CONSTANT100; - - f_CONSTANT100 = ConvertToFraction(100); - f_CONSTANT10 = ConvertToFraction(10); - - while (GreaterThan(A, f_CONSTANT100) || GreaterThan(B, f_CONSTANT100) || GreaterThan(C, f_CONSTANT100)) { - A = fDivide(A, f_CONSTANT10); - B = fDivide(B, f_CONSTANT10); - C = fDivide(C, f_CONSTANT10); - } - - temp = fMultiply(ConvertToFraction(4), A); /* root = 4*A */ - temp = fMultiply(temp, C); /* root = 4*A*C */ - temp = fSubtract(fGetSquare(B), temp); /* root = b^2 - 4AC */ - temp = fSqrt(temp); /*root = Sqrt (b^2 - 4AC); */ - - root_first = fSubtract(fNegate(B), temp); /* b - Sqrt(b^2 - 4AC) */ - root_second = fAdd(fNegate(B), temp); /* b + Sqrt(b^2 - 4AC) */ - - root_first = fDivide(root_first, ConvertToFraction(2)); /* [b +- Sqrt(b^2 - 4AC)]/[2] */ - root_first = fDivide(root_first, A); /*[b +- Sqrt(b^2 - 4AC)]/[2*A] */ - - root_second = fDivide(root_second, ConvertToFraction(2)); /* [b +- Sqrt(b^2 - 4AC)]/[2] */ - root_second = fDivide(root_second, A); /*[b +- Sqrt(b^2 - 4AC)]/[2*A] */ - - *(pRoots + 0) = root_first; - *(pRoots + 1) = root_second; -} - -/* ----------------------------------------------------------------------------- - * SUPPORT FUNCTIONS - * ----------------------------------------------------------------------------- - */ - -/* Conversion Functions */ -static int GetReal (fInt A) -{ - return (A.full >> SHIFT_AMOUNT); -} - -static fInt Divide (int X, int Y) -{ - fInt A, B, Quotient; - - A.full = X << SHIFT_AMOUNT; - B.full = Y << SHIFT_AMOUNT; - - Quotient = fDivide(A, B); - - return Quotient; -} - -static int uGetScaledDecimal (fInt A) /*Converts the fractional portion to whole integers - Costly function */ -{ - int dec[PRECISION]; - int i, scaledDecimal = 0, tmp = A.partial.decimal; - - for (i = 0; i < PRECISION; i++) { - dec[i] = tmp / (1 << SHIFT_AMOUNT); - tmp = tmp - ((1 << SHIFT_AMOUNT)*dec[i]); - tmp *= 10; - scaledDecimal = scaledDecimal + dec[i]*uPow(10, PRECISION - 1 - i); - } - - return scaledDecimal; -} - -static int uPow(int base, int power) -{ - if (power == 0) - return 1; - else - return (base)*uPow(base, power - 1); -} - -static int uAbs(int X) -{ - if (X < 0) - return (X * -1); - else - return X; -} - -static fInt fRoundUpByStepSize(fInt A, fInt fStepSize, bool error_term) -{ - fInt solution; - - solution = fDivide(A, fStepSize); - solution.partial.decimal = 0; /*All fractional digits changes to 0 */ - - if (error_term) - solution.partial.real += 1; /*Error term of 1 added */ - - solution = fMultiply(solution, fStepSize); - solution = fAdd(solution, fStepSize); - - return solution; -} - diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_processpptables.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_processpptables.c index 79c817752a33..2b446f8866ba 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_processpptables.c +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_processpptables.c @@ -62,578 +62,6 @@ static const void *get_powerplay_table(struct pp_hwmgr *hwmgr) return table_address; } -#if 0 -static void dump_pptable(PPTable_t *pptable) -{ - int i; - - pr_info("Version = 0x%08x\n", pptable->Version); - - pr_info("FeaturesToRun[0] = 0x%08x\n", pptable->FeaturesToRun[0]); - pr_info("FeaturesToRun[1] = 0x%08x\n", pptable->FeaturesToRun[1]); - - pr_info("SocketPowerLimitAc0 = %d\n", pptable->SocketPowerLimitAc0); - pr_info("SocketPowerLimitAc0Tau = %d\n", pptable->SocketPowerLimitAc0Tau); - pr_info("SocketPowerLimitAc1 = %d\n", pptable->SocketPowerLimitAc1); - pr_info("SocketPowerLimitAc1Tau = %d\n", pptable->SocketPowerLimitAc1Tau); - pr_info("SocketPowerLimitAc2 = %d\n", pptable->SocketPowerLimitAc2); - pr_info("SocketPowerLimitAc2Tau = %d\n", pptable->SocketPowerLimitAc2Tau); - pr_info("SocketPowerLimitAc3 = %d\n", pptable->SocketPowerLimitAc3); - pr_info("SocketPowerLimitAc3Tau = %d\n", pptable->SocketPowerLimitAc3Tau); - pr_info("SocketPowerLimitDc = %d\n", pptable->SocketPowerLimitDc); - pr_info("SocketPowerLimitDcTau = %d\n", pptable->SocketPowerLimitDcTau); - pr_info("TdcLimitSoc = %d\n", pptable->TdcLimitSoc); - pr_info("TdcLimitSocTau = %d\n", pptable->TdcLimitSocTau); - pr_info("TdcLimitGfx = %d\n", pptable->TdcLimitGfx); - pr_info("TdcLimitGfxTau = %d\n", pptable->TdcLimitGfxTau); - - pr_info("TedgeLimit = %d\n", pptable->TedgeLimit); - pr_info("ThotspotLimit = %d\n", pptable->ThotspotLimit); - pr_info("ThbmLimit = %d\n", pptable->ThbmLimit); - pr_info("Tvr_gfxLimit = %d\n", pptable->Tvr_gfxLimit); - pr_info("Tvr_memLimit = %d\n", pptable->Tvr_memLimit); - pr_info("Tliquid1Limit = %d\n", pptable->Tliquid1Limit); - pr_info("Tliquid2Limit = %d\n", pptable->Tliquid2Limit); - pr_info("TplxLimit = %d\n", pptable->TplxLimit); - pr_info("FitLimit = %d\n", pptable->FitLimit); - - pr_info("PpmPowerLimit = %d\n", pptable->PpmPowerLimit); - pr_info("PpmTemperatureThreshold = %d\n", pptable->PpmTemperatureThreshold); - - pr_info("MemoryOnPackage = 0x%02x\n", pptable->MemoryOnPackage); - pr_info("padding8_limits = 0x%02x\n", pptable->padding8_limits); - pr_info("Tvr_SocLimit = %d\n", pptable->Tvr_SocLimit); - - pr_info("UlvVoltageOffsetSoc = %d\n", pptable->UlvVoltageOffsetSoc); - pr_info("UlvVoltageOffsetGfx = %d\n", pptable->UlvVoltageOffsetGfx); - - pr_info("UlvSmnclkDid = %d\n", pptable->UlvSmnclkDid); - pr_info("UlvMp1clkDid = %d\n", pptable->UlvMp1clkDid); - pr_info("UlvGfxclkBypass = %d\n", pptable->UlvGfxclkBypass); - pr_info("Padding234 = 0x%02x\n", pptable->Padding234); - - pr_info("MinVoltageGfx = %d\n", pptable->MinVoltageGfx); - pr_info("MinVoltageSoc = %d\n", pptable->MinVoltageSoc); - pr_info("MaxVoltageGfx = %d\n", pptable->MaxVoltageGfx); - pr_info("MaxVoltageSoc = %d\n", pptable->MaxVoltageSoc); - - pr_info("LoadLineResistanceGfx = %d\n", pptable->LoadLineResistanceGfx); - pr_info("LoadLineResistanceSoc = %d\n", pptable->LoadLineResistanceSoc); - - pr_info("[PPCLK_GFXCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n", - pptable->DpmDescriptor[PPCLK_GFXCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_GFXCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_GFXCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_GFXCLK].padding, - pptable->DpmDescriptor[PPCLK_GFXCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_GFXCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsCurve.c); - - pr_info("[PPCLK_VCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n", - pptable->DpmDescriptor[PPCLK_VCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_VCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_VCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_VCLK].padding, - pptable->DpmDescriptor[PPCLK_VCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_VCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_VCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_VCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_VCLK].SsCurve.c); - - pr_info("[PPCLK_DCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n", - pptable->DpmDescriptor[PPCLK_DCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_DCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_DCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_DCLK].padding, - pptable->DpmDescriptor[PPCLK_DCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_DCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_DCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_DCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_DCLK].SsCurve.c); - - pr_info("[PPCLK_ECLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n", - pptable->DpmDescriptor[PPCLK_ECLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_ECLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_ECLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_ECLK].padding, - pptable->DpmDescriptor[PPCLK_ECLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_ECLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_ECLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_ECLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_ECLK].SsCurve.c); - - pr_info("[PPCLK_SOCCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n", - pptable->DpmDescriptor[PPCLK_SOCCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_SOCCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_SOCCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_SOCCLK].padding, - pptable->DpmDescriptor[PPCLK_SOCCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_SOCCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsCurve.c); - - pr_info("[PPCLK_UCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n", - pptable->DpmDescriptor[PPCLK_UCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_UCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_UCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_UCLK].padding, - pptable->DpmDescriptor[PPCLK_UCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_UCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_UCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_UCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_UCLK].SsCurve.c); - - pr_info("[PPCLK_DCEFCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n", - pptable->DpmDescriptor[PPCLK_DCEFCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_DCEFCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_DCEFCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_DCEFCLK].padding, - pptable->DpmDescriptor[PPCLK_DCEFCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_DCEFCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_DCEFCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_DCEFCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_DCEFCLK].SsCurve.c); - - pr_info("[PPCLK_DISPCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n", - pptable->DpmDescriptor[PPCLK_DISPCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_DISPCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_DISPCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_DISPCLK].padding, - pptable->DpmDescriptor[PPCLK_DISPCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_DISPCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_DISPCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_DISPCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_DISPCLK].SsCurve.c); - - pr_info("[PPCLK_PIXCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n", - pptable->DpmDescriptor[PPCLK_PIXCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_PIXCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_PIXCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_PIXCLK].padding, - pptable->DpmDescriptor[PPCLK_PIXCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_PIXCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_PIXCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_PIXCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_PIXCLK].SsCurve.c); - - pr_info("[PPCLK_PHYCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n", - pptable->DpmDescriptor[PPCLK_PHYCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_PHYCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_PHYCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_PHYCLK].padding, - pptable->DpmDescriptor[PPCLK_PHYCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_PHYCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_PHYCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_PHYCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_PHYCLK].SsCurve.c); - - pr_info("[PPCLK_FCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n", - pptable->DpmDescriptor[PPCLK_FCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_FCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_FCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_FCLK].padding, - pptable->DpmDescriptor[PPCLK_FCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_FCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_FCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_FCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_FCLK].SsCurve.c); - - - pr_info("FreqTableGfx\n"); - for (i = 0; i < NUM_GFXCLK_DPM_LEVELS; i++) - pr_info(" .[%02d] = %d\n", i, pptable->FreqTableGfx[i]); - - pr_info("FreqTableVclk\n"); - for (i = 0; i < NUM_VCLK_DPM_LEVELS; i++) - pr_info(" .[%02d] = %d\n", i, pptable->FreqTableVclk[i]); - - pr_info("FreqTableDclk\n"); - for (i = 0; i < NUM_DCLK_DPM_LEVELS; i++) - pr_info(" .[%02d] = %d\n", i, pptable->FreqTableDclk[i]); - - pr_info("FreqTableEclk\n"); - for (i = 0; i < NUM_ECLK_DPM_LEVELS; i++) - pr_info(" .[%02d] = %d\n", i, pptable->FreqTableEclk[i]); - - pr_info("FreqTableSocclk\n"); - for (i = 0; i < NUM_SOCCLK_DPM_LEVELS; i++) - pr_info(" .[%02d] = %d\n", i, pptable->FreqTableSocclk[i]); - - pr_info("FreqTableUclk\n"); - for (i = 0; i < NUM_UCLK_DPM_LEVELS; i++) - pr_info(" .[%02d] = %d\n", i, pptable->FreqTableUclk[i]); - - pr_info("FreqTableFclk\n"); - for (i = 0; i < NUM_FCLK_DPM_LEVELS; i++) - pr_info(" .[%02d] = %d\n", i, pptable->FreqTableFclk[i]); - - pr_info("FreqTableDcefclk\n"); - for (i = 0; i < NUM_DCEFCLK_DPM_LEVELS; i++) - pr_info(" .[%02d] = %d\n", i, pptable->FreqTableDcefclk[i]); - - pr_info("FreqTableDispclk\n"); - for (i = 0; i < NUM_DISPCLK_DPM_LEVELS; i++) - pr_info(" .[%02d] = %d\n", i, pptable->FreqTableDispclk[i]); - - pr_info("FreqTablePixclk\n"); - for (i = 0; i < NUM_PIXCLK_DPM_LEVELS; i++) - pr_info(" .[%02d] = %d\n", i, pptable->FreqTablePixclk[i]); - - pr_info("FreqTablePhyclk\n"); - for (i = 0; i < NUM_PHYCLK_DPM_LEVELS; i++) - pr_info(" .[%02d] = %d\n", i, pptable->FreqTablePhyclk[i]); - - pr_info("DcModeMaxFreq[PPCLK_GFXCLK] = %d\n", pptable->DcModeMaxFreq[PPCLK_GFXCLK]); - pr_info("DcModeMaxFreq[PPCLK_VCLK] = %d\n", pptable->DcModeMaxFreq[PPCLK_VCLK]); - pr_info("DcModeMaxFreq[PPCLK_DCLK] = %d\n", pptable->DcModeMaxFreq[PPCLK_DCLK]); - pr_info("DcModeMaxFreq[PPCLK_ECLK] = %d\n", pptable->DcModeMaxFreq[PPCLK_ECLK]); - pr_info("DcModeMaxFreq[PPCLK_SOCCLK] = %d\n", pptable->DcModeMaxFreq[PPCLK_SOCCLK]); - pr_info("DcModeMaxFreq[PPCLK_UCLK] = %d\n", pptable->DcModeMaxFreq[PPCLK_UCLK]); - pr_info("DcModeMaxFreq[PPCLK_DCEFCLK] = %d\n", pptable->DcModeMaxFreq[PPCLK_DCEFCLK]); - pr_info("DcModeMaxFreq[PPCLK_DISPCLK] = %d\n", pptable->DcModeMaxFreq[PPCLK_DISPCLK]); - pr_info("DcModeMaxFreq[PPCLK_PIXCLK] = %d\n", pptable->DcModeMaxFreq[PPCLK_PIXCLK]); - pr_info("DcModeMaxFreq[PPCLK_PHYCLK] = %d\n", pptable->DcModeMaxFreq[PPCLK_PHYCLK]); - pr_info("DcModeMaxFreq[PPCLK_FCLK] = %d\n", pptable->DcModeMaxFreq[PPCLK_FCLK]); - pr_info("Padding8_Clks = %d\n", pptable->Padding8_Clks); - - pr_info("Mp0clkFreq\n"); - for (i = 0; i < NUM_MP0CLK_DPM_LEVELS; i++) - pr_info(" .[%d] = %d\n", i, pptable->Mp0clkFreq[i]); - - pr_info("Mp0DpmVoltage\n"); - for (i = 0; i < NUM_MP0CLK_DPM_LEVELS; i++) - pr_info(" .[%d] = %d\n", i, pptable->Mp0DpmVoltage[i]); - - pr_info("GfxclkFidle = 0x%x\n", pptable->GfxclkFidle); - pr_info("GfxclkSlewRate = 0x%x\n", pptable->GfxclkSlewRate); - pr_info("CksEnableFreq = 0x%x\n", pptable->CksEnableFreq); - pr_info("Padding789 = 0x%x\n", pptable->Padding789); - pr_info("CksVoltageOffset[a = 0x%08x b = 0x%08x c = 0x%08x]\n", - pptable->CksVoltageOffset.a, - pptable->CksVoltageOffset.b, - pptable->CksVoltageOffset.c); - pr_info("Padding567[0] = 0x%x\n", pptable->Padding567[0]); - pr_info("Padding567[1] = 0x%x\n", pptable->Padding567[1]); - pr_info("Padding567[2] = 0x%x\n", pptable->Padding567[2]); - pr_info("Padding567[3] = 0x%x\n", pptable->Padding567[3]); - pr_info("GfxclkDsMaxFreq = %d\n", pptable->GfxclkDsMaxFreq); - pr_info("GfxclkSource = 0x%x\n", pptable->GfxclkSource); - pr_info("Padding456 = 0x%x\n", pptable->Padding456); - - pr_info("LowestUclkReservedForUlv = %d\n", pptable->LowestUclkReservedForUlv); - pr_info("Padding8_Uclk[0] = 0x%x\n", pptable->Padding8_Uclk[0]); - pr_info("Padding8_Uclk[1] = 0x%x\n", pptable->Padding8_Uclk[1]); - pr_info("Padding8_Uclk[2] = 0x%x\n", pptable->Padding8_Uclk[2]); - - pr_info("PcieGenSpeed\n"); - for (i = 0; i < NUM_LINK_LEVELS; i++) - pr_info(" .[%d] = %d\n", i, pptable->PcieGenSpeed[i]); - - pr_info("PcieLaneCount\n"); - for (i = 0; i < NUM_LINK_LEVELS; i++) - pr_info(" .[%d] = %d\n", i, pptable->PcieLaneCount[i]); - - pr_info("LclkFreq\n"); - for (i = 0; i < NUM_LINK_LEVELS; i++) - pr_info(" .[%d] = %d\n", i, pptable->LclkFreq[i]); - - pr_info("EnableTdpm = %d\n", pptable->EnableTdpm); - pr_info("TdpmHighHystTemperature = %d\n", pptable->TdpmHighHystTemperature); - pr_info("TdpmLowHystTemperature = %d\n", pptable->TdpmLowHystTemperature); - pr_info("GfxclkFreqHighTempLimit = %d\n", pptable->GfxclkFreqHighTempLimit); - - pr_info("FanStopTemp = %d\n", pptable->FanStopTemp); - pr_info("FanStartTemp = %d\n", pptable->FanStartTemp); - - pr_info("FanGainEdge = %d\n", pptable->FanGainEdge); - pr_info("FanGainHotspot = %d\n", pptable->FanGainHotspot); - pr_info("FanGainLiquid = %d\n", pptable->FanGainLiquid); - pr_info("FanGainVrGfx = %d\n", pptable->FanGainVrGfx); - pr_info("FanGainVrSoc = %d\n", pptable->FanGainVrSoc); - pr_info("FanGainPlx = %d\n", pptable->FanGainPlx); - pr_info("FanGainHbm = %d\n", pptable->FanGainHbm); - pr_info("FanPwmMin = %d\n", pptable->FanPwmMin); - pr_info("FanAcousticLimitRpm = %d\n", pptable->FanAcousticLimitRpm); - pr_info("FanThrottlingRpm = %d\n", pptable->FanThrottlingRpm); - pr_info("FanMaximumRpm = %d\n", pptable->FanMaximumRpm); - pr_info("FanTargetTemperature = %d\n", pptable->FanTargetTemperature); - pr_info("FanTargetGfxclk = %d\n", pptable->FanTargetGfxclk); - pr_info("FanZeroRpmEnable = %d\n", pptable->FanZeroRpmEnable); - pr_info("FanTachEdgePerRev = %d\n", pptable->FanTachEdgePerRev); - - pr_info("FuzzyFan_ErrorSetDelta = %d\n", pptable->FuzzyFan_ErrorSetDelta); - pr_info("FuzzyFan_ErrorRateSetDelta = %d\n", pptable->FuzzyFan_ErrorRateSetDelta); - pr_info("FuzzyFan_PwmSetDelta = %d\n", pptable->FuzzyFan_PwmSetDelta); - pr_info("FuzzyFan_Reserved = %d\n", pptable->FuzzyFan_Reserved); - - pr_info("OverrideAvfsGb[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->OverrideAvfsGb[AVFS_VOLTAGE_GFX]); - pr_info("OverrideAvfsGb[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->OverrideAvfsGb[AVFS_VOLTAGE_SOC]); - pr_info("Padding8_Avfs[0] = %d\n", pptable->Padding8_Avfs[0]); - pr_info("Padding8_Avfs[1] = %d\n", pptable->Padding8_Avfs[1]); - - pr_info("qAvfsGb[AVFS_VOLTAGE_GFX]{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->qAvfsGb[AVFS_VOLTAGE_GFX].a, - pptable->qAvfsGb[AVFS_VOLTAGE_GFX].b, - pptable->qAvfsGb[AVFS_VOLTAGE_GFX].c); - pr_info("qAvfsGb[AVFS_VOLTAGE_SOC]{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->qAvfsGb[AVFS_VOLTAGE_SOC].a, - pptable->qAvfsGb[AVFS_VOLTAGE_SOC].b, - pptable->qAvfsGb[AVFS_VOLTAGE_SOC].c); - pr_info("dBtcGbGfxCksOn{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->dBtcGbGfxCksOn.a, - pptable->dBtcGbGfxCksOn.b, - pptable->dBtcGbGfxCksOn.c); - pr_info("dBtcGbGfxCksOff{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->dBtcGbGfxCksOff.a, - pptable->dBtcGbGfxCksOff.b, - pptable->dBtcGbGfxCksOff.c); - pr_info("dBtcGbGfxAfll{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->dBtcGbGfxAfll.a, - pptable->dBtcGbGfxAfll.b, - pptable->dBtcGbGfxAfll.c); - pr_info("dBtcGbSoc{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->dBtcGbSoc.a, - pptable->dBtcGbSoc.b, - pptable->dBtcGbSoc.c); - pr_info("qAgingGb[AVFS_VOLTAGE_GFX]{m = 0x%x b = 0x%x}\n", - pptable->qAgingGb[AVFS_VOLTAGE_GFX].m, - pptable->qAgingGb[AVFS_VOLTAGE_GFX].b); - pr_info("qAgingGb[AVFS_VOLTAGE_SOC]{m = 0x%x b = 0x%x}\n", - pptable->qAgingGb[AVFS_VOLTAGE_SOC].m, - pptable->qAgingGb[AVFS_VOLTAGE_SOC].b); - - pr_info("qStaticVoltageOffset[AVFS_VOLTAGE_GFX]{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_GFX].a, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_GFX].b, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_GFX].c); - pr_info("qStaticVoltageOffset[AVFS_VOLTAGE_SOC]{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_SOC].a, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_SOC].b, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_SOC].c); - - pr_info("DcTol[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcTol[AVFS_VOLTAGE_GFX]); - pr_info("DcTol[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcTol[AVFS_VOLTAGE_SOC]); - - pr_info("DcBtcEnabled[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcEnabled[AVFS_VOLTAGE_GFX]); - pr_info("DcBtcEnabled[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcEnabled[AVFS_VOLTAGE_SOC]); - pr_info("Padding8_GfxBtc[0] = 0x%x\n", pptable->Padding8_GfxBtc[0]); - pr_info("Padding8_GfxBtc[1] = 0x%x\n", pptable->Padding8_GfxBtc[1]); - - pr_info("DcBtcMin[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcMin[AVFS_VOLTAGE_GFX]); - pr_info("DcBtcMin[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcMin[AVFS_VOLTAGE_SOC]); - pr_info("DcBtcMax[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcMax[AVFS_VOLTAGE_GFX]); - pr_info("DcBtcMax[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcMax[AVFS_VOLTAGE_SOC]); - - pr_info("XgmiLinkSpeed\n"); - for (i = 0; i < NUM_XGMI_LEVELS; i++) - pr_info(" .[%d] = %d\n", i, pptable->XgmiLinkSpeed[i]); - pr_info("XgmiLinkWidth\n"); - for (i = 0; i < NUM_XGMI_LEVELS; i++) - pr_info(" .[%d] = %d\n", i, pptable->XgmiLinkWidth[i]); - pr_info("XgmiFclkFreq\n"); - for (i = 0; i < NUM_XGMI_LEVELS; i++) - pr_info(" .[%d] = %d\n", i, pptable->XgmiFclkFreq[i]); - pr_info("XgmiUclkFreq\n"); - for (i = 0; i < NUM_XGMI_LEVELS; i++) - pr_info(" .[%d] = %d\n", i, pptable->XgmiUclkFreq[i]); - pr_info("XgmiSocclkFreq\n"); - for (i = 0; i < NUM_XGMI_LEVELS; i++) - pr_info(" .[%d] = %d\n", i, pptable->XgmiSocclkFreq[i]); - pr_info("XgmiSocVoltage\n"); - for (i = 0; i < NUM_XGMI_LEVELS; i++) - pr_info(" .[%d] = %d\n", i, pptable->XgmiSocVoltage[i]); - - pr_info("DebugOverrides = 0x%x\n", pptable->DebugOverrides); - pr_info("ReservedEquation0{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation0.a, - pptable->ReservedEquation0.b, - pptable->ReservedEquation0.c); - pr_info("ReservedEquation1{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation1.a, - pptable->ReservedEquation1.b, - pptable->ReservedEquation1.c); - pr_info("ReservedEquation2{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation2.a, - pptable->ReservedEquation2.b, - pptable->ReservedEquation2.c); - pr_info("ReservedEquation3{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation3.a, - pptable->ReservedEquation3.b, - pptable->ReservedEquation3.c); - - pr_info("MinVoltageUlvGfx = %d\n", pptable->MinVoltageUlvGfx); - pr_info("MinVoltageUlvSoc = %d\n", pptable->MinVoltageUlvSoc); - - pr_info("MGpuFanBoostLimitRpm = %d\n", pptable->MGpuFanBoostLimitRpm); - pr_info("padding16_Fan = %d\n", pptable->padding16_Fan); - - pr_info("FanGainVrMem0 = %d\n", pptable->FanGainVrMem0); - pr_info("FanGainVrMem0 = %d\n", pptable->FanGainVrMem0); - - pr_info("DcBtcGb[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcGb[AVFS_VOLTAGE_GFX]); - pr_info("DcBtcGb[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcGb[AVFS_VOLTAGE_SOC]); - - for (i = 0; i < 11; i++) - pr_info("Reserved[%d] = 0x%x\n", i, pptable->Reserved[i]); - - for (i = 0; i < 3; i++) - pr_info("Padding32[%d] = 0x%x\n", i, pptable->Padding32[i]); - - pr_info("MaxVoltageStepGfx = 0x%x\n", pptable->MaxVoltageStepGfx); - pr_info("MaxVoltageStepSoc = 0x%x\n", pptable->MaxVoltageStepSoc); - - pr_info("VddGfxVrMapping = 0x%x\n", pptable->VddGfxVrMapping); - pr_info("VddSocVrMapping = 0x%x\n", pptable->VddSocVrMapping); - pr_info("VddMem0VrMapping = 0x%x\n", pptable->VddMem0VrMapping); - pr_info("VddMem1VrMapping = 0x%x\n", pptable->VddMem1VrMapping); - - pr_info("GfxUlvPhaseSheddingMask = 0x%x\n", pptable->GfxUlvPhaseSheddingMask); - pr_info("SocUlvPhaseSheddingMask = 0x%x\n", pptable->SocUlvPhaseSheddingMask); - pr_info("ExternalSensorPresent = 0x%x\n", pptable->ExternalSensorPresent); - pr_info("Padding8_V = 0x%x\n", pptable->Padding8_V); - - pr_info("GfxMaxCurrent = 0x%x\n", pptable->GfxMaxCurrent); - pr_info("GfxOffset = 0x%x\n", pptable->GfxOffset); - pr_info("Padding_TelemetryGfx = 0x%x\n", pptable->Padding_TelemetryGfx); - - pr_info("SocMaxCurrent = 0x%x\n", pptable->SocMaxCurrent); - pr_info("SocOffset = 0x%x\n", pptable->SocOffset); - pr_info("Padding_TelemetrySoc = 0x%x\n", pptable->Padding_TelemetrySoc); - - pr_info("Mem0MaxCurrent = 0x%x\n", pptable->Mem0MaxCurrent); - pr_info("Mem0Offset = 0x%x\n", pptable->Mem0Offset); - pr_info("Padding_TelemetryMem0 = 0x%x\n", pptable->Padding_TelemetryMem0); - - pr_info("Mem1MaxCurrent = 0x%x\n", pptable->Mem1MaxCurrent); - pr_info("Mem1Offset = 0x%x\n", pptable->Mem1Offset); - pr_info("Padding_TelemetryMem1 = 0x%x\n", pptable->Padding_TelemetryMem1); - - pr_info("AcDcGpio = %d\n", pptable->AcDcGpio); - pr_info("AcDcPolarity = %d\n", pptable->AcDcPolarity); - pr_info("VR0HotGpio = %d\n", pptable->VR0HotGpio); - pr_info("VR0HotPolarity = %d\n", pptable->VR0HotPolarity); - - pr_info("VR1HotGpio = %d\n", pptable->VR1HotGpio); - pr_info("VR1HotPolarity = %d\n", pptable->VR1HotPolarity); - pr_info("Padding1 = 0x%x\n", pptable->Padding1); - pr_info("Padding2 = 0x%x\n", pptable->Padding2); - - pr_info("LedPin0 = %d\n", pptable->LedPin0); - pr_info("LedPin1 = %d\n", pptable->LedPin1); - pr_info("LedPin2 = %d\n", pptable->LedPin2); - pr_info("padding8_4 = 0x%x\n", pptable->padding8_4); - - pr_info("PllGfxclkSpreadEnabled = %d\n", pptable->PllGfxclkSpreadEnabled); - pr_info("PllGfxclkSpreadPercent = %d\n", pptable->PllGfxclkSpreadPercent); - pr_info("PllGfxclkSpreadFreq = %d\n", pptable->PllGfxclkSpreadFreq); - - pr_info("UclkSpreadEnabled = %d\n", pptable->UclkSpreadEnabled); - pr_info("UclkSpreadPercent = %d\n", pptable->UclkSpreadPercent); - pr_info("UclkSpreadFreq = %d\n", pptable->UclkSpreadFreq); - - pr_info("FclkSpreadEnabled = %d\n", pptable->FclkSpreadEnabled); - pr_info("FclkSpreadPercent = %d\n", pptable->FclkSpreadPercent); - pr_info("FclkSpreadFreq = %d\n", pptable->FclkSpreadFreq); - - pr_info("FllGfxclkSpreadEnabled = %d\n", pptable->FllGfxclkSpreadEnabled); - pr_info("FllGfxclkSpreadPercent = %d\n", pptable->FllGfxclkSpreadPercent); - pr_info("FllGfxclkSpreadFreq = %d\n", pptable->FllGfxclkSpreadFreq); - - for (i = 0; i < I2C_CONTROLLER_NAME_COUNT; i++) { - pr_info("I2cControllers[%d]:\n", i); - pr_info(" .Enabled = %d\n", - pptable->I2cControllers[i].Enabled); - pr_info(" .SlaveAddress = 0x%x\n", - pptable->I2cControllers[i].SlaveAddress); - pr_info(" .ControllerPort = %d\n", - pptable->I2cControllers[i].ControllerPort); - pr_info(" .ControllerName = %d\n", - pptable->I2cControllers[i].ControllerName); - pr_info(" .ThermalThrottler = %d\n", - pptable->I2cControllers[i].ThermalThrottler); - pr_info(" .I2cProtocol = %d\n", - pptable->I2cControllers[i].I2cProtocol); - pr_info(" .I2cSpeed = %d\n", - pptable->I2cControllers[i].I2cSpeed); - } - - for (i = 0; i < 10; i++) - pr_info("BoardReserved[%d] = 0x%x\n", i, pptable->BoardReserved[i]); - - for (i = 0; i < 8; i++) - pr_info("MmHubPadding[%d] = 0x%x\n", i, pptable->MmHubPadding[i]); -} -#endif - static int check_powerplay_tables( struct pp_hwmgr *hwmgr, const ATOM_Vega20_POWERPLAYTABLE *powerplay_table) @@ -652,8 +80,6 @@ static int check_powerplay_tables( return -EINVAL; } - //dump_pptable(&powerplay_table->smcPPTable); - return 0; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c index 80e60ea2d11e..64f917959576 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c +++ b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c @@ -140,7 +140,8 @@ int smu_set_soft_freq_range(struct smu_context *smu, ret = smu->ppt_funcs->set_soft_freq_limited_range(smu, clk_type, min, - max); + max, + false); return ret; } @@ -251,7 +252,7 @@ static int smu_dpm_set_vcn_enable(struct smu_context *smu, if (atomic_read(&power_gate->vcn_gated) ^ enable) return 0; - ret = smu->ppt_funcs->dpm_set_vcn_enable(smu, enable); + ret = smu->ppt_funcs->dpm_set_vcn_enable(smu, enable, 0xff); if (!ret) atomic_set(&power_gate->vcn_gated, !enable); @@ -549,7 +550,8 @@ bool is_support_sw_smu(struct amdgpu_device *adev) if (adev->asic_type == CHIP_VEGA20) return false; - if (amdgpu_ip_version(adev, MP1_HWIP, 0) >= IP_VERSION(11, 0, 0)) + if ((amdgpu_ip_version(adev, MP1_HWIP, 0) >= IP_VERSION(11, 0, 0)) && + amdgpu_device_ip_is_valid(adev, AMD_IP_BLOCK_TYPE_SMC)) return true; return false; @@ -741,9 +743,9 @@ static int smu_set_funcs(struct amdgpu_device *adev) return 0; } -static int smu_early_init(void *handle) +static int smu_early_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct smu_context *smu; int r; @@ -825,9 +827,9 @@ static int smu_apply_default_config_table_settings(struct smu_context *smu) return smu_set_config_table(smu, &adev->pm.config_table); } -static int smu_late_init(void *handle) +static int smu_late_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct smu_context *smu = adev->powerplay.pp_handle; int ret = 0; @@ -1242,9 +1244,9 @@ static bool smu_is_workload_profile_available(struct smu_context *smu, return smu->workload_map && smu->workload_map[profile].valid_mapping; } -static int smu_sw_init(void *handle) +static int smu_sw_init(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct smu_context *smu = adev->powerplay.pp_handle; int ret; @@ -1259,26 +1261,33 @@ static int smu_sw_init(void *handle) smu->watermarks_bitmap = 0; smu->power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT; smu->default_power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT; + smu->user_dpm_profile.user_workload_mask = 0; atomic_set(&smu->smu_power.power_gate.vcn_gated, 1); atomic_set(&smu->smu_power.power_gate.jpeg_gated, 1); atomic_set(&smu->smu_power.power_gate.vpe_gated, 1); atomic_set(&smu->smu_power.power_gate.umsch_mm_gated, 1); - smu->workload_prority[PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT] = 0; - smu->workload_prority[PP_SMC_POWER_PROFILE_FULLSCREEN3D] = 1; - smu->workload_prority[PP_SMC_POWER_PROFILE_POWERSAVING] = 2; - smu->workload_prority[PP_SMC_POWER_PROFILE_VIDEO] = 3; - smu->workload_prority[PP_SMC_POWER_PROFILE_VR] = 4; - smu->workload_prority[PP_SMC_POWER_PROFILE_COMPUTE] = 5; - smu->workload_prority[PP_SMC_POWER_PROFILE_CUSTOM] = 6; + smu->workload_priority[PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT] = 0; + smu->workload_priority[PP_SMC_POWER_PROFILE_FULLSCREEN3D] = 1; + smu->workload_priority[PP_SMC_POWER_PROFILE_POWERSAVING] = 2; + smu->workload_priority[PP_SMC_POWER_PROFILE_VIDEO] = 3; + smu->workload_priority[PP_SMC_POWER_PROFILE_VR] = 4; + smu->workload_priority[PP_SMC_POWER_PROFILE_COMPUTE] = 5; + smu->workload_priority[PP_SMC_POWER_PROFILE_CUSTOM] = 6; if (smu->is_apu || - !smu_is_workload_profile_available(smu, PP_SMC_POWER_PROFILE_FULLSCREEN3D)) - smu->workload_mask = 1 << smu->workload_prority[PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT]; - else - smu->workload_mask = 1 << smu->workload_prority[PP_SMC_POWER_PROFILE_FULLSCREEN3D]; + !smu_is_workload_profile_available(smu, PP_SMC_POWER_PROFILE_FULLSCREEN3D)) { + smu->driver_workload_mask = + 1 << smu->workload_priority[PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT]; + } else { + smu->driver_workload_mask = + 1 << smu->workload_priority[PP_SMC_POWER_PROFILE_FULLSCREEN3D]; + smu->default_power_profile_mode = PP_SMC_POWER_PROFILE_FULLSCREEN3D; + } + smu->workload_mask = smu->driver_workload_mask | + smu->user_dpm_profile.user_workload_mask; smu->workload_setting[0] = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT; smu->workload_setting[1] = PP_SMC_POWER_PROFILE_FULLSCREEN3D; smu->workload_setting[2] = PP_SMC_POWER_PROFILE_POWERSAVING; @@ -1326,9 +1335,9 @@ static int smu_sw_init(void *handle) return 0; } -static int smu_sw_fini(void *handle) +static int smu_sw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct smu_context *smu = adev->powerplay.pp_handle; int ret; @@ -1799,10 +1808,10 @@ static int smu_start_smc_engine(struct smu_context *smu) return ret; } -static int smu_hw_init(void *handle) +static int smu_hw_init(struct amdgpu_ip_block *ip_block) { int ret; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct smu_context *smu = adev->powerplay.pp_handle; if (amdgpu_sriov_vf(adev) && !amdgpu_sriov_is_pp_one_vf(adev)) { @@ -2021,9 +2030,9 @@ static int smu_reset_mp1_state(struct smu_context *smu) return ret; } -static int smu_hw_fini(void *handle) +static int smu_hw_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct smu_context *smu = adev->powerplay.pp_handle; int ret; @@ -2054,9 +2063,9 @@ static int smu_hw_fini(void *handle) return 0; } -static void smu_late_fini(void *handle) +static void smu_late_fini(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = handle; + struct amdgpu_device *adev = ip_block->adev; struct smu_context *smu = adev->powerplay.pp_handle; kfree(smu); @@ -2065,26 +2074,31 @@ static void smu_late_fini(void *handle) static int smu_reset(struct smu_context *smu) { struct amdgpu_device *adev = smu->adev; + struct amdgpu_ip_block *ip_block; int ret; - ret = smu_hw_fini(adev); + ip_block = amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_SMC); + if (!ip_block) + return -EINVAL; + + ret = smu_hw_fini(ip_block); if (ret) return ret; - ret = smu_hw_init(adev); + ret = smu_hw_init(ip_block); if (ret) return ret; - ret = smu_late_init(adev); + ret = smu_late_init(ip_block); if (ret) return ret; return 0; } -static int smu_suspend(void *handle) +static int smu_suspend(struct amdgpu_ip_block *ip_block) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct smu_context *smu = adev->powerplay.pp_handle; int ret; uint64_t count; @@ -2116,10 +2130,10 @@ static int smu_suspend(void *handle) return 0; } -static int smu_resume(void *handle) +static int smu_resume(struct amdgpu_ip_block *ip_block) { int ret; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_device *adev = ip_block->adev; struct smu_context *smu = adev->powerplay.pp_handle; if (amdgpu_sriov_vf(adev)&& !amdgpu_sriov_is_pp_one_vf(adev)) @@ -2348,17 +2362,20 @@ static int smu_switch_power_profile(void *handle, return -EINVAL; if (!en) { - smu->workload_mask &= ~(1 << smu->workload_prority[type]); + smu->driver_workload_mask &= ~(1 << smu->workload_priority[type]); index = fls(smu->workload_mask); index = index > 0 && index <= WORKLOAD_POLICY_MAX ? index - 1 : 0; workload[0] = smu->workload_setting[index]; } else { - smu->workload_mask |= (1 << smu->workload_prority[type]); + smu->driver_workload_mask |= (1 << smu->workload_priority[type]); index = fls(smu->workload_mask); index = index <= WORKLOAD_POLICY_MAX ? index - 1 : 0; workload[0] = smu->workload_setting[index]; } + smu->workload_mask = smu->driver_workload_mask | + smu->user_dpm_profile.user_workload_mask; + if (smu_dpm_ctx->dpm_level != AMD_DPM_FORCED_LEVEL_MANUAL && smu_dpm_ctx->dpm_level != AMD_DPM_FORCED_LEVEL_PERF_DETERMINISM) smu_bump_power_profile_mode(smu, workload, 0); @@ -2878,6 +2895,10 @@ static enum smu_clk_type smu_convert_to_smuclk(enum pp_clock_type type) clk_type = SMU_OD_FAN_TARGET_TEMPERATURE; break; case OD_FAN_MINIMUM_PWM: clk_type = SMU_OD_FAN_MINIMUM_PWM; break; + case OD_FAN_ZERO_RPM_ENABLE: + clk_type = SMU_OD_FAN_ZERO_RPM_ENABLE; break; + case OD_FAN_ZERO_RPM_STOP_TEMP: + clk_type = SMU_OD_FAN_ZERO_RPM_STOP_TEMP; break; default: clk_type = SMU_CLK_COUNT; break; } @@ -3049,12 +3070,23 @@ static int smu_set_power_profile_mode(void *handle, uint32_t param_size) { struct smu_context *smu = handle; + int ret; if (!smu->pm_enabled || !smu->adev->pm.dpm_enabled || !smu->ppt_funcs->set_power_profile_mode) return -EOPNOTSUPP; - return smu_bump_power_profile_mode(smu, param, param_size); + if (smu->user_dpm_profile.user_workload_mask & + (1 << smu->workload_priority[param[param_size]])) + return 0; + + smu->user_dpm_profile.user_workload_mask = + (1 << smu->workload_priority[param[param_size]]); + smu->workload_mask = smu->user_dpm_profile.user_workload_mask | + smu->driver_workload_mask; + ret = smu_bump_power_profile_mode(smu, param, param_size); + + return ret; } static int smu_get_fan_control_mode(void *handle, u32 *fan_mode) diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h index b44a185d07e8..d665c47f19b7 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h @@ -240,6 +240,7 @@ struct smu_user_dpm_profile { /* user clock state information */ uint32_t clk_mask[SMU_CLK_COUNT]; uint32_t clk_dependency; + uint32_t user_workload_mask; }; #define SMU_TABLE_INIT(tables, table_id, s, a, d) \ @@ -557,7 +558,8 @@ struct smu_context { bool disable_uclk_switch; uint32_t workload_mask; - uint32_t workload_prority[WORKLOAD_POLICY_MAX]; + uint32_t driver_workload_mask; + uint32_t workload_priority[WORKLOAD_POLICY_MAX]; uint32_t workload_setting[WORKLOAD_POLICY_MAX]; uint32_t power_profile_mode; uint32_t default_power_profile_mode; @@ -739,7 +741,7 @@ struct pptable_funcs { * @dpm_set_vcn_enable: Enable/disable VCN engine dynamic power * management. */ - int (*dpm_set_vcn_enable)(struct smu_context *smu, bool enable); + int (*dpm_set_vcn_enable)(struct smu_context *smu, bool enable, int inst); /** * @dpm_set_jpeg_enable: Enable/disable JPEG engine dynamic power @@ -858,11 +860,6 @@ struct pptable_funcs { */ int (*display_disable_memory_clock_switch)(struct smu_context *smu, bool disable_memory_clock_switch); - /** - * @dump_pptable: Print the power play table to the system log. - */ - void (*dump_pptable)(struct smu_context *smu); - /** * @get_power_limit: Get the device's power limits. */ @@ -1260,7 +1257,8 @@ struct pptable_funcs { * @set_soft_freq_limited_range: Set the soft frequency range of a clock * domain in MHz. */ - int (*set_soft_freq_limited_range)(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t min, uint32_t max); + int (*set_soft_freq_limited_range)(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t min, uint32_t max, + bool automatic); /** * @set_power_source: Notify the SMU of the current power source. diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_6_pmfw.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_6_pmfw.h index 822c6425d90e..0f96b8c59a0e 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_6_pmfw.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_6_pmfw.h @@ -123,7 +123,7 @@ typedef enum { VOLTAGE_GUARDBAND_COUNT } GFX_GUARDBAND_e; -#define SMU_METRICS_TABLE_VERSION 0xD +#define SMU_METRICS_TABLE_VERSION 0xE typedef struct __attribute__((packed, aligned(4))) { uint32_t AccumulationCounter; @@ -231,6 +231,9 @@ typedef struct __attribute__((packed, aligned(4))) { // PER XCD ACTIVITY uint32_t GfxBusy[8]; uint64_t GfxBusyAcc[8]; + + //PCIE BW Data and error count + uint32_t PCIeOtherEndRecoveryAcc; // The Pcie counter itself is accumulated } MetricsTableX_t; typedef struct __attribute__((packed, aligned(4))) { diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h index e71a721c12b9..a299dc4a8071 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h @@ -313,6 +313,8 @@ enum smu_clk_type { SMU_OD_ACOUSTIC_TARGET, SMU_OD_FAN_TARGET_TEMPERATURE, SMU_OD_FAN_MINIMUM_PWM, + SMU_OD_FAN_ZERO_RPM_ENABLE, + SMU_OD_FAN_ZERO_RPM_STOP_TEMP, SMU_CLK_COUNT, }; diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v11_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v11_0.h index c2ab336bb530..ed8304d82831 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v11_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v11_0.h @@ -255,7 +255,7 @@ int smu_v11_0_get_dpm_ultimate_freq(struct smu_context *smu, enum smu_clk_type c uint32_t *min, uint32_t *max); int smu_v11_0_set_soft_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type, - uint32_t min, uint32_t max); + uint32_t min, uint32_t max, bool automatic); int smu_v11_0_set_hard_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type, diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v12_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v12_0.h index 1ad2dff71090..0886d8cffbd0 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v12_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v12_0.h @@ -56,7 +56,7 @@ int smu_v12_0_set_default_dpm_tables(struct smu_context *smu); int smu_v12_0_mode2_reset(struct smu_context *smu); int smu_v12_0_set_soft_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type, - uint32_t min, uint32_t max); + uint32_t min, uint32_t max, bool automatic); int smu_v12_0_set_driver_table_location(struct smu_context *smu); diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h index e58220a7ee2f..ae3563d71fa0 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h @@ -219,7 +219,7 @@ int smu_v13_0_get_dpm_ultimate_freq(struct smu_context *smu, enum smu_clk_type c uint32_t *min, uint32_t *max); int smu_v13_0_set_soft_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type, - uint32_t min, uint32_t max); + uint32_t min, uint32_t max, bool automatic); int smu_v13_0_set_hard_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type, @@ -255,7 +255,8 @@ int smu_v13_0_wait_for_event(struct smu_context *smu, enum smu_event_type event, uint64_t event_arg); int smu_v13_0_set_vcn_enable(struct smu_context *smu, - bool enable); + bool enable, + int inst); int smu_v13_0_set_jpeg_enable(struct smu_context *smu, bool enable); diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h index 727d5b405435..0546b02e198d 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h @@ -186,7 +186,7 @@ int smu_v14_0_get_dpm_ultimate_freq(struct smu_context *smu, enum smu_clk_type c uint32_t *min, uint32_t *max); int smu_v14_0_set_soft_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type, - uint32_t min, uint32_t max); + uint32_t min, uint32_t max, bool automatic); int smu_v14_0_set_hard_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type, @@ -210,7 +210,8 @@ int smu_v14_0_wait_for_event(struct smu_context *smu, enum smu_event_type event, uint64_t event_arg); int smu_v14_0_set_vcn_enable(struct smu_context *smu, - bool enable); + bool enable, + int inst); int smu_v14_0_set_jpeg_enable(struct smu_context *smu, bool enable); diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c index c0f6b59369b7..4b36c230e43a 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c @@ -1455,7 +1455,6 @@ static int arcturus_set_power_profile_mode(struct smu_context *smu, return -EINVAL; } - if ((profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) && (smu->smc_fw_version >= 0x360d00)) { if (size != 10) @@ -1523,14 +1522,14 @@ static int arcturus_set_power_profile_mode(struct smu_context *smu, ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask, - 1 << workload_type, + smu->workload_mask, NULL); if (ret) { dev_err(smu->adev->dev, "Fail to set workload type %d\n", workload_type); return ret; } - smu->power_profile_mode = profile_mode; + smu_cmn_assign_power_profile(smu); return 0; } @@ -1559,437 +1558,6 @@ static int arcturus_set_performance_level(struct smu_context *smu, return smu_v11_0_set_performance_level(smu, level); } -static void arcturus_dump_pptable(struct smu_context *smu) -{ - struct smu_table_context *table_context = &smu->smu_table; - PPTable_t *pptable = table_context->driver_pptable; - int i; - - dev_info(smu->adev->dev, "Dumped PPTable:\n"); - - dev_info(smu->adev->dev, "Version = 0x%08x\n", pptable->Version); - - dev_info(smu->adev->dev, "FeaturesToRun[0] = 0x%08x\n", pptable->FeaturesToRun[0]); - dev_info(smu->adev->dev, "FeaturesToRun[1] = 0x%08x\n", pptable->FeaturesToRun[1]); - - for (i = 0; i < PPT_THROTTLER_COUNT; i++) { - dev_info(smu->adev->dev, "SocketPowerLimitAc[%d] = %d\n", i, pptable->SocketPowerLimitAc[i]); - dev_info(smu->adev->dev, "SocketPowerLimitAcTau[%d] = %d\n", i, pptable->SocketPowerLimitAcTau[i]); - } - - dev_info(smu->adev->dev, "TdcLimitSoc = %d\n", pptable->TdcLimitSoc); - dev_info(smu->adev->dev, "TdcLimitSocTau = %d\n", pptable->TdcLimitSocTau); - dev_info(smu->adev->dev, "TdcLimitGfx = %d\n", pptable->TdcLimitGfx); - dev_info(smu->adev->dev, "TdcLimitGfxTau = %d\n", pptable->TdcLimitGfxTau); - - dev_info(smu->adev->dev, "TedgeLimit = %d\n", pptable->TedgeLimit); - dev_info(smu->adev->dev, "ThotspotLimit = %d\n", pptable->ThotspotLimit); - dev_info(smu->adev->dev, "TmemLimit = %d\n", pptable->TmemLimit); - dev_info(smu->adev->dev, "Tvr_gfxLimit = %d\n", pptable->Tvr_gfxLimit); - dev_info(smu->adev->dev, "Tvr_memLimit = %d\n", pptable->Tvr_memLimit); - dev_info(smu->adev->dev, "Tvr_socLimit = %d\n", pptable->Tvr_socLimit); - dev_info(smu->adev->dev, "FitLimit = %d\n", pptable->FitLimit); - - dev_info(smu->adev->dev, "PpmPowerLimit = %d\n", pptable->PpmPowerLimit); - dev_info(smu->adev->dev, "PpmTemperatureThreshold = %d\n", pptable->PpmTemperatureThreshold); - - dev_info(smu->adev->dev, "ThrottlerControlMask = %d\n", pptable->ThrottlerControlMask); - - dev_info(smu->adev->dev, "UlvVoltageOffsetGfx = %d\n", pptable->UlvVoltageOffsetGfx); - dev_info(smu->adev->dev, "UlvPadding = 0x%08x\n", pptable->UlvPadding); - - dev_info(smu->adev->dev, "UlvGfxclkBypass = %d\n", pptable->UlvGfxclkBypass); - dev_info(smu->adev->dev, "Padding234[0] = 0x%02x\n", pptable->Padding234[0]); - dev_info(smu->adev->dev, "Padding234[1] = 0x%02x\n", pptable->Padding234[1]); - dev_info(smu->adev->dev, "Padding234[2] = 0x%02x\n", pptable->Padding234[2]); - - dev_info(smu->adev->dev, "MinVoltageGfx = %d\n", pptable->MinVoltageGfx); - dev_info(smu->adev->dev, "MinVoltageSoc = %d\n", pptable->MinVoltageSoc); - dev_info(smu->adev->dev, "MaxVoltageGfx = %d\n", pptable->MaxVoltageGfx); - dev_info(smu->adev->dev, "MaxVoltageSoc = %d\n", pptable->MaxVoltageSoc); - - dev_info(smu->adev->dev, "LoadLineResistanceGfx = %d\n", pptable->LoadLineResistanceGfx); - dev_info(smu->adev->dev, "LoadLineResistanceSoc = %d\n", pptable->LoadLineResistanceSoc); - - dev_info(smu->adev->dev, "[PPCLK_GFXCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_GFXCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_GFXCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_GFXCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_GFXCLK].padding, - pptable->DpmDescriptor[PPCLK_GFXCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_GFXCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsCurve.c, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsFmin, - pptable->DpmDescriptor[PPCLK_GFXCLK].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_VCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_VCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_VCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_VCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_VCLK].padding, - pptable->DpmDescriptor[PPCLK_VCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_VCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_VCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_VCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_VCLK].SsCurve.c, - pptable->DpmDescriptor[PPCLK_VCLK].SsFmin, - pptable->DpmDescriptor[PPCLK_VCLK].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_DCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_DCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_DCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_DCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_DCLK].padding, - pptable->DpmDescriptor[PPCLK_DCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_DCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_DCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_DCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_DCLK].SsCurve.c, - pptable->DpmDescriptor[PPCLK_DCLK].SsFmin, - pptable->DpmDescriptor[PPCLK_DCLK].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_SOCCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_SOCCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_SOCCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_SOCCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_SOCCLK].padding, - pptable->DpmDescriptor[PPCLK_SOCCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_SOCCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsCurve.c, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsFmin, - pptable->DpmDescriptor[PPCLK_SOCCLK].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_UCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_UCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_UCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_UCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_UCLK].padding, - pptable->DpmDescriptor[PPCLK_UCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_UCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_UCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_UCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_UCLK].SsCurve.c, - pptable->DpmDescriptor[PPCLK_UCLK].SsFmin, - pptable->DpmDescriptor[PPCLK_UCLK].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_FCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_FCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_FCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_FCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_FCLK].padding, - pptable->DpmDescriptor[PPCLK_FCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_FCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_FCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_FCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_FCLK].SsCurve.c, - pptable->DpmDescriptor[PPCLK_FCLK].SsFmin, - pptable->DpmDescriptor[PPCLK_FCLK].Padding16); - - - dev_info(smu->adev->dev, "FreqTableGfx\n"); - for (i = 0; i < NUM_GFXCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = %d\n", i, pptable->FreqTableGfx[i]); - - dev_info(smu->adev->dev, "FreqTableVclk\n"); - for (i = 0; i < NUM_VCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = %d\n", i, pptable->FreqTableVclk[i]); - - dev_info(smu->adev->dev, "FreqTableDclk\n"); - for (i = 0; i < NUM_DCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = %d\n", i, pptable->FreqTableDclk[i]); - - dev_info(smu->adev->dev, "FreqTableSocclk\n"); - for (i = 0; i < NUM_SOCCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = %d\n", i, pptable->FreqTableSocclk[i]); - - dev_info(smu->adev->dev, "FreqTableUclk\n"); - for (i = 0; i < NUM_UCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = %d\n", i, pptable->FreqTableUclk[i]); - - dev_info(smu->adev->dev, "FreqTableFclk\n"); - for (i = 0; i < NUM_FCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = %d\n", i, pptable->FreqTableFclk[i]); - - dev_info(smu->adev->dev, "Mp0clkFreq\n"); - for (i = 0; i < NUM_MP0CLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = %d\n", i, pptable->Mp0clkFreq[i]); - - dev_info(smu->adev->dev, "Mp0DpmVoltage\n"); - for (i = 0; i < NUM_MP0CLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = %d\n", i, pptable->Mp0DpmVoltage[i]); - - dev_info(smu->adev->dev, "GfxclkFidle = 0x%x\n", pptable->GfxclkFidle); - dev_info(smu->adev->dev, "GfxclkSlewRate = 0x%x\n", pptable->GfxclkSlewRate); - dev_info(smu->adev->dev, "Padding567[0] = 0x%x\n", pptable->Padding567[0]); - dev_info(smu->adev->dev, "Padding567[1] = 0x%x\n", pptable->Padding567[1]); - dev_info(smu->adev->dev, "Padding567[2] = 0x%x\n", pptable->Padding567[2]); - dev_info(smu->adev->dev, "Padding567[3] = 0x%x\n", pptable->Padding567[3]); - dev_info(smu->adev->dev, "GfxclkDsMaxFreq = %d\n", pptable->GfxclkDsMaxFreq); - dev_info(smu->adev->dev, "GfxclkSource = 0x%x\n", pptable->GfxclkSource); - dev_info(smu->adev->dev, "Padding456 = 0x%x\n", pptable->Padding456); - - dev_info(smu->adev->dev, "EnableTdpm = %d\n", pptable->EnableTdpm); - dev_info(smu->adev->dev, "TdpmHighHystTemperature = %d\n", pptable->TdpmHighHystTemperature); - dev_info(smu->adev->dev, "TdpmLowHystTemperature = %d\n", pptable->TdpmLowHystTemperature); - dev_info(smu->adev->dev, "GfxclkFreqHighTempLimit = %d\n", pptable->GfxclkFreqHighTempLimit); - - dev_info(smu->adev->dev, "FanStopTemp = %d\n", pptable->FanStopTemp); - dev_info(smu->adev->dev, "FanStartTemp = %d\n", pptable->FanStartTemp); - - dev_info(smu->adev->dev, "FanGainEdge = %d\n", pptable->FanGainEdge); - dev_info(smu->adev->dev, "FanGainHotspot = %d\n", pptable->FanGainHotspot); - dev_info(smu->adev->dev, "FanGainVrGfx = %d\n", pptable->FanGainVrGfx); - dev_info(smu->adev->dev, "FanGainVrSoc = %d\n", pptable->FanGainVrSoc); - dev_info(smu->adev->dev, "FanGainVrMem = %d\n", pptable->FanGainVrMem); - dev_info(smu->adev->dev, "FanGainHbm = %d\n", pptable->FanGainHbm); - - dev_info(smu->adev->dev, "FanPwmMin = %d\n", pptable->FanPwmMin); - dev_info(smu->adev->dev, "FanAcousticLimitRpm = %d\n", pptable->FanAcousticLimitRpm); - dev_info(smu->adev->dev, "FanThrottlingRpm = %d\n", pptable->FanThrottlingRpm); - dev_info(smu->adev->dev, "FanMaximumRpm = %d\n", pptable->FanMaximumRpm); - dev_info(smu->adev->dev, "FanTargetTemperature = %d\n", pptable->FanTargetTemperature); - dev_info(smu->adev->dev, "FanTargetGfxclk = %d\n", pptable->FanTargetGfxclk); - dev_info(smu->adev->dev, "FanZeroRpmEnable = %d\n", pptable->FanZeroRpmEnable); - dev_info(smu->adev->dev, "FanTachEdgePerRev = %d\n", pptable->FanTachEdgePerRev); - dev_info(smu->adev->dev, "FanTempInputSelect = %d\n", pptable->FanTempInputSelect); - - dev_info(smu->adev->dev, "FuzzyFan_ErrorSetDelta = %d\n", pptable->FuzzyFan_ErrorSetDelta); - dev_info(smu->adev->dev, "FuzzyFan_ErrorRateSetDelta = %d\n", pptable->FuzzyFan_ErrorRateSetDelta); - dev_info(smu->adev->dev, "FuzzyFan_PwmSetDelta = %d\n", pptable->FuzzyFan_PwmSetDelta); - dev_info(smu->adev->dev, "FuzzyFan_Reserved = %d\n", pptable->FuzzyFan_Reserved); - - dev_info(smu->adev->dev, "OverrideAvfsGb[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->OverrideAvfsGb[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "OverrideAvfsGb[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->OverrideAvfsGb[AVFS_VOLTAGE_SOC]); - dev_info(smu->adev->dev, "Padding8_Avfs[0] = %d\n", pptable->Padding8_Avfs[0]); - dev_info(smu->adev->dev, "Padding8_Avfs[1] = %d\n", pptable->Padding8_Avfs[1]); - - dev_info(smu->adev->dev, "dBtcGbGfxPll{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->dBtcGbGfxPll.a, - pptable->dBtcGbGfxPll.b, - pptable->dBtcGbGfxPll.c); - dev_info(smu->adev->dev, "dBtcGbGfxAfll{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->dBtcGbGfxAfll.a, - pptable->dBtcGbGfxAfll.b, - pptable->dBtcGbGfxAfll.c); - dev_info(smu->adev->dev, "dBtcGbSoc{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->dBtcGbSoc.a, - pptable->dBtcGbSoc.b, - pptable->dBtcGbSoc.c); - - dev_info(smu->adev->dev, "qAgingGb[AVFS_VOLTAGE_GFX]{m = 0x%x b = 0x%x}\n", - pptable->qAgingGb[AVFS_VOLTAGE_GFX].m, - pptable->qAgingGb[AVFS_VOLTAGE_GFX].b); - dev_info(smu->adev->dev, "qAgingGb[AVFS_VOLTAGE_SOC]{m = 0x%x b = 0x%x}\n", - pptable->qAgingGb[AVFS_VOLTAGE_SOC].m, - pptable->qAgingGb[AVFS_VOLTAGE_SOC].b); - - dev_info(smu->adev->dev, "qStaticVoltageOffset[AVFS_VOLTAGE_GFX]{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_GFX].a, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_GFX].b, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_GFX].c); - dev_info(smu->adev->dev, "qStaticVoltageOffset[AVFS_VOLTAGE_SOC]{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_SOC].a, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_SOC].b, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_SOC].c); - - dev_info(smu->adev->dev, "DcTol[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcTol[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcTol[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcTol[AVFS_VOLTAGE_SOC]); - - dev_info(smu->adev->dev, "DcBtcEnabled[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcEnabled[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcBtcEnabled[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcEnabled[AVFS_VOLTAGE_SOC]); - dev_info(smu->adev->dev, "Padding8_GfxBtc[0] = 0x%x\n", pptable->Padding8_GfxBtc[0]); - dev_info(smu->adev->dev, "Padding8_GfxBtc[1] = 0x%x\n", pptable->Padding8_GfxBtc[1]); - - dev_info(smu->adev->dev, "DcBtcMin[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcMin[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcBtcMin[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcMin[AVFS_VOLTAGE_SOC]); - dev_info(smu->adev->dev, "DcBtcMax[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcMax[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcBtcMax[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcMax[AVFS_VOLTAGE_SOC]); - - dev_info(smu->adev->dev, "DcBtcGb[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcGb[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcBtcGb[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcGb[AVFS_VOLTAGE_SOC]); - - dev_info(smu->adev->dev, "XgmiDpmPstates\n"); - for (i = 0; i < NUM_XGMI_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = %d\n", i, pptable->XgmiDpmPstates[i]); - dev_info(smu->adev->dev, "XgmiDpmSpare[0] = 0x%02x\n", pptable->XgmiDpmSpare[0]); - dev_info(smu->adev->dev, "XgmiDpmSpare[1] = 0x%02x\n", pptable->XgmiDpmSpare[1]); - - dev_info(smu->adev->dev, "VDDGFX_TVmin = %d\n", pptable->VDDGFX_TVmin); - dev_info(smu->adev->dev, "VDDSOC_TVmin = %d\n", pptable->VDDSOC_TVmin); - dev_info(smu->adev->dev, "VDDGFX_Vmin_HiTemp = %d\n", pptable->VDDGFX_Vmin_HiTemp); - dev_info(smu->adev->dev, "VDDGFX_Vmin_LoTemp = %d\n", pptable->VDDGFX_Vmin_LoTemp); - dev_info(smu->adev->dev, "VDDSOC_Vmin_HiTemp = %d\n", pptable->VDDSOC_Vmin_HiTemp); - dev_info(smu->adev->dev, "VDDSOC_Vmin_LoTemp = %d\n", pptable->VDDSOC_Vmin_LoTemp); - dev_info(smu->adev->dev, "VDDGFX_TVminHystersis = %d\n", pptable->VDDGFX_TVminHystersis); - dev_info(smu->adev->dev, "VDDSOC_TVminHystersis = %d\n", pptable->VDDSOC_TVminHystersis); - - dev_info(smu->adev->dev, "DebugOverrides = 0x%x\n", pptable->DebugOverrides); - dev_info(smu->adev->dev, "ReservedEquation0{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation0.a, - pptable->ReservedEquation0.b, - pptable->ReservedEquation0.c); - dev_info(smu->adev->dev, "ReservedEquation1{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation1.a, - pptable->ReservedEquation1.b, - pptable->ReservedEquation1.c); - dev_info(smu->adev->dev, "ReservedEquation2{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation2.a, - pptable->ReservedEquation2.b, - pptable->ReservedEquation2.c); - dev_info(smu->adev->dev, "ReservedEquation3{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation3.a, - pptable->ReservedEquation3.b, - pptable->ReservedEquation3.c); - - dev_info(smu->adev->dev, "MinVoltageUlvGfx = %d\n", pptable->MinVoltageUlvGfx); - dev_info(smu->adev->dev, "PaddingUlv = %d\n", pptable->PaddingUlv); - - dev_info(smu->adev->dev, "TotalPowerConfig = %d\n", pptable->TotalPowerConfig); - dev_info(smu->adev->dev, "TotalPowerSpare1 = %d\n", pptable->TotalPowerSpare1); - dev_info(smu->adev->dev, "TotalPowerSpare2 = %d\n", pptable->TotalPowerSpare2); - - dev_info(smu->adev->dev, "PccThresholdLow = %d\n", pptable->PccThresholdLow); - dev_info(smu->adev->dev, "PccThresholdHigh = %d\n", pptable->PccThresholdHigh); - - dev_info(smu->adev->dev, "Board Parameters:\n"); - dev_info(smu->adev->dev, "MaxVoltageStepGfx = 0x%x\n", pptable->MaxVoltageStepGfx); - dev_info(smu->adev->dev, "MaxVoltageStepSoc = 0x%x\n", pptable->MaxVoltageStepSoc); - - dev_info(smu->adev->dev, "VddGfxVrMapping = 0x%x\n", pptable->VddGfxVrMapping); - dev_info(smu->adev->dev, "VddSocVrMapping = 0x%x\n", pptable->VddSocVrMapping); - dev_info(smu->adev->dev, "VddMemVrMapping = 0x%x\n", pptable->VddMemVrMapping); - dev_info(smu->adev->dev, "BoardVrMapping = 0x%x\n", pptable->BoardVrMapping); - - dev_info(smu->adev->dev, "GfxUlvPhaseSheddingMask = 0x%x\n", pptable->GfxUlvPhaseSheddingMask); - dev_info(smu->adev->dev, "ExternalSensorPresent = 0x%x\n", pptable->ExternalSensorPresent); - - dev_info(smu->adev->dev, "GfxMaxCurrent = 0x%x\n", pptable->GfxMaxCurrent); - dev_info(smu->adev->dev, "GfxOffset = 0x%x\n", pptable->GfxOffset); - dev_info(smu->adev->dev, "Padding_TelemetryGfx = 0x%x\n", pptable->Padding_TelemetryGfx); - - dev_info(smu->adev->dev, "SocMaxCurrent = 0x%x\n", pptable->SocMaxCurrent); - dev_info(smu->adev->dev, "SocOffset = 0x%x\n", pptable->SocOffset); - dev_info(smu->adev->dev, "Padding_TelemetrySoc = 0x%x\n", pptable->Padding_TelemetrySoc); - - dev_info(smu->adev->dev, "MemMaxCurrent = 0x%x\n", pptable->MemMaxCurrent); - dev_info(smu->adev->dev, "MemOffset = 0x%x\n", pptable->MemOffset); - dev_info(smu->adev->dev, "Padding_TelemetryMem = 0x%x\n", pptable->Padding_TelemetryMem); - - dev_info(smu->adev->dev, "BoardMaxCurrent = 0x%x\n", pptable->BoardMaxCurrent); - dev_info(smu->adev->dev, "BoardOffset = 0x%x\n", pptable->BoardOffset); - dev_info(smu->adev->dev, "Padding_TelemetryBoardInput = 0x%x\n", pptable->Padding_TelemetryBoardInput); - - dev_info(smu->adev->dev, "VR0HotGpio = %d\n", pptable->VR0HotGpio); - dev_info(smu->adev->dev, "VR0HotPolarity = %d\n", pptable->VR0HotPolarity); - dev_info(smu->adev->dev, "VR1HotGpio = %d\n", pptable->VR1HotGpio); - dev_info(smu->adev->dev, "VR1HotPolarity = %d\n", pptable->VR1HotPolarity); - - dev_info(smu->adev->dev, "PllGfxclkSpreadEnabled = %d\n", pptable->PllGfxclkSpreadEnabled); - dev_info(smu->adev->dev, "PllGfxclkSpreadPercent = %d\n", pptable->PllGfxclkSpreadPercent); - dev_info(smu->adev->dev, "PllGfxclkSpreadFreq = %d\n", pptable->PllGfxclkSpreadFreq); - - dev_info(smu->adev->dev, "UclkSpreadEnabled = %d\n", pptable->UclkSpreadEnabled); - dev_info(smu->adev->dev, "UclkSpreadPercent = %d\n", pptable->UclkSpreadPercent); - dev_info(smu->adev->dev, "UclkSpreadFreq = %d\n", pptable->UclkSpreadFreq); - - dev_info(smu->adev->dev, "FclkSpreadEnabled = %d\n", pptable->FclkSpreadEnabled); - dev_info(smu->adev->dev, "FclkSpreadPercent = %d\n", pptable->FclkSpreadPercent); - dev_info(smu->adev->dev, "FclkSpreadFreq = %d\n", pptable->FclkSpreadFreq); - - dev_info(smu->adev->dev, "FllGfxclkSpreadEnabled = %d\n", pptable->FllGfxclkSpreadEnabled); - dev_info(smu->adev->dev, "FllGfxclkSpreadPercent = %d\n", pptable->FllGfxclkSpreadPercent); - dev_info(smu->adev->dev, "FllGfxclkSpreadFreq = %d\n", pptable->FllGfxclkSpreadFreq); - - for (i = 0; i < NUM_I2C_CONTROLLERS; i++) { - dev_info(smu->adev->dev, "I2cControllers[%d]:\n", i); - dev_info(smu->adev->dev, " .Enabled = %d\n", - pptable->I2cControllers[i].Enabled); - dev_info(smu->adev->dev, " .SlaveAddress = 0x%x\n", - pptable->I2cControllers[i].SlaveAddress); - dev_info(smu->adev->dev, " .ControllerPort = %d\n", - pptable->I2cControllers[i].ControllerPort); - dev_info(smu->adev->dev, " .ControllerName = %d\n", - pptable->I2cControllers[i].ControllerName); - dev_info(smu->adev->dev, " .ThermalThrottler = %d\n", - pptable->I2cControllers[i].ThermalThrotter); - dev_info(smu->adev->dev, " .I2cProtocol = %d\n", - pptable->I2cControllers[i].I2cProtocol); - dev_info(smu->adev->dev, " .Speed = %d\n", - pptable->I2cControllers[i].Speed); - } - - dev_info(smu->adev->dev, "MemoryChannelEnabled = %d\n", pptable->MemoryChannelEnabled); - dev_info(smu->adev->dev, "DramBitWidth = %d\n", pptable->DramBitWidth); - - dev_info(smu->adev->dev, "TotalBoardPower = %d\n", pptable->TotalBoardPower); - - dev_info(smu->adev->dev, "XgmiLinkSpeed\n"); - for (i = 0; i < NUM_XGMI_PSTATE_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = %d\n", i, pptable->XgmiLinkSpeed[i]); - dev_info(smu->adev->dev, "XgmiLinkWidth\n"); - for (i = 0; i < NUM_XGMI_PSTATE_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = %d\n", i, pptable->XgmiLinkWidth[i]); - dev_info(smu->adev->dev, "XgmiFclkFreq\n"); - for (i = 0; i < NUM_XGMI_PSTATE_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = %d\n", i, pptable->XgmiFclkFreq[i]); - dev_info(smu->adev->dev, "XgmiSocVoltage\n"); - for (i = 0; i < NUM_XGMI_PSTATE_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = %d\n", i, pptable->XgmiSocVoltage[i]); - -} - static bool arcturus_is_dpm_running(struct smu_context *smu) { int ret = 0; @@ -2002,7 +1570,9 @@ static bool arcturus_is_dpm_running(struct smu_context *smu) return !!