We skip loading the tile buffer from memory if the job has flagged
a clear (job->clears) for the buffer, however, this only tracks
clears emitted via the TLB. In some cases we may need to fallback
to clearing with a draw call, in which case we also want to skip
the load.
Reviewed-by: Jose Maria Casanova Crespo <jmcasanova@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30205>
We need to call iris_wait_syncobj() on both release and debug builds,
so take it out of the assert() call. Only assert the result.
With this patch, gnome-session finally works for me. Also steam.
Fixes: 665d30b544 ("iris: Wait for drm_xe_exec_queue to be idle before destroying it")
Reviewed-by: José Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30195>
Xe KMD renamed XE_PERF to XE_OBSERVATION to better match with Intel
specification and avoid confusion.
This uAPI rename will land in the same kernel version that added
the uAPI being renamed.
There is no uAPI change, just renames.
Sync xe_drm.h with 63347fe031e3 ("drm/xe/uapi: Rename xe perf layer as xe observation layer").
Acked-by: Caio Oliveira <caio.oliveira@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30027>
* swrast allocates images aligned to 64x64 tiles, which results in images
that are larger than the window. PutImage requests must be clamped on
the y-axis to avoid uploading/damaging out-of-bounds regions
* winsys coords are y-inverted
cc: mesa-stable
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29910>
this way we can use shader-db report.py to analyze shader-db changes for zink
with the vk driver of our choosing. requires corresponding report.py relaxations
to be useful, the idea was to change zink a bit and change report.py a bit and
meet in the middle with something useful for arbitrary vulkan drivers.
the output isn't too pretty but it works,
synthetic example on nvk:
Instruction count HURT: shaders/glmark/1-1.shader_test vertex: 42 -> 43 (2.38%)
total Instruction count in shared programs: 7135 -> 7136 (0.01%)
Instruction count in affected programs: 42 -> 43 (2.38%)
helped: 0
HURT: 1
total Code Size in shared programs: 114160 -> 114160 (0.00%)
Code Size in affected programs: 0 -> 0
helped: 0
HURT: 0
total Number of GPRs in shared programs: 2677 -> 2677 (0.00%)
Number of GPRs in affected programs: 0 -> 0
helped: 0
HURT: 0
total SLM Size in shared programs: 0 -> 0
SLM Size in affected programs: 0 -> 0
helped: 0
HURT: 0
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30194>
the newline is implicit for util debug messages. this fixes shaderdb output
being half blank lines and brings zink in line with hardware mesa drivers.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30194>
Use common db_alignment to calculate dpb_size for DPB_MAX_RES,
DPB_DYNAMIC_TIER_1 and DPB_DYNAMIC_TIER_2. This makes the db_pitch
in sync with all DPB types.
Remove the VCN5 hack of using 256 for H264 as 64 works.
Remove redundant codes for width and height as they were calculated
at the beginning in calc_dpb_size().
Reviewed-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: David (Ming Qiang) Wu <David.Wu3@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30186>
spirv_info_h was added to idep_vulkan_lite_runtime_headers but it's also
needed for building libvulkan_lite_runtime. Without this patch, from
a clean meson setup (tested with: -Dvulkan-drivers=freedreno -Dgallium-drivers= -Dplatforms=wayland)
ninja -C build src/vulkan/runtime/libvulkan_lite_runtime.a
In file included from ../src/vulkan/runtime/vk_physical_device.c:24:
../src/vulkan/runtime/vk_physical_device.h:32:10: fatal error: compiler/spirv/spirv_info.h: No such file or directory
Fixes: 1759c0eba7 ("vulkan: add helper to fill out spirv caps automatically")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30189>
Commit 94989b45a5 ("anv,driconf: Add fake non device local memory WA
for Total War: Warhammer 3") implemented a workaround to make
Warhammer 3 work on ADL, but the game still doesn't work on LNL, which
uses xe.ko, and MTL, which uses i915.ko: it still fails at launch
claiming it couldn't allocate memory.
So in this implementation, instead of clearing DEVICE_LOCAL_BIT we
just duplicate our memory types, one having the bit and one not
having.
v2:
- Check for VK_MAX_MEMORY_TYPES (José)
- Invert the order of the memory types (José)
- Fix white space issue (José)
v3:
- Comment our non-spec-compliance (José)
- Remove useless lines (José)
Link: https://gitlab.freedesktop.org/mesa/mesa/-/issues/8721
Fixes: 94989b45a5 ("anv,driconf: Add fake non device local memory WA for Total War: Warhammer 3")
Reviewed-by: José Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30162>
This decoration can now be used in Vulkan with
VK_KHR_shader_float_controls2.
Acked-by: Iván Briano <ivan.briano@intel.com>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30191>
VK_FORMAT_R10X6_UNORM_PACK16 and VK_FORMAT_R10X6G10X6_UNORM_2PACK16
were previously disabled because they were failing some CTS tests.
Re-enable them now as they are no longer causing crashes.
Signed-off-by: Valentine Burley <valentine.burley@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30157>
In practice these are equal but the old code was semantically wrong: that
dimension is "sources" not "components". Use the correct #define.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Suggested-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30214>
Some apps (like firefox and chromium) seem to like to create textures
with internal format GL_RGBA8 but then upload GL_BGRA to them. If the
app were a bit more clever it could hit the memcpy path. But fallback
sw conversion (convert_ubytes()) is slow. If we are going to have to
make this extra copy, do it on the gpu.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30064>
The blitter cannot do the needed swizzle gymnastics for blitting to
luminance and/or alpha formats. So fallback to the 3d path.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30064>
A sequence like:
1) set_framebuffer_state(A)
2) clears and/or draws
3) set_framebuffer_state(B)
4) context_flush(ASYNC)
would result in the fence for batch B being returned, instead of batch A.
Resulting in a fence that signals too early.
Instead, in context_flush() find the most recently modified batch, so
that in the example above batch A would be flushed.
There is one wrinkle, that we want to avoid a dependency loop. So the
'last_batch' can actually be a batch that depends on the most recently
modified batch. But in either case, the batch that is used for the
returned fence is one that directly or indirectly depends on all other
batches associated with the context.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30064>
These are for CPU access, so no point in having additional staging blit
to convert from/to linear. Depth formats are an exception, as normally
they cannot be linear.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30064>
Now that there is a util_blitter_clear_depth_stencil() helper, we can
use that to implement stencil blits in the fallback path.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30064>
The other callers were already doing the prepatory stencil clear before
calling util_blitter_stencil_fallback(). Doing the clear in the driver
avoids recursively entering u_blitter if the driver uses
util_blitter_clear_depth_stencil() to implement ->clear_depth_stencil().
Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30064>
It's 58, not 59. The original 58 feature was not
ready so it was dropped.
Fixes: 0bb83a4060 ("ac/surface: finish display DCC for gfx12")
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30212>
useful for shader-db, this isn't as simple as dividing the code size so it's
worth reporting.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: M Henning <drawoc@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30136>
This isn't as simple as dividing so we want a real shader info property for nvk
to consume. Plumb one through.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: M Henning <drawoc@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30136>
this is required for nvk to create a heap. fixes zink+nvk+drm-shim:
run: ../src/gallium/drivers/zink/zink_screen.c:3371: zink_internal_create_screen: Assertion `i == ZINK_HEAP_HOST_VISIBLE_COHERENT_CACHED || i == ZINK_HEAP_DEVICE_LOCAL_LAZY || i == ZINK_HEAP_DEVICE_LOCAL_VISIBLE' failed.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: M Henning <drawoc@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30136>
