linux/include/drm/drm_modeset_helper_vtables.h

1038 lines
41 KiB
C
Raw Normal View History

/*
* Copyright © 2006 Keith Packard
* Copyright © 2007-2008 Dave Airlie
* Copyright © 2007-2008 Intel Corporation
* Jesse Barnes <jesse.barnes@intel.com>
* Copyright © 2011-2013 Intel Corporation
* Copyright © 2015 Intel Corporation
* Daniel Vetter <daniel.vetter@ffwll.ch>
*
* 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.
*/
#ifndef __DRM_MODESET_HELPER_VTABLES_H__
#define __DRM_MODESET_HELPER_VTABLES_H__
#include <drm/drm_crtc.h>
#include <drm/drm_encoder.h>
/**
* DOC: overview
*
* The DRM mode setting helper functions are common code for drivers to use if
* they wish. Drivers are not forced to use this code in their
* implementations but it would be useful if the code they do use at least
* provides a consistent interface and operation to userspace. Therefore it is
* highly recommended to use the provided helpers as much as possible.
*
* Because there is only one pointer per modeset object to hold a vfunc table
* for helper libraries they are by necessity shared among the different
* helpers.
*
* To make this clear all the helper vtables are pulled together in this location here.
*/
enum mode_set_atomic;
/**
* struct drm_crtc_helper_funcs - helper operations for CRTCs
*
* These hooks are used by the legacy CRTC helpers, the transitional plane
* helpers and the new atomic modesetting helpers.
*/
struct drm_crtc_helper_funcs {
/**
* @dpms:
*
* Callback to control power levels on the CRTC. If the mode passed in
* is unsupported, the provider must use the next lowest power level.
* This is used by the legacy CRTC helpers to implement DPMS
* functionality in drm_helper_connector_dpms().
*
* This callback is also used to disable a CRTC by calling it with
* DRM_MODE_DPMS_OFF if the @disable hook isn't used.
*
* This callback is used by the legacy CRTC helpers. Atomic helpers
* also support using this hook for enabling and disabling a CRTC to
* facilitate transitions to atomic, but it is deprecated. Instead
* @enable and @disable should be used.
*/
void (*dpms)(struct drm_crtc *crtc, int mode);
/**
* @prepare:
*
* This callback should prepare the CRTC for a subsequent modeset, which
* in practice means the driver should disable the CRTC if it is
* running. Most drivers ended up implementing this by calling their
* @dpms hook with DRM_MODE_DPMS_OFF.
*
* This callback is used by the legacy CRTC helpers. Atomic helpers
* also support using this hook for disabling a CRTC to facilitate
* transitions to atomic, but it is deprecated. Instead @disable should
* be used.
*/
void (*prepare)(struct drm_crtc *crtc);
/**
* @commit:
*
* This callback should commit the new mode on the CRTC after a modeset,
* which in practice means the driver should enable the CRTC. Most
* drivers ended up implementing this by calling their @dpms hook with
* DRM_MODE_DPMS_ON.
*
* This callback is used by the legacy CRTC helpers. Atomic helpers
* also support using this hook for enabling a CRTC to facilitate
* transitions to atomic, but it is deprecated. Instead @enable should
* be used.
*/
void (*commit)(struct drm_crtc *crtc);
/**
* @mode_fixup:
*
* This callback is used to validate a mode. The parameter mode is the
* display mode that userspace requested, adjusted_mode is the mode the
* encoders need to be fed with. Note that this is the inverse semantics
* of the meaning for the &drm_encoder and &drm_bridge_funcs.mode_fixup
* vfunc. If the CRTC cannot support the requested conversion from mode
* to adjusted_mode it should reject the modeset.
*
* This function is used by both legacy CRTC helpers and atomic helpers.
* With atomic helpers it is optional.
*
* NOTE:
*
* This function is called in the check phase of atomic modesets, which
* can be aborted for any reason (including on userspace's request to
* just check whether a configuration would be possible). Atomic drivers
* MUST NOT touch any persistent state (hardware or software) or data
* structures except the passed in adjusted_mode parameter.
*
* This is in contrast to the legacy CRTC helpers where this was
* allowed.
*
* Atomic drivers which need to inspect and adjust more state should
* instead use the @atomic_check callback.
*
* Also beware that neither core nor helpers filter modes before
* passing them to the driver: While the list of modes that is
* advertised to userspace is filtered using the
* &drm_connector.mode_valid callback, neither the core nor the helpers
* do any filtering on modes passed in from userspace when setting a
* mode. It is therefore possible for userspace to pass in a mode that
* was previously filtered out using &drm_connector.mode_valid or add a
* custom mode that wasn't probed from EDID or similar to begin with.
* Even though this is an advanced feature and rarely used nowadays,
* some users rely on being able to specify modes manually so drivers
* must be prepared to deal with it. Specifically this means that all
* drivers need not only validate modes in &drm_connector.mode_valid but
* also in this or in the &drm_encoder_helper_funcs.mode_fixup callback
* to make sure invalid modes passed in from userspace are rejected.
*
* RETURNS:
*
* True if an acceptable configuration is possible, false if the modeset
* operation should be rejected.
*/
bool (*mode_fixup)(struct drm_crtc *crtc,
const struct drm_display_mode *mode,
struct drm_display_mode *adjusted_mode);
/**
* @mode_set:
*
* This callback is used by the legacy CRTC helpers to set a new mode,
* position and framebuffer. Since it ties the primary plane to every
* mode change it is incompatible with universal plane support. And
* since it can't update other planes it's incompatible with atomic
* modeset support.
