fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2026-05-24 19:18:11 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 1
209138 commits 66 branches 936 tags 995 MiB
Commit graph

193562 commits

Author SHA1 Message Date
Boyuan Zhang
9b158d0512 ci/fluster: remove 3 pass cases resulted by gaps_in_frame
Some checks are pending
macOS-CI / macOS-CI (dri) (push) Waiting to run
Details
macOS-CI / macOS-CI (xlib) (push) Waiting to run
Details 
These 3 failed tests are passing now by enabling the gaps_in_frame feature.
Therefore, remove all of them.

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36056>
 2025-07-17 12:44:51 +00:00
Boyuan Zhang
a63e5f015e radeon/vcn: add gaps_in_frame flag to h264 sps 
Implement gaps_in_frame_num_value_allowed_flag in h264 msg buffer.
Replace hardcoded flag values with defines.

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: David Rosca <david.rosca@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36056>
 2025-07-17 12:44:51 +00:00
Boyuan Zhang
a7567ec6f4 frontends/va: get gaps_in_frame for h264 dec 
Get gaps_in_frame_num_value_allowed_flag value from Va-api interface.

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: David Rosca <david.rosca@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36056>
 2025-07-17 12:44:51 +00:00
Boyuan Zhang
587102196c pipe: add gaps_in_frame for h264 
Add gaps_in_frame_num_value_allowed_flag to h264 sps struct.

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: David Rosca <david.rosca@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36056>
 2025-07-17 12:44:51 +00:00
Mike Blumenkrantz
ce09d80698 gallium/hud: set the framebuffer texture when drawing 
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13545

Fixes: 2eb45daa9c ("gallium: de-pointerize pipe_surface")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36167>
 2025-07-17 11:48:51 +00:00
Valentine Burley
34c969f767 freedreno,zink+tu/ci: Document Piglit bug
Some checks are pending
macOS-CI / macOS-CI (dri) (push) Waiting to run
Details
macOS-CI / macOS-CI (xlib) (push) Waiting to run
Details 
Document the Piglit bug exposed in 08b522d21e
("glsl: check against varying limits using NIR shader_info after nir_opt_varyings")

Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36027>
 2025-07-17 09:07:28 +00:00
Valentine Burley
bcc1a96a73 zink/ci: Run full zink-tu-a618 job pre-merge 
We can use one more device to increase parallelism, which allows us to
run the full nightly job pre-merge.

Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36027>
 2025-07-17 09:07:28 +00:00
Valentine Burley
0b677027a8 freedreno/ci: Re-enable a618-gl job 
Re-enables the previously disabled GL and EGL jobs on a618, combining
them into a single a618-gl job.

Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36027>
 2025-07-17 09:07:28 +00:00
Valentine Burley
2980a5913a turnip/ci: Increase coverage of a618-vk, reduce parallelism 
Thanks to the optimizations in the previous commits and further
increasing tests_per_group, we can now run more tests in a618-vk
while also reducing its parallelism.

Dropping `DEQP_FRACTION: 2` and only increasing the `fraction` of
the basic test set to 3 results in broader overall coverage.

Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36027>
 2025-07-17 09:07:26 +00:00
Valentine Burley
7ec9404971 freedreno/ci: Increase concurrency for a618 jobs 
The a618 jobs run on two different 8 thread devices, which is the optimal
value for FDO_CI_CONCURRENT on most of their jobs.

Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36027>
 2025-07-17 09:07:25 +00:00
Yiwei Zhang
afdb2c64de venus/ci: udpate expectations from venus-lavapipe-full runs
Some checks are pending
macOS-CI / macOS-CI (dri) (push) Waiting to run
Details
macOS-CI / macOS-CI (xlib) (push) Waiting to run
Details 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36180>
 2025-07-16 22:56:43 +00:00
Yiwei Zhang
442f242a49 venus: requests whole blob mem size for non-dedicated import 
For non-dedicated import, later import could have varied size against
the same bo, so we just request the whole blob mem size for those.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36128>
 2025-07-16 22:05:11 +00:00
Yiwei Zhang
b32a8d17a2 venus/virtgpu: use size zero to request mapping the entire blob mem 
For import, requested size might be smaller than the whole blob mem
size. Currently, venus virtgpu backend tracks the initial import size
and sanity checks with later re-import size. So let's add a way to
request whole blob mem size

No need vtest backend changes since dma_buf import is not supported.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36128>
 2025-07-16 22:05:11 +00:00
Yiwei Zhang
768375fede venus: refactor ahb import interface to take whole alloc info 
Prepare for adopting common runtime so that we don't leave unused vars
after dropping codes from vn_AllocateMemory. It's trying to minimize
code changes in the venus adoption since that'd be huge.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36128>
 2025-07-16 22:05:11 +00:00
Yiwei Zhang
e27ad2a021 venus: drop force_unmappable hack 
No longer needed since the vkMapMemory would fail properly instead of
failing the external memory import.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36128>
 2025-07-16 22:05:11 +00:00
Yiwei Zhang
6c8f075fb3 venus/virtgpu: drop mappable if blob size is smaller than requested 
Instead of failing the import, drop the mappable bit if blob size is
smaller than the requested alloc size, which defers the error to the
intended vkMapMemory api call if the client app requests Vulkan mapping
from the imported external memory.