(feature_enabled & SMC_DPM_FEATURE); } -static int arcturus_dpm_set_vcn_enable(struct smu_context *smu, bool enable) +static int arcturus_dpm_set_vcn_enable(struct smu_context *smu, + bool enable, + int inst) { int ret = 0; @@ -2365,8 +1935,6 @@ static const struct pptable_funcs arcturus_ppt_funcs = { .get_power_profile_mode = arcturus_get_power_profile_mode, .set_power_profile_mode = arcturus_set_power_profile_mode, .set_performance_level = arcturus_set_performance_level, - /* debug (internal used) */ - .dump_pptable = arcturus_dump_pptable, .get_power_limit = arcturus_get_power_limit, .is_dpm_running = arcturus_is_dpm_running, .dpm_set_vcn_enable = arcturus_dpm_set_vcn_enable, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c index 16af1a329621..211635dabed8 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c @@ -1135,7 +1135,9 @@ static int navi10_set_default_dpm_table(struct smu_context *smu) return 0; } -static int navi10_dpm_set_vcn_enable(struct smu_context *smu, bool enable) +static int navi10_dpm_set_vcn_enable(struct smu_context *smu, + bool enable, + int inst) { int ret = 0; @@ -1689,7 +1691,7 @@ static int navi10_force_clk_levels(struct smu_context *smu, if (ret) return 0; - ret = smu_v11_0_set_soft_freq_limited_range(smu, clk_type, min_freq, max_freq); + ret = smu_v11_0_set_soft_freq_limited_range(smu, clk_type, min_freq, max_freq, false); if (ret) return 0; break; @@ -2081,10 +2083,13 @@ static int navi10_set_power_profile_mode(struct smu_context *smu, long *input, u smu->power_profile_mode); if (workload_type < 0) return -EINVAL; + ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask, - 1 << workload_type, NULL); + smu->workload_mask, NULL); if (ret) dev_err(smu->adev->dev, "[%s] Failed to set work load mask!", __func__); + else + smu_cmn_assign_power_profile(smu); return ret; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c index 9c3c48297cba..d0ed0d060a8a 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c @@ -1152,7 +1152,9 @@ static int sienna_cichlid_set_default_dpm_table(struct smu_context *smu) return 0; } -static int sienna_cichlid_dpm_set_vcn_enable(struct smu_context *smu, bool enable) +static int sienna_cichlid_dpm_set_vcn_enable(struct smu_context *smu, + bool enable, + int inst) { struct amdgpu_device *adev = smu->adev; int i, ret = 0; @@ -1469,7 +1471,7 @@ static int sienna_cichlid_force_clk_levels(struct smu_context *smu, if (ret) goto forec_level_out; - ret = smu_v11_0_set_soft_freq_limited_range(smu, clk_type, min_freq, max_freq); + ret = smu_v11_0_set_soft_freq_limited_range(smu, clk_type, min_freq, max_freq, false); if (ret) goto forec_level_out; break; @@ -1786,10 +1788,13 @@ static int sienna_cichlid_set_power_profile_mode(struct smu_context *smu, long * smu->power_profile_mode); if (workload_type < 0) return -EINVAL; + ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask, - 1 << workload_type, NULL); + smu->workload_mask, NULL); if (ret) dev_err(smu->adev->dev, "[%s] Failed to set work load mask!", __func__); + else + smu_cmn_assign_power_profile(smu); return ret; } @@ -2493,1274 +2498,6 @@ static bool sienna_cichlid_is_mode1_reset_supported(struct smu_context *smu) return val != 0x0; } -static void beige_goby_dump_pptable(struct smu_context *smu) -{ - struct smu_table_context *table_context = &smu->smu_table; - PPTable_beige_goby_t *pptable = table_context->driver_pptable; - int i; - - dev_info(smu->adev->dev, "Dumped PPTable:\n"); - - dev_info(smu->adev->dev, "Version = 0x%08x\n", pptable->Version); - dev_info(smu->adev->dev, "FeaturesToRun[0] = 0x%08x\n", pptable->FeaturesToRun[0]); - dev_info(smu->adev->dev, "FeaturesToRun[1] = 0x%08x\n", pptable->FeaturesToRun[1]); - - for (i = 0; i < PPT_THROTTLER_COUNT; i++) { - dev_info(smu->adev->dev, "SocketPowerLimitAc[%d] = 0x%x\n", i, pptable->SocketPowerLimitAc[i]); - dev_info(smu->adev->dev, "SocketPowerLimitAcTau[%d] = 0x%x\n", i, pptable->SocketPowerLimitAcTau[i]); - dev_info(smu->adev->dev, "SocketPowerLimitDc[%d] = 0x%x\n", i, pptable->SocketPowerLimitDc[i]); - dev_info(smu->adev->dev, "SocketPowerLimitDcTau[%d] = 0x%x\n", i, pptable->SocketPowerLimitDcTau[i]); - } - - for (i = 0; i < TDC_THROTTLER_COUNT; i++) { - dev_info(smu->adev->dev, "TdcLimit[%d] = 0x%x\n", i, pptable->TdcLimit[i]); - dev_info(smu->adev->dev, "TdcLimitTau[%d] = 0x%x\n", i, pptable->TdcLimitTau[i]); - } - - for (i = 0; i < TEMP_COUNT; i++) { - dev_info(smu->adev->dev, "TemperatureLimit[%d] = 0x%x\n", i, pptable->TemperatureLimit[i]); - } - - dev_info(smu->adev->dev, "FitLimit = 0x%x\n", pptable->FitLimit); - dev_info(smu->adev->dev, "TotalPowerConfig = 0x%x\n", pptable->TotalPowerConfig); - dev_info(smu->adev->dev, "TotalPowerPadding[0] = 0x%x\n", pptable->TotalPowerPadding[0]); - dev_info(smu->adev->dev, "TotalPowerPadding[1] = 0x%x\n", pptable->TotalPowerPadding[1]); - dev_info(smu->adev->dev, "TotalPowerPadding[2] = 0x%x\n", pptable->TotalPowerPadding[2]); - - dev_info(smu->adev->dev, "ApccPlusResidencyLimit = 0x%x\n", pptable->ApccPlusResidencyLimit); - for (i = 0; i < NUM_SMNCLK_DPM_LEVELS; i++) { - dev_info(smu->adev->dev, "SmnclkDpmFreq[%d] = 0x%x\n", i, pptable->SmnclkDpmFreq[i]); - dev_info(smu->adev->dev, "SmnclkDpmVoltage[%d] = 0x%x\n", i, pptable->SmnclkDpmVoltage[i]); - } - dev_info(smu->adev->dev, "ThrottlerControlMask = 0x%x\n", pptable->ThrottlerControlMask); - - dev_info(smu->adev->dev, "FwDStateMask = 0x%x\n", pptable->FwDStateMask); - - dev_info(smu->adev->dev, "UlvVoltageOffsetSoc = 0x%x\n", pptable->UlvVoltageOffsetSoc); - dev_info(smu->adev->dev, "UlvVoltageOffsetGfx = 0x%x\n", pptable->UlvVoltageOffsetGfx); - dev_info(smu->adev->dev, "MinVoltageUlvGfx = 0x%x\n", pptable->MinVoltageUlvGfx); - dev_info(smu->adev->dev, "MinVoltageUlvSoc = 0x%x\n", pptable->MinVoltageUlvSoc); - - dev_info(smu->adev->dev, "SocLIVmin = 0x%x\n", pptable->SocLIVmin); - - dev_info(smu->adev->dev, "GceaLinkMgrIdleThreshold = 0x%x\n", pptable->GceaLinkMgrIdleThreshold); - - dev_info(smu->adev->dev, "MinVoltageGfx = 0x%x\n", pptable->MinVoltageGfx); - dev_info(smu->adev->dev, "MinVoltageSoc = 0x%x\n", pptable->MinVoltageSoc); - dev_info(smu->adev->dev, "MaxVoltageGfx = 0x%x\n", pptable->MaxVoltageGfx); - dev_info(smu->adev->dev, "MaxVoltageSoc = 0x%x\n", pptable->MaxVoltageSoc); - - dev_info(smu->adev->dev, "LoadLineResistanceGfx = 0x%x\n", pptable->LoadLineResistanceGfx); - dev_info(smu->adev->dev, "LoadLineResistanceSoc = 0x%x\n", pptable->LoadLineResistanceSoc); - - dev_info(smu->adev->dev, "VDDGFX_TVmin = 0x%x\n", pptable->VDDGFX_TVmin); - dev_info(smu->adev->dev, "VDDSOC_TVmin = 0x%x\n", pptable->VDDSOC_TVmin); - dev_info(smu->adev->dev, "VDDGFX_Vmin_HiTemp = 0x%x\n", pptable->VDDGFX_Vmin_HiTemp); - dev_info(smu->adev->dev, "VDDGFX_Vmin_LoTemp = 0x%x\n", pptable->VDDGFX_Vmin_LoTemp); - dev_info(smu->adev->dev, "VDDSOC_Vmin_HiTemp = 0x%x\n", pptable->VDDSOC_Vmin_HiTemp); - dev_info(smu->adev->dev, "VDDSOC_Vmin_LoTemp = 0x%x\n", pptable->VDDSOC_Vmin_LoTemp); - dev_info(smu->adev->dev, "VDDGFX_TVminHystersis = 0x%x\n", pptable->VDDGFX_TVminHystersis); - dev_info(smu->adev->dev, "VDDSOC_TVminHystersis = 0x%x\n", pptable->VDDSOC_TVminHystersis); - - dev_info(smu->adev->dev, "[PPCLK_GFXCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_GFXCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_GFXCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_GFXCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_GFXCLK].Padding, - pptable->DpmDescriptor[PPCLK_GFXCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_GFXCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsCurve.c, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsFmin, - pptable->DpmDescriptor[PPCLK_GFXCLK].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_SOCCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_SOCCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_SOCCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_SOCCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_SOCCLK].Padding, - pptable->DpmDescriptor[PPCLK_SOCCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_SOCCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsCurve.c, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsFmin, - pptable->DpmDescriptor[PPCLK_SOCCLK].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_UCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_UCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_UCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_UCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_UCLK].Padding, - pptable->DpmDescriptor[PPCLK_UCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_UCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_UCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_UCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_UCLK].SsCurve.c, - pptable->DpmDescriptor[PPCLK_UCLK].SsFmin, - pptable->DpmDescriptor[PPCLK_UCLK].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_FCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_FCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_FCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_FCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_FCLK].Padding, - pptable->DpmDescriptor[PPCLK_FCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_FCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_FCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_FCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_FCLK].SsCurve.c, - pptable->DpmDescriptor[PPCLK_FCLK].SsFmin, - pptable->DpmDescriptor[PPCLK_FCLK].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_DCLK_0]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_DCLK_0].VoltageMode, - pptable->DpmDescriptor[PPCLK_DCLK_0].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_DCLK_0].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_DCLK_0].Padding, - pptable->DpmDescriptor[PPCLK_DCLK_0].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_DCLK_0].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_DCLK_0].SsCurve.a, - pptable->DpmDescriptor[PPCLK_DCLK_0].SsCurve.b, - pptable->DpmDescriptor[PPCLK_DCLK_0].SsCurve.c, - pptable->DpmDescriptor[PPCLK_DCLK_0].SsFmin, - pptable->DpmDescriptor[PPCLK_DCLK_0].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_VCLK_0]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_VCLK_0].VoltageMode, - pptable->DpmDescriptor[PPCLK_VCLK_0].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_VCLK_0].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_VCLK_0].Padding, - pptable->DpmDescriptor[PPCLK_VCLK_0].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_VCLK_0].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_VCLK_0].SsCurve.a, - pptable->DpmDescriptor[PPCLK_VCLK_0].SsCurve.b, - pptable->DpmDescriptor[PPCLK_VCLK_0].SsCurve.c, - pptable->DpmDescriptor[PPCLK_VCLK_0].SsFmin, - pptable->DpmDescriptor[PPCLK_VCLK_0].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_DCLK_1]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_DCLK_1].VoltageMode, - pptable->DpmDescriptor[PPCLK_DCLK_1].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_DCLK_1].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_DCLK_1].Padding, - pptable->DpmDescriptor[PPCLK_DCLK_1].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_DCLK_1].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_DCLK_1].SsCurve.a, - pptable->DpmDescriptor[PPCLK_DCLK_1].SsCurve.b, - pptable->DpmDescriptor[PPCLK_DCLK_1].SsCurve.c, - pptable->DpmDescriptor[PPCLK_DCLK_1].SsFmin, - pptable->DpmDescriptor[PPCLK_DCLK_1].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_VCLK_1]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_VCLK_1].VoltageMode, - pptable->DpmDescriptor[PPCLK_VCLK_1].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_VCLK_1].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_VCLK_1].Padding, - pptable->DpmDescriptor[PPCLK_VCLK_1].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_VCLK_1].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_VCLK_1].SsCurve.a, - pptable->DpmDescriptor[PPCLK_VCLK_1].SsCurve.b, - pptable->DpmDescriptor[PPCLK_VCLK_1].SsCurve.c, - pptable->DpmDescriptor[PPCLK_VCLK_1].SsFmin, - pptable->DpmDescriptor[PPCLK_VCLK_1].Padding16); - - dev_info(smu->adev->dev, "FreqTableGfx\n"); - for (i = 0; i < NUM_GFXCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = 0x%x\n", i, pptable->FreqTableGfx[i]); - - dev_info(smu->adev->dev, "FreqTableVclk\n"); - for (i = 0; i < NUM_VCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = 0x%x\n", i, pptable->FreqTableVclk[i]); - - dev_info(smu->adev->dev, "FreqTableDclk\n"); - for (i = 0; i < NUM_DCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = 0x%x\n", i, pptable->FreqTableDclk[i]); - - dev_info(smu->adev->dev, "FreqTableSocclk\n"); - for (i = 0; i < NUM_SOCCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = 0x%x\n", i, pptable->FreqTableSocclk[i]); - - dev_info(smu->adev->dev, "FreqTableUclk\n"); - for (i = 0; i < NUM_UCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = 0x%x\n", i, pptable->FreqTableUclk[i]); - - dev_info(smu->adev->dev, "FreqTableFclk\n"); - for (i = 0; i < NUM_FCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = 0x%x\n", i, pptable->FreqTableFclk[i]); - - dev_info(smu->adev->dev, "DcModeMaxFreq\n"); - dev_info(smu->adev->dev, " .PPCLK_GFXCLK = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_GFXCLK]); - dev_info(smu->adev->dev, " .PPCLK_SOCCLK = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_SOCCLK]); - dev_info(smu->adev->dev, " .PPCLK_UCLK = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_UCLK]); - dev_info(smu->adev->dev, " .PPCLK_FCLK = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_FCLK]); - dev_info(smu->adev->dev, " .PPCLK_DCLK_0 = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_DCLK_0]); - dev_info(smu->adev->dev, " .PPCLK_VCLK_0 = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_VCLK_0]); - dev_info(smu->adev->dev, " .PPCLK_DCLK_1 = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_DCLK_1]); - dev_info(smu->adev->dev, " .PPCLK_VCLK_1 = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_VCLK_1]); - - dev_info(smu->adev->dev, "FreqTableUclkDiv\n"); - for (i = 0; i < NUM_UCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->FreqTableUclkDiv[i]); - - dev_info(smu->adev->dev, "FclkBoostFreq = 0x%x\n", pptable->FclkBoostFreq); - dev_info(smu->adev->dev, "FclkParamPadding = 0x%x\n", pptable->FclkParamPadding); - - dev_info(smu->adev->dev, "Mp0clkFreq\n"); - for (i = 0; i < NUM_MP0CLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->Mp0clkFreq[i]); - - dev_info(smu->adev->dev, "Mp0DpmVoltage\n"); - for (i = 0; i < NUM_MP0CLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->Mp0DpmVoltage[i]); - - dev_info(smu->adev->dev, "MemVddciVoltage\n"); - for (i = 0; i < NUM_UCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->MemVddciVoltage[i]); - - dev_info(smu->adev->dev, "MemMvddVoltage\n"); - for (i = 0; i < NUM_UCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->MemMvddVoltage[i]); - - dev_info(smu->adev->dev, "GfxclkFgfxoffEntry = 0x%x\n", pptable->GfxclkFgfxoffEntry); - dev_info(smu->adev->dev, "GfxclkFinit = 0x%x\n", pptable->GfxclkFinit); - dev_info(smu->adev->dev, "GfxclkFidle = 0x%x\n", pptable->GfxclkFidle); - dev_info(smu->adev->dev, "GfxclkSource = 0x%x\n", pptable->GfxclkSource); - dev_info(smu->adev->dev, "GfxclkPadding = 0x%x\n", pptable->GfxclkPadding); - - dev_info(smu->adev->dev, "GfxGpoSubFeatureMask = 0x%x\n", pptable->GfxGpoSubFeatureMask); - - dev_info(smu->adev->dev, "GfxGpoEnabledWorkPolicyMask = 0x%x\n", pptable->GfxGpoEnabledWorkPolicyMask); - dev_info(smu->adev->dev, "GfxGpoDisabledWorkPolicyMask = 0x%x\n", pptable->GfxGpoDisabledWorkPolicyMask); - dev_info(smu->adev->dev, "GfxGpoPadding[0] = 0x%x\n", pptable->GfxGpoPadding[0]); - dev_info(smu->adev->dev, "GfxGpoVotingAllow = 0x%x\n", pptable->GfxGpoVotingAllow); - dev_info(smu->adev->dev, "GfxGpoPadding32[0] = 0x%x\n", pptable->GfxGpoPadding32[0]); - dev_info(smu->adev->dev, "GfxGpoPadding32[1] = 0x%x\n", pptable->GfxGpoPadding32[1]); - dev_info(smu->adev->dev, "GfxGpoPadding32[2] = 0x%x\n", pptable->GfxGpoPadding32[2]); - dev_info(smu->adev->dev, "GfxGpoPadding32[3] = 0x%x\n", pptable->GfxGpoPadding32[3]); - dev_info(smu->adev->dev, "GfxDcsFopt = 0x%x\n", pptable->GfxDcsFopt); - dev_info(smu->adev->dev, "GfxDcsFclkFopt = 0x%x\n", pptable->GfxDcsFclkFopt); - dev_info(smu->adev->dev, "GfxDcsUclkFopt = 0x%x\n", pptable->GfxDcsUclkFopt); - - dev_info(smu->adev->dev, "DcsGfxOffVoltage = 0x%x\n", pptable->DcsGfxOffVoltage); - dev_info(smu->adev->dev, "DcsMinGfxOffTime = 0x%x\n", pptable->DcsMinGfxOffTime); - dev_info(smu->adev->dev, "DcsMaxGfxOffTime = 0x%x\n", pptable->DcsMaxGfxOffTime); - dev_info(smu->adev->dev, "DcsMinCreditAccum = 0x%x\n", pptable->DcsMinCreditAccum); - dev_info(smu->adev->dev, "DcsExitHysteresis = 0x%x\n", pptable->DcsExitHysteresis); - dev_info(smu->adev->dev, "DcsTimeout = 0x%x\n", pptable->DcsTimeout); - - dev_info(smu->adev->dev, "DcsParamPadding[0] = 0x%x\n", pptable->DcsParamPadding[0]); - dev_info(smu->adev->dev, "DcsParamPadding[1] = 0x%x\n", pptable->DcsParamPadding[1]); - dev_info(smu->adev->dev, "DcsParamPadding[2] = 0x%x\n", pptable->DcsParamPadding[2]); - dev_info(smu->adev->dev, "DcsParamPadding[3] = 0x%x\n", pptable->DcsParamPadding[3]); - dev_info(smu->adev->dev, "DcsParamPadding[4] = 0x%x\n", pptable->DcsParamPadding[4]); - - dev_info(smu->adev->dev, "FlopsPerByteTable\n"); - for (i = 0; i < RLC_PACE_TABLE_NUM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->FlopsPerByteTable[i]); - - dev_info(smu->adev->dev, "LowestUclkReservedForUlv = 0x%x\n", pptable->LowestUclkReservedForUlv); - dev_info(smu->adev->dev, "vddingMem[0] = 0x%x\n", pptable->PaddingMem[0]); - dev_info(smu->adev->dev, "vddingMem[1] = 0x%x\n", pptable->PaddingMem[1]); - dev_info(smu->adev->dev, "vddingMem[2] = 0x%x\n", pptable->PaddingMem[2]); - - dev_info(smu->adev->dev, "UclkDpmPstates\n"); - for (i = 0; i < NUM_UCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->UclkDpmPstates[i]); - - dev_info(smu->adev->dev, "UclkDpmSrcFreqRange\n"); - dev_info(smu->adev->dev, " .Fmin = 0x%x\n", - pptable->UclkDpmSrcFreqRange.Fmin); - dev_info(smu->adev->dev, " .Fmax = 0x%x\n", - pptable->UclkDpmSrcFreqRange.Fmax); - dev_info(smu->adev->dev, "UclkDpmTargFreqRange\n"); - dev_info(smu->adev->dev, " .Fmin = 0x%x\n", - pptable->UclkDpmTargFreqRange.Fmin); - dev_info(smu->adev->dev, " .Fmax = 0x%x\n", - pptable->UclkDpmTargFreqRange.Fmax); - dev_info(smu->adev->dev, "UclkDpmMidstepFreq = 0x%x\n", pptable->UclkDpmMidstepFreq); - dev_info(smu->adev->dev, "UclkMidstepPadding = 0x%x\n", pptable->UclkMidstepPadding); - - dev_info(smu->adev->dev, "PcieGenSpeed\n"); - for (i = 0; i < NUM_LINK_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->PcieGenSpeed[i]); - - dev_info(smu->adev->dev, "PcieLaneCount\n"); - for (i = 0; i < NUM_LINK_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->PcieLaneCount[i]); - - dev_info(smu->adev->dev, "LclkFreq\n"); - for (i = 0; i < NUM_LINK_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->LclkFreq[i]); - - dev_info(smu->adev->dev, "FanStopTemp = 0x%x\n", pptable->FanStopTemp); - dev_info(smu->adev->dev, "FanStartTemp = 0x%x\n", pptable->FanStartTemp); - - dev_info(smu->adev->dev, "FanGain\n"); - for (i = 0; i < TEMP_COUNT; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->FanGain[i]); - - dev_info(smu->adev->dev, "FanPwmMin = 0x%x\n", pptable->FanPwmMin); - dev_info(smu->adev->dev, "FanAcousticLimitRpm = 0x%x\n", pptable->FanAcousticLimitRpm); - dev_info(smu->adev->dev, "FanThrottlingRpm = 0x%x\n", pptable->FanThrottlingRpm); - dev_info(smu->adev->dev, "FanMaximumRpm = 0x%x\n", pptable->FanMaximumRpm); - dev_info(smu->adev->dev, "MGpuFanBoostLimitRpm = 0x%x\n", pptable->MGpuFanBoostLimitRpm); - dev_info(smu->adev->dev, "FanTargetTemperature = 0x%x\n", pptable->FanTargetTemperature); - dev_info(smu->adev->dev, "FanTargetGfxclk = 0x%x\n", pptable->FanTargetGfxclk); - dev_info(smu->adev->dev, "FanPadding16 = 0x%x\n", pptable->FanPadding16); - dev_info(smu->adev->dev, "FanTempInputSelect = 0x%x\n", pptable->FanTempInputSelect); - dev_info(smu->adev->dev, "FanPadding = 0x%x\n", pptable->FanPadding); - dev_info(smu->adev->dev, "FanZeroRpmEnable = 0x%x\n", pptable->FanZeroRpmEnable); - dev_info(smu->adev->dev, "FanTachEdgePerRev = 0x%x\n", pptable->FanTachEdgePerRev); - - dev_info(smu->adev->dev, "FuzzyFan_ErrorSetDelta = 0x%x\n", pptable->FuzzyFan_ErrorSetDelta); - dev_info(smu->adev->dev, "FuzzyFan_ErrorRateSetDelta = 0x%x\n", pptable->FuzzyFan_ErrorRateSetDelta); - dev_info(smu->adev->dev, "FuzzyFan_PwmSetDelta = 0x%x\n", pptable->FuzzyFan_PwmSetDelta); - dev_info(smu->adev->dev, "FuzzyFan_Reserved = 0x%x\n", pptable->FuzzyFan_Reserved); - - dev_info(smu->adev->dev, "OverrideAvfsGb[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->OverrideAvfsGb[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "OverrideAvfsGb[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->OverrideAvfsGb[AVFS_VOLTAGE_SOC]); - dev_info(smu->adev->dev, "dBtcGbGfxDfllModelSelect = 0x%x\n", pptable->dBtcGbGfxDfllModelSelect); - dev_info(smu->adev->dev, "Padding8_Avfs = 0x%x\n", pptable->Padding8_Avfs); - - dev_info(smu->adev->dev, "qAvfsGb[AVFS_VOLTAGE_GFX]{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->qAvfsGb[AVFS_VOLTAGE_GFX].a, - pptable->qAvfsGb[AVFS_VOLTAGE_GFX].b, - pptable->qAvfsGb[AVFS_VOLTAGE_GFX].c); - dev_info(smu->adev->dev, "qAvfsGb[AVFS_VOLTAGE_SOC]{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->qAvfsGb[AVFS_VOLTAGE_SOC].a, - pptable->qAvfsGb[AVFS_VOLTAGE_SOC].b, - pptable->qAvfsGb[AVFS_VOLTAGE_SOC].c); - dev_info(smu->adev->dev, "dBtcGbGfxPll{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->dBtcGbGfxPll.a, - pptable->dBtcGbGfxPll.b, - pptable->dBtcGbGfxPll.c); - dev_info(smu->adev->dev, "dBtcGbGfxAfll{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->dBtcGbGfxDfll.a, - pptable->dBtcGbGfxDfll.b, - pptable->dBtcGbGfxDfll.c); - dev_info(smu->adev->dev, "dBtcGbSoc{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->dBtcGbSoc.a, - pptable->dBtcGbSoc.b, - pptable->dBtcGbSoc.c); - dev_info(smu->adev->dev, "qAgingGb[AVFS_VOLTAGE_GFX]{m = 0x%x b = 0x%x}\n", - pptable->qAgingGb[AVFS_VOLTAGE_GFX].m, - pptable->qAgingGb[AVFS_VOLTAGE_GFX].b); - dev_info(smu->adev->dev, "qAgingGb[AVFS_VOLTAGE_SOC]{m = 0x%x b = 0x%x}\n", - pptable->qAgingGb[AVFS_VOLTAGE_SOC].m, - pptable->qAgingGb[AVFS_VOLTAGE_SOC].b); - - dev_info(smu->adev->dev, "PiecewiseLinearDroopIntGfxDfll\n"); - for (i = 0; i < NUM_PIECE_WISE_LINEAR_DROOP_MODEL_VF_POINTS; i++) { - dev_info(smu->adev->dev, " Fset[%d] = 0x%x\n", - i, pptable->PiecewiseLinearDroopIntGfxDfll.Fset[i]); - dev_info(smu->adev->dev, " Vdroop[%d] = 0x%x\n", - i, pptable->PiecewiseLinearDroopIntGfxDfll.Vdroop[i]); - } - - dev_info(smu->adev->dev, "qStaticVoltageOffset[AVFS_VOLTAGE_GFX]{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_GFX].a, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_GFX].b, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_GFX].c); - dev_info(smu->adev->dev, "qStaticVoltageOffset[AVFS_VOLTAGE_SOC]{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_SOC].a, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_SOC].b, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_SOC].c); - - dev_info(smu->adev->dev, "DcTol[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcTol[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcTol[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcTol[AVFS_VOLTAGE_SOC]); - - dev_info(smu->adev->dev, "DcBtcEnabled[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcEnabled[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcBtcEnabled[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcEnabled[AVFS_VOLTAGE_SOC]); - dev_info(smu->adev->dev, "Padding8_GfxBtc[0] = 0x%x\n", pptable->Padding8_GfxBtc[0]); - dev_info(smu->adev->dev, "Padding8_GfxBtc[1] = 0x%x\n", pptable->Padding8_GfxBtc[1]); - - dev_info(smu->adev->dev, "DcBtcMin[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcMin[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcBtcMin[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcMin[AVFS_VOLTAGE_SOC]); - dev_info(smu->adev->dev, "DcBtcMax[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcMax[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcBtcMax[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcMax[AVFS_VOLTAGE_SOC]); - - dev_info(smu->adev->dev, "DcBtcGb[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcGb[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcBtcGb[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcGb[AVFS_VOLTAGE_SOC]); - - dev_info(smu->adev->dev, "XgmiDpmPstates\n"); - for (i = 0; i < NUM_XGMI_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->XgmiDpmPstates[i]); - dev_info(smu->adev->dev, "XgmiDpmSpare[0] = 0x%02x\n", pptable->XgmiDpmSpare[0]); - dev_info(smu->adev->dev, "XgmiDpmSpare[1] = 0x%02x\n", pptable->XgmiDpmSpare[1]); - - dev_info(smu->adev->dev, "DebugOverrides = 0x%x\n", pptable->DebugOverrides); - dev_info(smu->adev->dev, "ReservedEquation0{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation0.a, - pptable->ReservedEquation0.b, - pptable->ReservedEquation0.c); - dev_info(smu->adev->dev, "ReservedEquation1{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation1.a, - pptable->ReservedEquation1.b, - pptable->ReservedEquation1.c); - dev_info(smu->adev->dev, "ReservedEquation2{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation2.a, - pptable->ReservedEquation2.b, - pptable->ReservedEquation2.c); - dev_info(smu->adev->dev, "ReservedEquation3{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation3.a, - pptable->ReservedEquation3.b, - pptable->ReservedEquation3.c); - - dev_info(smu->adev->dev, "SkuReserved[0] = 0x%x\n", pptable->SkuReserved[0]); - dev_info(smu->adev->dev, "SkuReserved[1] = 0x%x\n", pptable->SkuReserved[1]); - dev_info(smu->adev->dev, "SkuReserved[2] = 0x%x\n", pptable->SkuReserved[2]); - dev_info(smu->adev->dev, "SkuReserved[3] = 0x%x\n", pptable->SkuReserved[3]); - dev_info(smu->adev->dev, "SkuReserved[4] = 0x%x\n", pptable->SkuReserved[4]); - dev_info(smu->adev->dev, "SkuReserved[5] = 0x%x\n", pptable->SkuReserved[5]); - dev_info(smu->adev->dev, "SkuReserved[6] = 0x%x\n", pptable->SkuReserved[6]); - dev_info(smu->adev->dev, "SkuReserved[7] = 0x%x\n", pptable->SkuReserved[7]); - - dev_info(smu->adev->dev, "GamingClk[0] = 0x%x\n", pptable->GamingClk[0]); - dev_info(smu->adev->dev, "GamingClk[1] = 0x%x\n", pptable->GamingClk[1]); - dev_info(smu->adev->dev, "GamingClk[2] = 0x%x\n", pptable->GamingClk[2]); - dev_info(smu->adev->dev, "GamingClk[3] = 0x%x\n", pptable->GamingClk[3]); - dev_info(smu->adev->dev, "GamingClk[4] = 0x%x\n", pptable->GamingClk[4]); - dev_info(smu->adev->dev, "GamingClk[5] = 0x%x\n", pptable->GamingClk[5]); - - for (i = 0; i < NUM_I2C_CONTROLLERS; i++) { - dev_info(smu->adev->dev, "I2cControllers[%d]:\n", i); - dev_info(smu->adev->dev, " .Enabled = 0x%x\n", - pptable->I2cControllers[i].Enabled); - dev_info(smu->adev->dev, " .Speed = 0x%x\n", - pptable->I2cControllers[i].Speed); - dev_info(smu->adev->dev, " .SlaveAddress = 0x%x\n", - pptable->I2cControllers[i].SlaveAddress); - dev_info(smu->adev->dev, " .ControllerPort = 0x%x\n", - pptable->I2cControllers[i].ControllerPort); - dev_info(smu->adev->dev, " .ControllerName = 0x%x\n", - pptable->I2cControllers[i].ControllerName); - dev_info(smu->adev->dev, " .ThermalThrottler = 0x%x\n", - pptable->I2cControllers[i].ThermalThrotter); - dev_info(smu->adev->dev, " .I2cProtocol = 0x%x\n", - pptable->I2cControllers[i].I2cProtocol); - dev_info(smu->adev->dev, " .PaddingConfig = 0x%x\n", - pptable->I2cControllers[i].PaddingConfig); - } - - dev_info(smu->adev->dev, "GpioScl = 0x%x\n", pptable->GpioScl); - dev_info(smu->adev->dev, "GpioSda = 0x%x\n", pptable->GpioSda); - dev_info(smu->adev->dev, "FchUsbPdSlaveAddr = 0x%x\n", pptable->FchUsbPdSlaveAddr); - dev_info(smu->adev->dev, "I2cSpare[0] = 0x%x\n", pptable->I2cSpare[0]); - - dev_info(smu->adev->dev, "Board Parameters:\n"); - dev_info(smu->adev->dev, "VddGfxVrMapping = 0x%x\n", pptable->VddGfxVrMapping); - dev_info(smu->adev->dev, "VddSocVrMapping = 0x%x\n", pptable->VddSocVrMapping); - dev_info(smu->adev->dev, "VddMem0VrMapping = 0x%x\n", pptable->VddMem0VrMapping); - dev_info(smu->adev->dev, "VddMem1VrMapping = 0x%x\n", pptable->VddMem1VrMapping); - dev_info(smu->adev->dev, "GfxUlvPhaseSheddingMask = 0x%x\n", pptable->GfxUlvPhaseSheddingMask); - dev_info(smu->adev->dev, "SocUlvPhaseSheddingMask = 0x%x\n", pptable->SocUlvPhaseSheddingMask); - dev_info(smu->adev->dev, "VddciUlvPhaseSheddingMask = 0x%x\n", pptable->VddciUlvPhaseSheddingMask); - dev_info(smu->adev->dev, "MvddUlvPhaseSheddingMask = 0x%x\n", pptable->MvddUlvPhaseSheddingMask); - - dev_info(smu->adev->dev, "GfxMaxCurrent = 0x%x\n", pptable->GfxMaxCurrent); - dev_info(smu->adev->dev, "GfxOffset = 0x%x\n", pptable->GfxOffset); - dev_info(smu->adev->dev, "Padding_TelemetryGfx = 0x%x\n", pptable->Padding_TelemetryGfx); - - dev_info(smu->adev->dev, "SocMaxCurrent = 0x%x\n", pptable->SocMaxCurrent); - dev_info(smu->adev->dev, "SocOffset = 0x%x\n", pptable->SocOffset); - dev_info(smu->adev->dev, "Padding_TelemetrySoc = 0x%x\n", pptable->Padding_TelemetrySoc); - - dev_info(smu->adev->dev, "Mem0MaxCurrent = 0x%x\n", pptable->Mem0MaxCurrent); - dev_info(smu->adev->dev, "Mem0Offset = 0x%x\n", pptable->Mem0Offset); - dev_info(smu->adev->dev, "Padding_TelemetryMem0 = 0x%x\n", pptable->Padding_TelemetryMem0); - - dev_info(smu->adev->dev, "Mem1MaxCurrent = 0x%x\n", pptable->Mem1MaxCurrent); - dev_info(smu->adev->dev, "Mem1Offset = 0x%x\n", pptable->Mem1Offset); - dev_info(smu->adev->dev, "Padding_TelemetryMem1 = 0x%x\n", pptable->Padding_TelemetryMem1); - - dev_info(smu->adev->dev, "MvddRatio = 0x%x\n", pptable->MvddRatio); - - dev_info(smu->adev->dev, "AcDcGpio = 0x%x\n", pptable->AcDcGpio); - dev_info(smu->adev->dev, "AcDcPolarity = 0x%x\n", pptable->AcDcPolarity); - dev_info(smu->adev->dev, "VR0HotGpio = 0x%x\n", pptable->VR0HotGpio); - dev_info(smu->adev->dev, "VR0HotPolarity = 0x%x\n", pptable->VR0HotPolarity); - dev_info(smu->adev->dev, "VR1HotGpio = 0x%x\n", pptable->VR1HotGpio); - dev_info(smu->adev->dev, "VR1HotPolarity = 0x%x\n", pptable->VR1HotPolarity); - dev_info(smu->adev->dev, "GthrGpio = 0x%x\n", pptable->GthrGpio); - dev_info(smu->adev->dev, "GthrPolarity = 0x%x\n", pptable->GthrPolarity); - dev_info(smu->adev->dev, "LedPin0 = 0x%x\n", pptable->LedPin0); - dev_info(smu->adev->dev, "LedPin1 = 0x%x\n", pptable->LedPin1); - dev_info(smu->adev->dev, "LedPin2 = 0x%x\n", pptable->LedPin2); - dev_info(smu->adev->dev, "LedEnableMask = 0x%x\n", pptable->LedEnableMask); - dev_info(smu->adev->dev, "LedPcie = 0x%x\n", pptable->LedPcie); - dev_info(smu->adev->dev, "LedError = 0x%x\n", pptable->LedError); - dev_info(smu->adev->dev, "LedSpare1[0] = 0x%x\n", pptable->LedSpare1[0]); - dev_info(smu->adev->dev, "LedSpare1[1] = 0x%x\n", pptable->LedSpare1[1]); - - dev_info(smu->adev->dev, "PllGfxclkSpreadEnabled = 0x%x\n", pptable->PllGfxclkSpreadEnabled); - dev_info(smu->adev->dev, "PllGfxclkSpreadPercent = 0x%x\n", pptable->PllGfxclkSpreadPercent); - dev_info(smu->adev->dev, "PllGfxclkSpreadFreq = 0x%x\n", pptable->PllGfxclkSpreadFreq); - - dev_info(smu->adev->dev, "DfllGfxclkSpreadEnabled = 0x%x\n", pptable->DfllGfxclkSpreadEnabled); - dev_info(smu->adev->dev, "DfllGfxclkSpreadPercent = 0x%x\n", pptable->DfllGfxclkSpreadPercent); - dev_info(smu->adev->dev, "DfllGfxclkSpreadFreq = 0x%x\n", pptable->DfllGfxclkSpreadFreq); - - dev_info(smu->adev->dev, "UclkSpreadPadding = 0x%x\n", pptable->UclkSpreadPadding); - dev_info(smu->adev->dev, "UclkSpreadFreq = 0x%x\n", pptable->UclkSpreadFreq); - - dev_info(smu->adev->dev, "FclkSpreadEnabled = 0x%x\n", pptable->FclkSpreadEnabled); - dev_info(smu->adev->dev, "FclkSpreadPercent = 0x%x\n", pptable->FclkSpreadPercent); - dev_info(smu->adev->dev, "FclkSpreadFreq = 0x%x\n", pptable->FclkSpreadFreq); - - dev_info(smu->adev->dev, "MemoryChannelEnabled = 0x%x\n", pptable->MemoryChannelEnabled); - dev_info(smu->adev->dev, "DramBitWidth = 0x%x\n", pptable->DramBitWidth); - dev_info(smu->adev->dev, "PaddingMem1[0] = 0x%x\n", pptable->PaddingMem1[0]); - dev_info(smu->adev->dev, "PaddingMem1[1] = 0x%x\n", pptable->PaddingMem1[1]); - dev_info(smu->adev->dev, "PaddingMem1[2] = 0x%x\n", pptable->PaddingMem1[2]); - - dev_info(smu->adev->dev, "TotalBoardPower = 0x%x\n", pptable->TotalBoardPower); - dev_info(smu->adev->dev, "BoardPowerPadding = 0x%x\n", pptable->BoardPowerPadding); - - dev_info(smu->adev->dev, "XgmiLinkSpeed\n"); - for (i = 0; i < NUM_XGMI_PSTATE_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->XgmiLinkSpeed[i]); - dev_info(smu->adev->dev, "XgmiLinkWidth\n"); - for (i = 0; i < NUM_XGMI_PSTATE_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->XgmiLinkWidth[i]); - dev_info(smu->adev->dev, "XgmiFclkFreq\n"); - for (i = 0; i < NUM_XGMI_PSTATE_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->XgmiFclkFreq[i]); - dev_info(smu->adev->dev, "XgmiSocVoltage\n"); - for (i = 0; i < NUM_XGMI_PSTATE_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->XgmiSocVoltage[i]); - - dev_info(smu->adev->dev, "HsrEnabled = 0x%x\n", pptable->HsrEnabled); - dev_info(smu->adev->dev, "VddqOffEnabled = 0x%x\n", pptable->VddqOffEnabled); - dev_info(smu->adev->dev, "PaddingUmcFlags[0] = 0x%x\n", pptable->PaddingUmcFlags[0]); - dev_info(smu->adev->dev, "PaddingUmcFlags[1] = 0x%x\n", pptable->PaddingUmcFlags[1]); - - dev_info(smu->adev->dev, "BoardReserved[0] = 0x%x\n", pptable->BoardReserved[0]); - dev_info(smu->adev->dev, "BoardReserved[1] = 0x%x\n", pptable->BoardReserved[1]); - dev_info(smu->adev->dev, "BoardReserved[2] = 0x%x\n", pptable->BoardReserved[2]); - dev_info(smu->adev->dev, "BoardReserved[3] = 0x%x\n", pptable->BoardReserved[3]); - dev_info(smu->adev->dev, "BoardReserved[4] = 0x%x\n", pptable->BoardReserved[4]); - dev_info(smu->adev->dev, "BoardReserved[5] = 0x%x\n", pptable->BoardReserved[5]); - dev_info(smu->adev->dev, "BoardReserved[6] = 0x%x\n", pptable->BoardReserved[6]); - dev_info(smu->adev->dev, "BoardReserved[7] = 0x%x\n", pptable->BoardReserved[7]); - dev_info(smu->adev->dev, "BoardReserved[8] = 0x%x\n", pptable->BoardReserved[8]); - dev_info(smu->adev->dev, "BoardReserved[9] = 0x%x\n", pptable->BoardReserved[9]); - dev_info(smu->adev->dev, "BoardReserved[10] = 0x%x\n", pptable->BoardReserved[10]); - - dev_info(smu->adev->dev, "MmHubPadding[0] = 0x%x\n", pptable->MmHubPadding[0]); - dev_info(smu->adev->dev, "MmHubPadding[1] = 0x%x\n", pptable->MmHubPadding[1]); - dev_info(smu->adev->dev, "MmHubPadding[2] = 0x%x\n", pptable->MmHubPadding[2]); - dev_info(smu->adev->dev, "MmHubPadding[3] = 0x%x\n", pptable->MmHubPadding[3]); - dev_info(smu->adev->dev, "MmHubPadding[4] = 0x%x\n", pptable->MmHubPadding[4]); - dev_info(smu->adev->dev, "MmHubPadding[5] = 0x%x\n", pptable->MmHubPadding[5]); - dev_info(smu->adev->dev, "MmHubPadding[6] = 0x%x\n", pptable->MmHubPadding[6]); - dev_info(smu->adev->dev, "MmHubPadding[7] = 0x%x\n", pptable->MmHubPadding[7]); -} - -static void sienna_cichlid_dump_pptable(struct smu_context *smu) -{ - struct smu_table_context *table_context = &smu->smu_table; - PPTable_t *pptable = table_context->driver_pptable; - int i; - - if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == - IP_VERSION(11, 0, 13)) { - beige_goby_dump_pptable(smu); - return; - } - - dev_info(smu->adev->dev, "Dumped PPTable:\n"); - - dev_info(smu->adev->dev, "Version = 0x%08x\n", pptable->Version); - dev_info(smu->adev->dev, "FeaturesToRun[0] = 0x%08x\n", pptable->FeaturesToRun[0]); - dev_info(smu->adev->dev, "FeaturesToRun[1] = 0x%08x\n", pptable->FeaturesToRun[1]); - - for (i = 0; i < PPT_THROTTLER_COUNT; i++) { - dev_info(smu->adev->dev, "SocketPowerLimitAc[%d] = 0x%x\n", i, pptable->SocketPowerLimitAc[i]); - dev_info(smu->adev->dev, "SocketPowerLimitAcTau[%d] = 0x%x\n", i, pptable->SocketPowerLimitAcTau[i]); - dev_info(smu->adev->dev, "SocketPowerLimitDc[%d] = 0x%x\n", i, pptable->SocketPowerLimitDc[i]); - dev_info(smu->adev->dev, "SocketPowerLimitDcTau[%d] = 0x%x\n", i, pptable->SocketPowerLimitDcTau[i]); - } - - for (i = 0; i < TDC_THROTTLER_COUNT; i++) { - dev_info(smu->adev->dev, "TdcLimit[%d] = 0x%x\n", i, pptable->TdcLimit[i]); - dev_info(smu->adev->dev, "TdcLimitTau[%d] = 0x%x\n", i, pptable->TdcLimitTau[i]); - } - - for (i = 0; i < TEMP_COUNT; i++) { - dev_info(smu->adev->dev, "TemperatureLimit[%d] = 0x%x\n", i, pptable->TemperatureLimit[i]); - } - - dev_info(smu->adev->dev, "FitLimit = 0x%x\n", pptable->FitLimit); - dev_info(smu->adev->dev, "TotalPowerConfig = 0x%x\n", pptable->TotalPowerConfig); - dev_info(smu->adev->dev, "TotalPowerPadding[0] = 0x%x\n", pptable->TotalPowerPadding[0]); - dev_info(smu->adev->dev, "TotalPowerPadding[1] = 0x%x\n", pptable->TotalPowerPadding[1]); - dev_info(smu->adev->dev, "TotalPowerPadding[2] = 0x%x\n", pptable->TotalPowerPadding[2]); - - dev_info(smu->adev->dev, "ApccPlusResidencyLimit = 0x%x\n", pptable->ApccPlusResidencyLimit); - for (i = 0; i < NUM_SMNCLK_DPM_LEVELS; i++) { - dev_info(smu->adev->dev, "SmnclkDpmFreq[%d] = 0x%x\n", i, pptable->SmnclkDpmFreq[i]); - dev_info(smu->adev->dev, "SmnclkDpmVoltage[%d] = 0x%x\n", i, pptable->SmnclkDpmVoltage[i]); - } - dev_info(smu->adev->dev, "ThrottlerControlMask = 0x%x\n", pptable->ThrottlerControlMask); - - dev_info(smu->adev->dev, "FwDStateMask = 0x%x\n", pptable->FwDStateMask); - - dev_info(smu->adev->dev, "UlvVoltageOffsetSoc = 0x%x\n", pptable->UlvVoltageOffsetSoc); - dev_info(smu->adev->dev, "UlvVoltageOffsetGfx = 0x%x\n", pptable->UlvVoltageOffsetGfx); - dev_info(smu->adev->dev, "MinVoltageUlvGfx = 0x%x\n", pptable->MinVoltageUlvGfx); - dev_info(smu->adev->dev, "MinVoltageUlvSoc = 0x%x\n", pptable->MinVoltageUlvSoc); - - dev_info(smu->adev->dev, "SocLIVmin = 0x%x\n", pptable->SocLIVmin); - dev_info(smu->adev->dev, "PaddingLIVmin = 0x%x\n", pptable->PaddingLIVmin); - - dev_info(smu->adev->dev, "GceaLinkMgrIdleThreshold = 0x%x\n", pptable->GceaLinkMgrIdleThreshold); - dev_info(smu->adev->dev, "paddingRlcUlvParams[0] = 0x%x\n", pptable->paddingRlcUlvParams[0]); - dev_info(smu->adev->dev, "paddingRlcUlvParams[1] = 0x%x\n", pptable->paddingRlcUlvParams[1]); - dev_info(smu->adev->dev, "paddingRlcUlvParams[2] = 0x%x\n", pptable->paddingRlcUlvParams[2]); - - dev_info(smu->adev->dev, "MinVoltageGfx = 0x%x\n", pptable->MinVoltageGfx); - dev_info(smu->adev->dev, "MinVoltageSoc = 0x%x\n", pptable->MinVoltageSoc); - dev_info(smu->adev->dev, "MaxVoltageGfx = 0x%x\n", pptable->MaxVoltageGfx); - dev_info(smu->adev->dev, "MaxVoltageSoc = 0x%x\n", pptable->MaxVoltageSoc); - - dev_info(smu->adev->dev, "LoadLineResistanceGfx = 0x%x\n", pptable->LoadLineResistanceGfx); - dev_info(smu->adev->dev, "LoadLineResistanceSoc = 0x%x\n", pptable->LoadLineResistanceSoc); - - dev_info(smu->adev->dev, "VDDGFX_TVmin = 0x%x\n", pptable->VDDGFX_TVmin); - dev_info(smu->adev->dev, "VDDSOC_TVmin = 0x%x\n", pptable->VDDSOC_TVmin); - dev_info(smu->adev->dev, "VDDGFX_Vmin_HiTemp = 0x%x\n", pptable->VDDGFX_Vmin_HiTemp); - dev_info(smu->adev->dev, "VDDGFX_Vmin_LoTemp = 0x%x\n", pptable->VDDGFX_Vmin_LoTemp); - dev_info(smu->adev->dev, "VDDSOC_Vmin_HiTemp = 0x%x\n", pptable->VDDSOC_Vmin_HiTemp); - dev_info(smu->adev->dev, "VDDSOC_Vmin_LoTemp = 0x%x\n", pptable->VDDSOC_Vmin_LoTemp); - dev_info(smu->adev->dev, "VDDGFX_TVminHystersis = 0x%x\n", pptable->VDDGFX_TVminHystersis); - dev_info(smu->adev->dev, "VDDSOC_TVminHystersis = 0x%x\n", pptable->VDDSOC_TVminHystersis); - - dev_info(smu->adev->dev, "[PPCLK_GFXCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_GFXCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_GFXCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_GFXCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_GFXCLK].Padding, - pptable->DpmDescriptor[PPCLK_GFXCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_GFXCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsCurve.c, - pptable->DpmDescriptor[PPCLK_GFXCLK].SsFmin, - pptable->DpmDescriptor[PPCLK_GFXCLK].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_SOCCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_SOCCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_SOCCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_SOCCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_SOCCLK].Padding, - pptable->DpmDescriptor[PPCLK_SOCCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_SOCCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsCurve.c, - pptable->DpmDescriptor[PPCLK_SOCCLK].SsFmin, - pptable->DpmDescriptor[PPCLK_SOCCLK].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_UCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_UCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_UCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_UCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_UCLK].Padding, - pptable->DpmDescriptor[PPCLK_UCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_UCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_UCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_UCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_UCLK].SsCurve.c, - pptable->DpmDescriptor[PPCLK_UCLK].SsFmin, - pptable->DpmDescriptor[PPCLK_UCLK].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_FCLK]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_FCLK].VoltageMode, - pptable->DpmDescriptor[PPCLK_FCLK].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_FCLK].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_FCLK].Padding, - pptable->DpmDescriptor[PPCLK_FCLK].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_FCLK].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_FCLK].SsCurve.a, - pptable->DpmDescriptor[PPCLK_FCLK].SsCurve.b, - pptable->DpmDescriptor[PPCLK_FCLK].SsCurve.c, - pptable->DpmDescriptor[PPCLK_FCLK].SsFmin, - pptable->DpmDescriptor[PPCLK_FCLK].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_DCLK_0]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_DCLK_0].VoltageMode, - pptable->DpmDescriptor[PPCLK_DCLK_0].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_DCLK_0].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_DCLK_0].Padding, - pptable->DpmDescriptor[PPCLK_DCLK_0].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_DCLK_0].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_DCLK_0].SsCurve.a, - pptable->DpmDescriptor[PPCLK_DCLK_0].SsCurve.b, - pptable->DpmDescriptor[PPCLK_DCLK_0].SsCurve.c, - pptable->DpmDescriptor[PPCLK_DCLK_0].SsFmin, - pptable->DpmDescriptor[PPCLK_DCLK_0].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_VCLK_0]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_VCLK_0].VoltageMode, - pptable->DpmDescriptor[PPCLK_VCLK_0].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_VCLK_0].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_VCLK_0].Padding, - pptable->DpmDescriptor[PPCLK_VCLK_0].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_VCLK_0].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_VCLK_0].SsCurve.a, - pptable->DpmDescriptor[PPCLK_VCLK_0].SsCurve.b, - pptable->DpmDescriptor[PPCLK_VCLK_0].SsCurve.c, - pptable->DpmDescriptor[PPCLK_VCLK_0].SsFmin, - pptable->DpmDescriptor[PPCLK_VCLK_0].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_DCLK_1]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_DCLK_1].VoltageMode, - pptable->DpmDescriptor[PPCLK_DCLK_1].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_DCLK_1].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_DCLK_1].Padding, - pptable->DpmDescriptor[PPCLK_DCLK_1].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_DCLK_1].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_DCLK_1].SsCurve.a, - pptable->DpmDescriptor[PPCLK_DCLK_1].SsCurve.b, - pptable->DpmDescriptor[PPCLK_DCLK_1].SsCurve.c, - pptable->DpmDescriptor[PPCLK_DCLK_1].SsFmin, - pptable->DpmDescriptor[PPCLK_DCLK_1].Padding16); - - dev_info(smu->adev->dev, "[PPCLK_VCLK_1]\n" - " .VoltageMode = 0x%02x\n" - " .SnapToDiscrete = 0x%02x\n" - " .NumDiscreteLevels = 0x%02x\n" - " .padding = 0x%02x\n" - " .ConversionToAvfsClk{m = 0x%08x b = 0x%08x}\n" - " .SsCurve {a = 0x%08x b = 0x%08x c = 0x%08x}\n" - " .SsFmin = 0x%04x\n" - " .Padding_16 = 0x%04x\n", - pptable->DpmDescriptor[PPCLK_VCLK_1].VoltageMode, - pptable->DpmDescriptor[PPCLK_VCLK_1].SnapToDiscrete, - pptable->DpmDescriptor[PPCLK_VCLK_1].NumDiscreteLevels, - pptable->DpmDescriptor[PPCLK_VCLK_1].Padding, - pptable->DpmDescriptor[PPCLK_VCLK_1].ConversionToAvfsClk.m, - pptable->DpmDescriptor[PPCLK_VCLK_1].ConversionToAvfsClk.b, - pptable->DpmDescriptor[PPCLK_VCLK_1].SsCurve.a, - pptable->DpmDescriptor[PPCLK_VCLK_1].SsCurve.b, - pptable->DpmDescriptor[PPCLK_VCLK_1].SsCurve.c, - pptable->DpmDescriptor[PPCLK_VCLK_1].SsFmin, - pptable->DpmDescriptor[PPCLK_VCLK_1].Padding16); - - dev_info(smu->adev->dev, "FreqTableGfx\n"); - for (i = 0; i < NUM_GFXCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = 0x%x\n", i, pptable->FreqTableGfx[i]); - - dev_info(smu->adev->dev, "FreqTableVclk\n"); - for (i = 0; i < NUM_VCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = 0x%x\n", i, pptable->FreqTableVclk[i]); - - dev_info(smu->adev->dev, "FreqTableDclk\n"); - for (i = 0; i < NUM_DCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = 0x%x\n", i, pptable->FreqTableDclk[i]); - - dev_info(smu->adev->dev, "FreqTableSocclk\n"); - for (i = 0; i < NUM_SOCCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = 0x%x\n", i, pptable->FreqTableSocclk[i]); - - dev_info(smu->adev->dev, "FreqTableUclk\n"); - for (i = 0; i < NUM_UCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = 0x%x\n", i, pptable->FreqTableUclk[i]); - - dev_info(smu->adev->dev, "FreqTableFclk\n"); - for (i = 0; i < NUM_FCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%02d] = 0x%x\n", i, pptable->FreqTableFclk[i]); - - dev_info(smu->adev->dev, "DcModeMaxFreq\n"); - dev_info(smu->adev->dev, " .PPCLK_GFXCLK = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_GFXCLK]); - dev_info(smu->adev->dev, " .PPCLK_SOCCLK = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_SOCCLK]); - dev_info(smu->adev->dev, " .PPCLK_UCLK = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_UCLK]); - dev_info(smu->adev->dev, " .PPCLK_FCLK = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_FCLK]); - dev_info(smu->adev->dev, " .PPCLK_DCLK_0 = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_DCLK_0]); - dev_info(smu->adev->dev, " .PPCLK_VCLK_0 = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_VCLK_0]); - dev_info(smu->adev->dev, " .PPCLK_DCLK_1 = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_DCLK_1]); - dev_info(smu->adev->dev, " .PPCLK_VCLK_1 = 0x%x\n", pptable->DcModeMaxFreq[PPCLK_VCLK_1]); - - dev_info(smu->adev->dev, "FreqTableUclkDiv\n"); - for (i = 0; i < NUM_UCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->FreqTableUclkDiv[i]); - - dev_info(smu->adev->dev, "FclkBoostFreq = 0x%x\n", pptable->FclkBoostFreq); - dev_info(smu->adev->dev, "FclkParamPadding = 0x%x\n", pptable->FclkParamPadding); - - dev_info(smu->adev->dev, "Mp0clkFreq\n"); - for (i = 0; i < NUM_MP0CLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->Mp0clkFreq[i]); - - dev_info(smu->adev->dev, "Mp0DpmVoltage\n"); - for (i = 0; i < NUM_MP0CLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->Mp0DpmVoltage[i]); - - dev_info(smu->adev->dev, "MemVddciVoltage\n"); - for (i = 0; i < NUM_UCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->MemVddciVoltage[i]); - - dev_info(smu->adev->dev, "MemMvddVoltage\n"); - for (i = 0; i < NUM_UCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->MemMvddVoltage[i]); - - dev_info(smu->adev->dev, "GfxclkFgfxoffEntry = 0x%x\n", pptable->GfxclkFgfxoffEntry); - dev_info(smu->adev->dev, "GfxclkFinit = 0x%x\n", pptable->GfxclkFinit); - dev_info(smu->adev->dev, "GfxclkFidle = 0x%x\n", pptable->GfxclkFidle); - dev_info(smu->adev->dev, "GfxclkSource = 0x%x\n", pptable->GfxclkSource); - dev_info(smu->adev->dev, "GfxclkPadding = 0x%x\n", pptable->GfxclkPadding); - - dev_info(smu->adev->dev, "GfxGpoSubFeatureMask = 0x%x\n", pptable->GfxGpoSubFeatureMask); - - dev_info(smu->adev->dev, "GfxGpoEnabledWorkPolicyMask = 0x%x\n", pptable->GfxGpoEnabledWorkPolicyMask); - dev_info(smu->adev->dev, "GfxGpoDisabledWorkPolicyMask = 0x%x\n", pptable->GfxGpoDisabledWorkPolicyMask); - dev_info(smu->adev->dev, "GfxGpoPadding[0] = 0x%x\n", pptable->GfxGpoPadding[0]); - dev_info(smu->adev->dev, "GfxGpoVotingAllow = 0x%x\n", pptable->GfxGpoVotingAllow); - dev_info(smu->adev->dev, "GfxGpoPadding32[0] = 0x%x\n", pptable->GfxGpoPadding32[0]); - dev_info(smu->adev->dev, "GfxGpoPadding32[1] = 0x%x\n", pptable->GfxGpoPadding32[1]); - dev_info(smu->adev->dev, "GfxGpoPadding32[2] = 0x%x\n", pptable->GfxGpoPadding32[2]); - dev_info(smu->adev->dev, "GfxGpoPadding32[3] = 0x%x\n", pptable->GfxGpoPadding32[3]); - dev_info(smu->adev->dev, "GfxDcsFopt = 0x%x\n", pptable->GfxDcsFopt); - dev_info(smu->adev->dev, "GfxDcsFclkFopt = 0x%x\n", pptable->GfxDcsFclkFopt); - dev_info(smu->adev->dev, "GfxDcsUclkFopt = 0x%x\n", pptable->GfxDcsUclkFopt); - - dev_info(smu->adev->dev, "DcsGfxOffVoltage = 0x%x\n", pptable->DcsGfxOffVoltage); - dev_info(smu->adev->dev, "DcsMinGfxOffTime = 0x%x\n", pptable->DcsMinGfxOffTime); - dev_info(smu->adev->dev, "DcsMaxGfxOffTime = 0x%x\n", pptable->DcsMaxGfxOffTime); - dev_info(smu->adev->dev, "DcsMinCreditAccum = 0x%x\n", pptable->DcsMinCreditAccum); - dev_info(smu->adev->dev, "DcsExitHysteresis = 0x%x\n", pptable->DcsExitHysteresis); - dev_info(smu->adev->dev, "DcsTimeout = 0x%x\n", pptable->DcsTimeout); - - dev_info(smu->adev->dev, "DcsParamPadding[0] = 0x%x\n", pptable->DcsParamPadding[0]); - dev_info(smu->adev->dev, "DcsParamPadding[1] = 0x%x\n", pptable->DcsParamPadding[1]); - dev_info(smu->adev->dev, "DcsParamPadding[2] = 0x%x\n", pptable->DcsParamPadding[2]); - dev_info(smu->adev->dev, "DcsParamPadding[3] = 0x%x\n", pptable->DcsParamPadding[3]); - dev_info(smu->adev->dev, "DcsParamPadding[4] = 0x%x\n", pptable->DcsParamPadding[4]); - - dev_info(smu->adev->dev, "FlopsPerByteTable\n"); - for (i = 0; i < RLC_PACE_TABLE_NUM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->FlopsPerByteTable[i]); - - dev_info(smu->adev->dev, "LowestUclkReservedForUlv = 0x%x\n", pptable->LowestUclkReservedForUlv); - dev_info(smu->adev->dev, "vddingMem[0] = 0x%x\n", pptable->PaddingMem[0]); - dev_info(smu->adev->dev, "vddingMem[1] = 0x%x\n", pptable->PaddingMem[1]); - dev_info(smu->adev->dev, "vddingMem[2] = 0x%x\n", pptable->PaddingMem[2]); - - dev_info(smu->adev->dev, "UclkDpmPstates\n"); - for (i = 0; i < NUM_UCLK_DPM_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->UclkDpmPstates[i]); - - dev_info(smu->adev->dev, "UclkDpmSrcFreqRange\n"); - dev_info(smu->adev->dev, " .Fmin = 0x%x\n", - pptable->UclkDpmSrcFreqRange.Fmin); - dev_info(smu->adev->dev, " .Fmax = 0x%x\n", - pptable->UclkDpmSrcFreqRange.Fmax); - dev_info(smu->adev->dev, "UclkDpmTargFreqRange\n"); - dev_info(smu->adev->dev, " .Fmin = 0x%x\n", - pptable->UclkDpmTargFreqRange.Fmin); - dev_info(smu->adev->dev, " .Fmax = 0x%x\n", - pptable->UclkDpmTargFreqRange.Fmax); - dev_info(smu->adev->dev, "UclkDpmMidstepFreq = 0x%x\n", pptable->UclkDpmMidstepFreq); - dev_info(smu->adev->dev, "UclkMidstepPadding = 0x%x\n", pptable->UclkMidstepPadding); - - dev_info(smu->adev->dev, "PcieGenSpeed\n"); - for (i = 0; i < NUM_LINK_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->PcieGenSpeed[i]); - - dev_info(smu->adev->dev, "PcieLaneCount\n"); - for (i = 0; i < NUM_LINK_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->PcieLaneCount[i]); - - dev_info(smu->adev->dev, "LclkFreq\n"); - for (i = 0; i < NUM_LINK_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->LclkFreq[i]); - - dev_info(smu->adev->dev, "FanStopTemp = 0x%x\n", pptable->FanStopTemp); - dev_info(smu->adev->dev, "FanStartTemp = 0x%x\n", pptable->FanStartTemp); - - dev_info(smu->adev->dev, "FanGain\n"); - for (i = 0; i < TEMP_COUNT; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->FanGain[i]); - - dev_info(smu->adev->dev, "FanPwmMin = 0x%x\n", pptable->FanPwmMin); - dev_info(smu->adev->dev, "FanAcousticLimitRpm = 0x%x\n", pptable->FanAcousticLimitRpm); - dev_info(smu->adev->dev, "FanThrottlingRpm = 0x%x\n", pptable->FanThrottlingRpm); - dev_info(smu->adev->dev, "FanMaximumRpm = 0x%x\n", pptable->FanMaximumRpm); - dev_info(smu->adev->dev, "MGpuFanBoostLimitRpm = 0x%x\n", pptable->MGpuFanBoostLimitRpm); - dev_info(smu->adev->dev, "FanTargetTemperature = 0x%x\n", pptable->FanTargetTemperature); - dev_info(smu->adev->dev, "FanTargetGfxclk = 0x%x\n", pptable->FanTargetGfxclk); - dev_info(smu->adev->dev, "FanPadding16 = 0x%x\n", pptable->FanPadding16); - dev_info(smu->adev->dev, "FanTempInputSelect = 0x%x\n", pptable->FanTempInputSelect); - dev_info(smu->adev->dev, "FanPadding = 0x%x\n", pptable->FanPadding); - dev_info(smu->adev->dev, "FanZeroRpmEnable = 0x%x\n", pptable->FanZeroRpmEnable); - dev_info(smu->adev->dev, "FanTachEdgePerRev = 0x%x\n", pptable->FanTachEdgePerRev); - - dev_info(smu->adev->dev, "FuzzyFan_ErrorSetDelta = 0x%x\n", pptable->FuzzyFan_ErrorSetDelta); - dev_info(smu->adev->dev, "FuzzyFan_ErrorRateSetDelta = 0x%x\n", pptable->FuzzyFan_ErrorRateSetDelta); - dev_info(smu->adev->dev, "FuzzyFan_PwmSetDelta = 0x%x\n", pptable->FuzzyFan_PwmSetDelta); - dev_info(smu->adev->dev, "FuzzyFan_Reserved = 0x%x\n", pptable->FuzzyFan_Reserved); - - dev_info(smu->adev->dev, "OverrideAvfsGb[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->OverrideAvfsGb[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "OverrideAvfsGb[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->OverrideAvfsGb[AVFS_VOLTAGE_SOC]); - dev_info(smu->adev->dev, "dBtcGbGfxDfllModelSelect = 0x%x\n", pptable->dBtcGbGfxDfllModelSelect); - dev_info(smu->adev->dev, "Padding8_Avfs = 0x%x\n", pptable->Padding8_Avfs); - - dev_info(smu->adev->dev, "qAvfsGb[AVFS_VOLTAGE_GFX]{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->qAvfsGb[AVFS_VOLTAGE_GFX].a, - pptable->qAvfsGb[AVFS_VOLTAGE_GFX].b, - pptable->qAvfsGb[AVFS_VOLTAGE_GFX].c); - dev_info(smu->adev->dev, "qAvfsGb[AVFS_VOLTAGE_SOC]{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->qAvfsGb[AVFS_VOLTAGE_SOC].a, - pptable->qAvfsGb[AVFS_VOLTAGE_SOC].b, - pptable->qAvfsGb[AVFS_VOLTAGE_SOC].c); - dev_info(smu->adev->dev, "dBtcGbGfxPll{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->dBtcGbGfxPll.a, - pptable->dBtcGbGfxPll.b, - pptable->dBtcGbGfxPll.c); - dev_info(smu->adev->dev, "dBtcGbGfxAfll{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->dBtcGbGfxDfll.a, - pptable->dBtcGbGfxDfll.b, - pptable->dBtcGbGfxDfll.c); - dev_info(smu->adev->dev, "dBtcGbSoc{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->dBtcGbSoc.a, - pptable->dBtcGbSoc.b, - pptable->dBtcGbSoc.c); - dev_info(smu->adev->dev, "qAgingGb[AVFS_VOLTAGE_GFX]{m = 0x%x b = 0x%x}\n", - pptable->qAgingGb[AVFS_VOLTAGE_GFX].m, - pptable->qAgingGb[AVFS_VOLTAGE_GFX].b); - dev_info(smu->adev->dev, "qAgingGb[AVFS_VOLTAGE_SOC]{m = 0x%x b = 0x%x}\n", - pptable->qAgingGb[AVFS_VOLTAGE_SOC].m, - pptable->qAgingGb[AVFS_VOLTAGE_SOC].b); - - dev_info(smu->adev->dev, "PiecewiseLinearDroopIntGfxDfll\n"); - for (i = 0; i < NUM_PIECE_WISE_LINEAR_DROOP_MODEL_VF_POINTS; i++) { - dev_info(smu->adev->dev, " Fset[%d] = 0x%x\n", - i, pptable->PiecewiseLinearDroopIntGfxDfll.Fset[i]); - dev_info(smu->adev->dev, " Vdroop[%d] = 0x%x\n", - i, pptable->PiecewiseLinearDroopIntGfxDfll.Vdroop[i]); - } - - dev_info(smu->adev->dev, "qStaticVoltageOffset[AVFS_VOLTAGE_GFX]{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_GFX].a, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_GFX].b, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_GFX].c); - dev_info(smu->adev->dev, "qStaticVoltageOffset[AVFS_VOLTAGE_SOC]{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_SOC].a, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_SOC].b, - pptable->qStaticVoltageOffset[AVFS_VOLTAGE_SOC].c); - - dev_info(smu->adev->dev, "DcTol[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcTol[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcTol[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcTol[AVFS_VOLTAGE_SOC]); - - dev_info(smu->adev->dev, "DcBtcEnabled[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcEnabled[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcBtcEnabled[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcEnabled[AVFS_VOLTAGE_SOC]); - dev_info(smu->adev->dev, "Padding8_GfxBtc[0] = 0x%x\n", pptable->Padding8_GfxBtc[0]); - dev_info(smu->adev->dev, "Padding8_GfxBtc[1] = 0x%x\n", pptable->Padding8_GfxBtc[1]); - - dev_info(smu->adev->dev, "DcBtcMin[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcMin[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcBtcMin[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcMin[AVFS_VOLTAGE_SOC]); - dev_info(smu->adev->dev, "DcBtcMax[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcMax[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcBtcMax[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcMax[AVFS_VOLTAGE_SOC]); - - dev_info(smu->adev->dev, "DcBtcGb[AVFS_VOLTAGE_GFX] = 0x%x\n", pptable->DcBtcGb[AVFS_VOLTAGE_GFX]); - dev_info(smu->adev->dev, "DcBtcGb[AVFS_VOLTAGE_SOC] = 0x%x\n", pptable->DcBtcGb[AVFS_VOLTAGE_SOC]); - - dev_info(smu->adev->dev, "XgmiDpmPstates\n"); - for (i = 0; i < NUM_XGMI_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->XgmiDpmPstates[i]); - dev_info(smu->adev->dev, "XgmiDpmSpare[0] = 0x%02x\n", pptable->XgmiDpmSpare[0]); - dev_info(smu->adev->dev, "XgmiDpmSpare[1] = 0x%02x\n", pptable->XgmiDpmSpare[1]); - - dev_info(smu->adev->dev, "DebugOverrides = 0x%x\n", pptable->DebugOverrides); - dev_info(smu->adev->dev, "ReservedEquation0{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation0.a, - pptable->ReservedEquation0.b, - pptable->ReservedEquation0.c); - dev_info(smu->adev->dev, "ReservedEquation1{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation1.a, - pptable->ReservedEquation1.b, - pptable->ReservedEquation1.c); - dev_info(smu->adev->dev, "ReservedEquation2{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation2.a, - pptable->ReservedEquation2.b, - pptable->ReservedEquation2.c); - dev_info(smu->adev->dev, "ReservedEquation3{a = 0x%x b = 0x%x c = 0x%x}\n", - pptable->ReservedEquation3.a, - pptable->ReservedEquation3.b, - pptable->ReservedEquation3.c); - - dev_info(smu->adev->dev, "SkuReserved[0] = 0x%x\n", pptable->SkuReserved[0]); - dev_info(smu->adev->dev, "SkuReserved[1] = 0x%x\n", pptable->SkuReserved[1]); - dev_info(smu->adev->dev, "SkuReserved[2] = 0x%x\n", pptable->SkuReserved[2]); - dev_info(smu->adev->dev, "SkuReserved[3] = 0x%x\n", pptable->SkuReserved[3]); - dev_info(smu->adev->dev, "SkuReserved[4] = 0x%x\n", pptable->SkuReserved[4]); - dev_info(smu->adev->dev, "SkuReserved[5] = 0x%x\n", pptable->SkuReserved[5]); - dev_info(smu->adev->dev, "SkuReserved[6] = 0x%x\n", pptable->SkuReserved[6]); - dev_info(smu->adev->dev, "SkuReserved[7] = 0x%x\n", pptable->SkuReserved[7]); - - dev_info(smu->adev->dev, "GamingClk[0] = 0x%x\n", pptable->GamingClk[0]); - dev_info(smu->adev->dev, "GamingClk[1] = 0x%x\n", pptable->GamingClk[1]); - dev_info(smu->adev->dev, "GamingClk[2] = 0x%x\n", pptable->GamingClk[2]); - dev_info(smu->adev->dev, "GamingClk[3] = 