*
* This callback is only used by CRTC helpers and deprecated.
*
* RETURNS:
*
* 0 on success or a negative error code on failure.
*/
int (*mode_set)(struct drm_crtc *crtc, struct drm_display_mode *mode,
struct drm_display_mode *adjusted_mode, int x, int y,
struct drm_framebuffer *old_fb);
/**
* @mode_set_nofb:
*
* This callback is used to update the display mode of a CRTC without
* changing anything of the primary plane configuration. This fits the
* requirement of atomic and hence is used by the atomic helpers. It is
* also used by the transitional plane helpers to implement a
* @mode_set hook in drm_helper_crtc_mode_set().
*
* Note that the display pipe is completely off when this function is
* called. Atomic drivers which need hardware to be running before they
* program the new display mode (e.g. because they implement runtime PM)
* should not use this hook. This is because the helper library calls
* this hook only once per mode change and not every time the display
* pipeline is suspended using either DPMS or the new "ACTIVE" property.
* Which means register values set in this callback might get reset when
* the CRTC is suspended, but not restored. Such drivers should instead
* move all their CRTC setup into the @enable callback.
*
* This callback is optional.
*/
void (*mode_set_nofb)(struct drm_crtc *crtc);
/**
* @mode_set_base:
*
* This callback is used by the legacy CRTC helpers to set a new
* framebuffer and scanout position. It is optional and used as an
* optimized fast-path instead of a full mode set operation with all the
* resulting flickering. If it is not present
* drm_crtc_helper_set_config() will fall back to a full modeset, using
* the @mode_set callback. Since it can't update other planes it's
* incompatible with atomic modeset support.
*
* This callback is only used by the CRTC helpers and deprecated.
*
* RETURNS:
*
* 0 on success or a negative error code on failure.
*/
int (*mode_set_base)(struct drm_crtc *crtc, int x, int y,
struct drm_framebuffer *old_fb);
/**
* @mode_set_base_atomic:
*
* This callback is used by the fbdev helpers to set a new framebuffer
* and scanout without sleeping, i.e. from an atomic calling context. It
* is only used to implement kgdb support.
*
* This callback is optional and only needed for kgdb support in the fbdev
* helpers.
*
* RETURNS:
*
* 0 on success or a negative error code on failure.
*/
int (*mode_set_base_atomic)(struct drm_crtc *crtc,
struct drm_framebuffer *fb, int x, int y,
enum mode_set_atomic);
/**
* @load_lut:
*
* Load a LUT prepared with the &drm_fb_helper_funcs.gamma_set vfunc.
*
* This callback is optional and is only used by the fbdev emulation
* helpers.
*
* FIXME:
*
* This callback is functionally redundant with the core gamma table
* support and simply exists because the fbdev hasn't yet been
* refactored to use the core gamma table interfaces.
*/
void (*load_lut)(struct drm_crtc *crtc);
/**
* @disable:
*
* This callback should be used to disable the CRTC. With the atomic
* drivers it is called after all encoders connected to this CRTC have
* been shut off already using their own
* &drm_encoder_helper_funcs.disable hook. If that sequence is too
* simple drivers can just add their own hooks and call it from this
* CRTC callback here by looping over all encoders connected to it using
* for_each_encoder_on_crtc().
*
* This hook is used both by legacy CRTC helpers and atomic helpers.
* Atomic drivers don't need to implement it if there's no need to
* disable anything at the CRTC level. To ensure that runtime PM
* handling (using either DPMS or the new "ACTIVE" property) works
* @disable must be the inverse of @enable for atomic drivers.
* Atomic drivers should consider to use @atomic_disable instead of
* this one.
*
* NOTE:
*
* With legacy CRTC helpers there's a big semantic difference between
* @disable and other hooks (like @prepare or @dpms) used to shut down a
* CRTC: @disable is only called when also logically disabling the
* display pipeline and needs to release any resources acquired in
* @mode_set (like shared PLLs, or again release pinned framebuffers).
*
* Therefore @disable must be the inverse of @mode_set plus @commit for
* drivers still using legacy CRTC helpers, which is different from the
* rules under atomic.
*/
void (*disable)(struct drm_crtc *crtc);
/**
* @enable:
*
* This callback should be used to enable the CRTC. With the atomic
* drivers it is called before all encoders connected to this CRTC are
* enabled through the encoder's own &drm_encoder_helper_funcs.enable
* hook. If that sequence is too simple drivers can just add their own
* hooks and call it from this CRTC callback here by looping over all
* encoders connected to it using for_each_encoder_on_crtc().
*
* This hook is used only by atomic helpers, for symmetry with @disable.
* Atomic drivers don't need to implement it if there's no need to
* enable anything at the CRTC level. To ensure that runtime PM handling
* (using either DPMS or the new "ACTIVE" property) works
* @enable must be the inverse of @disable for atomic drivers.
*/
void (*enable)(struct drm_crtc *crtc);
/**
* @atomic_check:
*
* Drivers should check plane-update related CRTC constraints in this
* hook. They can also check mode related limitations but need to be
* aware of the calling order, since this hook is used by
* drm_atomic_helper_check_planes() whereas the preparations needed to
* check output routing and the display mode is done in
* drm_atomic_helper_check_modeset(). Therefore drivers that want to
* check output routing and display mode constraints in this callback
* must ensure that drm_atomic_helper_check_modeset() has been called
* beforehand. This is calling order used by the default helper
* implementation in drm_atomic_helper_check().