Normally this won't occur if the app obeys the spec with a properly
implemented blob mem allocator. However, legacy allocators like minigbm
virtgpu_virgl backend could allocate via virgl w/o knowing the real blob
size from the guest side. The unsatified cases are external gralloc
images, which won't be mapped in any meaningful way after being
imported. This is to prepare dropping the force_unmappable workaround,
and to further prepare adopting common Android runtime.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36128>
 2025-07-16 22:05:10 +00:00
Eric Engestrom
e39a41e104 gfxstream: move variables into the #ifdef that uses them 
../src/gfxstream/guest/vulkan_enc/ResourceTracker.cpp:7585:14: warning: unused variable ‘actualFormat’ [-Wunused-variable]
 7585 |     VkFormat actualFormat = imageInfo.createInfo.format;
      |              ^~~~~~~~~~~~

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36171>
 2025-07-16 20:30:40 +00:00
Alyssa Rosenzweig
2308960bed treewide: use nir_mov_scalar
Some checks are pending
macOS-CI / macOS-CI (dri) (push) Waiting to run
Details
macOS-CI / macOS-CI (xlib) (push) Waiting to run
Details 
Via Coccinelle patch:

    @@
    expression builder, scalar;
    @@

    -nir_channel(builder, scalar.def, scalar.comp)
    +nir_mov_scalar(builder, scalar)

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36142>
 2025-07-16 18:59:16 +00:00
Alyssa Rosenzweig
186db0ebfe nir: add nir_mov_scalar helper 
I keep reaching for this helper but it doesn't exist. So I fixed that.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36142>
 2025-07-16 18:59:16 +00:00
Alyssa Rosenzweig
a5e9669a78 hk: only pass sampler heap if needed 
I'm guessing the hardware needs to prefetch the whole sampler heap, so if we're
not gonna use it, let's omit it. I don't know if this helps, but it can't hurt.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36127>
 2025-07-16 18:27:22 +00:00
Alyssa Rosenzweig
74c32c2357 hk: optimize desc set addr push 
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36127>
 2025-07-16 18:27:22 +00:00
Alyssa Rosenzweig
642c6c6f62 hk,agx: promote bindless samplers 
via the bindless_sampler_agx intrinsic.

Totals from 29771 (55.11% of 54019) affected shaders:
MaxWaves: 28934080 -> 28938304 (+0.01%); split: +0.02%, -0.00%
Instrs: 16623874 -> 16369120 (-1.53%); split: -1.54%, +0.01%
CodeSize: 117532138 -> 115994992 (-1.31%); split: -1.32%, +0.01%
Spills: 12721 -> 12652 (-0.54%); split: -0.72%, +0.17%
Fills: 6733 -> 6636 (-1.44%); split: -1.96%, +0.52%
Scratch: 132994 -> 132712 (-0.21%); split: -0.22%, +0.01%
ALU: 13054253 -> 12803059 (-1.92%); split: -1.93%, +0.01%
FSCIB: 13054138 -> 12802912 (-1.92%); split: -1.94%, +0.01%
IC: 3916012 -> 3915588 (-0.01%); split: -0.01%, +0.00%
GPRs: 2290907 -> 2289519 (-0.06%); split: -0.07%, +0.01%
Uniforms: 6794773 -> 6696943 (-1.44%); split: -1.44%, +0.00%
Preamble instrs: 6953594 -> 7024455 (+1.02%); split: -0.37%, +1.39%

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36127>
 2025-07-16 18:27:21 +00:00
Alyssa Rosenzweig
49f042c5e8 hk: plumb sampler state counts 
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36127>
 2025-07-16 18:27:21 +00:00
Alyssa Rosenzweig
4a9ae295cd agx: report sampler state count 
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36127>
 2025-07-16 18:27:20 +00:00
Alyssa Rosenzweig
2340541eca asahi: drop image heap decode 
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36127>
 2025-07-16 18:27:20 +00:00
Alyssa Rosenzweig
5598d6aa10 hk: drop image heap 
now everything is just inline in the set. saves 24MiB. fossil changes from start
of series to here:

Totals from 45270 (83.80% of 54019) affected shaders:
MaxWaves: 44151552 -> 44167808 (+0.04%); split: +0.05%, -0.01%
Instrs: 22630059 -> 22599471 (-0.14%); split: -0.31%, +0.17%
CodeSize: 160981702 -> 160987912 (+0.00%); split: -0.29%, +0.29%
Spills: 66105 -> 66276 (+0.26%); split: -1.09%, +1.35%
Fills: 43894 -> 43987 (+0.21%); split: -1.82%, +2.03%
Scratch: 375176 -> 374896 (-0.07%); split: -0.20%, +0.13%
ALU: 17656552 -> 17640750 (-0.09%); split: -0.29%, +0.20%
FSCIB: 17601880 -> 17586092 (-0.09%); split: -0.29%, +0.20%
IC: 4990178 -> 4995334 (+0.10%); split: -0.05%, +0.15%
GPRs: 3417893 -> 3413589 (-0.13%); split: -0.17%, +0.04%
Uniforms: 9260606 -> 9474203 (+2.31%); split: -0.22%, +2.52%
Preamble instrs: 9679627 -> 9282629 (-4.10%); split: -4.22%, +0.12%

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36127>
 2025-07-16 18:27:20 +00:00
Alyssa Rosenzweig
1ea1b2d360 hk: stop reserving uniform for image heap 
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36127>
 2025-07-16 18:27:19 +00:00
Alyssa Rosenzweig
97ff0ada8b hk: stop pushing image heap 
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36127>
 2025-07-16 18:27:19 +00:00
Alyssa Rosenzweig
5a6c5eed65 hk: embed texture desc in set 
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36127>
 2025-07-16 18:27:18 +00:00
Alyssa Rosenzweig
98aad84d73 hk: push descriptor set addresses 
saves an indirection and sets us up for more goodness.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36127>
 2025-07-16 18:27:18 +00:00
Alyssa Rosenzweig
899f001936 hk: dedupe hk_buffer_view_descriptor 
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36127>
 2025-07-16 18:27:17 +00:00
Alyssa Rosenzweig
24c708564f nir: add bindless_sampler_agx intrinsic 
to facilitate pushing on AGX.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36127>
 2025-07-16 18:27:17 +00:00
Alyssa Rosenzweig
58cc66238a nir/opt_preamble: add sampler class 
AGX will use.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36127>
 2025-07-16 18:27:17 +00:00
Bo Hu
f146a3a13a gfxstream: update codegen for event save and load
Some checks are pending
macOS-CI / macOS-CI (dri) (push) Waiting to run
Details
macOS-CI / macOS-CI (xlib) (push) Waiting to run
Details 
Event save and load codegen.

Reviewed-by: Aaron Ruby <aruby@qnx.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36047>
 2025-07-16 17:03:43 +00:00
Serdar Kocdemir
650691e93c gfxstream: fix warnings about unused parameters 
Fix warnings.

Reviewed-by: Aaron Ruby <aruby@qnx.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36047>
 2025-07-16 17:03:43 +00:00
Dave Airlie
da61149b2a nak: disable imma 8x8x16 on Blackwell+ 
It's not supported anymore

Fixes: 669c8a5145 ("nvk: Advertise VK_KHR_cooperative_matrix")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36163>
 2025-07-16 14:53:50 +00:00
Faith Ekstrand
4bb67cacba nak: Wire up the mma predicate on Hopper+ 
Fixes: 90438bae51 ("nir: Add NVIDIA-specific muladd intrinsics")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36163>
 2025-07-16 14:53:50 +00:00
David Rosca
bc11dc72c1 Revert "radeonsi/vcn: Stop using stream handle for decode" 
Caused issues on VCN5.

This reverts commit 46d5926d83.

Fixes: 46d5926d83 ("radeonsi/vcn: Stop using stream handle for decode")
Reviewed-by: Leo Liu <leo.liu@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36154>
 2025-07-16 13:25:31 +00:00
Eric Engestrom
8d61c2751f ci: move script: override from .piglit-traces-test to llvmpipe-traces
Some checks are pending
macOS-CI / macOS-CI (dri) (push) Waiting to run
Details
macOS-CI / macOS-CI (xlib) (push) Waiting to run
Details 
This is already being set as needed everywhere else, and would cause
issues in future work.

