OpenHarmony/third_party_mesa3d
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
200,131 Commits 9 Branches 1 Tag
3c03bae20c0e609f22e53ec7c1f49338034fc1ea
Commit Graph

200131 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Valentine Burley
4fd4358ba9 ci: Add Valentine to the restricted traces access list 
Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32864>
 2025-01-09 12:24:34 +00:00
Valentine Burley
372973cd2a zink/ci: Increase zink-anv-tgl parallelism to 4 
Thanks to the migrations, we now have enough 1160g7-volteer DUTs
to increase the parallelism of pre-merge zink TGL testing. This
allows us to reduce the fraction of Piglit tests and introduce
fractional GLESCTS testing.

Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32864>
 2025-01-09 12:24:34 +00:00
Valentine Burley
1f34bc5e42 zink/ci: Migrate the two TGL traces jobs to 1130g7-volteer 
Move zink-anv-tgl-traces and zink-anv-tgl-traces-restricted to
the smaller 1130g7-volteer DUT. These jobs are quick and short,
allowing us to use the 1160g7-volteer device for more
performance-sensitive tasks.

Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32864>
 2025-01-09 12:24:34 +00:00
Valentine Burley
288249811d anv/ci: Increase anv-tgl-angle parallelism to 2 
We have enough DUTs available, so increase parallelism to ensure
that we stay within the 10-minute time limit.

Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32864>
 2025-01-09 12:24:34 +00:00
Valentine Burley
71704ac2c6 virgl/ci: Migrate the two iris jobs to 1130g7-volteer 
Both virgl-iris-traces and virgl-iris-traces-performance jobs are
currently disabled for being broken, but we'll want to use the
smaller volteer DUT for them when they are re-enabled.

Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32864>
 2025-01-09 12:24:34 +00:00
Valentine Burley
c71a7ffc02 virgl/ci: Disable virgl-iris-traces-performance 
This job has never passed a run in a long time. It fails with
the following error when triggered:

head: cannot open '/dev/dri/renderD128' for reading: No such file or directory

Disable it until it's fixed.

Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32864>
 2025-01-09 12:24:34 +00:00
Samuel Pitoiset
f09f31d093 ac/nir: fix a comment typo in load_subgroup_id_lowered() 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32940>
 2025-01-09 08:02:19 +00:00
Samuel Pitoiset
44ba856089 ac/nir: fix lowering subgroup ID for compute shaders on GFX12 
This is lowered in backend compilers (LLVM or ACO) because it needs
to access ttmp registers which aren't exposed to NIR.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32940>
 2025-01-09 08:02:19 +00:00
Samuel Pitoiset
bc1374355b radv: program DB_RENDER_OVERRIDE correctly on GFX12 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32941>
 2025-01-09 07:39:23 +00:00
Vinson Lee
83809f06a7 intel/elk: Fix assert with side effect 
Fix defect reported by Coverity Scan.

Side effect in assertion (ASSERT_SIDE_EFFECT)
assert_side_effect: Argument ++eot_count of assert() has a side effect.
The containing function might work differently in a non-debug build.

Fixes: ebd6738260 ("intel/elk/chv: Implement WaClearArfDependenciesBeforeEot")
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Caio Oliveira <caio.oliveira@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32884>
 2025-01-09 04:07:42 +00:00
Vinson Lee
9b0fc62221 vulkan: Add missing va_end 
Fix defect reported by Coverity Scan.

Missing varargs init or cleanup (VARARGS)
missing_va_end: va_end was not called for ap.

Fixes: f8b584d6a5 ("vulkan/runtime,radv: Add shared BVH building framework")
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32858>
 2025-01-08 19:18:32 -08:00
Aleksi Sapon
60e97e16aa meson: fix finding Python on Windows 
On Windows, a venv has no versioned aliased to `python`.
`python3` will never be found from a venv.

Reviewed-by: Dylan Baker <None>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32934>
 2025-01-08 22:47:42 +00:00
David Heidelberg
5774f2c66e drm-uapi: update licenses statement 
Replace deprecated GPL-1.0+ with clear GPL-1.0-or-later.