*
* When using drm_atomic_helper_check_planes() this hook is called
* after the &drm_plane_helper_funcs.atomc_check hook for planes, which
* allows drivers to assign shared resources requested by planes in this
* callback here. For more complicated dependencies the driver can call
* the provided check helpers multiple times until the computed state
* has a final configuration and everything has been checked.
*
* This function is also allowed to inspect any other object's state and
* can add more state objects to the atomic commit if needed. Care must
* be taken though to ensure that state check and compute functions for
* these added states are all called, and derived state in other objects
* all updated. Again the recommendation is to just call check helpers
* until a maximal configuration is reached.
*
* This callback is used by the atomic modeset helpers and by the
* transitional plane helpers, but it is optional.
*
* NOTE:
*
* This function is called in the check phase of an atomic update. The
* driver is not allowed to change anything outside of the free-standing
* state objects passed-in or assembled in the overall &drm_atomic_state
* update tracking structure.
*
* RETURNS:
*
* 0 on success, -EINVAL if the state or the transition can't be
* supported, -ENOMEM on memory allocation failure and -EDEADLK if an
* attempt to obtain another state object ran into a &drm_modeset_lock
* deadlock.
*/
int (*atomic_check)(struct drm_crtc *crtc,
struct drm_crtc_state *state);
/**
* @atomic_begin:
*
* Drivers should prepare for an atomic update of multiple planes on
* a CRTC in this hook. Depending upon hardware this might be vblank
* evasion, blocking updates by setting bits or doing preparatory work
* for e.g. manual update display.
*
* This hook is called before any plane commit functions are called.
*
* Note that the power state of the display pipe when this function is
* called depends upon the exact helpers and calling sequence the driver
* has picked. See drm_atomic_helper_commit_planes() for a discussion of
* the tradeoffs and variants of plane commit helpers.
*
* This callback is used by the atomic modeset helpers and by the
* transitional plane helpers, but it is optional.
*/
void (*atomic_begin)(struct drm_crtc *crtc,
struct drm_crtc_state *old_crtc_state);
/**
* @atomic_flush:
*
* Drivers should finalize an atomic update of multiple planes on
* a CRTC in this hook. Depending upon hardware this might include
* checking that vblank evasion was successful, unblocking updates by
* setting bits or setting the GO bit to flush out all updates.
*
* Simple hardware or hardware with special requirements can commit and
* flush out all updates for all planes from this hook and forgo all the
* other commit hooks for plane updates.
*
* This hook is called after any plane commit functions are called.
*
* Note that the power state of the display pipe when this function is
* called depends upon the exact helpers and calling sequence the driver
* has picked. See drm_atomic_helper_commit_planes() for a discussion of
* the tradeoffs and variants of plane commit helpers.
*
* This callback is used by the atomic modeset helpers and by the
* transitional plane helpers, but it is optional.
*/
void (*atomic_flush)(struct drm_crtc *crtc,
struct drm_crtc_state *old_crtc_state);
/**
* @atomic_disable:
*
* This callback should be used to disable the CRTC. With the atomic
* drivers it is called after all encoders connected to this CRTC have
* been shut off already using their own
* &drm_encoder_helper_funcs.disable hook. If that sequence is too
* simple drivers can just add their own hooks and call it from this
* CRTC callback here by looping over all encoders connected to it using
* for_each_encoder_on_crtc().
*
* This hook is used only by atomic helpers. Atomic drivers don't
* need to implement it if there's no need to disable anything at the
* CRTC level.
*
* Comparing to @disable, this one provides the additional input
* parameter @old_crtc_state which could be used to access the old
* state. Atomic drivers should consider to use this one instead
* of @disable.
*/
void (*atomic_disable)(struct drm_crtc *crtc,
struct drm_crtc_state *old_crtc_state);
};
/**
* drm_crtc_helper_add - sets the helper vtable for a crtc
* @crtc: DRM CRTC
* @funcs: helper vtable to set for @crtc
*/
static inline void drm_crtc_helper_add(struct drm_crtc *crtc,
const struct drm_crtc_helper_funcs *funcs)
{
crtc->helper_private = funcs;
}
/**
* struct drm_encoder_helper_funcs - helper operations for encoders
*
* These hooks are used by the legacy CRTC helpers, the transitional plane
* helpers and the new atomic modesetting helpers.
*/
struct drm_encoder_helper_funcs {
/**
* @dpms:
*
* Callback to control power levels on the encoder. If the mode passed in
* is unsupported, the provider must use the next lowest power level.
* This is used by the legacy encoder helpers to implement DPMS
* functionality in drm_helper_connector_dpms().
*
* This callback is also used to disable an encoder by calling it with
* DRM_MODE_DPMS_OFF if the @disable hook isn't used.
*
* This callback is used by the legacy CRTC helpers. Atomic helpers
* also support using this hook for enabling and disabling an encoder to
* facilitate transitions to atomic, but it is deprecated. Instead
* @enable and @disable should be used.
*/
void (*dpms)(struct drm_encoder *encoder, int mode);
/**
* @mode_fixup:
*
* This callback is used to validate and adjust a mode. The parameter
* mode is the display mode that should be fed to the next element in
* the display chain, either the final &drm_connector or a &drm_bridge.
* The parameter adjusted_mode is the input mode the encoder requires. It
* can be modified by this callback and does not need to match mode.