Use the relative `install/` path for `HWCI_TEST_SCRIPT` as that's
supported by both HW runners and FDo runners.
A separate MR will fix the `/install/` vs `install/` mess.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36145>
 2025-07-16 12:54:06 +00:00
Eric Engestrom
5ebe02db30 ci/piglit: provide default results file name 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36145>
 2025-07-16 12:54:06 +00:00
Eric Engestrom
9c42e66de1 ci/piglit: provide default device name 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36145>
 2025-07-16 12:54:06 +00:00
Eric Engestrom
3dc28c9e55 zink+radv/ci: document recent flakes 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36160>
 2025-07-16 12:39:53 +00:00
Eric Engestrom
8808f039cc freedreno/ci: document recent flakes 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36160>
 2025-07-16 12:39:53 +00:00
Eric Engestrom
e1277100df zink+nvk/ci: fix mistake in yesterday's crash->fail improvement update 
Fixes: e703847410 ("zink+nvk/ci: document crash->fail change from !36031")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36160>
 2025-07-16 12:39:53 +00:00
Eric Engestrom
6f9fcfb0ad nvk/ci: document vkd3d regression 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36160>
 2025-07-16 12:39:53 +00:00
Eric R. Smith
65bcae096a panfrost: fix SSA register allocation 
We were allocating a fixed number of temporary registers; this isn't
always enough, and in fact we should have calculated the number of
temporaries required.

Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com>
Fixes: 6c64ad934f ("panfrost: spill registers in SSA form")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36135>
 2025-07-16 12:16:50 +00:00
Georg Lehmann
497f607c8e radv/nir/lower_cmat: vectorize GFX11 B -> ACC conversion 
Foz-DB Navi31:
Totals from 7 out of 14 FSR4 shaders:
MaxWaves: 50 -> 52 (+4.00%)
Instrs: 44951 -> 44516 (-0.97%); split: -1.00%, +0.03%
CodeSize: 309176 -> 305500 (-1.19%); split: -1.23%, +0.04%
VGPRs: 1464 -> 1416 (-3.28%)
SpillVGPRs: 188 -> 92 (-51.06%)
Scratch: 24064 -> 11776 (-51.06%)
Latency: 171318 -> 163663 (-4.47%); split: -4.51%, +0.04%
InvThroughput: 178796 -> 178956 (+0.09%); split: -0.04%, +0.13%
VClause: 769 -> 730 (-5.07%); split: -6.50%, +1.43%
Copies: 3149 -> 3261 (+3.56%); split: -1.21%, +4.76%
PreVGPRs: 1607 -> 1467 (-8.71%)
VALU: 37715 -> 37744 (+0.08%); split: -0.11%, +0.18%
SALU: 754 -> 753 (-0.13%)
VMEM: 2813 -> 2621 (-6.83%)
VOPD: 1674 -> 1685 (+0.66%); split: +1.55%, -0.90%

Acked-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36115>
 2025-07-16 11:46:52 +00:00
Georg Lehmann
7546169e1c radv/nir/lower_cmat: vectorize GFX11 ACC -> B conversion 
Foz-DB Navi31:
Totals from 10 out of 14 FSR4 shaders:
Instrs: 64204 -> 60749 (-5.38%)
CodeSize: 439052 -> 417668 (-4.87%)
SpillVGPRs: 186 -> 188 (+1.08%)
Scratch: 23808 -> 24064 (+1.08%)
Latency: 208878 -> 202903 (-2.86%)
InvThroughput: 232898 -> 225688 (-3.10%)
VClause: 902 -> 907 (+0.55%); split: -1.55%, +2.11%
Copies: 6418 -> 3762 (-41.38%)
Branches: 55 -> 37 (-32.73%)
PreSGPRs: 297 -> 298 (+0.34%)
PreVGPRs: 2299 -> 2303 (+0.17%)
VALU: 54762 -> 51489 (-5.98%)
SALU: 956 -> 938 (-1.88%)
VMEM: 3469 -> 3473 (+0.12%)
VOPD: 3895 -> 2126 (-45.42%)

Acked-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36115>
 2025-07-16 11:46:52 +00:00
Georg Lehmann
d672737372 nir,aco: add byte_perm_amd 
Acked-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36115>
 2025-07-16 11:46:52 +00:00
Georg Lehmann
56d93c40ea radv/nir/lower_cmat: convert matrix use in smaller type 
Less conversions, and less data to move around.

Foz-DB Navi31:
Totals from 10 out of 14 FSR4 shaders:
Instrs: 65443 -> 64204 (-1.89%); split: -1.93%, +0.04%
CodeSize: 441884 -> 439052 (-0.64%); split: -1.21%, +0.57%
Latency: 213374 -> 208878 (-2.11%); split: -2.17%, +0.07%
InvThroughput: 236922 -> 232898 (-1.70%); split: -1.77%, +0.08%
VClause: 935 -> 902 (-3.53%); split: -3.74%, +0.21%
Copies: 5064 -> 6418 (+26.74%); split: -13.35%, +40.09%
Branches: 54 -> 55 (+1.85%)
VALU: 55700 -> 54762 (-1.68%); split: -1.85%, +0.16%
VOPD: 3459 -> 3895 (+12.60%); split: +16.88%, -4.28%

Acked-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36115>
 2025-07-16 11:46:52 +00:00