0x%x\n", pptable->GamingClk[3]); - dev_info(smu->adev->dev, "GamingClk[4] = 0x%x\n", pptable->GamingClk[4]); - dev_info(smu->adev->dev, "GamingClk[5] = 0x%x\n", pptable->GamingClk[5]); - - for (i = 0; i < NUM_I2C_CONTROLLERS; i++) { - dev_info(smu->adev->dev, "I2cControllers[%d]:\n", i); - dev_info(smu->adev->dev, " .Enabled = 0x%x\n", - pptable->I2cControllers[i].Enabled); - dev_info(smu->adev->dev, " .Speed = 0x%x\n", - pptable->I2cControllers[i].Speed); - dev_info(smu->adev->dev, " .SlaveAddress = 0x%x\n", - pptable->I2cControllers[i].SlaveAddress); - dev_info(smu->adev->dev, " .ControllerPort = 0x%x\n", - pptable->I2cControllers[i].ControllerPort); - dev_info(smu->adev->dev, " .ControllerName = 0x%x\n", - pptable->I2cControllers[i].ControllerName); - dev_info(smu->adev->dev, " .ThermalThrottler = 0x%x\n", - pptable->I2cControllers[i].ThermalThrotter); - dev_info(smu->adev->dev, " .I2cProtocol = 0x%x\n", - pptable->I2cControllers[i].I2cProtocol); - dev_info(smu->adev->dev, " .PaddingConfig = 0x%x\n", - pptable->I2cControllers[i].PaddingConfig); - } - - dev_info(smu->adev->dev, "GpioScl = 0x%x\n", pptable->GpioScl); - dev_info(smu->adev->dev, "GpioSda = 0x%x\n", pptable->GpioSda); - dev_info(smu->adev->dev, "FchUsbPdSlaveAddr = 0x%x\n", pptable->FchUsbPdSlaveAddr); - dev_info(smu->adev->dev, "I2cSpare[0] = 0x%x\n", pptable->I2cSpare[0]); - - dev_info(smu->adev->dev, "Board Parameters:\n"); - dev_info(smu->adev->dev, "VddGfxVrMapping = 0x%x\n", pptable->VddGfxVrMapping); - dev_info(smu->adev->dev, "VddSocVrMapping = 0x%x\n", pptable->VddSocVrMapping); - dev_info(smu->adev->dev, "VddMem0VrMapping = 0x%x\n", pptable->VddMem0VrMapping); - dev_info(smu->adev->dev, "VddMem1VrMapping = 0x%x\n", pptable->VddMem1VrMapping); - dev_info(smu->adev->dev, "GfxUlvPhaseSheddingMask = 0x%x\n", pptable->GfxUlvPhaseSheddingMask); - dev_info(smu->adev->dev, "SocUlvPhaseSheddingMask = 0x%x\n", pptable->SocUlvPhaseSheddingMask); - dev_info(smu->adev->dev, "VddciUlvPhaseSheddingMask = 0x%x\n", pptable->VddciUlvPhaseSheddingMask); - dev_info(smu->adev->dev, "MvddUlvPhaseSheddingMask = 0x%x\n", pptable->MvddUlvPhaseSheddingMask); - - dev_info(smu->adev->dev, "GfxMaxCurrent = 0x%x\n", pptable->GfxMaxCurrent); - dev_info(smu->adev->dev, "GfxOffset = 0x%x\n", pptable->GfxOffset); - dev_info(smu->adev->dev, "Padding_TelemetryGfx = 0x%x\n", pptable->Padding_TelemetryGfx); - - dev_info(smu->adev->dev, "SocMaxCurrent = 0x%x\n", pptable->SocMaxCurrent); - dev_info(smu->adev->dev, "SocOffset = 0x%x\n", pptable->SocOffset); - dev_info(smu->adev->dev, "Padding_TelemetrySoc = 0x%x\n", pptable->Padding_TelemetrySoc); - - dev_info(smu->adev->dev, "Mem0MaxCurrent = 0x%x\n", pptable->Mem0MaxCurrent); - dev_info(smu->adev->dev, "Mem0Offset = 0x%x\n", pptable->Mem0Offset); - dev_info(smu->adev->dev, "Padding_TelemetryMem0 = 0x%x\n", pptable->Padding_TelemetryMem0); - - dev_info(smu->adev->dev, "Mem1MaxCurrent = 0x%x\n", pptable->Mem1MaxCurrent); - dev_info(smu->adev->dev, "Mem1Offset = 0x%x\n", pptable->Mem1Offset); - dev_info(smu->adev->dev, "Padding_TelemetryMem1 = 0x%x\n", pptable->Padding_TelemetryMem1); - - dev_info(smu->adev->dev, "MvddRatio = 0x%x\n", pptable->MvddRatio); - - dev_info(smu->adev->dev, "AcDcGpio = 0x%x\n", pptable->AcDcGpio); - dev_info(smu->adev->dev, "AcDcPolarity = 0x%x\n", pptable->AcDcPolarity); - dev_info(smu->adev->dev, "VR0HotGpio = 0x%x\n", pptable->VR0HotGpio); - dev_info(smu->adev->dev, "VR0HotPolarity = 0x%x\n", pptable->VR0HotPolarity); - dev_info(smu->adev->dev, "VR1HotGpio = 0x%x\n", pptable->VR1HotGpio); - dev_info(smu->adev->dev, "VR1HotPolarity = 0x%x\n", pptable->VR1HotPolarity); - dev_info(smu->adev->dev, "GthrGpio = 0x%x\n", pptable->GthrGpio); - dev_info(smu->adev->dev, "GthrPolarity = 0x%x\n", pptable->GthrPolarity); - dev_info(smu->adev->dev, "LedPin0 = 0x%x\n", pptable->LedPin0); - dev_info(smu->adev->dev, "LedPin1 = 0x%x\n", pptable->LedPin1); - dev_info(smu->adev->dev, "LedPin2 = 0x%x\n", pptable->LedPin2); - dev_info(smu->adev->dev, "LedEnableMask = 0x%x\n", pptable->LedEnableMask); - dev_info(smu->adev->dev, "LedPcie = 0x%x\n", pptable->LedPcie); - dev_info(smu->adev->dev, "LedError = 0x%x\n", pptable->LedError); - dev_info(smu->adev->dev, "LedSpare1[0] = 0x%x\n", pptable->LedSpare1[0]); - dev_info(smu->adev->dev, "LedSpare1[1] = 0x%x\n", pptable->LedSpare1[1]); - - dev_info(smu->adev->dev, "PllGfxclkSpreadEnabled = 0x%x\n", pptable->PllGfxclkSpreadEnabled); - dev_info(smu->adev->dev, "PllGfxclkSpreadPercent = 0x%x\n", pptable->PllGfxclkSpreadPercent); - dev_info(smu->adev->dev, "PllGfxclkSpreadFreq = 0x%x\n", pptable->PllGfxclkSpreadFreq); - - dev_info(smu->adev->dev, "DfllGfxclkSpreadEnabled = 0x%x\n", pptable->DfllGfxclkSpreadEnabled); - dev_info(smu->adev->dev, "DfllGfxclkSpreadPercent = 0x%x\n", pptable->DfllGfxclkSpreadPercent); - dev_info(smu->adev->dev, "DfllGfxclkSpreadFreq = 0x%x\n", pptable->DfllGfxclkSpreadFreq); - - dev_info(smu->adev->dev, "UclkSpreadPadding = 0x%x\n", pptable->UclkSpreadPadding); - dev_info(smu->adev->dev, "UclkSpreadFreq = 0x%x\n", pptable->UclkSpreadFreq); - - dev_info(smu->adev->dev, "FclkSpreadEnabled = 0x%x\n", pptable->FclkSpreadEnabled); - dev_info(smu->adev->dev, "FclkSpreadPercent = 0x%x\n", pptable->FclkSpreadPercent); - dev_info(smu->adev->dev, "FclkSpreadFreq = 0x%x\n", pptable->FclkSpreadFreq); - - dev_info(smu->adev->dev, "MemoryChannelEnabled = 0x%x\n", pptable->MemoryChannelEnabled); - dev_info(smu->adev->dev, "DramBitWidth = 0x%x\n", pptable->DramBitWidth); - dev_info(smu->adev->dev, "PaddingMem1[0] = 0x%x\n", pptable->PaddingMem1[0]); - dev_info(smu->adev->dev, "PaddingMem1[1] = 0x%x\n", pptable->PaddingMem1[1]); - dev_info(smu->adev->dev, "PaddingMem1[2] = 0x%x\n", pptable->PaddingMem1[2]); - - dev_info(smu->adev->dev, "TotalBoardPower = 0x%x\n", pptable->TotalBoardPower); - dev_info(smu->adev->dev, "BoardPowerPadding = 0x%x\n", pptable->BoardPowerPadding); - - dev_info(smu->adev->dev, "XgmiLinkSpeed\n"); - for (i = 0; i < NUM_XGMI_PSTATE_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->XgmiLinkSpeed[i]); - dev_info(smu->adev->dev, "XgmiLinkWidth\n"); - for (i = 0; i < NUM_XGMI_PSTATE_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->XgmiLinkWidth[i]); - dev_info(smu->adev->dev, "XgmiFclkFreq\n"); - for (i = 0; i < NUM_XGMI_PSTATE_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->XgmiFclkFreq[i]); - dev_info(smu->adev->dev, "XgmiSocVoltage\n"); - for (i = 0; i < NUM_XGMI_PSTATE_LEVELS; i++) - dev_info(smu->adev->dev, " .[%d] = 0x%x\n", i, pptable->XgmiSocVoltage[i]); - - dev_info(smu->adev->dev, "HsrEnabled = 0x%x\n", pptable->HsrEnabled); - dev_info(smu->adev->dev, "VddqOffEnabled = 0x%x\n", pptable->VddqOffEnabled); - dev_info(smu->adev->dev, "PaddingUmcFlags[0] = 0x%x\n", pptable->PaddingUmcFlags[0]); - dev_info(smu->adev->dev, "PaddingUmcFlags[1] = 0x%x\n", pptable->PaddingUmcFlags[1]); - - dev_info(smu->adev->dev, "BoardReserved[0] = 0x%x\n", pptable->BoardReserved[0]); - dev_info(smu->adev->dev, "BoardReserved[1] = 0x%x\n", pptable->BoardReserved[1]); - dev_info(smu->adev->dev, "BoardReserved[2] = 0x%x\n", pptable->BoardReserved[2]); - dev_info(smu->adev->dev, "BoardReserved[3] = 0x%x\n", pptable->BoardReserved[3]); - dev_info(smu->adev->dev, "BoardReserved[4] = 0x%x\n", pptable->BoardReserved[4]); - dev_info(smu->adev->dev, "BoardReserved[5] = 0x%x\n", pptable->BoardReserved[5]); - dev_info(smu->adev->dev, "BoardReserved[6] = 0x%x\n", pptable->BoardReserved[6]); - dev_info(smu->adev->dev, "BoardReserved[7] = 0x%x\n", pptable->BoardReserved[7]); - dev_info(smu->adev->dev, "BoardReserved[8] = 0x%x\n", pptable->BoardReserved[8]); - dev_info(smu->adev->dev, "BoardReserved[9] = 0x%x\n", pptable->BoardReserved[9]); - dev_info(smu->adev->dev, "BoardReserved[10] = 0x%x\n", pptable->BoardReserved[10]); - - dev_info(smu->adev->dev, "MmHubPadding[0] = 0x%x\n", pptable->MmHubPadding[0]); - dev_info(smu->adev->dev, "MmHubPadding[1] = 0x%x\n", pptable->MmHubPadding[1]); - dev_info(smu->adev->dev, "MmHubPadding[2] = 0x%x\n", pptable->MmHubPadding[2]); - dev_info(smu->adev->dev, "MmHubPadding[3] = 0x%x\n", pptable->MmHubPadding[3]); - dev_info(smu->adev->dev, "MmHubPadding[4] = 0x%x\n", pptable->MmHubPadding[4]); - dev_info(smu->adev->dev, "MmHubPadding[5] = 0x%x\n", pptable->MmHubPadding[5]); - dev_info(smu->adev->dev, "MmHubPadding[6] = 0x%x\n", pptable->MmHubPadding[6]); - dev_info(smu->adev->dev, "MmHubPadding[7] = 0x%x\n", pptable->MmHubPadding[7]); -} - static int sienna_cichlid_i2c_xfer(struct i2c_adapter *i2c_adap, struct i2c_msg *msg, int num_msgs) { @@ -4397,7 +3134,6 @@ static const struct pptable_funcs sienna_cichlid_ppt_funcs = { .display_disable_memory_clock_switch = sienna_cichlid_display_disable_memory_clock_switch, .get_power_limit = sienna_cichlid_get_power_limit, .update_pcie_parameters = sienna_cichlid_update_pcie_parameters, - .dump_pptable = sienna_cichlid_dump_pptable, .init_microcode = smu_v11_0_init_microcode, .load_microcode = smu_v11_0_load_microcode, .fini_microcode = smu_v11_0_fini_microcode, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c index 16fcd9dcd202..480cf3cb204d 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c @@ -1616,7 +1616,8 @@ int smu_v11_0_baco_set_state(struct smu_context *smu, enum smu_baco_state state) break; default: if (!ras || !adev->ras_enabled || - adev->gmc.xgmi.pending_reset) { + (adev->init_lvl->level == + AMDGPU_INIT_LEVEL_MINIMAL_XGMI)) { if (amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(11, 0, 2)) { data = RREG32_SOC15(THM, 0, mmTHM_BACO_CNTL_ARCT); @@ -1763,7 +1764,8 @@ failed: int smu_v11_0_set_soft_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t min, - uint32_t max) + uint32_t max, + bool automatic) { int ret = 0, clk_id = 0; uint32_t param; @@ -1778,7 +1780,10 @@ int smu_v11_0_set_soft_freq_limited_range(struct smu_context *smu, return clk_id; if (max > 0) { - param = (uint32_t)((clk_id << 16) | (max & 0xffff)); + if (automatic) + param = (uint32_t)((clk_id << 16) | 0xffff); + else + param = (uint32_t)((clk_id << 16) | (max & 0xffff)); ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMaxByFreq, param, NULL); if (ret) @@ -1786,7 +1791,10 @@ int smu_v11_0_set_soft_freq_limited_range(struct smu_context *smu, } if (min > 0) { - param = (uint32_t)((clk_id << 16) | (min & 0xffff)); + if (automatic) + param = (uint32_t)((clk_id << 16) | 0); + else + param = (uint32_t)((clk_id << 16) | (min & 0xffff)); ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMinByFreq, param, NULL); if (ret) @@ -1854,6 +1862,7 @@ int smu_v11_0_set_performance_level(struct smu_context *smu, uint32_t mclk_min = 0, mclk_max = 0; uint32_t socclk_min = 0, socclk_max = 0; int ret = 0; + bool auto_level = false; switch (level) { case AMD_DPM_FORCED_LEVEL_HIGH: @@ -1873,6 +1882,7 @@ int smu_v11_0_set_performance_level(struct smu_context *smu, mclk_max = mem_table->max; socclk_min = soc_table->min; socclk_max = soc_table->max; + auto_level = true; break; case AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD: sclk_min = sclk_max = pstate_table->gfxclk_pstate.standard; @@ -1905,13 +1915,15 @@ int smu_v11_0_set_performance_level(struct smu_context *smu, if (amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(11, 0, 2)) { mclk_min = mclk_max = 0; socclk_min = socclk_max = 0; + auto_level = false; } if (sclk_min && sclk_max) { ret = smu_v11_0_set_soft_freq_limited_range(smu, SMU_GFXCLK, sclk_min, - sclk_max); + sclk_max, + auto_level); if (ret) return ret; } @@ -1920,7 +1932,8 @@ int smu_v11_0_set_performance_level(struct smu_context *smu, ret = smu_v11_0_set_soft_freq_limited_range(smu, SMU_MCLK, mclk_min, - mclk_max); + mclk_max, + auto_level); if (ret) return ret; } @@ -1929,7 +1942,8 @@ int smu_v11_0_set_performance_level(struct smu_context *smu, ret = smu_v11_0_set_soft_freq_limited_range(smu, SMU_SOCCLK, socclk_min, - socclk_max); + socclk_max, + auto_level); if (ret) return ret; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c index 1fe020f1f4db..f89c487dce72 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c @@ -461,7 +461,9 @@ static int vangogh_init_smc_tables(struct smu_context *smu) return smu_v11_0_init_smc_tables(smu); } -static int vangogh_dpm_set_vcn_enable(struct smu_context *smu, bool enable) +static int vangogh_dpm_set_vcn_enable(struct smu_context *smu, + bool enable, + int inst) { int ret = 0; @@ -1079,7 +1081,7 @@ static int vangogh_set_power_profile_mode(struct smu_context *smu, long *input, } ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_ActiveProcessNotify, - 1 << workload_type, + smu->workload_mask, NULL); if (ret) { dev_err_once(smu->adev->dev, "Fail to set workload type %d\n", @@ -1087,15 +1089,16 @@ static int vangogh_set_power_profile_mode(struct smu_context *smu, long *input, return ret; } - smu->power_profile_mode = profile_mode; + smu_cmn_assign_power_profile(smu); return 0; } static int vangogh_set_soft_freq_limited_range(struct smu_context *smu, - enum smu_clk_type clk_type, - uint32_t min, - uint32_t max) + enum smu_clk_type clk_type, + uint32_t min, + uint32_t max, + bool automatic) { int ret = 0; @@ -1301,7 +1304,7 @@ static int vangogh_force_dpm_limit_value(struct smu_context *smu, bool highest) return ret; force_freq = highest ? max_freq : min_freq; - ret = vangogh_set_soft_freq_limited_range(smu, clk_type, force_freq, force_freq); + ret = vangogh_set_soft_freq_limited_range(smu, clk_type, force_freq, force_freq, false); if (ret) return ret; } @@ -1337,7 +1340,7 @@ static int vangogh_unforce_dpm_levels(struct smu_context *smu) if (ret) return ret; - ret = vangogh_set_soft_freq_limited_range(smu, clk_type, min_freq, max_freq); + ret = vangogh_set_soft_freq_limited_range(smu, clk_type, min_freq, max_freq, false); if (ret) return ret; @@ -1356,7 +1359,7 @@ static int vangogh_set_peak_clock_by_device(struct smu_context *smu) if (ret) return ret; - ret = vangogh_set_soft_freq_limited_range(smu, SMU_FCLK, fclk_freq, fclk_freq); + ret = vangogh_set_soft_freq_limited_range(smu, SMU_FCLK, fclk_freq, fclk_freq, false); if (ret) return ret; @@ -1364,7 +1367,7 @@ static int vangogh_set_peak_clock_by_device(struct smu_context *smu) if (ret) return ret; - ret = vangogh_set_soft_freq_limited_range(smu, SMU_SOCCLK, socclk_freq, socclk_freq); + ret = vangogh_set_soft_freq_limited_range(smu, SMU_SOCCLK, socclk_freq, socclk_freq, false); if (ret) return ret; @@ -1372,7 +1375,7 @@ static int vangogh_set_peak_clock_by_device(struct smu_context *smu) if (ret) return ret; - ret = vangogh_set_soft_freq_limited_range(smu, SMU_VCLK, vclk_freq, vclk_freq); + ret = vangogh_set_soft_freq_limited_range(smu, SMU_VCLK, vclk_freq, vclk_freq, false); if (ret) return ret; @@ -1380,7 +1383,7 @@ static int vangogh_set_peak_clock_by_device(struct smu_context *smu) if (ret) return ret; - ret = vangogh_set_soft_freq_limited_range(smu, SMU_DCLK, dclk_freq, dclk_freq); + ret = vangogh_set_soft_freq_limited_range(smu, SMU_DCLK, dclk_freq, dclk_freq, false); if (ret) return ret; diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c index cc0504b063fa..75a9ea87f419 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c @@ -645,7 +645,9 @@ static enum amd_pm_state_type renoir_get_current_power_state(struct smu_context return pm_type; } -static int renoir_dpm_set_vcn_enable(struct smu_context *smu, bool enable) +static int renoir_dpm_set_vcn_enable(struct smu_context *smu, + bool enable, + int inst) { int ret = 0; @@ -707,7 +709,7 @@ static int renoir_force_dpm_limit_value(struct smu_context *smu, bool highest) return ret; force_freq = highest ? max_freq : min_freq; - ret = smu_v12_0_set_soft_freq_limited_range(smu, clk_type, force_freq, force_freq); + ret = smu_v12_0_set_soft_freq_limited_range(smu, clk_type, force_freq, force_freq, false); if (ret) return ret; } @@ -740,7 +742,7 @@ static int renoir_unforce_dpm_levels(struct smu_context *smu) { if (ret) return ret; - ret = smu_v12_0_set_soft_freq_limited_range(smu, clk_type, min_freq, max_freq); + ret = smu_v12_0_set_soft_freq_limited_range(smu, clk_type, min_freq, max_freq, false); if (ret) return ret; } @@ -890,14 +892,14 @@ static int renoir_set_power_profile_mode(struct smu_context *smu, long *input, u } ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_ActiveProcessNotify, - 1 << workload_type, + smu->workload_mask, NULL); if (ret) { dev_err_once(smu->adev->dev, "Fail to set workload type %d\n", workload_type); return ret; } - smu->power_profile_mode = profile_mode; + smu_cmn_assign_power_profile(smu); return 0; } @@ -911,7 +913,7 @@ static int renoir_set_peak_clock_by_device(struct smu_context *smu) if (ret) return ret; - ret = smu_v12_0_set_soft_freq_limited_range(smu, SMU_SCLK, sclk_freq, sclk_freq); + ret = smu_v12_0_set_soft_freq_limited_range(smu, SMU_SCLK, sclk_freq, sclk_freq, false); if (ret) return ret; @@ -919,7 +921,7 @@ static int renoir_set_peak_clock_by_device(struct smu_context *smu) if (ret) return ret; - ret = smu_v12_0_set_soft_freq_limited_range(smu, SMU_UCLK, uclk_freq, uclk_freq); + ret = smu_v12_0_set_soft_freq_limited_range(smu, SMU_UCLK, uclk_freq, uclk_freq, false); if (ret) return ret; @@ -961,13 +963,13 @@ static int renior_set_dpm_profile_freq(struct smu_context *smu, } if (sclk) - ret = smu_v12_0_set_soft_freq_limited_range(smu, SMU_SCLK, sclk, sclk); + ret = smu_v12_0_set_soft_freq_limited_range(smu, SMU_SCLK, sclk, sclk, false); if (socclk) - ret = smu_v12_0_set_soft_freq_limited_range(smu, SMU_SOCCLK, socclk, socclk); + ret = smu_v12_0_set_soft_freq_limited_range(smu, SMU_SOCCLK, socclk, socclk, false); if (fclk) - ret = smu_v12_0_set_soft_freq_limited_range(smu, SMU_FCLK, fclk, fclk); + ret = smu_v12_0_set_soft_freq_limited_range(smu, SMU_FCLK, fclk, fclk, false); return ret; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu12/smu_v12_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu12/smu_v12_0.c index ed15f5a0fd11..3d3cd546f0ad 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu12/smu_v12_0.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu12/smu_v12_0.c @@ -211,7 +211,7 @@ int smu_v12_0_mode2_reset(struct smu_context *smu) } int smu_v12_0_set_soft_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type, - uint32_t min, uint32_t max) + uint32_t min, uint32_t max, bool automatic) { int ret = 0; diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c index 2c35eb31475a..f6b029354327 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c @@ -1297,9 +1297,10 @@ static int aldebaran_set_performance_level(struct smu_context *smu, } static int aldebaran_set_soft_freq_limited_range(struct smu_context *smu, - enum smu_clk_type clk_type, - uint32_t min, - uint32_t max) + enum smu_clk_type clk_type, + uint32_t min, + uint32_t max, + bool automatic) { struct smu_dpm_context *smu_dpm = &(smu->smu_dpm); struct smu_13_0_dpm_context *dpm_context = smu_dpm->dpm_context; @@ -1328,7 +1329,7 @@ static int aldebaran_set_soft_freq_limited_range(struct smu_context *smu, return 0; ret = smu_v13_0_set_soft_freq_limited_range(smu, SMU_GFXCLK, - min, max); + min, max, false); if (!ret) { pstate_table->gfxclk_pstate.curr.min = min; pstate_table->gfxclk_pstate.curr.max = max; @@ -1348,7 +1349,7 @@ static int aldebaran_set_soft_freq_limited_range(struct smu_context *smu, /* Restore default min/max clocks and enable determinism */ min_clk = dpm_context->dpm_tables.gfx_table.min; max_clk = dpm_context->dpm_tables.gfx_table.max; - ret = smu_v13_0_set_soft_freq_limited_range(smu, SMU_GFXCLK, min_clk, max_clk); + ret = smu_v13_0_set_soft_freq_limited_range(smu, SMU_GFXCLK, min_clk, max_clk, false); if (!ret) { usleep_range(500, 1000); ret = smu_cmn_send_smc_msg_with_param(smu, @@ -1422,7 +1423,7 @@ static int aldebaran_usr_edit_dpm_table(struct smu_context *smu, enum PP_OD_DPM_ min_clk = dpm_context->dpm_tables.gfx_table.min; max_clk = dpm_context->dpm_tables.gfx_table.max; - return aldebaran_set_soft_freq_limited_range(smu, SMU_GFXCLK, min_clk, max_clk); + return aldebaran_set_soft_freq_limited_range(smu, SMU_GFXCLK, min_clk, max_clk, false); } break; case PP_OD_COMMIT_DPM_TABLE: @@ -1441,7 +1442,7 @@ static int aldebaran_usr_edit_dpm_table(struct smu_context *smu, enum PP_OD_DPM_ min_clk = pstate_table->gfxclk_pstate.custom.min; max_clk = pstate_table->gfxclk_pstate.custom.max; - return aldebaran_set_soft_freq_limited_range(smu, SMU_GFXCLK, min_clk, max_clk); + return aldebaran_set_soft_freq_limited_range(smu, SMU_GFXCLK, min_clk, max_clk, false); } break; default: diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c index e17466cc1952..2bfea740dace 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c @@ -1608,7 +1608,8 @@ failed: int smu_v13_0_set_soft_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t min, - uint32_t max) + uint32_t max, + bool automatic) { int ret = 0, clk_id = 0; uint32_t param; @@ -1623,7 +1624,10 @@ int smu_v13_0_set_soft_freq_limited_range(struct smu_context *smu, return clk_id; if (max > 0) { - param = (uint32_t)((clk_id << 16) | (max & 0xffff)); + if (automatic) + param = (uint32_t)((clk_id << 16) | 0xffff); + else + param = (uint32_t)((clk_id << 16) | (max & 0xffff)); ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMaxByFreq, param, NULL); if (ret) @@ -1631,7 +1635,10 @@ int smu_v13_0_set_soft_freq_limited_range(struct smu_context *smu, } if (min > 0) { - param = (uint32_t)((clk_id << 16) | (min & 0xffff)); + if (automatic) + param = (uint32_t)((clk_id << 16) | 0); + else + param = (uint32_t)((clk_id << 16) | (min & 0xffff)); ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMinByFreq, param, NULL); if (ret) @@ -1708,6 +1715,7 @@ int smu_v13_0_set_performance_level(struct smu_context *smu, uint32_t dclk_min = 0, dclk_max = 0; uint32_t fclk_min = 0, fclk_max = 0; int ret = 0, i; + bool auto_level = false; switch (level) { case AMD_DPM_FORCED_LEVEL_HIGH: @@ -1739,6 +1747,7 @@ int smu_v13_0_set_performance_level(struct smu_context *smu, dclk_max = dclk_table->max; fclk_min = fclk_table->min; fclk_max = fclk_table->max; + auto_level = true; break; case AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD: sclk_min = sclk_max = pstate_table->gfxclk_pstate.standard; @@ -1780,13 +1789,15 @@ int smu_v13_0_set_performance_level(struct smu_context *smu, vclk_min = vclk_max = 0; dclk_min = dclk_max = 0; fclk_min = fclk_max = 0; + auto_level = false; } if (sclk_min && sclk_max) { ret = smu_v13_0_set_soft_freq_limited_range(smu, SMU_GFXCLK, sclk_min, - sclk_max); + sclk_max, + auto_level); if (ret) return ret; @@ -1798,7 +1809,8 @@ int smu_v13_0_set_performance_level(struct smu_context *smu, ret = smu_v13_0_set_soft_freq_limited_range(smu, SMU_MCLK, mclk_min, - mclk_max); + mclk_max, + auto_level); if (ret) return ret; @@ -1810,7 +1822,8 @@ int smu_v13_0_set_performance_level(struct smu_context *smu, ret = smu_v13_0_set_soft_freq_limited_range(smu, SMU_SOCCLK, socclk_min, - socclk_max); + socclk_max, + auto_level); if (ret) return ret; @@ -1825,7 +1838,8 @@ int smu_v13_0_set_performance_level(struct smu_context *smu, ret = smu_v13_0_set_soft_freq_limited_range(smu, i ? SMU_VCLK1 : SMU_VCLK, vclk_min, - vclk_max); + vclk_max, + auto_level); if (ret) return ret; } @@ -1840,7 +1854,8 @@ int smu_v13_0_set_performance_level(struct smu_context *smu, ret = smu_v13_0_set_soft_freq_limited_range(smu, i ? SMU_DCLK1 : SMU_DCLK, dclk_min, - dclk_max); + dclk_max, + auto_level); if (ret) return ret; } @@ -1852,7 +1867,8 @@ int smu_v13_0_set_performance_level(struct smu_context *smu, ret = smu_v13_0_set_soft_freq_limited_range(smu, SMU_FCLK, fclk_min, - fclk_max); + fclk_max, + auto_level); if (ret) return ret; @@ -2088,7 +2104,8 @@ int smu_v13_0_get_current_pcie_link_speed(struct smu_context *smu) } int smu_v13_0_set_vcn_enable(struct smu_context *smu, - bool enable) + bool enable, + int inst) { struct amdgpu_device *adev = smu->adev; int i, ret = 0; diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c index d53e162dcd8d..80c6b1e523aa 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c @@ -107,6 +107,8 @@ #define PP_OD_FEATURE_FAN_ACOUSTIC_TARGET 8 #define PP_OD_FEATURE_FAN_TARGET_TEMPERATURE 9 #define PP_OD_FEATURE_FAN_MINIMUM_PWM 10 +#define PP_OD_FEATURE_FAN_ZERO_RPM_ENABLE 11 +#define PP_OD_FEATURE_FAN_ZERO_RPM_STOP_TEMP 12 #define LINK_SPEED_MAX 3 @@ -736,19 +738,6 @@ static bool smu_v13_0_0_is_dpm_running(struct smu_context *smu) return !!(feature_enabled & SMC_DPM_FEATURE); } -static void smu_v13_0_0_dump_pptable(struct smu_context *smu) -{ - struct smu_table_context *table_context = &smu->smu_table; - PPTable_t *pptable = table_context->driver_pptable; - SkuTable_t *skutable = &pptable->SkuTable; - - dev_info(smu->adev->dev, "Dumped PPTable:\n"); - - dev_info(smu->adev->dev, "Version = 0x%08x\n", skutable->Version); - dev_info(smu->adev->dev, "FeaturesToRun[0] = 0x%08x\n", skutable->FeaturesToRun[0]); - dev_info(smu->adev->dev, "FeaturesToRun[1] = 0x%08x\n", skutable->FeaturesToRun[1]); -} - static int smu_v13_0_0_system_features_control(struct smu_context *smu, bool en) { @@ -1143,6 +1132,14 @@ static void smu_v13_0_0_get_od_setting_limits(struct smu_context *smu, od_min_setting = overdrive_lowerlimits->FanMinimumPwm; od_max_setting = overdrive_upperlimits->FanMinimumPwm; break; + case PP_OD_FEATURE_FAN_ZERO_RPM_ENABLE: + od_min_setting = overdrive_lowerlimits->FanZeroRpmEnable; + od_max_setting = overdrive_upperlimits->FanZeroRpmEnable; + break; + case PP_OD_FEATURE_FAN_ZERO_RPM_STOP_TEMP: + od_min_setting = overdrive_lowerlimits->FanZeroRpmStopTemp; + od_max_setting = overdrive_upperlimits->FanZeroRpmStopTemp; + break; default: od_min_setting = od_max_setting = INT_MAX; break; @@ -1463,6 +1460,42 @@ static int smu_v13_0_0_print_clk_levels(struct smu_context *smu, min_value, max_value); break; + case SMU_OD_FAN_ZERO_RPM_ENABLE: + if (!smu_v13_0_0_is_od_feature_supported(smu, + PP_OD_FEATURE_ZERO_FAN_BIT)) + break; + + size += sysfs_emit_at(buf, size, "FAN_ZERO_RPM_ENABLE:\n"); + size += sysfs_emit_at(buf, size, "%d\n", + (int)od_table->OverDriveTable.FanZeroRpmEnable); + + size += sysfs_emit_at(buf, size, "%s:\n", "OD_RANGE"); + smu_v13_0_0_get_od_setting_limits(smu, + PP_OD_FEATURE_FAN_ZERO_RPM_ENABLE, + &min_value, + &max_value); + size += sysfs_emit_at(buf, size, "ZERO_RPM_ENABLE: %u %u\n", + min_value, max_value); + break; + + case SMU_OD_FAN_ZERO_RPM_STOP_TEMP: + if (!smu_v13_0_0_is_od_feature_supported(smu, + PP_OD_FEATURE_ZERO_FAN_BIT)) + break; + + size += sysfs_emit_at(buf, size, "FAN_ZERO_RPM_STOP_TEMPERATURE:\n"); + size += sysfs_emit_at(buf, size, "%d\n", + (int)od_table->OverDriveTable.FanZeroRpmStopTemp); + + size += sysfs_emit_at(buf, size, "%s:\n", "OD_RANGE"); + smu_v13_0_0_get_od_setting_limits(smu, + PP_OD_FEATURE_FAN_ZERO_RPM_STOP_TEMP, + &min_value, + &max_value); + size += sysfs_emit_at(buf, size, "ZERO_RPM_STOP_TEMPERATURE: %u %u\n", + min_value, max_value); + break; + case SMU_OD_RANGE: if (!smu_v13_0_0_is_od_feature_supported(smu, PP_OD_FEATURE_GFXCLK_BIT) && !smu_v13_0_0_is_od_feature_supported(smu, PP_OD_FEATURE_UCLK_BIT) && @@ -1560,6 +1593,16 @@ static int smu_v13_0_0_od_restore_table_single(struct smu_context *smu, long inp od_table->OverDriveTable.FanMode = FAN_MODE_AUTO; od_table->OverDriveTable.FeatureCtrlMask |= BIT(PP_OD_FEATURE_FAN_CURVE_BIT); break; + case PP_OD_EDIT_FAN_ZERO_RPM_ENABLE: + od_table->OverDriveTable.FanZeroRpmEnable = + boot_overdrive_table->OverDriveTable.FanZeroRpmEnable; + od_table->OverDriveTable.FeatureCtrlMask |= BIT(PP_OD_FEATURE_ZERO_FAN_BIT); + break; + case PP_OD_EDIT_FAN_ZERO_RPM_STOP_TEMP: + od_table->OverDriveTable.FanZeroRpmStopTemp = + boot_overdrive_table->OverDriveTable.FanZeroRpmStopTemp; + od_table->OverDriveTable.FeatureCtrlMask |= BIT(PP_OD_FEATURE_ZERO_FAN_BIT); + break; default: dev_info(adev->dev, "Invalid table index: %ld\n", input); return -EINVAL; @@ -1853,6 +1896,48 @@ static int smu_v13_0_0_od_edit_dpm_table(struct smu_context *smu, od_table->OverDriveTable.FeatureCtrlMask |= BIT(PP_OD_FEATURE_FAN_CURVE_BIT); break; + case PP_OD_EDIT_FAN_ZERO_RPM_ENABLE: + if (!smu_v13_0_0_is_od_feature_supported(smu, PP_OD_FEATURE_ZERO_FAN_BIT)) { + dev_warn(adev->dev, "Zero RPM setting not supported!\n"); + return -ENOTSUPP; + } + + smu_v13_0_0_get_od_setting_limits(smu, + PP_OD_FEATURE_FAN_ZERO_RPM_ENABLE, + &minimum, + &maximum); + if (input[0] < minimum || + input[0] > maximum) { + dev_info(adev->dev, "zero RPM enable setting(%ld) must be within [%d, %d]!\n", + input[0], minimum, maximum); + return -EINVAL; + } + + od_table->OverDriveTable.FanZeroRpmEnable = input[0]; + od_table->OverDriveTable.FeatureCtrlMask |= BIT(PP_OD_FEATURE_ZERO_FAN_BIT); + break; + + case PP_OD_EDIT_FAN_ZERO_RPM_STOP_TEMP: + if (!smu_v13_0_0_is_od_feature_supported(smu, PP_OD_FEATURE_ZERO_FAN_BIT)) { + dev_warn(adev->dev, "Zero RPM setting not supported!\n"); + return -ENOTSUPP; + } + + smu_v13_0_0_get_od_setting_limits(smu, + PP_OD_FEATURE_FAN_ZERO_RPM_STOP_TEMP, + &minimum, + &maximum); + if (input[0] < minimum || + input[0] > maximum) { + dev_info(adev->dev, "zero RPM stop temperature setting(%ld) must be within [%d, %d]!