*
* This function is used by both legacy CRTC helpers and atomic helpers.
* This hook is optional.
*
* NOTE:
*
* This function is called in the check phase of atomic modesets, which
* can be aborted for any reason (including on userspace's request to
* just check whether a configuration would be possible). Atomic drivers
* MUST NOT touch any persistent state (hardware or software) or data
* structures except the passed in adjusted_mode parameter.
*
* This is in contrast to the legacy CRTC helpers where this was
* allowed.
*
* Atomic drivers which need to inspect and adjust more state should
* instead use the @atomic_check callback.
*
* Also beware that neither core nor helpers filter modes before
* passing them to the driver: While the list of modes that is
* advertised to userspace is filtered using the connector's
* &drm_connector_helper_funcs.mode_valid callback, neither the core nor
* the helpers do any filtering on modes passed in from userspace when
* setting a mode. It is therefore possible for userspace to pass in a
* mode that was previously filtered out using
* &drm_connector_helper_funcs.mode_valid or add a custom mode that
* wasn't probed from EDID or similar to begin with. Even though this
* is an advanced feature and rarely used nowadays, some users rely on
* being able to specify modes manually so drivers must be prepared to
* deal with it. Specifically this means that all drivers need not only
* validate modes in &drm_connector.mode_valid but also in this or in
* the &drm_crtc_helper_funcs.mode_fixup callback to make sure
* invalid modes passed in from userspace are rejected.
*
* RETURNS:
*
* True if an acceptable configuration is possible, false if the modeset
* operation should be rejected.
*/
bool (*mode_fixup)(struct drm_encoder *encoder,
const struct drm_display_mode *mode,
struct drm_display_mode *adjusted_mode);
/**
* @prepare:
*
* This callback should prepare the encoder for a subsequent modeset,
* which in practice means the driver should disable the encoder if it
* is running. Most drivers ended up implementing this by calling their
* @dpms hook with DRM_MODE_DPMS_OFF.
*
* This callback is used by the legacy CRTC helpers. Atomic helpers
* also support using this hook for disabling an encoder to facilitate
* transitions to atomic, but it is deprecated. Instead @disable should
* be used.
*/
void (*prepare)(struct drm_encoder *encoder);
/**
* @commit:
*
* This callback should commit the new mode on the encoder after a modeset,
* which in practice means the driver should enable the encoder. Most
* drivers ended up implementing this by calling their @dpms hook with
* DRM_MODE_DPMS_ON.
*
* This callback is used by the legacy CRTC helpers. Atomic helpers
* also support using this hook for enabling an encoder to facilitate
* transitions to atomic, but it is deprecated. Instead @enable should
* be used.
*/
void (*commit)(struct drm_encoder *encoder);
/**
* @mode_set:
*
* This callback is used to update the display mode of an encoder.
*
* Note that the display pipe is completely off when this function is
* called. Drivers which need hardware to be running before they program
* the new display mode (because they implement runtime PM) should not
* use this hook, because the helper library calls it only once and not
* every time the display pipeline is suspend using either DPMS or the
* new "ACTIVE" property. Such drivers should instead move all their
* encoder setup into the @enable callback.
*
* This callback is used both by the legacy CRTC helpers and the atomic
* modeset helpers. It is optional in the atomic helpers.
*
* NOTE:
*
* If the driver uses the atomic modeset helpers and needs to inspect
* the connector state or connector display info during mode setting,
* @atomic_mode_set can be used instead.
*/
void (*mode_set)(struct drm_encoder *encoder,
struct drm_display_mode *mode,
struct drm_display_mode *adjusted_mode);
/**
* @atomic_mode_set:
*
* This callback is used to update the display mode of an encoder.
*
* Note that the display pipe is completely off when this function is
* called. Drivers which need hardware to be running before they program
* the new display mode (because they implement runtime PM) should not
* use this hook, because the helper library calls it only once and not
* every time the display pipeline is suspended using either DPMS or the
* new "ACTIVE" property. Such drivers should instead move all their
* encoder setup into the @enable callback.
*
* This callback is used by the atomic modeset helpers in place of the
* @mode_set callback, if set by the driver. It is optional and should
* be used instead of @mode_set if the driver needs to inspect the
* connector state or display info, since there is no direct way to
* go from the encoder to the current connector.
*/
void (*atomic_mode_set)(struct drm_encoder *encoder,
struct drm_crtc_state *crtc_state,
struct drm_connector_state *conn_state);
/**
* @get_crtc:
*
* This callback is used by the legacy CRTC helpers to work around
* deficiencies in its own book-keeping.
*
* Do not use, use atomic helpers instead, which get the book keeping
* right.
*
* FIXME:
*
* Currently only nouveau is using this, and as soon as nouveau is
* atomic we can ditch this hook.
*/
struct drm_crtc *(*get_crtc)(struct drm_encoder *encoder);
/**
* @detect:
*
* This callback can be used by drivers who want to do detection on the
* encoder object instead of in connector functions.
*
* It is not used by any helper and therefore has purely driver-specific
* semantics. New drivers shouldn't use this and instead just implement
* their own private callbacks.
*
* FIXME:
*
* This should just be converted into a pile of driver vfuncs.
* Currently radeon, amdgpu and nouveau are using it.