Cc: mesa-stable
Signed-off-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32447>
 2025-01-08 20:37:51 +00:00
David Heidelberg
cc85d0c156 licenses: add missing licenses 
Cc: mesa-stable
Signed-off-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32447>
 2025-01-08 20:37:51 +00:00
David Heidelberg
fa2fddbf3a c11: use SPDX-License-Identifier header 
Cc: mesa-stable
Signed-off-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32447>
 2025-01-08 20:37:51 +00:00
David Heidelberg
c22d640fe9 docs: move license(s) to licenses directory 
Cc: mesa-stable
Signed-off-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32447>
 2025-01-08 20:37:51 +00:00
David Heidelberg
6f08f921bf docs: Clarify project name and include Mesa3D 
Mesa3D is commonly used, as evidenced by the use of the Mesa3D.org domain.

Additionally, it is unnecessary to advise against using "MesaGL"
since we do not use it ourselves.

Cc: mesa-stable
Signed-off-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32447>
 2025-01-08 20:37:51 +00:00
David Heidelberg
69849bc4d1 docs: remove deprecated component list and licenses 
Cc: mesa-stable
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/12180
Signed-off-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32447>
 2025-01-08 20:37:51 +00:00
Matt Turner
89da5a9626 intel/decoder: Avoid duplicate symbols when expat is not available 
Fixes: 0669210ef4 ("intel/decoder: Add ELK support")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/12335
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32732>
 2025-01-08 18:58:35 +00:00
Eric Engestrom
731138386a docs/release-calendar: fixup sed fail 
Fixes: 1a41743764 ("docs/release-calendar: add 25.0 branchpoint and RCs schedule")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32949>
 2025-01-08 17:39:58 +00:00
Eric R. Smith
d0a95541fe panvk: re-enable fragmentStoresAndAtomics for v10 
This feature actually worked, but accidentally triggered some
CTS failures due do an unrelated sample_mask bug. That bug
is fixed in the previous commits in this series, so re-enable
fragmentStoresAndAtomics.

There are two spurious CI failures created by this change; one is in
a ycbcr test (which we're still working on) and the other is a CTS
bug (it's using a feature we don't advertise).

Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32879>
 2025-01-08 16:45:21 +00:00
Eric R. Smith
03a1a5009f panfrost: ensure sample_mask is written before color 
The ATEST instruction needs sample_mask as an input, but if the
shader writes to color before sample_mask we could emit them
in the wrong order. Fix this in pan_nir_lower_zs_store by
deferring the color write until after the sample_mask write.

Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Benjamin Lee <benjamin.lee@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32879>
 2025-01-08 16:45:21 +00:00
Eric R. Smith
115eb1d5f3 panfrost: Allow ATEST input to be a FAU index 
We have an assert requiring that the sample_mask source for
ATEST be a register. However, the sample mask code can
sometimes produce a FAU index, and this actually works
correctly (tested with the vk and gl CTSes).

Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32879>
 2025-01-08 16:45:21 +00:00
Eric Engestrom
1a41743764 docs/release-calendar: add 25.0 branchpoint and RCs schedule 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32948>
 2025-01-08 16:36:46 +00:00
Alyssa Rosenzweig
d9b4867e2a nir/lower_robust_access: fix robustness with atomic swap 
this was missed in the original v3d pass, and then the common code port
inherited the bug. (so strictly this fix "should" be backported even farther
back but it won't apply before the Fixes here, and I don't think we do LTS that
far back anyway).

in theory this should fix a corner case with robustness on the gl (but not
vulkan, at least for apple) drivers on broadcom & apple.

Fixes: f0fb8d05e3 ("nir: Add nir_lower_robust_access pass")
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32907>
 2025-01-08 15:59:05 +00:00
Alyssa Rosenzweig
7a4469681e nir: pass a callback to nir_lower_robust_access 
rather than try to enumerate everything a driver might want with an unmanageable
collection of booleans, just do a filter callback + data. this ends up simpler
overall, and will allow Intel to use this pass for just 64-bit images without
needing to add even more booleans.

while we're churning the pass signature, also do a quick port to
nir_shader_intrinsics_pass