\n", + input[0], minimum, maximum); + return -EINVAL; + } + + od_table->OverDriveTable.FanZeroRpmStopTemp = input[0]; + od_table->OverDriveTable.FeatureCtrlMask |= BIT(PP_OD_FEATURE_ZERO_FAN_BIT); + break; + case PP_OD_RESTORE_DEFAULT_TABLE: if (size == 1) { ret = smu_v13_0_0_od_restore_table_single(smu, input[0]); @@ -1975,7 +2060,8 @@ static int smu_v13_0_0_force_clk_levels(struct smu_context *smu, ret = smu_v13_0_set_soft_freq_limited_range(smu, clk_type, min_freq, - max_freq); + max_freq, + false); break; case SMU_DCEFCLK: case SMU_PCIE: @@ -2122,7 +2208,11 @@ static void smu_v13_0_0_set_supported_od_feature_mask(struct smu_context *smu) OD_OPS_SUPPORT_FAN_TARGET_TEMPERATURE_RETRIEVE | OD_OPS_SUPPORT_FAN_TARGET_TEMPERATURE_SET | OD_OPS_SUPPORT_FAN_MINIMUM_PWM_RETRIEVE | - OD_OPS_SUPPORT_FAN_MINIMUM_PWM_SET; + OD_OPS_SUPPORT_FAN_MINIMUM_PWM_SET | + OD_OPS_SUPPORT_FAN_ZERO_RPM_ENABLE_RETRIEVE | + OD_OPS_SUPPORT_FAN_ZERO_RPM_ENABLE_SET | + OD_OPS_SUPPORT_FAN_ZERO_RPM_STOP_TEMP_RETRIEVE | + OD_OPS_SUPPORT_FAN_ZERO_RPM_STOP_TEMP_SET; } static int smu_v13_0_0_set_default_od_settings(struct smu_context *smu) @@ -2188,6 +2278,10 @@ static int smu_v13_0_0_set_default_od_settings(struct smu_context *smu) user_od_table_bak.OverDriveTable.FanTargetTemperature; user_od_table->OverDriveTable.FanMinimumPwm = user_od_table_bak.OverDriveTable.FanMinimumPwm; + user_od_table->OverDriveTable.FanZeroRpmEnable = + user_od_table_bak.OverDriveTable.FanZeroRpmEnable; + user_od_table->OverDriveTable.FanZeroRpmStopTemp = + user_od_table_bak.OverDriveTable.FanZeroRpmStopTemp; } smu_v13_0_0_set_supported_od_feature_mask(smu); @@ -2485,7 +2579,7 @@ static int smu_v13_0_0_set_power_profile_mode(struct smu_context *smu, DpmActivityMonitorCoeffInt_t *activity_monitor = &(activity_monitor_external.DpmActivityMonitorCoeffInt); int workload_type, ret = 0; - u32 workload_mask, selected_workload_mask; + u32 workload_mask; smu->power_profile_mode = input[size]; @@ -2552,7 +2646,7 @@ static int smu_v13_0_0_set_power_profile_mode(struct smu_context *smu, if (workload_type < 0) return -EINVAL; - selected_workload_mask = workload_mask = 1 << workload_type; + workload_mask = 1 << workload_type; /* Add optimizations for SMU13.0.0/10. Reuse the power saving profile */ if ((amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(13, 0, 0) && @@ -2567,12 +2661,22 @@ static int smu_v13_0_0_set_power_profile_mode(struct smu_context *smu, workload_mask |= 1 << workload_type; } + smu->workload_mask |= workload_mask; ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask, - workload_mask, + smu->workload_mask, NULL); - if (!ret) - smu->workload_mask = selected_workload_mask; + if (!ret) { + smu_cmn_assign_power_profile(smu); + if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_POWERSAVING) { + workload_type = smu_cmn_to_asic_specific_index(smu, + CMN2ASIC_MAPPING_WORKLOAD, + PP_SMC_POWER_PROFILE_FULLSCREEN3D); + smu->power_profile_mode = smu->workload_mask & (1 << workload_type) + ? PP_SMC_POWER_PROFILE_FULLSCREEN3D + : PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT; + } + } return ret; } @@ -3024,7 +3128,6 @@ static const struct pptable_funcs smu_v13_0_0_ppt_funcs = { .i2c_init = smu_v13_0_0_i2c_control_init, .i2c_fini = smu_v13_0_0_i2c_control_fini, .is_dpm_running = smu_v13_0_0_is_dpm_running, - .dump_pptable = smu_v13_0_0_dump_pptable, .init_microcode = smu_v13_0_init_microcode, .load_microcode = smu_v13_0_load_microcode, .fini_microcode = smu_v13_0_fini_microcode, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_5_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_5_ppt.c index 9c2c43bfed0b..f5db181ef489 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_5_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_5_ppt.c @@ -193,7 +193,9 @@ static int smu_v13_0_5_system_features_control(struct smu_context *smu, bool en) return ret; } -static int smu_v13_0_5_dpm_set_vcn_enable(struct smu_context *smu, bool enable) +static int smu_v13_0_5_dpm_set_vcn_enable(struct smu_context *smu, + bool enable, + int inst) { int ret = 0; @@ -811,9 +813,10 @@ failed: } static int smu_v13_0_5_set_soft_freq_limited_range(struct smu_context *smu, - enum smu_clk_type clk_type, - uint32_t min, - uint32_t max) + enum smu_clk_type clk_type, + uint32_t min, + uint32_t max, + bool automatic) { enum smu_message_type msg_set_min, msg_set_max; uint32_t min_clk = min; @@ -950,7 +953,7 @@ static int smu_v13_0_5_force_clk_levels(struct smu_context *smu, if (ret) goto force_level_out; - ret = smu_v13_0_5_set_soft_freq_limited_range(smu, clk_type, min_freq, max_freq); + ret = smu_v13_0_5_set_soft_freq_limited_range(smu, clk_type, min_freq, max_freq, false); if (ret) goto force_level_out; break; @@ -1046,9 +1049,10 @@ static int smu_v13_0_5_set_performance_level(struct smu_context *smu, if (sclk_min && sclk_max) { ret = smu_v13_0_5_set_soft_freq_limited_range(smu, - SMU_SCLK, - sclk_min, - sclk_max); + SMU_SCLK, + sclk_min, + sclk_max, + false); if (ret) return ret; @@ -1060,7 +1064,8 @@ static int smu_v13_0_5_set_performance_level(struct smu_context *smu, ret = smu_v13_0_5_set_soft_freq_limited_range(smu, SMU_VCLK, vclk_min, - vclk_max); + vclk_max, + false); if (ret) return ret; } @@ -1069,7 +1074,8 @@ static int smu_v13_0_5_set_performance_level(struct smu_context *smu, ret = smu_v13_0_5_set_soft_freq_limited_range(smu, SMU_DCLK, dclk_min, - dclk_max); + dclk_max, + false); if (ret) return ret; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c index 55ed6247eb61..fa30a9e1f27a 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c @@ -102,6 +102,24 @@ MODULE_FIRMWARE("amdgpu/smu_13_0_14.bin"); #define MCA_BANK_IPID(_ip, _hwid, _type) \ [AMDGPU_MCA_IP_##_ip] = { .hwid = _hwid, .mcatype = _type, } +static inline bool smu_v13_0_6_is_unified_metrics(struct smu_context *smu) +{ + return (smu->adev->flags & AMD_IS_APU) && + smu->smc_fw_version <= 0x4556900; +} + +static inline bool smu_v13_0_6_is_other_end_count_available(struct smu_context *smu) +{ + switch (amdgpu_ip_version(smu->adev, MP1_HWIP, 0)) { + case IP_VERSION(13, 0, 6): + return smu->smc_fw_version >= 0x557600; + case IP_VERSION(13, 0, 14): + return smu->smc_fw_version >= 0x05550E00; + default: + return false; + } +} + struct mca_bank_ipid { enum amdgpu_mca_ip ip; uint16_t hwid; @@ -253,7 +271,7 @@ struct PPTable_t { #define SMUQ10_TO_UINT(x) ((x) >> 10) #define SMUQ10_FRAC(x) ((x) & 0x3ff) #define SMUQ10_ROUND(x) ((SMUQ10_TO_UINT(x)) + ((SMUQ10_FRAC(x)) >= 0x200)) -#define GET_METRIC_FIELD(field) ((adev->flags & AMD_IS_APU) ?\ +#define GET_METRIC_FIELD(field, flag) ((flag) ?\ (metrics_a->field) : (metrics_x->field)) struct smu_v13_0_6_dpm_map { @@ -352,7 +370,7 @@ static int smu_v13_0_6_tables_init(struct smu_context *smu) return -ENOMEM; smu_table->metrics_time = 0; - smu_table->gpu_metrics_table_size = sizeof(struct gpu_metrics_v1_5); + smu_table->gpu_metrics_table_size = sizeof(struct gpu_metrics_v1_6); smu_table->gpu_metrics_table = kzalloc(smu_table->gpu_metrics_table_size, GFP_KERNEL); if (!smu_table->gpu_metrics_table) { @@ -583,7 +601,7 @@ static int smu_v13_0_6_setup_driver_pptable(struct smu_context *smu) MetricsTableA_t *metrics_a = (MetricsTableA_t *)smu_table->metrics_table; struct PPTable_t *pptable = (struct PPTable_t *)smu_table->driver_pptable; - struct amdgpu_device *adev = smu->adev; + bool flag = smu_v13_0_6_is_unified_metrics(smu); int ret, i, retry = 100; uint32_t table_version; @@ -595,7 +613,7 @@ static int smu_v13_0_6_setup_driver_pptable(struct smu_context *smu) return ret; /* Ensure that metrics have been updated */ - if (GET_METRIC_FIELD(AccumulationCounter)) + if (GET_METRIC_FIELD(AccumulationCounter, flag)) break; usleep_range(1000, 1100); @@ -612,29 +630,29 @@ static int smu_v13_0_6_setup_driver_pptable(struct smu_context *smu) table_version; pptable->MaxSocketPowerLimit = - SMUQ10_ROUND(GET_METRIC_FIELD(MaxSocketPowerLimit)); + SMUQ10_ROUND(GET_METRIC_FIELD(MaxSocketPowerLimit, flag)); pptable->MaxGfxclkFrequency = - SMUQ10_ROUND(GET_METRIC_FIELD(MaxGfxclkFrequency)); + SMUQ10_ROUND(GET_METRIC_FIELD(MaxGfxclkFrequency, flag)); pptable->MinGfxclkFrequency = - SMUQ10_ROUND(GET_METRIC_FIELD(MinGfxclkFrequency)); + SMUQ10_ROUND(GET_METRIC_FIELD(MinGfxclkFrequency, flag)); for (i = 0; i < 4; ++i) { pptable->FclkFrequencyTable[i] = - SMUQ10_ROUND(GET_METRIC_FIELD(FclkFrequencyTable)[i]); + SMUQ10_ROUND(GET_METRIC_FIELD(FclkFrequencyTable, flag)[i]); pptable->UclkFrequencyTable[i] = - SMUQ10_ROUND(GET_METRIC_FIELD(UclkFrequencyTable)[i]); + SMUQ10_ROUND(GET_METRIC_FIELD(UclkFrequencyTable, flag)[i]); pptable->SocclkFrequencyTable[i] = SMUQ10_ROUND( - GET_METRIC_FIELD(SocclkFrequencyTable)[i]); + GET_METRIC_FIELD(SocclkFrequencyTable, flag)[i]); pptable->VclkFrequencyTable[i] = - SMUQ10_ROUND(GET_METRIC_FIELD(VclkFrequencyTable)[i]); + SMUQ10_ROUND(GET_METRIC_FIELD(VclkFrequencyTable, flag)[i]); pptable->DclkFrequencyTable[i] = - SMUQ10_ROUND(GET_METRIC_FIELD(DclkFrequencyTable)[i]); + SMUQ10_ROUND(GET_METRIC_FIELD(DclkFrequencyTable, flag)[i]); pptable->LclkFrequencyTable[i] = - SMUQ10_ROUND(GET_METRIC_FIELD(LclkFrequencyTable)[i]); + SMUQ10_ROUND(GET_METRIC_FIELD(LclkFrequencyTable, flag)[i]); } /* use AID0 serial number by default */ - pptable->PublicSerialNumber_AID = GET_METRIC_FIELD(PublicSerialNumber_AID)[0]; + pptable->PublicSerialNumber_AID = GET_METRIC_FIELD(PublicSerialNumber_AID, flag)[0]; pptable->Init = true; } @@ -957,6 +975,7 @@ static int smu_v13_0_6_get_smu_metrics_data(struct smu_context *smu, struct smu_table_context *smu_table = &smu->smu_table; MetricsTableX_t *metrics_x = (MetricsTableX_t *)smu_table->metrics_table; MetricsTableA_t *metrics_a = (MetricsTableA_t *)smu_table->metrics_table; + bool flag = smu_v13_0_6_is_unified_metrics(smu); struct amdgpu_device *adev = smu->adev; int ret = 0; int xcc_id; @@ -971,50 +990,50 @@ static int smu_v13_0_6_get_smu_metrics_data(struct smu_context *smu, case METRICS_AVERAGE_GFXCLK: if (smu->smc_fw_version >= 0x552F00) { xcc_id = GET_INST(GC, 0); - *value = SMUQ10_ROUND(GET_METRIC_FIELD(GfxclkFrequency)[xcc_id]); + *value = SMUQ10_ROUND(GET_METRIC_FIELD(GfxclkFrequency, flag)[xcc_id]); } else { *value = 0; } break; case METRICS_CURR_SOCCLK: case METRICS_AVERAGE_SOCCLK: - *value = SMUQ10_ROUND(GET_METRIC_FIELD(SocclkFrequency)[0]); + *value = SMUQ10_ROUND(GET_METRIC_FIELD(SocclkFrequency, flag)[0]); break; case METRICS_CURR_UCLK: case METRICS_AVERAGE_UCLK: - *value = SMUQ10_ROUND(GET_METRIC_FIELD(UclkFrequency)); + *value = SMUQ10_ROUND(GET_METRIC_FIELD(UclkFrequency, flag)); break; case METRICS_CURR_VCLK: - *value = SMUQ10_ROUND(GET_METRIC_FIELD(VclkFrequency)[0]); + *value = SMUQ10_ROUND(GET_METRIC_FIELD(VclkFrequency, flag)[0]); break; case METRICS_CURR_DCLK: - *value = SMUQ10_ROUND(GET_METRIC_FIELD(DclkFrequency)[0]); + *value = SMUQ10_ROUND(GET_METRIC_FIELD(DclkFrequency, flag)[0]); break; case METRICS_CURR_FCLK: - *value = SMUQ10_ROUND(GET_METRIC_FIELD(FclkFrequency)); + *value = SMUQ10_ROUND(GET_METRIC_FIELD(FclkFrequency, flag)); break; case METRICS_AVERAGE_GFXACTIVITY: - *value = SMUQ10_ROUND(GET_METRIC_FIELD(SocketGfxBusy)); + *value = SMUQ10_ROUND(GET_METRIC_FIELD(SocketGfxBusy, flag)); break; case METRICS_AVERAGE_MEMACTIVITY: - *value = SMUQ10_ROUND(GET_METRIC_FIELD(DramBandwidthUtilization)); + *value = SMUQ10_ROUND(GET_METRIC_FIELD(DramBandwidthUtilization, flag)); break; case METRICS_CURR_SOCKETPOWER: - *value = SMUQ10_ROUND(GET_METRIC_FIELD(SocketPower)) << 8; + *value = SMUQ10_ROUND(GET_METRIC_FIELD(SocketPower, flag)) << 8; break; case METRICS_TEMPERATURE_HOTSPOT: - *value = SMUQ10_ROUND(GET_METRIC_FIELD(MaxSocketTemperature)) * + *value = SMUQ10_ROUND(GET_METRIC_FIELD(MaxSocketTemperature, flag)) * SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; break; case METRICS_TEMPERATURE_MEM: - *value = SMUQ10_ROUND(GET_METRIC_FIELD(MaxHbmTemperature)) * + *value = SMUQ10_ROUND(GET_METRIC_FIELD(MaxHbmTemperature, flag)) * SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; break; /* This is the max of all VRs and not just SOC VR. * No need to define another data type for the same. */ case METRICS_TEMPERATURE_VRSOC: - *value = SMUQ10_ROUND(GET_METRIC_FIELD(MaxVrTemperature)) * + *value = SMUQ10_ROUND(GET_METRIC_FIELD(MaxVrTemperature, flag)) * SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; break; default: @@ -1739,7 +1758,7 @@ static int smu_v13_0_6_set_performance_level(struct smu_context *smu, if (uclk_table->max != pstate_table->uclk_pstate.curr.max) { /* Min UCLK is not expected to be changed */ ret = smu_v13_0_set_soft_freq_limited_range( - smu, SMU_UCLK, 0, uclk_table->max); + smu, SMU_UCLK, 0, uclk_table->max, false); if (ret) return ret; pstate_table->uclk_pstate.curr.max = uclk_table->max; @@ -1758,7 +1777,8 @@ static int smu_v13_0_6_set_performance_level(struct smu_context *smu, static int smu_v13_0_6_set_soft_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type, - uint32_t min, uint32_t max) + uint32_t min, uint32_t max, + bool automatic) { struct smu_dpm_context *smu_dpm = &(smu->smu_dpm); struct smu_13_0_dpm_context *dpm_context = smu_dpm->dpm_context; @@ -1806,7 +1826,7 @@ static int smu_v13_0_6_set_soft_freq_limited_range(struct smu_context *smu, return -EOPNOTSUPP; /* Only max clock limiting is allowed for UCLK */ ret = smu_v13_0_set_soft_freq_limited_range( - smu, SMU_UCLK, 0, max); + smu, SMU_UCLK, 0, max, false); if (!ret) pstate_table->uclk_pstate.curr.max = max; } @@ -1946,7 +1966,7 @@ static int smu_v13_0_6_usr_edit_dpm_table(struct smu_context *smu, max_clk = dpm_context->dpm_tables.gfx_table.max; ret = smu_v13_0_6_set_soft_freq_limited_range( - smu, SMU_GFXCLK, min_clk, max_clk); + smu, SMU_GFXCLK, min_clk, max_clk, false); if (ret) return ret; @@ -1954,7 +1974,7 @@ static int smu_v13_0_6_usr_edit_dpm_table(struct smu_context *smu, min_clk = dpm_context->dpm_tables.uclk_table.min; max_clk = dpm_context->dpm_tables.uclk_table.max; ret = smu_v13_0_6_set_soft_freq_limited_range( - smu, SMU_UCLK, min_clk, max_clk); + smu, SMU_UCLK, min_clk, max_clk, false); if (ret) return ret; pstate_table->uclk_pstate.custom.max = 0; @@ -1978,7 +1998,7 @@ static int smu_v13_0_6_usr_edit_dpm_table(struct smu_context *smu, max_clk = pstate_table->gfxclk_pstate.custom.max; ret = smu_v13_0_6_set_soft_freq_limited_range( - smu, SMU_GFXCLK, min_clk, max_clk); + smu, SMU_GFXCLK, min_clk, max_clk, false); if (ret) return ret; @@ -1989,7 +2009,7 @@ static int smu_v13_0_6_usr_edit_dpm_table(struct smu_context *smu, min_clk = pstate_table->uclk_pstate.curr.min; max_clk = pstate_table->uclk_pstate.custom.max; return smu_v13_0_6_set_soft_freq_limited_range( - smu, SMU_UCLK, min_clk, max_clk); + smu, SMU_UCLK, min_clk, max_clk, false); } break; default: @@ -2299,14 +2319,18 @@ static int smu_v13_0_6_get_current_pcie_link_speed(struct smu_context *smu) static ssize_t smu_v13_0_6_get_gpu_metrics(struct smu_context *smu, void **table) { + bool per_inst, smu_13_0_6_per_inst, smu_13_0_14_per_inst, apu_per_inst; struct smu_table_context *smu_table = &smu->smu_table; - struct gpu_metrics_v1_5 *gpu_metrics = - (struct gpu_metrics_v1_5 *)smu_table->gpu_metrics_table; + struct gpu_metrics_v1_6 *gpu_metrics = + (struct gpu_metrics_v1_6 *)smu_table->gpu_metrics_table; + bool flag = smu_v13_0_6_is_unified_metrics(smu); + int ret = 0, xcc_id, inst, i, j, k, idx; struct amdgpu_device *adev = smu->adev; - int ret = 0, xcc_id, inst, i, j; MetricsTableX_t *metrics_x; MetricsTableA_t *metrics_a; + struct amdgpu_xcp *xcp; u16 link_width_level; + u32 inst_mask; metrics_x = kzalloc(max(sizeof(MetricsTableX_t), sizeof(MetricsTableA_t)), GFP_KERNEL); ret = smu_v13_0_6_get_metrics_table(smu, metrics_x, true); @@ -2317,53 +2341,60 @@ static ssize_t smu_v13_0_6_get_gpu_metrics(struct smu_context *smu, void **table metrics_a = (MetricsTableA_t *)metrics_x; - smu_cmn_init_soft_gpu_metrics(gpu_metrics, 1, 5); + smu_cmn_init_soft_gpu_metrics(gpu_metrics, 1, 6); gpu_metrics->temperature_hotspot = - SMUQ10_ROUND(GET_METRIC_FIELD(MaxSocketTemperature)); + SMUQ10_ROUND(GET_METRIC_FIELD(MaxSocketTemperature, flag)); /* Individual HBM stack temperature is not reported */ gpu_metrics->temperature_mem = - SMUQ10_ROUND(GET_METRIC_FIELD(MaxHbmTemperature)); + SMUQ10_ROUND(GET_METRIC_FIELD(MaxHbmTemperature, flag)); /* Reports max temperature of all voltage rails */ gpu_metrics->temperature_vrsoc = - SMUQ10_ROUND(GET_METRIC_FIELD(MaxVrTemperature)); + SMUQ10_ROUND(GET_METRIC_FIELD(MaxVrTemperature, flag)); gpu_metrics->average_gfx_activity = - SMUQ10_ROUND(GET_METRIC_FIELD(SocketGfxBusy)); + SMUQ10_ROUND(GET_METRIC_FIELD(SocketGfxBusy, flag)); gpu_metrics->average_umc_activity = - SMUQ10_ROUND(GET_METRIC_FIELD(DramBandwidthUtilization)); + SMUQ10_ROUND(GET_METRIC_FIELD(DramBandwidthUtilization, flag)); gpu_metrics->curr_socket_power = - SMUQ10_ROUND(GET_METRIC_FIELD(SocketPower)); + SMUQ10_ROUND(GET_METRIC_FIELD(SocketPower, flag)); /* Energy counter reported in 15.259uJ (2^-16) units */ - gpu_metrics->energy_accumulator = GET_METRIC_FIELD(SocketEnergyAcc); + gpu_metrics->energy_accumulator = GET_METRIC_FIELD(SocketEnergyAcc, flag); for (i = 0; i < MAX_GFX_CLKS; i++) { xcc_id = GET_INST(GC, i); if (xcc_id >= 0) gpu_metrics->current_gfxclk[i] = - SMUQ10_ROUND(GET_METRIC_FIELD(GfxclkFrequency)[xcc_id]); + SMUQ10_ROUND(GET_METRIC_FIELD(GfxclkFrequency, flag)[xcc_id]); if (i < MAX_CLKS) { gpu_metrics->current_socclk[i] = - SMUQ10_ROUND(GET_METRIC_FIELD(SocclkFrequency)[i]); + SMUQ10_ROUND(GET_METRIC_FIELD(SocclkFrequency, flag)[i]); inst = GET_INST(VCN, i); if (inst >= 0) { gpu_metrics->current_vclk0[i] = - SMUQ10_ROUND(GET_METRIC_FIELD(VclkFrequency)[inst]); + SMUQ10_ROUND(GET_METRIC_FIELD(VclkFrequency, flag)[inst]); gpu_metrics->current_dclk0[i] = - SMUQ10_ROUND(GET_METRIC_FIELD(DclkFrequency)[inst]); + SMUQ10_ROUND(GET_METRIC_FIELD(DclkFrequency, flag)[inst]); } } } - gpu_metrics->current_uclk = SMUQ10_ROUND(GET_METRIC_FIELD(UclkFrequency)); + gpu_metrics->current_uclk = SMUQ10_ROUND(GET_METRIC_FIELD(UclkFrequency, flag)); - /* Throttle status is not reported through metrics now */ - gpu_metrics->throttle_status = 0; + /* Total accumulated cycle counter */ + gpu_metrics->accumulation_counter = GET_METRIC_FIELD(AccumulationCounter, flag); + + /* Accumulated throttler residencies */ + gpu_metrics->prochot_residency_acc = GET_METRIC_FIELD(ProchotResidencyAcc, flag); + gpu_metrics->ppt_residency_acc = GET_METRIC_FIELD(PptResidencyAcc, flag); + gpu_metrics->socket_thm_residency_acc = GET_METRIC_FIELD(SocketThmResidencyAcc, flag); + gpu_metrics->vr_thm_residency_acc = GET_METRIC_FIELD(VrThmResidencyAcc, flag); + gpu_metrics->hbm_thm_residency_acc = GET_METRIC_FIELD(HbmThmResidencyAcc, flag); /* Clock Lock Status. Each bit corresponds to each GFXCLK instance */ - gpu_metrics->gfxclk_lock_status = GET_METRIC_FIELD(GfxLockXCDMak) >> GET_INST(GC, 0); + gpu_metrics->gfxclk_lock_status = GET_METRIC_FIELD(GfxLockXCDMak, flag) >> GET_INST(GC, 0); if (!(adev->flags & AMD_IS_APU)) { /*Check smu version, PCIE link speed and width will be reported from pmfw metric @@ -2399,41 +2430,77 @@ static ssize_t smu_v13_0_6_get_gpu_metrics(struct smu_context *smu, void **table metrics_x->PCIeNAKSentCountAcc; gpu_metrics->pcie_nak_rcvd_count_acc = metrics_x->PCIeNAKReceivedCountAcc; + if (smu_v13_0_6_is_other_end_count_available(smu)) + gpu_metrics->pcie_lc_perf_other_end_recovery = + metrics_x->PCIeOtherEndRecoveryAcc; + } gpu_metrics->system_clock_counter = ktime_get_boottime_ns(); gpu_metrics->gfx_activity_acc = - SMUQ10_ROUND(GET_METRIC_FIELD(SocketGfxBusyAcc)); + SMUQ10_ROUND(GET_METRIC_FIELD(SocketGfxBusyAcc, flag)); gpu_metrics->mem_activity_acc = - SMUQ10_ROUND(GET_METRIC_FIELD(DramBandwidthUtilizationAcc)); + SMUQ10_ROUND(GET_METRIC_FIELD(DramBandwidthUtilizationAcc, flag)); for (i = 0; i < NUM_XGMI_LINKS; i++) { gpu_metrics->xgmi_read_data_acc[i] = - SMUQ10_ROUND(GET_METRIC_FIELD(XgmiReadDataSizeAcc)[i]); + SMUQ10_ROUND(GET_METRIC_FIELD(XgmiReadDataSizeAcc, flag)[i]); gpu_metrics->xgmi_write_data_acc[i] = - SMUQ10_ROUND(GET_METRIC_FIELD(XgmiWriteDataSizeAcc)[i]); + SMUQ10_ROUND(GET_METRIC_FIELD(XgmiWriteDataSizeAcc, flag)[i]); } - for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) { - inst = GET_INST(JPEG, i); - for (j = 0; j < adev->jpeg.num_jpeg_rings; ++j) { - gpu_metrics->jpeg_activity[(i * adev->jpeg.num_jpeg_rings) + j] = - SMUQ10_ROUND(GET_METRIC_FIELD(JpegBusy) - [(inst * adev->jpeg.num_jpeg_rings) + j]); + gpu_metrics->num_partition = adev->xcp_mgr->num_xcps; + + apu_per_inst = (adev->flags & AMD_IS_APU) && (smu->smc_fw_version >= 0x04556A00); + smu_13_0_6_per_inst = !(adev->flags & AMD_IS_APU) && + (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) + == IP_VERSION(13, 0, 6)) && + (smu->smc_fw_version >= 0x556F00); + smu_13_0_14_per_inst = !(adev->flags & AMD_IS_APU) && + (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) + == IP_VERSION(13, 0, 14)) && + (smu->smc_fw_version >= 0x05550B00); + + per_inst = apu_per_inst || smu_13_0_6_per_inst || smu_13_0_14_per_inst; + + for_each_xcp(adev->xcp_mgr, xcp, i) { + amdgpu_xcp_get_inst_details(xcp, AMDGPU_XCP_VCN, &inst_mask); + idx = 0; + for_each_inst(k, inst_mask) { + /* Both JPEG and VCN has same instances */ + inst = GET_INST(VCN, k); + + for (j = 0; j < adev->jpeg.num_jpeg_rings; ++j) { + gpu_metrics->xcp_stats[i].jpeg_busy + [(idx * adev->jpeg.num_jpeg_rings) + j] = + SMUQ10_ROUND(GET_METRIC_FIELD(JpegBusy, flag) + [(inst * adev->jpeg.num_jpeg_rings) + j]); + } + gpu_metrics->xcp_stats[i].vcn_busy[idx] = + SMUQ10_ROUND(GET_METRIC_FIELD(VcnBusy, flag)[inst]); + idx++; + + } + + if (per_inst) { + amdgpu_xcp_get_inst_details(xcp, AMDGPU_XCP_GFX, &inst_mask); + idx = 0; + for_each_inst(k, inst_mask) { + inst = GET_INST(GC, k); + gpu_metrics->xcp_stats[i].gfx_busy_inst[idx] = + SMUQ10_ROUND(metrics_x->GfxBusy[inst]); + gpu_metrics->xcp_stats[i].gfx_busy_acc[idx] = + SMUQ10_ROUND(metrics_x->GfxBusyAcc[inst]); + idx++; + } } } - for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { - inst = GET_INST(VCN, i); - gpu_metrics->vcn_activity[i] = - SMUQ10_ROUND(GET_METRIC_FIELD(VcnBusy)[inst]); - } + gpu_metrics->xgmi_link_width = SMUQ10_ROUND(GET_METRIC_FIELD(XgmiWidth, flag)); + gpu_metrics->xgmi_link_speed = SMUQ10_ROUND(GET_METRIC_FIELD(XgmiBitrate, flag)); - gpu_metrics->xgmi_link_width = SMUQ10_ROUND(GET_METRIC_FIELD(XgmiWidth)); - gpu_metrics->xgmi_link_speed = SMUQ10_ROUND(GET_METRIC_FIELD(XgmiBitrate)); - - gpu_metrics->firmware_timestamp = GET_METRIC_FIELD(Timestamp); + gpu_metrics->firmware_timestamp = GET_METRIC_FIELD(Timestamp, flag); *table = (void *)gpu_metrics; kfree(metrics_x); @@ -2974,6 +3041,16 @@ static int mmhub_err_codes[] = { CODE_VML2, CODE_VML2_WALKER, CODE_MMCANE, }; +static int vcn_err_codes[] = { + CODE_VIDD, CODE_VIDV, +}; +static int jpeg_err_codes[] = { + CODE_JPEG0S, CODE_JPEG0D, CODE_JPEG1S, CODE_JPEG1D, + CODE_JPEG2S, CODE_JPEG2D, CODE_JPEG3S, CODE_JPEG3D, + CODE_JPEG4S, CODE_JPEG4D, CODE_JPEG5S, CODE_JPEG5D, + CODE_JPEG6S, CODE_JPEG6D, CODE_JPEG7S, CODE_JPEG7D, +}; + static const struct mca_ras_info mca_ras_table[] = { { .blkid = AMDGPU_RAS_BLOCK__UMC, @@ -3002,6 +3079,20 @@ static const struct mca_ras_info mca_ras_table[] = { .blkid = AMDGPU_RAS_BLOCK__XGMI_WAFL, .ip = AMDGPU_MCA_IP_PCS_XGMI, .get_err_count = mca_pcs_xgmi_mca_get_err_count, + }, { + .blkid = AMDGPU_RAS_BLOCK__VCN, + .ip = AMDGPU_MCA_IP_SMU, + .err_code_array = vcn_err_codes, + .err_code_count = ARRAY_SIZE(vcn_err_codes), + .get_err_count = mca_smu_mca_get_err_count, + .bank_is_valid = mca_smu_bank_is_valid, + }, { + .blkid = AMDGPU_RAS_BLOCK__JPEG, + .ip = AMDGPU_MCA_IP_SMU, + .err_code_array = jpeg_err_codes, + .err_code_count = ARRAY_SIZE(jpeg_err_codes), + .get_err_count = mca_smu_mca_get_err_count, + .bank_is_valid = mca_smu_bank_is_valid, }, }; diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c index b891a5e0a396..c5d3e25cc967 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c @@ -83,6 +83,8 @@ #define PP_OD_FEATURE_FAN_ACOUSTIC_TARGET 8 #define PP_OD_FEATURE_FAN_TARGET_TEMPERATURE 9 #define PP_OD_FEATURE_FAN_MINIMUM_PWM 10 +#define PP_OD_FEATURE_FAN_ZERO_RPM_ENABLE 11 +#define PP_OD_FEATURE_FAN_ZERO_RPM_STOP_TEMP 12 #define LINK_SPEED_MAX 3 @@ -734,19 +736,6 @@ static bool smu_v13_0_7_is_dpm_running(struct smu_context *smu) return !!(feature_enabled & SMC_DPM_FEATURE); } -static void smu_v13_0_7_dump_pptable(struct smu_context *smu) -{ - struct smu_table_context *table_context = &smu->smu_table; - PPTable_t *pptable = table_context->driver_pptable; - SkuTable_t *skutable = &pptable->SkuTable; - - dev_info(smu->adev->dev, "Dumped PPTable:\n"); - - dev_info(smu->adev->dev, "Version = 0x%08x\n", skutable->Version); - dev_info(smu->adev->dev, "FeaturesToRun[0] = 0x%08x\n", skutable->FeaturesToRun[0]); - dev_info(smu->adev->dev, "FeaturesToRun[1] = 0x%08x\n", skutable->FeaturesToRun[1]); -} - static uint32_t smu_v13_0_7_get_throttler_status(SmuMetrics_t *metrics) { uint32_t throttler_status = 0; @@ -1132,6 +1121,14 @@ static void smu_v13_0_7_get_od_setting_limits(struct smu_context *smu, od_min_setting = overdrive_lowerlimits->FanMinimumPwm; od_max_setting = overdrive_upperlimits->FanMinimumPwm; break; + case PP_OD_FEATURE_FAN_ZERO_RPM_ENABLE: + od_min_setting = overdrive_lowerlimits->FanZeroRpmEnable; + od_max_setting = overdrive_upperlimits->FanZeroRpmEnable; + break; + case PP_OD_FEATURE_FAN_ZERO_RPM_STOP_TEMP: + od_min_setting = overdrive_lowerlimits->FanZeroRpmStopTemp; + od_max_setting = overdrive_upperlimits->FanZeroRpmStopTemp; + break; default: od_min_setting = od_max_setting = INT_MAX; break; @@ -1452,6 +1449,42 @@ static int smu_v13_0_7_print_clk_levels(struct smu_context *smu, min_value, max_value); break; + case SMU_OD_FAN_ZERO_RPM_ENABLE: + if (!smu_v13_0_7_is_od_feature_supported(smu, + PP_OD_FEATURE_ZERO_FAN_BIT)) + break; + + size += sysfs_emit_at(buf, size, "FAN_ZERO_RPM_ENABLE:\n"); + size += sysfs_emit_at(buf, size, "%d\n", + (int)od_table->OverDriveTable.FanZeroRpmEnable); + + size += sysfs_emit_at(buf, size, "%s:\n", "OD_RANGE"); + smu_v13_0_7_get_od_setting_limits(smu, + PP_OD_FEATURE_FAN_ZERO_RPM_ENABLE, + &min_value, + &max_value); + size += sysfs_emit_at(buf, size, "ZERO_RPM_ENABLE: %u %u\n", + min_value, max_value); + break; + + case SMU_OD_FAN_ZERO_RPM_STOP_TEMP: + if (!smu_v13_0_7_is_od_feature_supported(smu, + PP_OD_FEATURE_ZERO_FAN_BIT)) + break; + + size += sysfs_emit_at(buf, size, "FAN_ZERO_RPM_STOP_TEMPERATURE:\n"); + size += sysfs_emit_at(buf, size, "%d\n", + (int)od_table->OverDriveTable.FanZeroRpmStopTemp); + + size += sysfs_emit_at(buf, size, "%s:\n", "OD_RANGE"); + smu_v13_0_7_get_od_setting_limits(smu, + PP_OD_FEATURE_FAN_ZERO_RPM_STOP_TEMP, + &min_value, + &max_value); + size += sysfs_emit_at(buf, size, "ZERO_RPM_STOP_TEMPERATURE: %u %u\n", + min_value, max_value); + break; + case SMU_OD_RANGE: if (!smu_v13_0_7_is_od_feature_supported(smu, PP_OD_FEATURE_GFXCLK_BIT) && !smu_v13_0_7_is_od_feature_supported(smu, PP_OD_FEATURE_UCLK_BIT) && @@ -1548,6 +1581,16 @@ static int smu_v13_0_7_od_restore_table_single(struct smu_context *smu, long inp od_table->OverDriveTable.FanMode = FAN_MODE_AUTO; od_table->OverDriveTable.FeatureCtrlMask |= BIT(PP_OD_FEATURE_FAN_CURVE_BIT); break; + case PP_OD_EDIT_FAN_ZERO_RPM_ENABLE: + od_table->OverDriveTable.FanZeroRpmEnable = + boot_overdrive_table->OverDriveTable.FanZeroRpmEnable; + od_table->OverDriveTable.FeatureCtrlMask |= BIT(PP_OD_FEATURE_ZERO_FAN_BIT); + break; + case PP_OD_EDIT_FAN_ZERO_RPM_STOP_TEMP: + od_table->OverDriveTable.FanZeroRpmStopTemp = + boot_overdrive_table->OverDriveTable.FanZeroRpmStopTemp; + od_table->OverDriveTable.FeatureCtrlMask |= BIT(PP_OD_FEATURE_ZERO_FAN_BIT); + break; default: dev_info(adev->dev, "Invalid table index: %ld\n", input); return -EINVAL; @@ -1841,6 +1884,48 @@ static int smu_v13_0_7_od_edit_dpm_table(struct smu_context *smu, od_table->OverDriveTable.FeatureCtrlMask |= BIT(PP_OD_FEATURE_FAN_CURVE_BIT); break; + case PP_OD_EDIT_FAN_ZERO_RPM_ENABLE: + if (!smu_v13_0_7_is_od_feature_supported(smu, PP_OD_FEATURE_ZERO_FAN_BIT)) { + dev_warn(adev->dev, "Zero RPM setting not supported!\n"); + return -ENOTSUPP; + } + + smu_v13_0_7_get_od_setting_limits(smu, + PP_OD_FEATURE_FAN_ZERO_RPM_ENABLE, + &minimum, + &maximum); + if (input[0] < minimum || + input[0] > maximum) { + dev_info(adev->dev, "zero RPM enable setting(%ld) must be within [%d, %d]!\n", + input[0], minimum, maximum); + return -EINVAL; + } + + od_table->OverDriveTable.FanZeroRpmEnable = input[0]; + od_table->OverDriveTable.FeatureCtrlMask |= BIT(PP_OD_FEATURE_ZERO_FAN_BIT); + break; + + case PP_OD_EDIT_FAN_ZERO_RPM_STOP_TEMP: + if (!smu_v13_0_7_is_od_feature_supported(smu, PP_OD_FEATURE_ZERO_FAN_BIT)) { + dev_warn(adev->dev, "Zero RPM setting not supported!\n"); + return -ENOTSUPP; + } + + smu_v13_0_7_get_od_setting_limits(smu, + PP_OD_FEATURE_FAN_ZERO_RPM_STOP_TEMP, + &minimum, + &maximum); + if (input[0] < minimum || + input[0] > maximum) { + dev_info(adev->dev, "zero RPM stop temperature setting(%ld) must be within [%d, %d]!