*/
enum drm_connector_status (*detect)(struct drm_encoder *encoder,
struct drm_connector *connector);
/**
* @disable:
*
* This callback should be used to disable the encoder. With the atomic
* drivers it is called before this encoder's CRTC has been shut off
* using their own &drm_crtc_helper_funcs.disable hook. If that
* sequence is too simple drivers can just add their own driver private
* encoder hooks and call them from CRTC's callback by looping over all
* encoders connected to it using for_each_encoder_on_crtc().
*
* This hook is used both by legacy CRTC helpers and atomic helpers.
* Atomic drivers don't need to implement it if there's no need to
* disable anything at the encoder level. To ensure that runtime PM
* handling (using either DPMS or the new "ACTIVE" property) works
* @disable must be the inverse of @enable for atomic drivers.
*
* NOTE:
*
* With legacy CRTC helpers there's a big semantic difference between
* @disable and other hooks (like @prepare or @dpms) used to shut down a
* encoder: @disable is only called when also logically disabling the
* display pipeline and needs to release any resources acquired in
* @mode_set (like shared PLLs, or again release pinned framebuffers).
*
* Therefore @disable must be the inverse of @mode_set plus @commit for
* drivers still using legacy CRTC helpers, which is different from the
* rules under atomic.
*/
void (*disable)(struct drm_encoder *encoder);
/**
* @enable:
*
* This callback should be used to enable the encoder. With the atomic
* drivers it is called after this encoder's CRTC has been enabled using
* their own &drm_crtc_helper_funcs.enable hook. If that sequence is
* too simple drivers can just add their own driver private encoder
* hooks and call them from CRTC's callback by looping over all encoders
* connected to it using for_each_encoder_on_crtc().
*
* This hook is used only by atomic helpers, for symmetry with @disable.
* Atomic drivers don't need to implement it if there's no need to
* enable anything at the encoder level. To ensure that runtime PM handling
* (using either DPMS or the new "ACTIVE" property) works
* @enable must be the inverse of @disable for atomic drivers.
*/
void (*enable)(struct drm_encoder *encoder);
/**
* @atomic_check:
*
* This callback is used to validate encoder state for atomic drivers.
* Since the encoder is the object connecting the CRTC and connector it
* gets passed both states, to be able to validate interactions and
* update the CRTC to match what the encoder needs for the requested
* connector.
*
* This function is used by the atomic helpers, but it is optional.
*
* NOTE:
*
* This function is called in the check phase of an atomic update. The
* driver is not allowed to change anything outside of the free-standing
* state objects passed-in or assembled in the overall &drm_atomic_state
* update tracking structure.
*
* RETURNS:
*
* 0 on success, -EINVAL if the state or the transition can't be
* supported, -ENOMEM on memory allocation failure and -EDEADLK if an
* attempt to obtain another state object ran into a &drm_modeset_lock
* deadlock.
*/
int (*atomic_check)(struct drm_encoder *encoder,
struct drm_crtc_state *crtc_state,
struct drm_connector_state *conn_state);
};
/**
* drm_encoder_helper_add - sets the helper vtable for an encoder
* @encoder: DRM encoder
* @funcs: helper vtable to set for @encoder
*/
static inline void drm_encoder_helper_add(struct drm_encoder *encoder,
const struct drm_encoder_helper_funcs *funcs)
{
encoder->helper_private = funcs;
}
/**
* struct drm_connector_helper_funcs - helper operations for connectors
*
* These functions are used by the atomic and legacy modeset helpers and by the
* probe helpers.
*/
struct drm_connector_helper_funcs {
/**
* @get_modes:
*
* This function should fill in all modes currently valid for the sink
* into the &drm_connector.probed_modes list. It should also update the
* EDID property by calling drm_mode_connector_update_edid_property().
*
* The usual way to implement this is to cache the EDID retrieved in the
* probe callback somewhere in the driver-private connector structure.
* In this function drivers then parse the modes in the EDID and add
* them by calling drm_add_edid_modes(). But connectors that driver a
* fixed panel can also manually add specific modes using
* drm_mode_probed_add(). Drivers which manually add modes should also
* make sure that the &drm_connector.display_info,
* &drm_connector.width_mm and &drm_connector.height_mm fields are
* filled in.
*
* Virtual drivers that just want some standard VESA mode with a given
* resolution can call drm_add_modes_noedid(), and mark the preferred
* one using drm_set_preferred_mode().
*
* Finally drivers that support audio probably want to update the ELD
* data, too, using drm_edid_to_eld().
*
* This function is only called after the @detect hook has indicated
* that a sink is connected and when the EDID isn't overridden through
* sysfs or the kernel commandline.
*
* This callback is used by the probe helpers in e.g.
* drm_helper_probe_single_connector_modes().
*
* RETURNS:
*
* The number of modes added by calling drm_mode_probed_add().
*/
int (*get_modes)(struct drm_connector *connector);
/**
* @mode_valid:
*
* Callback to validate a mode for a connector, irrespective of the
* specific display configuration.
*
* This callback is used by the probe helpers to filter the mode list
* (which is usually derived from the EDID data block from the sink).
* See e.g. drm_helper_probe_single_connector_modes().
*
* NOTE:
*
* This only filters the mode list supplied to userspace in the
* GETCONNECOTR IOCTL. Userspace is free to create modes of its own and
* ask the kernel to use them. It this case the atomic helpers or legacy
* CRTC helpers will not call this function. Drivers therefore must
* still fully validate any mode passed in in a modeset request.
*
* RETURNS:
*
* Either &drm_mode_status.MODE_OK or one of the failure reasons in &enum
* drm_mode_status.