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com> [NIR and V3D]
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32907>
 2025-01-08 15:59:05 +00:00
Rhys Perry
8ac4744706 aco/tests: fix skip_lines=True with remaining characters in matches 
If the remaining character check fails, we should try a later line if
skip_lines=True. So the check has to be done earlier.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32902>
 2025-01-08 15:28:37 +00:00
Lucas Stach
8a55de3338 ci/etnaviv: drop GC2000 flat shading fails 
Now that the varying use is properly set for color varyings, tests
related to flatshading are fixed.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32922>
 2025-01-08 13:55:26 +00:00
Lucas Stach
ef2d7a5066 etnaviv: emit full varying component use 
Emit usages for varying components 32-63.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32922>
 2025-01-08 13:55:26 +00:00
Lucas Stach
ca5a9a3861 etnaviv: fix flatshading 
color varyings must be properly annoted, so they don't get interpolated
when the rasterizer is configured for flatshading. For whatever reason
the etnaviv NIR compiler failed to do so from its inception.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32922>
 2025-01-08 13:55:26 +00:00
Lucas Stach
2483d28d0b etnaviv: clean up component use setting in linker 
Use a trivial loop instead of duplicating the code for each
component.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32922>
 2025-01-08 13:55:26 +00:00
Lucas Stach
cca43e76e1 etnaviv: Update headers from rnndb 
Update to rnndb commit 1b944df41e12.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32922>
 2025-01-08 13:55:26 +00:00
Daniel Schürmann
d2f52e61c2 nir/divergence: change nir_has_divergent_loop() to return true only for divergent breaks 
The important information is whether a loop has a uniform number
of iterations.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28627>
 2025-01-08 13:33:54 +01:00
Mary Guillemard
42f6bb0456 libcl: Add VkQueryType and VkQueryResultFlagBits definitions 
Useful for query pool copy/clear meta shaders.

Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32939>
 2025-01-08 11:37:27 +00:00
Mary Guillemard
9911e313d0 pan/genxml: Move pack_header to an external file 
Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32939>
 2025-01-08 11:37:27 +00:00
Mary Guillemard
27d94a16c7 panforst: Update ForEachMacros 
We have various clang-format issues around on some common code macros.

This should fix this in panvk at least

Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32939>
 2025-01-08 11:37:27 +00:00
Mary Guillemard
2e38a15070 libcl: Respect NDEBUG for assert 
In C, NDEBUG allows disabling the assert macro, let's follow this
behaviour.

Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Eric R. Smith <eric.smith@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32939>
 2025-01-08 11:37:27 +00:00
Mary Guillemard
ecdccae990 nir,agx: Allow nir_precomp_print_blob to print a static array 
This makes it stop leaking shader binary blobs definition and is
required for panfrost clc.

Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Eric R. Smith <eric.smith@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32939>
 2025-01-08 11:37:27 +00:00
Mary Guillemard
5f8addfd99 util/bitpack_helpers: Make fixed packs CL safe 
We emulate roundf and llroundf for compatibility.

Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Acked-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Eric R. Smith <eric.smith@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32939>
 2025-01-08 11:37:27 +00:00
Mary Guillemard
bd872e2aaa util/bitpack_helpers: Use UINT64_MAX instead of ~0ULL 
This fixes issues with LLVM on OpenCL C failing to represent 128-bit
integers.

Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32939>
 2025-01-08 11:37:27 +00:00
Boris Brezillon
9ba38350de panvk/jm: Fix depth clipping with small viewport depth range 
Same as 7ca01506c9 ("panvk: hack to improve depth clipping with
small viewport depth range") but applied to the JM backend.

Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Benjamin Lee <benjamin.lee@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32905>
 2025-01-08 11:13:18 +00:00
Boris Brezillon
3b6f0ce032 panvk: Don't invalidate the viewport on cull mode updates 
vk_rasterization_state_depth_clip_enable() checks the clip and clamp
modes, not the cull mode. RS_DEPTH_CLIP_ENABLE got confused with
RS_CULL_MODE in 7ca01506c9 ("panvk: hack to improve depth clipping
with small viewport depth range").

Fixes: 7ca01506c9 ("panvk: hack to improve depth clipping with small viewport depth range")
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Benjamin Lee <benjamin.lee@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32905>
 2025-01-08 11:13:18 +00:00
Friedrich Vock
71392fff25 aco: Fix dead instruction/index handling for try_insert_saveexec_out_of_loop 
The loop checking if exec is overwritten didn't check for NULL
instructions, and didn't fix up reg write indices after inserting
instructions.