\n", + input[0], minimum, maximum); + return -EINVAL; + } + + od_table->OverDriveTable.FanZeroRpmStopTemp = input[0]; + od_table->OverDriveTable.FeatureCtrlMask |= BIT(PP_OD_FEATURE_ZERO_FAN_BIT); + break; + case PP_OD_RESTORE_DEFAULT_TABLE: if (size == 1) { ret = smu_v13_0_7_od_restore_table_single(smu, input[0]); @@ -1964,7 +2049,8 @@ static int smu_v13_0_7_force_clk_levels(struct smu_context *smu, ret = smu_v13_0_set_soft_freq_limited_range(smu, clk_type, min_freq, - max_freq); + max_freq, + false); break; case SMU_DCEFCLK: case SMU_PCIE: @@ -2106,7 +2192,11 @@ static void smu_v13_0_7_set_supported_od_feature_mask(struct smu_context *smu) OD_OPS_SUPPORT_FAN_TARGET_TEMPERATURE_RETRIEVE | OD_OPS_SUPPORT_FAN_TARGET_TEMPERATURE_SET | OD_OPS_SUPPORT_FAN_MINIMUM_PWM_RETRIEVE | - OD_OPS_SUPPORT_FAN_MINIMUM_PWM_SET; + OD_OPS_SUPPORT_FAN_MINIMUM_PWM_SET | + OD_OPS_SUPPORT_FAN_ZERO_RPM_ENABLE_RETRIEVE | + OD_OPS_SUPPORT_FAN_ZERO_RPM_ENABLE_SET | + OD_OPS_SUPPORT_FAN_ZERO_RPM_STOP_TEMP_RETRIEVE | + OD_OPS_SUPPORT_FAN_ZERO_RPM_STOP_TEMP_SET; } static int smu_v13_0_7_set_default_od_settings(struct smu_context *smu) @@ -2172,6 +2262,10 @@ static int smu_v13_0_7_set_default_od_settings(struct smu_context *smu) user_od_table_bak.OverDriveTable.FanTargetTemperature; user_od_table->OverDriveTable.FanMinimumPwm = user_od_table_bak.OverDriveTable.FanMinimumPwm; + user_od_table->OverDriveTable.FanZeroRpmEnable = + user_od_table_bak.OverDriveTable.FanZeroRpmEnable; + user_od_table->OverDriveTable.FanZeroRpmStopTemp = + user_od_table_bak.OverDriveTable.FanZeroRpmStopTemp; } smu_v13_0_7_set_supported_od_feature_mask(smu); @@ -2499,13 +2593,14 @@ static int smu_v13_0_7_set_power_profile_mode(struct smu_context *smu, long *inp smu->power_profile_mode); if (workload_type < 0) return -EINVAL; + ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask, - 1 << workload_type, NULL); + smu->workload_mask, NULL); if (ret) dev_err(smu->adev->dev, "[%s] Failed to set work load mask!", __func__); else - smu->workload_mask = (1 << workload_type); + smu_cmn_assign_power_profile(smu); return ret; } @@ -2605,7 +2700,6 @@ static const struct pptable_funcs smu_v13_0_7_ppt_funcs = { .get_allowed_feature_mask = smu_v13_0_7_get_allowed_feature_mask, .set_default_dpm_table = smu_v13_0_7_set_default_dpm_table, .is_dpm_running = smu_v13_0_7_is_dpm_running, - .dump_pptable = smu_v13_0_7_dump_pptable, .init_microcode = smu_v13_0_init_microcode, .load_microcode = smu_v13_0_load_microcode, .fini_microcode = smu_v13_0_fini_microcode, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/yellow_carp_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/yellow_carp_ppt.c index 260c339f89c5..73b4506ef5a8 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/yellow_carp_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/yellow_carp_ppt.c @@ -220,7 +220,9 @@ static int yellow_carp_system_features_control(struct smu_context *smu, bool en) return ret; } -static int yellow_carp_dpm_set_vcn_enable(struct smu_context *smu, bool enable) +static int yellow_carp_dpm_set_vcn_enable(struct smu_context *smu, + bool enable, + int inst) { int ret = 0; @@ -945,9 +947,10 @@ failed: } static int yellow_carp_set_soft_freq_limited_range(struct smu_context *smu, - enum smu_clk_type clk_type, - uint32_t min, - uint32_t max) + enum smu_clk_type clk_type, + uint32_t min, + uint32_t max, + bool automatic) { enum smu_message_type msg_set_min, msg_set_max; uint32_t min_clk = min; @@ -1134,7 +1137,7 @@ static int yellow_carp_force_clk_levels(struct smu_context *smu, if (ret) goto force_level_out; - ret = yellow_carp_set_soft_freq_limited_range(smu, clk_type, min_freq, max_freq); + ret = yellow_carp_set_soft_freq_limited_range(smu, clk_type, min_freq, max_freq, false); if (ret) goto force_level_out; break; @@ -1254,9 +1257,10 @@ static int yellow_carp_set_performance_level(struct smu_context *smu, if (sclk_min && sclk_max) { ret = yellow_carp_set_soft_freq_limited_range(smu, - SMU_SCLK, - sclk_min, - sclk_max); + SMU_SCLK, + sclk_min, + sclk_max, + false); if (ret) return ret; @@ -1266,18 +1270,20 @@ static int yellow_carp_set_performance_level(struct smu_context *smu, if (fclk_min && fclk_max) { ret = yellow_carp_set_soft_freq_limited_range(smu, - SMU_FCLK, - fclk_min, - fclk_max); + SMU_FCLK, + fclk_min, + fclk_max, + false); if (ret) return ret; } if (socclk_min && socclk_max) { ret = yellow_carp_set_soft_freq_limited_range(smu, - SMU_SOCCLK, - socclk_min, - socclk_max); + SMU_SOCCLK, + socclk_min, + socclk_max, + false); if (ret) return ret; } @@ -1286,7 +1292,8 @@ static int yellow_carp_set_performance_level(struct smu_context *smu, ret = yellow_carp_set_soft_freq_limited_range(smu, SMU_VCLK, vclk_min, - vclk_max); + vclk_max, + false); if (ret) return ret; } @@ -1295,7 +1302,8 @@ static int yellow_carp_set_performance_level(struct smu_context *smu, ret = yellow_carp_set_soft_freq_limited_range(smu, SMU_DCLK, dclk_min, - dclk_max); + dclk_max, + false); if (ret) return ret; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0.c index 865e916fc425..ecb0164d533e 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0.c @@ -1102,7 +1102,8 @@ failed: int smu_v14_0_set_soft_freq_limited_range(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t min, - uint32_t max) + uint32_t max, + bool automatic) { int ret = 0, clk_id = 0; uint32_t param; @@ -1117,7 +1118,10 @@ int smu_v14_0_set_soft_freq_limited_range(struct smu_context *smu, return clk_id; if (max > 0) { - param = (uint32_t)((clk_id << 16) | (max & 0xffff)); + if (automatic) + param = (uint32_t)((clk_id << 16) | 0xffff); + else + param = (uint32_t)((clk_id << 16) | (max & 0xffff)); ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMaxByFreq, param, NULL); if (ret) @@ -1125,7 +1129,10 @@ int smu_v14_0_set_soft_freq_limited_range(struct smu_context *smu, } if (min > 0) { - param = (uint32_t)((clk_id << 16) | (min & 0xffff)); + if (automatic) + param = (uint32_t)((clk_id << 16) | 0); + else + param = (uint32_t)((clk_id << 16) | (min & 0xffff)); ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetSoftMinByFreq, param, NULL); if (ret) @@ -1202,6 +1209,7 @@ int smu_v14_0_set_performance_level(struct smu_context *smu, uint32_t dclk_min = 0, dclk_max = 0; uint32_t fclk_min = 0, fclk_max = 0; int ret = 0, i; + bool auto_level = false; switch (level) { case AMD_DPM_FORCED_LEVEL_HIGH: @@ -1233,6 +1241,7 @@ int smu_v14_0_set_performance_level(struct smu_context *smu, dclk_max = dclk_table->max; fclk_min = fclk_table->min; fclk_max = fclk_table->max; + auto_level = true; break; case AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD: sclk_min = sclk_max = pstate_table->gfxclk_pstate.standard; @@ -1268,7 +1277,8 @@ int smu_v14_0_set_performance_level(struct smu_context *smu, ret = smu_v14_0_set_soft_freq_limited_range(smu, SMU_GFXCLK, sclk_min, - sclk_max); + sclk_max, + auto_level); if (ret) return ret; @@ -1280,7 +1290,8 @@ int smu_v14_0_set_performance_level(struct smu_context *smu, ret = smu_v14_0_set_soft_freq_limited_range(smu, SMU_MCLK, mclk_min, - mclk_max); + mclk_max, + auto_level); if (ret) return ret; @@ -1292,7 +1303,8 @@ int smu_v14_0_set_performance_level(struct smu_context *smu, ret = smu_v14_0_set_soft_freq_limited_range(smu, SMU_SOCCLK, socclk_min, - socclk_max); + socclk_max, + auto_level); if (ret) return ret; @@ -1307,7 +1319,8 @@ int smu_v14_0_set_performance_level(struct smu_context *smu, ret = smu_v14_0_set_soft_freq_limited_range(smu, i ? SMU_VCLK1 : SMU_VCLK, vclk_min, - vclk_max); + vclk_max, + auto_level); if (ret) return ret; } @@ -1322,7 +1335,8 @@ int smu_v14_0_set_performance_level(struct smu_context *smu, ret = smu_v14_0_set_soft_freq_limited_range(smu, i ? SMU_DCLK1 : SMU_DCLK, dclk_min, - dclk_max); + dclk_max, + auto_level); if (ret) return ret; } @@ -1334,7 +1348,8 @@ int smu_v14_0_set_performance_level(struct smu_context *smu, ret = smu_v14_0_set_soft_freq_limited_range(smu, SMU_FCLK, fclk_min, - fclk_max); + fclk_max, + auto_level); if (ret) return ret; @@ -1492,7 +1507,8 @@ int smu_v14_0_set_single_dpm_table(struct smu_context *smu, } int smu_v14_0_set_vcn_enable(struct smu_context *smu, - bool enable) + bool enable, + int inst) { struct amdgpu_device *adev = smu->adev; int i, ret = 0; diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c index 1e16a281f2dc..59b369eff30f 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c @@ -670,19 +670,6 @@ static bool smu_v14_0_2_is_dpm_running(struct smu_context *smu) return !!(feature_enabled & SMC_DPM_FEATURE); } -static void smu_v14_0_2_dump_pptable(struct smu_context *smu) -{ - struct smu_table_context *table_context = &smu->smu_table; - PPTable_t *pptable = table_context->driver_pptable; - PFE_Settings_t *PFEsettings = &pptable->PFE_Settings; - - dev_info(smu->adev->dev, "Dumped PPTable:\n"); - - dev_info(smu->adev->dev, "Version = 0x%08x\n", PFEsettings->Version); - dev_info(smu->adev->dev, "FeaturesToRun[0] = 0x%08x\n", PFEsettings->FeaturesToRun[0]); - dev_info(smu->adev->dev, "FeaturesToRun[1] = 0x%08x\n", PFEsettings->FeaturesToRun[1]); -} - static uint32_t smu_v14_0_2_get_throttler_status(SmuMetrics_t *metrics) { uint32_t throttler_status = 0; @@ -1457,7 +1444,8 @@ static int smu_v14_0_2_force_clk_levels(struct smu_context *smu, ret = smu_v14_0_set_soft_freq_limited_range(smu, clk_type, min_freq, - max_freq); + max_freq, + false); break; case SMU_DCEFCLK: case SMU_PCIE: @@ -1807,12 +1795,11 @@ static int smu_v14_0_2_set_power_profile_mode(struct smu_context *smu, if (workload_type < 0) return -EINVAL; - ret = smu_cmn_send_smc_msg_with_param(smu, - SMU_MSG_SetWorkloadMask, - 1 << workload_type, - NULL); + ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask, + smu->workload_mask, NULL); + if (!ret) - smu->workload_mask = 1 << workload_type; + smu_cmn_assign_power_profile(smu); return ret; } @@ -2726,7 +2713,6 @@ static const struct pptable_funcs smu_v14_0_2_ppt_funcs = { .i2c_init = smu_v14_0_2_i2c_control_init, .i2c_fini = smu_v14_0_2_i2c_control_fini, .is_dpm_running = smu_v14_0_2_is_dpm_running, - .dump_pptable = smu_v14_0_2_dump_pptable, .init_microcode = smu_v14_0_init_microcode, .load_microcode = smu_v14_0_load_microcode, .fini_microcode = smu_v14_0_fini_microcode, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c index 91ad434bcdae..f1ab1a6bb467 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c @@ -1078,6 +1078,9 @@ void smu_cmn_init_soft_gpu_metrics(void *table, uint8_t frev, uint8_t crev) case METRICS_VERSION(1, 5): structure_size = sizeof(struct gpu_metrics_v1_5); break; + case METRICS_VERSION(1, 6): + structure_size = sizeof(struct gpu_metrics_v1_6); + break; case METRICS_VERSION(2, 0): structure_size = sizeof(struct gpu_metrics_v2_0); break; @@ -1138,6 +1141,14 @@ int smu_cmn_set_mp1_state(struct smu_context *smu, return ret; } +void smu_cmn_assign_power_profile(struct smu_context *smu) +{ + uint32_t index; + index = fls(smu->workload_mask); + index = index > 0 && index <= WORKLOAD_POLICY_MAX ? index - 1 : 0; + smu->power_profile_mode = smu->workload_setting[index]; +} + bool smu_cmn_is_audio_func_enabled(struct amdgpu_device *adev) { struct pci_dev *p = NULL; diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h index 1de685defe85..8a801e389659 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h +++ b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h @@ -130,6 +130,8 @@ void smu_cmn_init_soft_gpu_metrics(void *table, uint8_t frev, uint8_t crev); int smu_cmn_set_mp1_state(struct smu_context *smu, enum pp_mp1_state mp1_state); +void smu_cmn_assign_power_profile(struct smu_context *smu); + /* * Helper function to make sysfs_emit_at() happy. Align buf to * the current page boundary and record the offset. diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_internal.h b/drivers/gpu/drm/amd/pm/swsmu/smu_internal.h index 6f4d212607d7..c09ecf1a68a0 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu_internal.h +++ b/drivers/gpu/drm/amd/pm/swsmu/smu_internal.h @@ -78,7 +78,6 @@ #define smu_register_irq_handler(smu) smu_ppt_funcs(register_irq_handler, 0, smu) #define smu_get_dpm_ultimate_freq(smu, param, min, max) smu_ppt_funcs(get_dpm_ultimate_freq, 0, smu, param, min, max) #define smu_asic_set_performance_level(smu, level) smu_ppt_funcs(set_performance_level, -EINVAL, smu, level) -#define smu_dump_pptable(smu) smu_ppt_funcs(dump_pptable, 0, smu) #define smu_update_pcie_parameters(smu, pcie_gen_cap, pcie_width_cap) smu_ppt_funcs(update_pcie_parameters, 0, smu, pcie_gen_cap, pcie_width_cap) #define smu_set_power_source(smu, power_src) smu_ppt_funcs(set_power_source, 0, smu, power_src) #define smu_i2c_init(smu) smu_ppt_funcs(i2c_init, 0, smu) diff --git a/drivers/gpu/drm/arm/Kconfig b/drivers/gpu/drm/arm/Kconfig index ddf20708370f..c901ac00c0c3 100644 --- a/drivers/gpu/drm/arm/Kconfig +++ b/drivers/gpu/drm/arm/Kconfig @@ -6,6 +6,7 @@ config DRM_HDLCD tristate "ARM HDLCD" depends on DRM && OF && (ARM || ARM64 || COMPILE_TEST) depends on COMMON_CLK + select DRM_CLIENT_SELECTION select DRM_KMS_HELPER select DRM_GEM_DMA_HELPER help @@ -27,6 +28,7 @@ config DRM_MALI_DISPLAY tristate "ARM Mali Display Processor" depends on DRM && OF && (ARM || ARM64 || COMPILE_TEST) depends on COMMON_CLK + select DRM_CLIENT_SELECTION select DRM_KMS_HELPER select DRM_GEM_DMA_HELPER select VIDEOMODE_HELPERS diff --git a/drivers/gpu/drm/arm/display/Kconfig b/drivers/gpu/drm/arm/display/Kconfig index 4acc4285a4eb..415c10a6374b 100644 --- a/drivers/gpu/drm/arm/display/Kconfig +++ b/drivers/gpu/drm/arm/display/Kconfig @@ -3,6 +3,7 @@ config DRM_KOMEDA tristate "ARM Komeda display driver" depends on DRM && OF depends on COMMON_CLK + select DRM_CLIENT_SELECTION select DRM_KMS_HELPER select DRM_GEM_DMA_HELPER select VIDEOMODE_HELPERS diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_drv.c b/drivers/gpu/drm/arm/display/komeda/komeda_drv.c index 55c3773befde..6d475bb34002 100644 --- a/drivers/gpu/drm/arm/display/komeda/komeda_drv.c +++ b/drivers/gpu/drm/arm/display/komeda/komeda_drv.c @@ -9,7 +9,7 @@ #include #include #include -#include +#include #include #include #include "komeda_dev.h" @@ -84,7 +84,7 @@ static int komeda_platform_probe(struct platform_device *pdev) } dev_set_drvdata(dev, mdrv); - drm_fbdev_dma_setup(&mdrv->kms->base, 32); + drm_client_setup(&mdrv->kms->base, NULL); return 0; diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_kms.c b/drivers/gpu/drm/arm/display/komeda/komeda_kms.c index e5eb5d672bcd..1e7b1fcb2848 100644 --- a/drivers/gpu/drm/arm/display/komeda/komeda_kms.c +++ b/drivers/gpu/drm/arm/display/komeda/komeda_kms.c @@ -9,6 +9,7 @@ #include #include #include +#include #include #include #include @@ -58,6 +59,7 @@ static irqreturn_t komeda_kms_irq_handler(int irq, void *data) static const struct drm_driver komeda_kms_driver = { .driver_features = DRIVER_GEM | DRIVER_MODESET | DRIVER_ATOMIC, DRM_GEM_DMA_DRIVER_OPS_WITH_DUMB_CREATE(komeda_gem_dma_dumb_create), + DRM_FBDEV_DMA_DRIVER_OPS, .fops = &komeda_cma_fops, .name = "komeda", .desc = "Arm Komeda Display Processor driver", diff --git a/drivers/gpu/drm/arm/hdlcd_drv.c b/drivers/gpu/drm/arm/hdlcd_drv.c index 32be9e370049..cd4389809d42 100644 --- a/drivers/gpu/drm/arm/hdlcd_drv.c +++ b/drivers/gpu/drm/arm/hdlcd_drv.c @@ -9,6 +9,7 @@ * ARM HDLCD Driver */ +#include #include #include #include @@ -21,8 +22,8 @@ #include #include -#include #include +#include #include #include #include @@ -228,6 +229,7 @@ DEFINE_DRM_GEM_DMA_FOPS(fops); static const struct drm_driver hdlcd_driver = { .driver_features = DRIVER_GEM | DRIVER_MODESET | DRIVER_ATOMIC, DRM_GEM_DMA_DRIVER_OPS, + DRM_FBDEV_DMA_DRIVER_OPS, .fops = &fops, .name = "hdlcd", .desc = "ARM HDLCD Controller DRM", @@ -285,7 +287,7 @@ static int hdlcd_drm_bind(struct device *dev) */ if (hdlcd_read(hdlcd, HDLCD_REG_COMMAND)) { hdlcd_write(hdlcd, HDLCD_REG_COMMAND, 0); - drm_aperture_remove_framebuffers(&hdlcd_driver); + aperture_remove_all_conflicting_devices(hdlcd_driver.name); } drm_mode_config_reset(drm); @@ -299,7 +301,7 @@ static int hdlcd_drm_bind(struct device *dev) if (ret) goto err_register; - drm_fbdev_dma_setup(drm, 32); + drm_client_setup(drm, NULL); return 0; diff --git a/drivers/gpu/drm/arm/malidp_drv.c b/drivers/gpu/drm/arm/malidp_drv.c index 6682131d2910..4cb25004b84f 100644 --- a/drivers/gpu/drm/arm/malidp_drv.c +++ b/drivers/gpu/drm/arm/malidp_drv.c @@ -18,6 +18,7 @@ #include #include +#include #include #include #include @@ -562,6 +563,7 @@ static void malidp_debugfs_init(struct drm_minor *minor) static const struct drm_driver malidp_driver = { .driver_features = DRIVER_GEM | DRIVER_MODESET | DRIVER_ATOMIC, DRM_GEM_DMA_DRIVER_OPS_WITH_DUMB_CREATE(malidp_dumb_create), + DRM_FBDEV_DMA_DRIVER_OPS, #ifdef CONFIG_DEBUG_FS .debugfs_init = malidp_debugfs_init, #endif @@ -852,7 +854,7 @@ static int malidp_bind(struct device *dev) if (ret) goto register_fail; - drm_fbdev_dma_setup(drm, 32); + drm_client_setup(drm, NULL); return 0; diff --git a/drivers/gpu/drm/armada/Kconfig b/drivers/gpu/drm/armada/Kconfig index e5597d7c9ae1..b22c891a670b 100644 --- a/drivers/gpu/drm/armada/Kconfig +++ b/drivers/gpu/drm/armada/Kconfig @@ -2,6 +2,7 @@ config DRM_ARMADA tristate "DRM support for Marvell Armada SoCs" depends on DRM && HAVE_CLK && ARM && MMU + select DRM_CLIENT_SELECTION select DRM_KMS_HELPER select FB_IOMEM_HELPERS if DRM_FBDEV_EMULATION help diff --git a/drivers/gpu/drm/armada/armada_drm.h b/drivers/gpu/drm/armada/armada_drm.h index c303e8c7ff6c..3c0ff221a43b 100644 --- a/drivers/gpu/drm/armada/armada_drm.h +++ b/drivers/gpu/drm/armada/armada_drm.h @@ -16,6 +16,8 @@ struct armada_crtc; struct armada_gem_object; struct clk; struct drm_display_mode; +struct drm_fb_helper; +struct drm_fb_helper_surface_size; static inline void armada_updatel(uint32_t val, uint32_t mask, void __iomem *ptr) @@ -74,10 +76,13 @@ struct armada_private { #define drm_to_armada_dev(dev) container_of(dev, struct armada_private, drm) #if defined(CONFIG_DRM_FBDEV_EMULATION) -void armada_fbdev_setup(struct drm_device *dev); +int armada_fbdev_driver_fbdev_probe(struct drm_fb_helper *fbh, + struct drm_fb_helper_surface_size *sizes); +#define ARMADA_FBDEV_DRIVER_OPS \ + .fbdev_probe = armada_fbdev_driver_fbdev_probe #else -static inline void armada_fbdev_setup(struct drm_device *dev) -{ } +#define ARMADA_FBDEV_DRIVER_OPS \ + .fbdev_probe = NULL #endif int armada_overlay_plane_create(struct drm_device *, unsigned long); diff --git a/drivers/gpu/drm/armada/armada_drv.c b/drivers/gpu/drm/armada/armada_drv.c index e51ecc4f7ef4..5c26f0409478 100644 --- a/drivers/gpu/drm/armada/armada_drv.c +++ b/drivers/gpu/drm/armada/armada_drv.c @@ -3,6 +3,7 @@ * Copyright (C) 2012 Russell King */ +#include #include #include #include @@ -10,8 +11,8 @@ #include #include -#include #include +#include #include #include #include @@ -39,6 +40,7 @@ DEFINE_DRM_GEM_FOPS(armada_drm_fops); static const struct drm_driver armada_drm_driver = { .gem_prime_import = armada_gem_prime_import, .dumb_create = armada_gem_dumb_create, + ARMADA_FBDEV_DRIVER_OPS, .major = 1, .minor = 0, .name = "armada-drm", @@ -91,7 +93,7 @@ static int armada_drm_bind(struct device *dev) } /* Remove early framebuffers */ - ret = drm_aperture_remove_framebuffers(&armada_drm_driver); + ret = aperture_remove_all_conflicting_devices(armada_drm_driver.name); if (ret) { dev_err(dev, "[" DRM_NAME ":%s] can't kick out simple-fb: %d\n", __func__, ret); @@ -137,7 +139,7 @@ static int armada_drm_bind(struct device *dev) armada_drm_debugfs_init(priv->drm.primary); #endif - armada_fbdev_setup(&priv->drm); + drm_client_setup(&priv->drm, NULL); return 0; diff --git a/drivers/gpu/drm/armada/armada_fbdev.c b/drivers/gpu/drm/armada/armada_fbdev.c index d223176912b6..6ee7ce04ee71 100644 --- a/drivers/gpu/drm/armada/armada_fbdev.c +++ b/drivers/gpu/drm/armada/armada_fbdev.c @@ -39,8 +39,10 @@ static const struct fb_ops armada_fb_ops = { .fb_destroy = armada_fbdev_fb_destroy, }; -static int armada_fbdev_create(struct drm_fb_helper *fbh, - struct drm_fb_helper_surface_size *sizes) +static const struct drm_fb_helper_funcs armada_fbdev_helper_funcs; + +int armada_fbdev_driver_fbdev_probe(struct drm_fb_helper *fbh, + struct drm_fb_helper_surface_size *sizes) { struct drm_device *dev = fbh->dev; struct drm_mode_fb_cmd2 mode; @@ -98,6 +100,7 @@ static int armada_fbdev_create(struct drm_fb_helper *fbh, info->fix.smem_len = obj->obj.size; info->screen_size = obj->obj.size; info->screen_base = ptr; + fbh->funcs = &armada_fbdev_helper_funcs; fbh->fb = &dfb->fb; drm_fb_helper_fill_info(info, fbh, sizes); @@ -112,109 +115,3 @@ static int armada_fbdev_create(struct drm_fb_helper *fbh, dfb->fb.funcs->destroy(&dfb->fb); return ret; } - -static int armada_fb_probe(struct drm_fb_helper *fbh, - struct drm_fb_helper_surface_size *sizes) -{ - int ret = 0; - - if (!fbh->fb) { - ret = armada_fbdev_create(fbh, sizes); - if (ret == 0) - ret = 1; - } - return ret; -} - -static const struct drm_fb_helper_funcs armada_fb_helper_funcs = { - .fb_probe = armada_fb_probe, -}; - -/* - * Fbdev client and struct drm_client_funcs - */ - -static void armada_fbdev_client_unregister(struct drm_client_dev *client) -{ - struct drm_fb_helper *fbh = drm_fb_helper_from_client(client); - - if (fbh->info) { - drm_fb_helper_unregister_info(fbh); - } else { - drm_client_release(&fbh->client); - drm_fb_helper_unprepare(fbh); - kfree(fbh); - } -} - -static int armada_fbdev_client_restore(struct drm_client_dev *client) -{ - drm_fb_helper_lastclose(client->dev); - - return 0; -} - -static int armada_fbdev_client_hotplug(struct drm_client_dev *client) -{ - struct drm_fb_helper *fbh = drm_fb_helper_from_client(client); - struct drm_device *dev = client->dev; - int ret; - - if (dev->fb_helper) - return drm_fb_helper_hotplug_event(dev->fb_helper); - - ret = drm_fb_helper_init(dev, fbh); - if (ret) - goto err_drm_err; - - if (!drm_drv_uses_atomic_modeset(dev)) - drm_helper_disable_unused_functions(dev); - - ret = drm_fb_helper_initial_config(fbh); - if (ret) - goto err_drm_fb_helper_fini; - - return 0; - -err_drm_fb_helper_fini: - drm_fb_helper_fini(fbh); -err_drm_err: - drm_err(dev, "armada: Failed to setup fbdev emulation (ret=%d)\n", ret); - return ret; -} - -static const struct drm_client_funcs armada_fbdev_client_funcs = { - .owner = THIS_MODULE, - .unregister = armada_fbdev_client_unregister, - .restore = armada_fbdev_client_restore, - .hotplug = armada_fbdev_client_hotplug, -}; - -void armada_fbdev_setup(struct drm_device *dev) -{ - struct drm_fb_helper *fbh; - int ret; - - drm_WARN(dev, !dev->registered, "Device has not been registered.\n"); - drm_WARN(dev, dev->fb_helper, "fb_helper is already set!\n"); - - fbh = kzalloc(sizeof(*fbh), GFP_KERNEL); - if (!fbh) - return; - drm_fb_helper_prepare(dev, fbh, 32, &armada_fb_helper_funcs); - - ret = drm_client_init(dev, &fbh->client, "fbdev", &armada_fbdev_client_funcs); - if (ret) { - drm_err(dev, "Failed to register client: %d\n", ret); - goto err_drm_client_init; - } - - drm_client_register(&fbh->client); - - return; - -err_drm_client_init: - drm_fb_helper_unprepare(fbh); - kfree(fbh); - return; -} diff --git a/drivers/gpu/drm/aspeed/Kconfig b/drivers/gpu/drm/aspeed/Kconfig index 8137c39b057b..6e68f20aac21 100644 --- a/drivers/gpu/drm/aspeed/Kconfig +++ b/drivers/gpu/drm/aspeed/Kconfig @@ -4,6 +4,7 @@ config DRM_ASPEED_GFX depends on DRM && OF depends on (COMPILE_TEST || ARCH_ASPEED) depends on MMU + select DRM_CLIENT_SELECTION select DRM_KMS_HELPER select DRM_GEM_DMA_HELPER select DMA_CMA if HAVE_DMA_CONTIGUOUS diff --git a/drivers/gpu/drm/aspeed/aspeed_gfx_drv.c b/drivers/gpu/drm/aspeed/aspeed_gfx_drv.c index a7a6b70220eb..109023815fa2 100644 --- a/drivers/gpu/drm/aspeed/aspeed_gfx_drv.c +++ b/drivers/gpu/drm/aspeed/aspeed_gfx_drv.c @@ -14,6 +14,7 @@ #include #include +#include #include #include #include @@ -247,6 +248,7 @@ DEFINE_DRM_GEM_DMA_FOPS(fops); static const struct drm_driver aspeed_gfx_driver = { .driver_features = DRIVER_GEM | DRIVER_MODESET | DRIVER_ATOMIC, DRM_GEM_DMA_DRIVER_OPS, + DRM_FBDEV_DMA_DRIVER_OPS, .fops = &fops, .name = "aspeed-gfx-drm", .desc = "ASPEED GFX DRM", @@ -339,7 +341,7 @@ static int aspeed_gfx_probe(struct platform_device *pdev) if (ret) goto err_unload; - drm_fbdev_dma_setup(&priv->drm, 32); + drm_client_setup(&priv->drm, NULL); return 0; err_unload: diff --git a/drivers/gpu/drm/ast/Kconfig b/drivers/gpu/drm/ast/Kconfig index 563fa7a3b546..da0663542e8a 100644 --- a/drivers/gpu/drm/ast/Kconfig +++ b/drivers/gpu/drm/ast/Kconfig @@ -2,6 +2,7 @@ config DRM_AST tristate "AST server chips" depends on DRM && PCI && MMU + select DRM_CLIENT_SELECTION select DRM_GEM_SHMEM_HELPER select DRM_KMS_HELPER select I2C diff --git a/drivers/gpu/drm/ast/ast_dp.c b/drivers/gpu/drm/ast/ast_dp.c index 00b364f9a71e..0e282b7b167c 100644 --- a/drivers/gpu/drm/ast/ast_dp.c +++ b/drivers/gpu/drm/ast/ast_dp.c @@ -149,28 +149,22 @@ int ast_dp_launch(struct ast_device *ast) return 0; } -static bool ast_dp_power_is_on(struct ast_device *ast) +static bool ast_dp_get_phy_sleep(struct ast_device *ast) { - u8 vgacre3; + u8 vgacre3 = ast_get_index_reg(ast, AST_IO_VGACRI, 0xe3); - vgacre3 = ast_get_index_reg(ast, AST_IO_VGACRI, 0xe3); - - return !(vgacre3 & AST_DP_PHY_SLEEP); + return (vgacre3 & AST_IO_VGACRE3_DP_PHY_SLEEP); } -static void ast_dp_power_on_off(struct drm_device *dev, bool on) +static void ast_dp_set_phy_sleep(struct ast_device *ast, bool sleep) { - struct ast_device *ast = to_ast_device(dev); - // Read and Turn off DP PHY sleep - u8 bE3 = ast_get_index_reg_mask(ast, AST_IO_VGACRI, 0xE3, AST_DP_VIDEO_ENABLE); + u8 vgacre3 = 0x00; - // Turn on DP PHY sleep - if (!on) - bE3 |= AST_DP_PHY_SLEEP; - - // DP Power on/off - ast_set_index_reg_mask(ast, AST_IO_VGACRI, 0xE3, (u8) ~AST_DP_PHY_SLEEP, bE3); + if (sleep) + vgacre3 |= AST_IO_VGACRE3_DP_PHY_SLEEP; + ast_set_index_reg_mask(ast, AST_IO_VGACRI, 0xe3, (u8)~AST_IO_VGACRE3_DP_PHY_SLEEP, + vgacre3); msleep(50); } @@ -192,23 +186,39 @@ static void ast_dp_link_training(struct ast_device *ast) drm_err(dev, "Link training failed\n"); } -static void ast_dp_set_on_off(struct drm_device *dev, bool on) +static bool __ast_dp_wait_enable(struct ast_device *ast, bool enabled) { - struct ast_device *ast = to_ast_device(dev); - u8 video_on_off = on; - u32 i = 0; + u8 vgacrdf_test = 0x00; + u8 vgacrdf; + unsigned int i; - // Video On/Off - ast_set_index_reg_mask(ast, AST_IO_VGACRI, 0xE3, (u8) ~AST_DP_VIDEO_ENABLE, on); + if (enabled) + vgacrdf_test |= AST_IO_VGACRDF_DP_VIDEO_ENABLE; - video_on_off <<= 4; - while (ast_get_index_reg_mask(ast, AST_IO_VGACRI, 0xDF, - ASTDP_MIRROR_VIDEO_ENABLE) != video_on_off) { - // wait 1 ms - mdelay(1); - if (++i > 200) - break; + for (i = 0; i < 200; ++i) { + if (i) + mdelay(1); + vgacrdf = ast_get_index_reg_mask(ast, AST_IO_VGACRI, 0xdf, + AST_IO_VGACRDF_DP_VIDEO_ENABLE); + if (vgacrdf == vgacrdf_test) + return true; } + + return false; +} + +static void ast_dp_set_enable(struct ast_device *ast, bool enabled) +{ + struct drm_device *dev = &ast->base; + u8 vgacre3 = 0x00; + + if (enabled) + vgacre3 |= AST_IO_VGACRE3_DP_VIDEO_ENABLE; + + ast_set_index_reg_mask(ast, AST_IO_VGACRI, 0xe3, (u8)~AST_IO_VGACRE3_DP_VIDEO_ENABLE, + vgacre3); + + drm_WARN_ON(dev, !__ast_dp_wait_enable(ast, enabled)); } static void ast_dp_set_mode(struct drm_crtc *crtc, struct ast_vbios_mode_info *vbios_mode) @@ -317,26 +327,25 @@ static void ast_astdp_encoder_helper_atomic_mode_set(struct drm_encoder *encoder static void ast_astdp_encoder_helper_atomic_enable(struct drm_encoder *encoder, struct drm_atomic_state *state) { - struct drm_device *dev = encoder->dev; - struct ast_device *ast = to_ast_device(dev); + struct ast_device *ast = to_ast_device(encoder->dev); struct ast_connector *ast_connector = &ast->output.astdp.connector; if (ast_connector->physical_status == connector_status_connected) { - ast_dp_power_on_off(dev, AST_DP_POWER_ON); + ast_dp_set_phy_sleep(ast, false); ast_dp_link_training(ast); ast_wait_for_vretrace(ast); - ast_dp_set_on_off(dev, 1); + ast_dp_set_enable(ast, true); } } static void ast_astdp_encoder_helper_atomic_disable(struct drm_encoder *encoder, struct drm_atomic_state *state) { - struct drm_device *dev = encoder->dev; + struct ast_device *ast = to_ast_device(encoder->dev); - ast_dp_set_on_off(dev, 0); - ast_dp_power_on_off(dev, AST_DP_POWER_OFF); + ast_dp_set_enable(ast, false); + ast_dp_set_phy_sleep(ast, true); } static const struct drm_encoder_helper_funcs ast_astdp_encoder_helper_funcs = { @@ -383,22 +392,21 @@ static int ast_astdp_connector_helper_detect_ctx(struct drm_connector *connector bool force) { struct ast_connector *ast_connector = to_ast_connector(connector); - struct drm_device *dev = connector->dev; struct ast_device *ast = to_ast_device(connector->dev); enum drm_connector_status status = connector_status_disconnected; - bool power_is_on; + bool phy_sleep; mutex_lock(&ast->modeset_lock); - power_is_on = ast_dp_power_is_on(ast); - if (!power_is_on) - ast_dp_power_on_off(dev, true); + phy_sleep = ast_dp_get_phy_sleep(ast); + if (phy_sleep) + ast_dp_set_phy_sleep(ast, false); if (ast_astdp_is_connected(ast)) status = connector_status_connected; - if (!power_is_on && status == connector_status_disconnected) - ast_dp_power_on_off(dev, false); + if (phy_sleep && status == connector_status_disconnected) + ast_dp_set_phy_sleep(ast, true); mutex_unlock(&ast->modeset_lock); @@ -414,6 +422,10 @@ static const struct drm_connector_helper_funcs ast_astdp_connector_helper_funcs .detect_ctx = ast_astdp_connector_helper_detect_ctx, }; +/* + * Output + */ + static const struct drm_connector_funcs ast_astdp_connector_funcs = { .reset = drm_atomic_helper_connector_reset, .fill_modes = drm_helper_probe_single_connector_modes, @@ -422,34 +434,18 @@ static const struct drm_connector_funcs ast_astdp_connector_funcs = { .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, }; -static int ast_astdp_connector_init(struct drm_device *dev, struct drm_connector *connector) -{ - int ret; - - ret = drm_connector_init(dev, connector, &ast_astdp_connector_funcs, - DRM_MODE_CONNECTOR_DisplayPort); - if (ret) - return ret; - - drm_connector_helper_add(connector, &ast_astdp_connector_helper_funcs); - - connector->interlace_allowed = 0; - connector->doublescan_allowed = 0; - - connector->polled = DRM_CONNECTOR_POLL_CONNECT | DRM_CONNECTOR_POLL_DISCONNECT; - - return 0; -} - int ast_astdp_output_init(struct ast_device *ast) { struct drm_device *dev = &ast->base; struct drm_crtc *crtc = &ast->crtc; - struct drm_encoder *encoder = &ast->output.astdp.encoder; - struct ast_connector *ast_connector = &ast->output.astdp.connector; - struct drm_connector *connector = &ast_connector->base; + struct drm_encoder *encoder; + struct ast_connector *ast_connector; + struct drm_connector *connector; int ret; + /* encoder */ + + encoder = &ast->output.astdp.encoder; ret = drm_encoder_init(dev, encoder, &ast_astdp_encoder_funcs, DRM_MODE_ENCODER_TMDS, NULL); if (ret) @@ -458,9 +454,20 @@ int ast_astdp_output_init(struct ast_device *ast) encoder->possible_crtcs = drm_crtc_mask(crtc); - ret = ast_astdp_connector_init(dev, connector); + /* connector */ + + ast_connector = &ast->output.astdp.connector; + connector = &ast_connector->base; + ret = drm_connector_init(dev, connector, &ast_astdp_connector_funcs, + DRM_MODE_CONNECTOR_DisplayPort); if (ret) return ret; + drm_connector_helper_add(connector, &ast_astdp_connector_helper_funcs); + + connector->interlace_allowed = 0; + connector->doublescan_allowed = 0; + connector->polled = DRM_CONNECTOR_POLL_CONNECT | DRM_CONNECTOR_POLL_DISCONNECT; + ast_connector->physical_status = connector->status; ret = drm_connector_attach_encoder(connector, encoder); diff --git a/drivers/gpu/drm/ast/ast_dp501.c b/drivers/gpu/drm/ast/ast_dp501.c index e4c636f45082..9e19d8c17730 100644 --- a/drivers/gpu/drm/ast/ast_dp501.c +++ b/drivers/gpu/drm/ast/ast_dp501.c @@ -21,9 +21,9 @@ static void ast_release_firmware(void *data) ast->dp501_fw = NULL; } -static int ast_load_dp501_microcode(struct drm_device *dev) +static int ast_load_dp501_microcode(struct ast_device *ast) { - struct ast_device *ast = to_ast_device(dev); + struct drm_device *dev = &ast->base; int ret; ret = request_firmware(&ast->dp501_fw, "ast_dp501_fw.bin", dev->dev); @@ -109,10 +109,10 @@ static bool wait_fw_ready(struct ast_device *ast) } #endif -static bool ast_write_cmd(struct drm_device *dev, u8 data) +static bool ast_write_cmd(struct ast_device *ast, u8 data) { - struct ast_device *ast = to_ast_device(dev); int retry = 0; + if (wait_nack(ast)) { send_nack(ast); ast_set_index_reg_mask(ast, AST_IO_VGACRI, 0x9a, 0x00, data); @@ -131,10 +131,8 @@ static bool ast_write_cmd(struct drm_device *dev, u8 data) return false; } -static bool ast_write_data(struct drm_device *dev, u8 data) +static bool ast_write_data(struct ast_device *ast, u8 data) { - struct ast_device *ast = to_ast_device(dev); - if (wait_nack(ast)) { send_nack(ast); ast_set_index_reg_mask(ast, AST_IO_VGACRI, 0x9a, 0x00, data); @@ -175,10 +173,10 @@ static void clear_cmd(struct ast_device *ast) } #endif -static void ast_set_dp501_video_output(struct drm_device *dev, u8 mode) +static void ast_set_dp501_video_output(struct ast_device *ast, u8 mode) { - ast_write_cmd(dev, 0x40); - ast_write_data(dev, mode); + ast_write_cmd(ast, 0x40); + ast_write_data(ast, mode); msleep(10); } @@ -188,9 +186,8 @@ static u32 get_fw_base(struct ast_device *ast) return ast_mindwm(ast, 0x1e6e2104) & 0x7fffffff; } -bool ast_backup_fw(struct drm_device *dev, u8 *addr, u32 size) +bool ast_backup_fw(struct ast_device *ast, u8 *addr, u32 size) { - struct ast_device *ast = to_ast_device(dev); u32 i, data; u32 boot_address; @@ -207,9 +204,8 @@ bool ast_backup_fw(struct drm_device *dev, u8 *addr, u32 size) return false; } -static bool ast_launch_m68k(struct drm_device *dev) +static bool ast_launch_m68k(struct ast_device *ast) { - struct ast_device *ast = to_ast_device(dev); u32 i, data, len = 0; u32 boot_address; u8 *fw_addr = NULL; @@ -226,7 +222,7 @@ static bool ast_launch_m68k(struct drm_device *dev) len = 32*1024; } else { if (!ast->dp501_fw && - ast_load_dp501_microcode(dev) < 0) + ast_load_dp501_microcode(ast) < 0) return false; fw_addr = (u8 *)ast->dp501_fw->data; @@ -348,9 +344,8 @@ static int ast_dp512_read_edid_block(void *data, u8 *buf, unsigned int block, si return true; } -static bool ast_init_dvo(struct drm_device *dev) +static bool ast_init_dvo(struct ast_device *ast) { - struct ast_device *ast = to_ast_device(dev); u8 jreg; u32 data; ast_write32(ast, 0xf004, 0x1e6e0000); @@ -421,9 +416,8 @@ static bool ast_init_dvo(struct drm_device *dev) } -static void ast_init_analog(struct drm_device *dev) +static void ast_init_analog(struct ast_device *ast) { - struct ast_device *ast = to_ast_device(dev); u32 data; /* @@ -448,28 +442,28 @@ static void ast_init_analog(struct drm_device *dev) ast_set_index_reg_mask(ast, AST_IO_VGACRI, 0xa3, 0xcf, 0x00); } -void ast_init_3rdtx(struct drm_device *dev) +void ast_init_3rdtx(struct ast_device *ast) { - struct ast_device *ast = to_ast_device(dev); - u8 jreg; + u8 vgacrd1; if (IS_AST_GEN4(ast) || IS_AST_GEN5(ast)) { - jreg = ast_get_index_reg_mask(ast, AST_IO_VGACRI, 0xd1, 0xff); - switch (jreg & 0x0e) { - case 0x04: - ast_init_dvo(dev); + vgacrd1 = ast_get_index_reg_mask(ast, AST_IO_VGACRI, 0xd1, + AST_IO_VGACRD1_TX_TYPE_MASK); + switch (vgacrd1) { + case AST_IO_VGACRD1_TX_SIL164_VBIOS: + ast_init_dvo(ast); break; - case 0x08: - ast_launch_m68k(dev); + case AST_IO_VGACRD1_TX_DP501_VBIOS: + ast_launch_m68k(ast); break; - case 0x0c: - ast_init_dvo(dev); + case AST_IO_VGACRD1_TX_FW_EMBEDDED_FW: + ast_init_dvo(ast); break; default: - if (ast->tx_chip_types & BIT(AST_TX_SIL164)) - ast_init_dvo(dev); + if (ast->tx_chip == AST_TX_SIL164) + ast_init_dvo(ast); else - ast_init_analog(dev); + ast_init_analog(ast); } } } @@ -485,17 +479,17 @@ static const struct drm_encoder_funcs ast_dp501_encoder_funcs = { static void ast_dp501_encoder_helper_atomic_enable(struct drm_encoder *encoder, struct drm_atomic_state *state) { - struct drm_device *dev = encoder->dev; + struct ast_device *ast = to_ast_device(encoder->dev); - ast_set_dp501_video_output(dev, 1); + ast_set_dp501_video_output(ast, 1); } static void ast_dp501_encoder_helper_atomic_disable(struct drm_encoder *encoder, struct drm_atomic_state *state) { - struct drm_device *dev = encoder->dev; + struct ast_device *ast = to_ast_device(encoder->dev); - ast_set_dp501_video_output(dev, 0); + ast_set_dp501_video_output(ast, 0); } static const struct drm_encoder_helper_funcs ast_dp501_encoder_helper_funcs = { @@ -567,34 +561,22 @@ static const struct drm_connector_funcs ast_dp501_connector_funcs = { .