*/
enum drm_mode_status (*mode_valid)(struct drm_connector *connector,
struct drm_display_mode *mode);
/**
* @best_encoder:
*
* This function should select the best encoder for the given connector.
*
* This function is used by both the atomic helpers (in the
* drm_atomic_helper_check_modeset() function) and in the legacy CRTC
* helpers.
*
* NOTE:
*
* In atomic drivers this function is called in the check phase of an
* atomic update. The driver is not allowed to change or inspect
* anything outside of arguments passed-in. Atomic drivers which need to
* inspect dynamic configuration state should instead use
* @atomic_best_encoder.
*
* You can leave this function to NULL if the connector is only
* attached to a single encoder and you are using the atomic helpers.
* In this case, the core will call drm_atomic_helper_best_encoder()
* for you.
*
* RETURNS:
*
* Encoder that should be used for the given connector and connector
* state, or NULL if no suitable encoder exists. Note that the helpers
* will ensure that encoders aren't used twice, drivers should not check
* for this.
*/
struct drm_encoder *(*best_encoder)(struct drm_connector *connector);
/**
* @atomic_best_encoder:
*
* This is the atomic version of @best_encoder for atomic drivers which
* need to select the best encoder depending upon the desired
* configuration and can't select it statically.
*
* This function is used by drm_atomic_helper_check_modeset().
* If it is not implemented, the core will fallback to @best_encoder
* (or drm_atomic_helper_best_encoder() if @best_encoder is NULL).
*
* NOTE:
*
* This function is called in the check phase of an atomic update. The
* driver is not allowed to change anything outside of the free-standing
* state objects passed-in or assembled in the overall &drm_atomic_state
* update tracking structure.
*
* RETURNS:
*
* Encoder that should be used for the given connector and connector
* state, or NULL if no suitable encoder exists. Note that the helpers
* will ensure that encoders aren't used twice, drivers should not check
* for this.
*/
struct drm_encoder *(*atomic_best_encoder)(struct drm_connector *connector,
struct drm_connector_state *connector_state);
};
/**
* drm_connector_helper_add - sets the helper vtable for a connector
* @connector: DRM connector
* @funcs: helper vtable to set for @connector
*/
static inline void drm_connector_helper_add(struct drm_connector *connector,
const struct drm_connector_helper_funcs *funcs)
{
connector->helper_private = funcs;
}
/**
* struct drm_plane_helper_funcs - helper operations for planes
*
* These functions are used by the atomic helpers and by the transitional plane
* helpers.
*/
struct drm_plane_helper_funcs {
/**
* @prepare_fb:
*
* This hook is to prepare a framebuffer for scanout by e.g. pinning
* it's backing storage or relocating it into a contiguous block of
* VRAM. Other possible preparatory work includes flushing caches.
*
* This function must not block for outstanding rendering, since it is
* called in the context of the atomic IOCTL even for async commits to
* be able to return any errors to userspace. Instead the recommended
* way is to fill out the fence member of the passed-in
* &drm_plane_state. If the driver doesn't support native fences then
* equivalent functionality should be implemented through private
* members in the plane structure.
*
* The helpers will call @cleanup_fb with matching arguments for every
* successful call to this hook.
*
* This callback is used by the atomic modeset helpers and by the
* transitional plane helpers, but it is optional.
*
* RETURNS:
*
* 0 on success or one of the following negative error codes allowed by
* the &drm_mode_config_funcs.atomic_commit vfunc. When using helpers
* this callback is the only one which can fail an atomic commit,
* everything else must complete successfully.
*/
int (*prepare_fb)(struct drm_plane *plane,
struct drm_plane_state *new_state);
/**
* @cleanup_fb:
*
* This hook is called to clean up any resources allocated for the given
* framebuffer and plane configuration in @prepare_fb.
*
* This callback is used by the atomic modeset helpers and by the
* transitional plane helpers, but it is optional.
*/
void (*cleanup_fb)(struct drm_plane *plane,
struct drm_plane_state *old_state);
/**
* @atomic_check:
*
* Drivers should check plane specific constraints in this hook.
*
* When using drm_atomic_helper_check_planes() plane's @atomic_check
* hooks are called before the ones for CRTCs, which allows drivers to
* request shared resources that the CRTC controls here. For more
* complicated dependencies the driver can call the provided check helpers
* multiple times until the computed state has a final configuration and
* everything has been checked.
*
* This function is also allowed to inspect any other object's state and
* can add more state objects to the atomic commit if needed. Care must
* be taken though to ensure that state check and compute functions for
* these added states are all called, and derived state in other objects
* all updated. Again the recommendation is to just call check helpers
* until a maximal configuration is reached.
*
* This callback is used by the atomic modeset helpers and by the
* transitional plane helpers, but it is optional.
*
* NOTE:
*
* This function is called in the check phase of an atomic update. The
* driver is not allowed to change anything outside of the free-standing
* state objects passed-in or assembled in the overall &drm_atomic_state
* update tracking structure.
*
* RETURNS:
*
* 0 on success, -EINVAL if the state or the transition can't be
* supported, -ENOMEM on memory allocation failure and -EDEADLK if an
* attempt to obtain another state object ran into a &drm_modeset_lock
* deadlock.
*/
int (*atomic_check)(struct drm_plane *plane,
struct drm_plane_state *state);
/**
* @atomic_update:
*
* Drivers should use this function to update the plane state. This
* hook is called in-between the &drm_crtc_helper_funcs.atomic_begin and
* drm_crtc_helper_funcs.atomic_flush callbacks.
*
* Note that the power state of the display pipe when this function is
* called depends upon the exact helpers and calling sequence the driver
* has picked. See drm_atomic_helper_commit_planes() for a discussion of
* the tradeoffs and variants of plane commit helpers.
*
* This callback is used by the atomic modeset helpers and by the
* transitional plane helpers, but it is optional.
*/
void (*atomic_update)(struct drm_plane *plane,
struct drm_plane_state *old_state);
/**
* @atomic_disable:
*
* Drivers should use this function to unconditionally disable a plane.
* This hook is called in-between the
* &drm_crtc_helper_funcs.atomic_begin and
* drm_crtc_helper_funcs.atomic_flush callbacks. It is an alternative to
* @atomic_update, which will be called for disabling planes, too, if
* the @atomic_disable hook isn't implemented.
*
* This hook is also useful to disable planes in preparation of a modeset,
* by calling drm_atomic_helper_disable_planes_on_crtc() from the
* &drm_crtc_helper_funcs.disable hook.
*
* Note that the power state of the display pipe when this function is
* called depends upon the exact helpers and calling sequence the driver
* has picked. See drm_atomic_helper_commit_planes() for a discussion of
* the tradeoffs and variants of plane commit helpers.
*
* This callback is used by the atomic modeset helpers and by the
* transitional plane helpers, but it is optional.
*/
void (*atomic_disable)(struct drm_plane *plane,
struct drm_plane_state *old_state);
};
/**
* drm_plane_helper_add - sets the helper vtable for a plane
* @plane: DRM plane
* @funcs: helper vtable to set for @plane
*/
static inline void drm_plane_helper_add(struct drm_plane *plane,
const struct drm_plane_helper_funcs *funcs)
{
plane->helper_private = funcs;
}
drm/atomic-helper: nonblocking commit support Design ideas: - split up the actual commit into different phases, and have completions for each of them. This will be useful for the future when we want to interleave phases much more aggressively, for e.g. queue depth > 1. For not it's just a minimal optimization compared to current common nonblocking implementation patterns from drivers, which all stall for the entire commit to complete, including vblank waits and cleanups. - Extract a separate atomic_commit_hw hook since that's the part most drivers will need to overwrite, hopefully allowing even more shared code. - Enforce EBUSY seamntics by attaching one of the completions to the flip_done vblank event. Side benefit of forcing atomic drivers using these helpers to implement event handlign at least semi-correct. I'm evil that way ;-) - Ridiculously modular, as usual. - The main tracking unit for a commit stays struct drm_atomic_state, and the ownership rules for that are unchanged. Ownership still gets transferred to the driver (and subsequently to the worker) on successful commits. What is added is a small, per-crtc, refcounted structure to track pending commits called struct drm_crtc_commit. No actual state is attached to that though, it's purely for ordering and waiting. - Dependencies are implicitly handled by assuming that any CRTC part of &drm_atomic_state is a dependency, and that the current commit must wait for any commits to complete on those CRTC. This way drivers can easily add more depencies using drm_atomic_get_crtc_state(), which is very natural since in most case a dependency exists iff there's some bit of state that needs to be cross checked. Removing depencies is not possible, drivers simply need to be careful to not include every CRTC in a commit if that's not necessary. Which is a good idea anyway, since that also avoids ww_mutex lock contention. - Queue depth > 1 sees some prep work in this patch by adding a stall paramater to drm_atomic_helper_swap_states(). To be able to push commits entirely free-standing and in a deeper queue through the back-end the driver must not access any obj->state pointers. This means we need to track the old state in drm_atomic_state (much easier with the consolidated arrays), and pass them all explicitly to driver backends (this will be serious amounts of churn). Once that's done stall can be set to false in swap_states. v2: Dont ask for flip_done signalling when the CRTC is off and stays off: Drivers don't handle events in that case. Instead complete right away. This way future commits don't need to have special-case logic, but can keep blocking for the flip_done completion. v3: Tons of fixes: - Stall for preceeding commit for real, not the current one by accident. - Add WARN_ON in case drivers don't fire the drm event. - Don't double-free drm events. v4: Make legacy cursor not stall. v5: Extend the helper hook to cover the entire commit tail. Some drivers need special code for cleanup and vblank waiting, this makes it a bit more useful. Inspired by the rockchip driver. v6: Add WARN_ON to catch drivers who forget to send out the drm event. v7: Fixup the stalls in swap_state for real!! v8: - Fixup trailing whitespace, spotted by Maarten. - Actually wait for flip_done in cleanup_done, like the comment says we should do. Thanks a lot for Tomeu for helping with debugging this on. v9: Now with awesome kerneldoc! v10: Split out drm_crtc_commit tracking infrastructure. v: - Add missing static (Gustavo). - Split out the sync functions, only do the actual nonblocking logic in this patch (Maarten). Cc: Gustavo Padovan <gustavo.padovan@collabora.co.uk> Tested-by: Tomeu Vizoso <tomeu.vizoso@collabora.com> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Cc: Tomeu Vizoso <tomeu.vizoso@gmail.com> Cc: Daniel Stone <daniels@collabora.com> Tested-by: Liviu Dudau <Liviu.Dudau@arm.com> Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Testcase: igt/kms_flip/* Testcase: igt/kms_cursor* Testcase: igt/kms*plane* Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1465388359-8070-10-git-send-email-daniel.vetter@ffwll.ch
2016-06-08 20:19:02 +08:00
/**
* struct drm_mode_config_helper_funcs - global modeset helper operations
*
* These helper functions are used by the atomic helpers.