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, }; -static int ast_dp501_connector_init(struct drm_device *dev, struct drm_connector *connector) -{ - int ret; - - ret = drm_connector_init(dev, connector, &ast_dp501_connector_funcs, - DRM_MODE_CONNECTOR_DisplayPort); - if (ret) - return ret; - - drm_connector_helper_add(connector, &ast_dp501_connector_helper_funcs); - - connector->interlace_allowed = 0; - connector->doublescan_allowed = 0; - - connector->polled = DRM_CONNECTOR_POLL_CONNECT | DRM_CONNECTOR_POLL_DISCONNECT; - - return 0; -} +/* + * Output + */ int ast_dp501_output_init(struct ast_device *ast) { struct drm_device *dev = &ast->base; struct drm_crtc *crtc = &ast->crtc; - struct drm_encoder *encoder = &ast->output.dp501.encoder; - struct ast_connector *ast_connector = &ast->output.dp501.connector; - struct drm_connector *connector = &ast_connector->base; + struct drm_encoder *encoder; + struct ast_connector *ast_connector; + struct drm_connector *connector; int ret; + /* encoder */ + + encoder = &ast->output.dp501.encoder; ret = drm_encoder_init(dev, encoder, &ast_dp501_encoder_funcs, DRM_MODE_ENCODER_TMDS, NULL); if (ret) @@ -603,9 +585,20 @@ int ast_dp501_output_init(struct ast_device *ast) encoder->possible_crtcs = drm_crtc_mask(crtc); - ret = ast_dp501_connector_init(dev, connector); + /* connector */ + + ast_connector = &ast->output.dp501.connector; + connector = &ast_connector->base; + ret = drm_connector_init(dev, connector, &ast_dp501_connector_funcs, + DRM_MODE_CONNECTOR_DisplayPort); if (ret) return ret; + drm_connector_helper_add(connector, &ast_dp501_connector_helper_funcs); + + connector->interlace_allowed = 0; + connector->doublescan_allowed = 0; + connector->polled = DRM_CONNECTOR_POLL_CONNECT | DRM_CONNECTOR_POLL_DISCONNECT; + ast_connector->physical_status = connector->status; ret = drm_connector_attach_encoder(connector, encoder); diff --git a/drivers/gpu/drm/ast/ast_drv.c b/drivers/gpu/drm/ast/ast_drv.c index 3a908bb015fe..4afe4be072ef 100644 --- a/drivers/gpu/drm/ast/ast_drv.c +++ b/drivers/gpu/drm/ast/ast_drv.c @@ -26,12 +26,13 @@ * Authors: Dave Airlie */ +#include #include #include #include -#include #include +#include #include #include #include @@ -64,7 +65,8 @@ static const struct drm_driver ast_driver = { .minor = DRIVER_MINOR, .patchlevel = DRIVER_PATCHLEVEL, - DRM_GEM_SHMEM_DRIVER_OPS + DRM_GEM_SHMEM_DRIVER_OPS, + DRM_FBDEV_SHMEM_DRIVER_OPS, }; /* @@ -279,7 +281,7 @@ static int ast_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent) struct drm_device *drm; bool need_post = false; - ret = drm_aperture_remove_conflicting_pci_framebuffers(pdev, &ast_driver); + ret = aperture_remove_conflicting_pci_devices(pdev, ast_driver.name); if (ret) return ret; @@ -360,7 +362,7 @@ static int ast_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent) if (ret) return ret; - drm_fbdev_shmem_setup(drm, 32); + drm_client_setup(drm, NULL); return 0; } @@ -396,7 +398,7 @@ static int ast_drm_thaw(struct drm_device *dev) ast_enable_vga(ast->ioregs); ast_open_key(ast->ioregs); ast_enable_mmio(dev->dev, ast->ioregs); - ast_post_gpu(dev); + ast_post_gpu(ast); return drm_mode_config_helper_resume(dev); } diff --git a/drivers/gpu/drm/ast/ast_drv.h b/drivers/gpu/drm/ast/ast_drv.h index 91fe07cf7b07..21ce3769bf0d 100644 --- a/drivers/gpu/drm/ast/ast_drv.h +++ b/drivers/gpu/drm/ast/ast_drv.h @@ -91,11 +91,6 @@ enum ast_tx_chip { AST_TX_ASTDP, }; -#define AST_TX_NONE_BIT BIT(AST_TX_NONE) -#define AST_TX_SIL164_BIT BIT(AST_TX_SIL164) -#define AST_TX_DP501_BIT BIT(AST_TX_DP501) -#define AST_TX_ASTDP_BIT BIT(AST_TX_ASTDP) - enum ast_config_mode { ast_use_p2a, ast_use_dt, @@ -187,10 +182,12 @@ struct ast_device { struct mutex modeset_lock; /* Protects access to modeset I/O registers in ioregs */ + enum ast_tx_chip tx_chip; + struct ast_plane primary_plane; struct ast_plane cursor_plane; struct drm_crtc crtc; - struct { + union { struct { struct drm_encoder encoder; struct ast_connector connector; @@ -211,7 +208,6 @@ struct ast_device { bool support_wide_screen; - unsigned long tx_chip_types; /* bitfield of enum ast_chip_type */ u8 *dp501_fw_addr; const struct firmware *dp501_fw; /* dp501 fw */ }; @@ -407,9 +403,6 @@ int ast_mode_config_init(struct ast_device *ast); #define AST_DP501_LINKRATE 0xf014 #define AST_DP501_EDID_DATA 0xf020 -#define AST_DP_POWER_ON true -#define AST_DP_POWER_OFF false - /* * ASTDP resoultion table: * EX: ASTDP_A_B_C: @@ -453,7 +446,7 @@ int ast_mode_config_init(struct ast_device *ast); int ast_mm_init(struct ast_device *ast); /* ast post */ -void ast_post_gpu(struct drm_device *dev); +void ast_post_gpu(struct ast_device *ast); u32 ast_mindwm(struct ast_device *ast, u32 r); void ast_moutdwm(struct ast_device *ast, u32 r, u32 v); void ast_patch_ahb_2500(void __iomem *regs); @@ -462,8 +455,8 @@ int ast_vga_output_init(struct ast_device *ast); int ast_sil164_output_init(struct ast_device *ast); /* ast dp501 */ -bool ast_backup_fw(struct drm_device *dev, u8 *addr, u32 size); -void ast_init_3rdtx(struct drm_device *dev); +bool ast_backup_fw(struct ast_device *ast, u8 *addr, u32 size); +void ast_init_3rdtx(struct ast_device *ast); int ast_dp501_output_init(struct ast_device *ast); /* aspeed DP */ diff --git a/drivers/gpu/drm/ast/ast_main.c b/drivers/gpu/drm/ast/ast_main.c index d836f2a4f9f3..bc37c65305d4 100644 --- a/drivers/gpu/drm/ast/ast_main.c +++ b/drivers/gpu/drm/ast/ast_main.c @@ -68,11 +68,33 @@ static void ast_detect_widescreen(struct ast_device *ast) static void ast_detect_tx_chip(struct ast_device *ast, bool need_post) { + static const char * const info_str[] = { + "analog VGA", + "Sil164 TMDS transmitter", + "DP501 DisplayPort transmitter", + "ASPEED DisplayPort transmitter", + }; + struct drm_device *dev = &ast->base; - u8 jreg; + u8 jreg, vgacrd1; + + /* + * Several of the listed TX chips are not explicitly supported + * by the ast driver. If these exist in real-world devices, they + * are most likely reported as VGA or SIL164 outputs. We warn here + * to get bug reports for these devices. If none come in for some + * time, we can begin to fail device probing on these values. + */ + vgacrd1 = ast_get_index_reg_mask(ast, AST_IO_VGACRI, 0xd1, AST_IO_VGACRD1_TX_TYPE_MASK); + drm_WARN(dev, vgacrd1 == AST_IO_VGACRD1_TX_ITE66121_VBIOS, + "ITE IT66121 detected, 0x%x, Gen%lu\n", vgacrd1, AST_GEN(ast)); + drm_WARN(dev, vgacrd1 == AST_IO_VGACRD1_TX_CH7003_VBIOS, + "Chrontel CH7003 detected, 0x%x, Gen%lu\n", vgacrd1, AST_GEN(ast)); + drm_WARN(dev, vgacrd1 == AST_IO_VGACRD1_TX_ANX9807_VBIOS, + "Analogix ANX9807 detected, 0x%x, Gen%lu\n", vgacrd1, AST_GEN(ast)); /* Check 3rd Tx option (digital output afaik) */ - ast->tx_chip_types |= AST_TX_NONE_BIT; + ast->tx_chip = AST_TX_NONE; /* * VGACRA3 Enhanced Color Mode Register, check if DVO is already @@ -85,7 +107,7 @@ static void ast_detect_tx_chip(struct ast_device *ast, bool need_post) if (!need_post) { jreg = ast_get_index_reg_mask(ast, AST_IO_VGACRI, 0xa3, 0xff); if (jreg & 0x80) - ast->tx_chip_types = AST_TX_SIL164_BIT; + ast->tx_chip = AST_TX_SIL164; } if (IS_AST_GEN4(ast) || IS_AST_GEN5(ast) || IS_AST_GEN6(ast)) { @@ -94,49 +116,42 @@ static void ast_detect_tx_chip(struct ast_device *ast, bool need_post) * the SOC scratch register #1 bits 11:8 (interestingly marked * as "reserved" in the spec) */ - jreg = ast_get_index_reg_mask(ast, AST_IO_VGACRI, 0xd1, 0xff); + jreg = ast_get_index_reg_mask(ast, AST_IO_VGACRI, 0xd1, + AST_IO_VGACRD1_TX_TYPE_MASK); switch (jreg) { - case 0x04: - ast->tx_chip_types = AST_TX_SIL164_BIT; + case AST_IO_VGACRD1_TX_SIL164_VBIOS: + ast->tx_chip = AST_TX_SIL164; break; - case 0x08: + case AST_IO_VGACRD1_TX_DP501_VBIOS: ast->dp501_fw_addr = drmm_kzalloc(dev, 32*1024, GFP_KERNEL); if (ast->dp501_fw_addr) { /* backup firmware */ - if (ast_backup_fw(dev, ast->dp501_fw_addr, 32*1024)) { + if (ast_backup_fw(ast, ast->dp501_fw_addr, 32*1024)) { drmm_kfree(dev, ast->dp501_fw_addr); ast->dp501_fw_addr = NULL; } } fallthrough; - case 0x0c: - ast->tx_chip_types = AST_TX_DP501_BIT; + case AST_IO_VGACRD1_TX_FW_EMBEDDED_FW: + ast->tx_chip = AST_TX_DP501; } } else if (IS_AST_GEN7(ast)) { - if (ast_get_index_reg_mask(ast, AST_IO_VGACRI, 0xD1, TX_TYPE_MASK) == - ASTDP_DPMCU_TX) { + if (ast_get_index_reg_mask(ast, AST_IO_VGACRI, 0xd1, AST_IO_VGACRD1_TX_TYPE_MASK) == + AST_IO_VGACRD1_TX_ASTDP) { int ret = ast_dp_launch(ast); if (!ret) - ast->tx_chip_types = AST_TX_ASTDP_BIT; + ast->tx_chip = AST_TX_ASTDP; } } - /* Print stuff for diagnostic purposes */ - if (ast->tx_chip_types & AST_TX_NONE_BIT) - drm_info(dev, "Using analog VGA\n"); - if (ast->tx_chip_types & AST_TX_SIL164_BIT) - drm_info(dev, "Using Sil164 TMDS transmitter\n"); - if (ast->tx_chip_types & AST_TX_DP501_BIT) - drm_info(dev, "Using DP501 DisplayPort transmitter\n"); - if (ast->tx_chip_types & AST_TX_ASTDP_BIT) - drm_info(dev, "Using ASPEED DisplayPort transmitter\n"); + drm_info(dev, "Using %s\n", info_str[ast->tx_chip]); } -static int ast_get_dram_info(struct drm_device *dev) +static int ast_get_dram_info(struct ast_device *ast) { + struct drm_device *dev = &ast->base; struct device_node *np = dev->dev->of_node; - struct ast_device *ast = to_ast_device(dev); uint32_t mcr_cfg, mcr_scu_mpll, mcr_scu_strap; uint32_t denum, num, div, ref_pll, dsel; @@ -278,7 +293,7 @@ struct drm_device *ast_device_create(struct pci_dev *pdev, ast_detect_widescreen(ast); ast_detect_tx_chip(ast, need_post); - ret = ast_get_dram_info(dev); + ret = ast_get_dram_info(ast); if (ret) return ERR_PTR(ret); @@ -286,7 +301,7 @@ struct drm_device *ast_device_create(struct pci_dev *pdev, ast->mclk, ast->dram_type, ast->dram_bus_width); if (need_post) - ast_post_gpu(dev); + ast_post_gpu(ast); ret = ast_mm_init(ast); if (ret) diff --git a/drivers/gpu/drm/ast/ast_mode.c b/drivers/gpu/drm/ast/ast_mode.c index ed496fb32bf3..9d5321c81e68 100644 --- a/drivers/gpu/drm/ast/ast_mode.c +++ b/drivers/gpu/drm/ast/ast_mode.c @@ -1287,9 +1287,9 @@ static const struct drm_crtc_funcs ast_crtc_funcs = { .atomic_destroy_state = ast_crtc_atomic_destroy_state, }; -static int ast_crtc_init(struct drm_device *dev) +static int ast_crtc_init(struct ast_device *ast) { - struct ast_device *ast = to_ast_device(dev); + struct drm_device *dev = &ast->base; struct drm_crtc *crtc = &ast->crtc; int ret; @@ -1396,28 +1396,26 @@ int ast_mode_config_init(struct ast_device *ast) if (ret) return ret; - ast_crtc_init(dev); + ret = ast_crtc_init(ast); + if (ret) + return ret; - if (ast->tx_chip_types & AST_TX_NONE_BIT) { + switch (ast->tx_chip) { + case AST_TX_NONE: ret = ast_vga_output_init(ast); - if (ret) - return ret; - } - if (ast->tx_chip_types & AST_TX_SIL164_BIT) { + break; + case AST_TX_SIL164: ret = ast_sil164_output_init(ast); - if (ret) - return ret; - } - if (ast->tx_chip_types & AST_TX_DP501_BIT) { + break; + case AST_TX_DP501: ret = ast_dp501_output_init(ast); - if (ret) - return ret; - } - if (ast->tx_chip_types & AST_TX_ASTDP_BIT) { + break; + case AST_TX_ASTDP: ret = ast_astdp_output_init(ast); - if (ret) - return ret; + break; } + if (ret) + return ret; drm_mode_config_reset(dev); diff --git a/drivers/gpu/drm/ast/ast_post.c b/drivers/gpu/drm/ast/ast_post.c index 65755798ab94..364030f97571 100644 --- a/drivers/gpu/drm/ast/ast_post.c +++ b/drivers/gpu/drm/ast/ast_post.c @@ -34,16 +34,14 @@ #include "ast_dram_tables.h" #include "ast_drv.h" -static void ast_post_chip_2300(struct drm_device *dev); -static void ast_post_chip_2500(struct drm_device *dev); +static void ast_post_chip_2300(struct ast_device *ast); +static void ast_post_chip_2500(struct ast_device *ast); static const u8 extreginfo[] = { 0x0f, 0x04, 0x1c, 0xff }; static const u8 extreginfo_ast2300[] = { 0x0f, 0x04, 0x1f, 0xff }; -static void -ast_set_def_ext_reg(struct drm_device *dev) +static void ast_set_def_ext_reg(struct ast_device *ast) { - struct ast_device *ast = to_ast_device(dev); u8 i, index, reg; const u8 *ext_reg_info; @@ -252,9 +250,8 @@ cbr_start: -static void ast_init_dram_reg(struct drm_device *dev) +static void ast_init_dram_reg(struct ast_device *ast) { - struct ast_device *ast = to_ast_device(dev); u8 j; u32 data, temp, i; const struct ast_dramstruct *dram_reg_info; @@ -343,26 +340,24 @@ static void ast_init_dram_reg(struct drm_device *dev) } while ((j & 0x40) == 0); } -void ast_post_gpu(struct drm_device *dev) +void ast_post_gpu(struct ast_device *ast) { - struct ast_device *ast = to_ast_device(dev); - - ast_set_def_ext_reg(dev); + ast_set_def_ext_reg(ast); if (IS_AST_GEN7(ast)) { - if (ast->tx_chip_types & AST_TX_ASTDP_BIT) + if (ast->tx_chip == AST_TX_ASTDP) ast_dp_launch(ast); } else if (ast->config_mode == ast_use_p2a) { if (IS_AST_GEN6(ast)) - ast_post_chip_2500(dev); + ast_post_chip_2500(ast); else if (IS_AST_GEN5(ast) || IS_AST_GEN4(ast)) - ast_post_chip_2300(dev); + ast_post_chip_2300(ast); else - ast_init_dram_reg(dev); + ast_init_dram_reg(ast); - ast_init_3rdtx(dev); + ast_init_3rdtx(ast); } else { - if (ast->tx_chip_types & AST_TX_SIL164_BIT) + if (ast->tx_chip == AST_TX_SIL164) ast_set_index_reg_mask(ast, AST_IO_VGACRI, 0xa3, 0xcf, 0x80); /* Enable DVO */ } } @@ -1569,9 +1564,8 @@ ddr2_init_start: } -static void ast_post_chip_2300(struct drm_device *dev) +static void ast_post_chip_2300(struct ast_device *ast) { - struct ast_device *ast = to_ast_device(dev); struct ast2300_dram_param param; u32 temp; u8 reg; @@ -2038,9 +2032,9 @@ void ast_patch_ahb_2500(void __iomem *regs) __ast_moutdwm(regs, 0x1e6e207c, 0x08000000); /* clear fast reset */ } -void ast_post_chip_2500(struct drm_device *dev) +void ast_post_chip_2500(struct ast_device *ast) { - struct ast_device *ast = to_ast_device(dev); + struct drm_device *dev = &ast->base; u32 temp; u8 reg; diff --git a/drivers/gpu/drm/ast/ast_reg.h b/drivers/gpu/drm/ast/ast_reg.h index 040961cc1a19..2aadf07d135a 100644 --- a/drivers/gpu/drm/ast/ast_reg.h +++ b/drivers/gpu/drm/ast/ast_reg.h @@ -37,28 +37,29 @@ #define AST_IO_VGACRCB_HWC_16BPP BIT(0) /* set: ARGB4444, cleared: 2bpp palette */ #define AST_IO_VGACRCB_HWC_ENABLED BIT(1) -#define AST_IO_VGACRD1_MCU_FW_EXECUTING BIT(5) +#define AST_IO_VGACRD1_MCU_FW_EXECUTING BIT(5) +/* Display Transmitter Type */ +#define AST_IO_VGACRD1_TX_TYPE_MASK GENMASK(3, 1) +#define AST_IO_VGACRD1_NO_TX 0x00 +#define AST_IO_VGACRD1_TX_ITE66121_VBIOS 0x02 +#define AST_IO_VGACRD1_TX_SIL164_VBIOS 0x04 +#define AST_IO_VGACRD1_TX_CH7003_VBIOS 0x06 +#define AST_IO_VGACRD1_TX_DP501_VBIOS 0x08 +#define AST_IO_VGACRD1_TX_ANX9807_VBIOS 0x0a +#define AST_IO_VGACRD1_TX_FW_EMBEDDED_FW 0x0c /* special case of DP501 */ +#define AST_IO_VGACRD1_TX_ASTDP 0x0e + #define AST_IO_VGACRD7_EDID_VALID_FLAG BIT(0) #define AST_IO_VGACRDC_LINK_SUCCESS BIT(0) #define AST_IO_VGACRDF_HPD BIT(0) +#define AST_IO_VGACRDF_DP_VIDEO_ENABLE BIT(4) /* mirrors AST_IO_VGACRE3_DP_VIDEO_ENABLE */ +#define AST_IO_VGACRE3_DP_VIDEO_ENABLE BIT(0) +#define AST_IO_VGACRE3_DP_PHY_SLEEP BIT(4) #define AST_IO_VGACRE5_EDID_READ_DONE BIT(0) #define AST_IO_VGAIR1_R (0x5A) #define AST_IO_VGAIR1_VREFRESH BIT(3) -/* - * Display Transmitter Type - */ - -#define TX_TYPE_MASK GENMASK(3, 1) -#define NO_TX (0 << 1) -#define ITE66121_VBIOS_TX (1 << 1) -#define SI164_VBIOS_TX (2 << 1) -#define CH7003_VBIOS_TX (3 << 1) -#define DP501_VBIOS_TX (4 << 1) -#define ANX9807_VBIOS_TX (5 << 1) -#define TX_FW_EMBEDDED_FW_TX (6 << 1) -#define ASTDP_DPMCU_TX (7 << 1) #define AST_VRAM_INIT_STATUS_MASK GENMASK(7, 6) //#define AST_VRAM_INIT_BY_BMC BIT(7) @@ -68,18 +69,6 @@ * AST DisplayPort */ -/* Define for Soc scratched reg used on ASTDP */ -#define AST_DP_PHY_SLEEP BIT(4) -#define AST_DP_VIDEO_ENABLE BIT(0) - -/* - * CRDF[b4]: Mirror of AST_DP_VIDEO_ENABLE - * Precondition: A. ~AST_DP_PHY_SLEEP && - * B. DP_HPD && - * C. DP_LINK_SUCCESS - */ -#define ASTDP_MIRROR_VIDEO_ENABLE BIT(4) - /* * ASTDP setmode registers: * CRE0[7:0]: MISC0 ((0x00: 18-bpp) or (0x20: 24-bpp) diff --git a/drivers/gpu/drm/ast/ast_sil164.c b/drivers/gpu/drm/ast/ast_sil164.c index c231389936bd..be01254dd48a 100644 --- a/drivers/gpu/drm/ast/ast_sil164.c +++ b/drivers/gpu/drm/ast/ast_sil164.c @@ -73,52 +73,49 @@ static const struct drm_connector_funcs ast_sil164_connector_funcs = { .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, }; -static int ast_sil164_connector_init(struct drm_device *dev, struct drm_connector *connector) -{ - struct ast_device *ast = to_ast_device(dev); - struct i2c_adapter *ddc; - int ret; - - ddc = ast_ddc_create(ast); - if (IS_ERR(ddc)) { - ret = PTR_ERR(ddc); - drm_err(dev, "failed to add DDC bus for connector; ret=%d\n", ret); - return ret; - } - - ret = drm_connector_init_with_ddc(dev, connector, &ast_sil164_connector_funcs, - DRM_MODE_CONNECTOR_DVII, ddc); - if (ret) - return ret; - - drm_connector_helper_add(connector, &ast_sil164_connector_helper_funcs); - - connector->interlace_allowed = 0; - connector->doublescan_allowed = 0; - - connector->polled = DRM_CONNECTOR_POLL_CONNECT | DRM_CONNECTOR_POLL_DISCONNECT; - - return 0; -} +/* + * Output + */ int ast_sil164_output_init(struct ast_device *ast) { struct drm_device *dev = &ast->base; struct drm_crtc *crtc = &ast->crtc; - struct drm_encoder *encoder = &ast->output.sil164.encoder; - struct ast_connector *ast_connector = &ast->output.sil164.connector; - struct drm_connector *connector = &ast_connector->base; + struct i2c_adapter *ddc; + struct drm_encoder *encoder; + struct ast_connector *ast_connector; + struct drm_connector *connector; int ret; + /* DDC */ + + ddc = ast_ddc_create(ast); + if (IS_ERR(ddc)) + return PTR_ERR(ddc); + + /* encoder */ + + encoder = &ast->output.sil164.encoder; ret = drm_encoder_init(dev, encoder, &ast_sil164_encoder_funcs, DRM_MODE_ENCODER_TMDS, NULL); if (ret) return ret; encoder->possible_crtcs = drm_crtc_mask(crtc); - ret = ast_sil164_connector_init(dev, connector); + /* connector */ + + ast_connector = &ast->output.sil164.connector; + connector = &ast_connector->base; + ret = drm_connector_init_with_ddc(dev, connector, &ast_sil164_connector_funcs, + DRM_MODE_CONNECTOR_DVII, ddc); if (ret) return ret; + drm_connector_helper_add(connector, &ast_sil164_connector_helper_funcs); + + connector->interlace_allowed = 0; + connector->doublescan_allowed = 0; + connector->polled = DRM_CONNECTOR_POLL_CONNECT | DRM_CONNECTOR_POLL_DISCONNECT; + ast_connector->physical_status = connector->status; ret = drm_connector_attach_encoder(connector, encoder); diff --git a/drivers/gpu/drm/ast/ast_vga.c b/drivers/gpu/drm/ast/ast_vga.c index dd389a0a8f4a..abe0fff8485c 100644 --- a/drivers/gpu/drm/ast/ast_vga.c +++ b/drivers/gpu/drm/ast/ast_vga.c @@ -73,52 +73,49 @@ static const struct drm_connector_funcs ast_vga_connector_funcs = { .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, }; -static int ast_vga_connector_init(struct drm_device *dev, struct drm_connector *connector) -{ - struct ast_device *ast = to_ast_device(dev); - struct i2c_adapter *ddc; - int ret; - - ddc = ast_ddc_create(ast); - if (IS_ERR(ddc)) { - ret = PTR_ERR(ddc); - drm_err(dev, "failed to add DDC bus for connector; ret=%d\n", ret); - return ret; - } - - ret = drm_connector_init_with_ddc(dev, connector, &ast_vga_connector_funcs, - DRM_MODE_CONNECTOR_VGA, ddc); - if (ret) - return ret; - - drm_connector_helper_add(connector, &ast_vga_connector_helper_funcs); - - connector->interlace_allowed = 0; - connector->doublescan_allowed = 0; - - connector->polled = DRM_CONNECTOR_POLL_CONNECT | DRM_CONNECTOR_POLL_DISCONNECT; - - return 0; -} +/* + * Output + */ int ast_vga_output_init(struct ast_device *ast) { struct drm_device *dev = &ast->base; struct drm_crtc *crtc = &ast->crtc; - struct drm_encoder *encoder = &ast->output.vga.encoder; - struct ast_connector *ast_connector = &ast->output.vga.connector; - struct drm_connector *connector = &ast_connector->base; + struct i2c_adapter *ddc; + struct drm_encoder *encoder; + struct ast_connector *ast_connector; + struct drm_connector *connector; int ret; + /* DDC */ + + ddc = ast_ddc_create(ast); + if (IS_ERR(ddc)) + return PTR_ERR(ddc); + + /* encoder */ + + encoder = &ast->output.vga.encoder; ret = drm_encoder_init(dev, encoder, &ast_vga_encoder_funcs, DRM_MODE_ENCODER_DAC, NULL); if (ret) return ret; encoder->possible_crtcs = drm_crtc_mask(crtc); - ret = ast_vga_connector_init(dev, connector); + /* connector */ + + ast_connector = &ast->output.vga.connector; + connector = &ast_connector->base; + ret = drm_connector_init_with_ddc(dev, connector, &ast_vga_connector_funcs, + DRM_MODE_CONNECTOR_VGA, ddc); if (ret) return ret; + drm_connector_helper_add(connector, &ast_vga_connector_helper_funcs); + + connector->interlace_allowed = 0; + connector->doublescan_allowed = 0; + connector->polled = DRM_CONNECTOR_POLL_CONNECT | DRM_CONNECTOR_POLL_DISCONNECT; + ast_connector->physical_status = connector->status; ret = drm_connector_attach_encoder(connector, encoder); diff --git a/drivers/gpu/drm/atmel-hlcdc/Kconfig b/drivers/gpu/drm/atmel-hlcdc/Kconfig index 945f3aa7bb24..f8b9c91907d8 100644 --- a/drivers/gpu/drm/atmel-hlcdc/Kconfig +++ b/drivers/gpu/drm/atmel-hlcdc/Kconfig @@ -2,6 +2,7 @@ config DRM_ATMEL_HLCDC tristate "DRM Support for ATMEL HLCDC Display Controller" depends on DRM && OF && COMMON_CLK && ((MFD_ATMEL_HLCDC && ARM) || COMPILE_TEST) + select DRM_CLIENT_SELECTION select DRM_GEM_DMA_HELPER select DRM_KMS_HELPER select DRM_PANEL diff --git a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_dc.c b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_dc.c index 9ce429f889ca..792dcc19e8e7 100644 --- a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_dc.c +++ b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_dc.c @@ -18,8 +18,10 @@ #include #include +#include #include #include +#include #include #include #include @@ -840,6 +842,7 @@ DEFINE_DRM_GEM_DMA_FOPS(fops); static const struct drm_driver atmel_hlcdc_dc_driver = { .driver_features = DRIVER_GEM | DRIVER_MODESET | DRIVER_ATOMIC, DRM_GEM_DMA_DRIVER_OPS, + DRM_FBDEV_DMA_DRIVER_OPS, .fops = &fops, .name = "atmel-hlcdc", .desc = "Atmel HLCD Controller DRM", @@ -865,7 +868,7 @@ static int atmel_hlcdc_dc_drm_probe(struct platform_device *pdev) if (ret) goto err_unload; - drm_fbdev_dma_setup(ddev, 24); + drm_client_setup_with_fourcc(ddev, DRM_FORMAT_RGB888); return 0; diff --git a/drivers/gpu/drm/bridge/Kconfig b/drivers/gpu/drm/bridge/Kconfig index 3eb955333c80..6b4664d91faa 100644 --- a/drivers/gpu/drm/bridge/Kconfig +++ b/drivers/gpu/drm/bridge/Kconfig @@ -90,6 +90,17 @@ config DRM_FSL_LDB help Support for i.MX8MP DPI-to-LVDS on-SoC encoder. +config DRM_ITE_IT6263 + tristate "ITE IT6263 LVDS/HDMI bridge" + depends on OF + select DRM_DISPLAY_HDMI_STATE_HELPER + select DRM_DISPLAY_HELPER + select DRM_BRIDGE_CONNECTOR + select DRM_KMS_HELPER + select REGMAP_I2C + help + ITE IT6263 LVDS to HDMI bridge chip driver. + config DRM_ITE_IT6505 tristate "ITE IT6505 DisplayPort bridge" depends on OF @@ -140,6 +151,8 @@ config DRM_LONTIUM_LT9611 select DRM_PANEL_BRIDGE select DRM_KMS_HELPER select DRM_MIPI_DSI + select DRM_DISPLAY_HELPER + select DRM_DISPLAY_HDMI_STATE_HELPER select REGMAP_I2C help Driver for Lontium LT9611 DSI to HDMI bridge @@ -368,6 +381,13 @@ config DRM_TI_DLPC3433 It supports up to 720p resolution with 60 and 120 Hz refresh rates. +config DRM_TI_TDP158 + tristate "TI TDP158 HDMI/TMDS bridge" + depends on OF + select DRM_PANEL_BRIDGE + help + Texas Instruments TDP158 HDMI/TMDS Bridge driver + config DRM_TI_TFP410 tristate "TI TFP410 DVI/HDMI bridge" depends on OF diff --git a/drivers/gpu/drm/bridge/Makefile b/drivers/gpu/drm/bridge/Makefile index 7df87b582dca..97304b429a53 100644 --- a/drivers/gpu/drm/bridge/Makefile +++ b/drivers/gpu/drm/bridge/Makefile @@ -6,6 +6,7 @@ obj-$(CONFIG_DRM_CHRONTEL_CH7033) += chrontel-ch7033.o obj-$(CONFIG_DRM_CROS_EC_ANX7688) += cros-ec-anx7688.o obj-$(CONFIG_DRM_DISPLAY_CONNECTOR) += display-connector.o obj-$(CONFIG_DRM_FSL_LDB) += fsl-ldb.o +obj-$(CONFIG_DRM_ITE_IT6263) += ite-it6263.o obj-$(CONFIG_DRM_ITE_IT6505) += ite-it6505.o obj-$(CONFIG_DRM_LONTIUM_LT8912B) += lontium-lt8912b.o obj-$(CONFIG_DRM_LONTIUM_LT9211) += lontium-lt9211.o @@ -32,6 +33,7 @@ obj-$(CONFIG_DRM_I2C_ADV7511) += adv7511/ obj-$(CONFIG_DRM_TI_DLPC3433) += ti-dlpc3433.o obj-$(CONFIG_DRM_TI_SN65DSI83) += ti-sn65dsi83.o obj-$(CONFIG_DRM_TI_SN65DSI86) += ti-sn65dsi86.o +obj-$(CONFIG_DRM_TI_TDP158) += ti-tdp158.o obj-$(CONFIG_DRM_TI_TFP410) += ti-tfp410.o obj-$(CONFIG_DRM_TI_TPD12S015) += ti-tpd12s015.o obj-$(CONFIG_DRM_NWL_MIPI_DSI) += nwl-dsi.o diff --git a/drivers/gpu/drm/bridge/analogix/anx7625.c b/drivers/gpu/drm/bridge/analogix/anx7625.c index a2e9bb485c36..a2675b121fe4 100644 --- a/drivers/gpu/drm/bridge/analogix/anx7625.c +++ b/drivers/gpu/drm/bridge/analogix/anx7625.c @@ -2551,6 +2551,8 @@ static int __maybe_unused anx7625_runtime_pm_suspend(struct device *dev) mutex_lock(&ctx->lock); anx7625_stop_dp_work(ctx); + if (!ctx->pdata.panel_bridge) + anx7625_remove_edid(ctx); anx7625_power_standby(ctx); mutex_unlock(&ctx->lock); diff --git a/drivers/gpu/drm/bridge/aux-bridge.c b/drivers/gpu/drm/bridge/aux-bridge.c index 295e9d031e2d..015983c015e5 100644 --- a/drivers/gpu/drm/bridge/aux-bridge.c +++ b/drivers/gpu/drm/bridge/aux-bridge.c @@ -121,6 +121,10 @@ static int drm_aux_bridge_probe(struct auxiliary_device *auxdev, data->bridge.funcs = &drm_aux_bridge_funcs; data->bridge.of_node = data->dev->of_node; + /* passthrough data, allow everything */ + data->bridge.interlace_allowed = true; + data->bridge.ycbcr_420_allowed = true; + return devm_drm_bridge_add(data->dev, &data->bridge); } diff --git a/drivers/gpu/drm/bridge/aux-hpd-bridge.c b/drivers/gpu/drm/bridge/aux-hpd-bridge.c index 6886db2d9e00..48f297c78ee6 100644 --- a/drivers/gpu/drm/bridge/aux-hpd-bridge.c +++ b/drivers/gpu/drm/bridge/aux-hpd-bridge.c @@ -180,6 +180,10 @@ static int drm_aux_hpd_bridge_probe(struct auxiliary_device *auxdev, data->bridge.ops = DRM_BRIDGE_OP_HPD; data->bridge.type = id->driver_data; + /* passthrough data, allow everything */ + data->bridge.interlace_allowed = true; + data->bridge.ycbcr_420_allowed = true; + auxiliary_set_drvdata(auxdev, data); return devm_drm_bridge_add(data->dev, &data->bridge); diff --git a/drivers/gpu/drm/bridge/display-connector.c b/drivers/gpu/drm/bridge/display-connector.c index ab8e00baf3f1..aab9ce7be94c 100644 --- a/drivers/gpu/drm/bridge/display-connector.c +++ b/drivers/gpu/drm/bridge/display-connector.c @@ -270,6 +270,10 @@ static int display_connector_probe(struct platform_device *pdev) /* All the supported connector types support interlaced modes. */ conn->bridge.interlace_allowed = true; + if (type == DRM_MODE_CONNECTOR_HDMIA || + type == DRM_MODE_CONNECTOR_DisplayPort) + conn->bridge.ycbcr_420_allowed = true; + /* Get the optional connector label. */ of_property_read_string(pdev->dev.of_node, "label", &label); diff --git a/drivers/gpu/drm/bridge/imx/Kconfig b/drivers/gpu/drm/bridge/imx/Kconfig index 8dd89efa8ea7..9a480c6abb85 100644 --- a/drivers/gpu/drm/bridge/imx/Kconfig +++ b/drivers/gpu/drm/bridge/imx/Kconfig @@ -3,6 +3,16 @@ if ARCH_MXC || COMPILE_TEST config DRM_IMX_LDB_HELPER tristate +config DRM_IMX_LEGACY_BRIDGE + tristate + depends on DRM_IMX + help + This is a DRM bridge implementation for the DRM i.MX IPUv3 driver, + that uses of_get_drm_display_mode to acquire display mode. + + Newer designs should not use this bridge and should use proper panel + driver instead. + config DRM_IMX8MP_DW_HDMI_BRIDGE tristate "Freescale i.MX8MP HDMI-TX bridge support" depends on OF diff --git a/drivers/gpu/drm/bridge/imx/Makefile b/drivers/gpu/drm/bridge/imx/Makefile index edb0a7b71b30..dd5d48584806 100644 --- a/drivers/gpu/drm/bridge/imx/Makefile +++ b/drivers/gpu/drm/bridge/imx/Makefile @@ -1,4 +1,5 @@ obj-$(CONFIG_DRM_IMX_LDB_HELPER) += imx-ldb-helper.o +obj-$(CONFIG_DRM_IMX_LEGACY_BRIDGE) += imx-legacy-bridge.o obj-$(CONFIG_DRM_IMX8MP_DW_HDMI_BRIDGE) += imx8mp-hdmi-tx.o obj-$(CONFIG_DRM_IMX8MP_HDMI_PVI) += imx8mp-hdmi-pvi.o obj-$(CONFIG_DRM_IMX8QM_LDB) += imx8qm-ldb.o diff --git a/drivers/gpu/drm/bridge/imx/imx-legacy-bridge.c b/drivers/gpu/drm/bridge/imx/imx-legacy-bridge.c new file mode 100644 index 000000000000..3ebf0b9866de --- /dev/null +++ b/drivers/gpu/drm/bridge/imx/imx-legacy-bridge.c @@ -0,0 +1,88 @@ +// SPDX-License-Identifier: GPL-2.0+ +/* + * Freescale i.MX drm driver + * + * bridge driver for legacy DT bindings, utilizing display-timings node + */ + +#include +#include +#include +#include + +#include