*/
struct drm_mode_config_helper_funcs {
/**
* @atomic_commit_tail:
*
* This hook is used by the default atomic_commit() hook implemented in
* drm_atomic_helper_commit() together with the nonblocking commit
* helpers (see drm_atomic_helper_setup_commit() for a starting point)
* to implement blocking and nonblocking commits easily. It is not used
* by the atomic helpers
*
* This function is called when the new atomic state has already been
* swapped into the various state pointers. The passed in state
* therefore contains copies of the old/previous state. This hook should
* commit the new state into hardware. Note that the helpers have
* already waited for preceeding atomic commits and fences, but drivers
* can add more waiting calls at the start of their implementation, e.g.
* to wait for driver-internal request for implicit syncing, before
* starting to commit the update to the hardware.
drm/atomic-helper: nonblocking commit support Design ideas: - split up the actual commit into different phases, and have completions for each of them. This will be useful for the future when we want to interleave phases much more aggressively, for e.g. queue depth > 1. For not it's just a minimal optimization compared to current common nonblocking implementation patterns from drivers, which all stall for the entire commit to complete, including vblank waits and cleanups. - Extract a separate atomic_commit_hw hook since that's the part most drivers will need to overwrite, hopefully allowing even more shared code. - Enforce EBUSY seamntics by attaching one of the completions to the flip_done vblank event. Side benefit of forcing atomic drivers using these helpers to implement event handlign at least semi-correct. I'm evil that way ;-) - Ridiculously modular, as usual. - The main tracking unit for a commit stays struct drm_atomic_state, and the ownership rules for that are unchanged. Ownership still gets transferred to the driver (and subsequently to the worker) on successful commits. What is added is a small, per-crtc, refcounted structure to track pending commits called struct drm_crtc_commit. No actual state is attached to that though, it's purely for ordering and waiting. - Dependencies are implicitly handled by assuming that any CRTC part of &drm_atomic_state is a dependency, and that the current commit must wait for any commits to complete on those CRTC. This way drivers can easily add more depencies using drm_atomic_get_crtc_state(), which is very natural since in most case a dependency exists iff there's some bit of state that needs to be cross checked. Removing depencies is not possible, drivers simply need to be careful to not include every CRTC in a commit if that's not necessary. Which is a good idea anyway, since that also avoids ww_mutex lock contention. - Queue depth > 1 sees some prep work in this patch by adding a stall paramater to drm_atomic_helper_swap_states(). To be able to push commits entirely free-standing and in a deeper queue through the back-end the driver must not access any obj->state pointers. This means we need to track the old state in drm_atomic_state (much easier with the consolidated arrays), and pass them all explicitly to driver backends (this will be serious amounts of churn). Once that's done stall can be set to false in swap_states. v2: Dont ask for flip_done signalling when the CRTC is off and stays off: Drivers don't handle events in that case. Instead complete right away. This way future commits don't need to have special-case logic, but can keep blocking for the flip_done completion. v3: Tons of fixes: - Stall for preceeding commit for real, not the current one by accident. - Add WARN_ON in case drivers don't fire the drm event. - Don't double-free drm events. v4: Make legacy cursor not stall. v5: Extend the helper hook to cover the entire commit tail. Some drivers need special code for cleanup and vblank waiting, this makes it a bit more useful. Inspired by the rockchip driver. v6: Add WARN_ON to catch drivers who forget to send out the drm event. v7: Fixup the stalls in swap_state for real!! v8: - Fixup trailing whitespace, spotted by Maarten. - Actually wait for flip_done in cleanup_done, like the comment says we should do. Thanks a lot for Tomeu for helping with debugging this on. v9: Now with awesome kerneldoc! v10: Split out drm_crtc_commit tracking infrastructure. v: - Add missing static (Gustavo). - Split out the sync functions, only do the actual nonblocking logic in this patch (Maarten). Cc: Gustavo Padovan <gustavo.padovan@collabora.co.uk> Tested-by: Tomeu Vizoso <tomeu.vizoso@collabora.com> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Cc: Tomeu Vizoso <tomeu.vizoso@gmail.com> Cc: Daniel Stone <daniels@collabora.com> Tested-by: Liviu Dudau <Liviu.Dudau@arm.com> Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Testcase: igt/kms_flip/* Testcase: igt/kms_cursor* Testcase: igt/kms*plane* Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1465388359-8070-10-git-send-email-daniel.vetter@ffwll.ch
2016-06-08 20:19:02 +08:00
*
* After the atomic update is committed to the hardware this hook needs
* to call drm_atomic_helper_commit_hw_done(). Then wait for the upate
* to be executed by the hardware, for example using
* drm_atomic_helper_wait_for_vblanks(), and then clean up the old
* framebuffers using drm_atomic_helper_cleanup_planes().
*
* When disabling a CRTC this hook _must_ stall for the commit to
* complete. Vblank waits don't work on disabled CRTC, hence the core
* can't take care of this. And it also can't rely on the vblank event,
* since that can be signalled already when the screen shows black,
* which can happen much earlier than the last hardware access needed to
* shut off the display pipeline completely.
*
* This hook is optional, the default implementation is
* drm_atomic_helper_commit_tail().
*/
void (*atomic_commit_tail)(struct drm_atomic_state *state);
};
#endif