Fixes: fcd94a8c ("aco: move try_optimize_branching_sequence() to postRA optimizations")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32746>
 2025-01-08 10:48:01 +00:00
Georg Lehmann
67d74a04b9 nir/peephole_select: allow load_vector/scalar_arg_amd 
Foz-DB Navi21:
Totals from 1507 (1.90% of 79395) affected shaders:
MaxWaves: 31830 -> 31870 (+0.13%); split: +0.20%, -0.08%
Instrs: 938704 -> 937232 (-0.16%); split: -0.19%, +0.03%
CodeSize: 4970860 -> 4964652 (-0.12%); split: -0.14%, +0.02%
VGPRs: 79536 -> 79512 (-0.03%); split: -0.08%, +0.05%
Latency: 5194524 -> 5218285 (+0.46%); split: -0.38%, +0.84%
InvThroughput: 1200152 -> 1207251 (+0.59%); split: -0.02%, +0.61%
VClause: 20728 -> 20741 (+0.06%); split: -0.11%, +0.17%
SClause: 33612 -> 32871 (-2.20%); split: -2.78%, +0.57%
Copies: 70601 -> 68847 (-2.48%); split: -2.62%, +0.13%
Branches: 20032 -> 17521 (-12.53%)
PreSGPRs: 47828 -> 47801 (-0.06%)
VALU: 637446 -> 638094 (+0.10%); split: -0.02%, +0.13%
SALU: 88627 -> 88462 (-0.19%); split: -1.08%, +0.90%
VMEM: 36664 -> 36659 (-0.01%)

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32792>
 2025-01-08 09:56:39 +00:00
Georg Lehmann
208d8cd715 radv: run peephole_select in optimize_nir_algebraic 
Foz-DB Navi21:
Totals from 451 (0.57% of 79395) affected shaders:
MaxWaves: 8680 -> 8616 (-0.74%)
Instrs: 689610 -> 688225 (-0.20%); split: -0.21%, +0.01%
CodeSize: 3524580 -> 3521740 (-0.08%); split: -0.11%, +0.03%
VGPRs: 28512 -> 28584 (+0.25%)
Latency: 1906219 -> 1892124 (-0.74%); split: -0.91%, +0.17%
InvThroughput: 481931 -> 483570 (+0.34%); split: -0.00%, +0.34%
VClause: 10317 -> 10296 (-0.20%)
SClause: 18105 -> 18088 (-0.09%); split: -0.17%, +0.07%
Copies: 69532 -> 67579 (-2.81%); split: -2.85%, +0.04%
Branches: 21353 -> 20501 (-3.99%)
PreSGPRs: 27004 -> 27005 (+0.00%)
VALU: 436235 -> 436334 (+0.02%); split: -0.01%, +0.03%
SALU: 102349 -> 101944 (-0.40%); split: -0.61%, +0.21%

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32792>
 2025-01-08 09:56:39 +00:00
Christian Gmeiner
b6ef9017f4 etnaviv: isa: Support src2 for texldb and texldl 
We need to add variants of these instructions, which are used with a shadow
samper and passed the shadow reference value via src2.

Fixes: abe5bd35 ("etnaviv: Switch to isa_assemble_instruction(..)")
Signed-off-by: Christian Gmeiner <cgmeiner@igalia.com>
Reviewed-by: Lucas Stach <l.stach@pengutronix.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32926>
 2025-01-08 07:57:39 +00:00
Christian Gmeiner
5daa47c1f8 etnaviv: isa: Support src2 for texld 
We need to add a variant of the texld instruction, which is used with a shadow
samper and passed the shadow reference value via src2.

Blob generates such texld's for deqp's GLES3.functional.texture.shadow.2d.* (GC3000).
Fixes spec@arb_depth_texture@texdepth.

Fixes: abe5bd35 ("etnaviv: Switch to isa_assemble_instruction(..)")
Signed-off-by: Christian Gmeiner <cgmeiner@igalia.com>
Reviewed-by: Lucas Stach <l.stach@pengutronix.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32926>
 2025-01-08 07:57:39 +00:00
Lucas Stach
9e71829bcf etnaviv: isa: fix typo in SRC2_USE map 
Fixes: b216fd044b ("etnaviv: isa: Add encode support")
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <cgmeiner@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32926>
 2025-01-08 07:57:39 +00:00
Kenneth Graunke
35f175301d brw: Fix vectorizer hole_size condition after signedness change 
Marek recently changed hole_size to be signed, rather than unsigned.
A negative hole_size means that the two loads overlap - and thus are
prime candidates to be combined.

My original hole_size handling was:

   if hole_size > 4 * (8 - low->num_components) then don't vectorize

For non-overlapping loads, this worked: NIR's largest vector is vec16,
and if low was already a vec16, combining it with anything would exceed
that, so it'd never be considered.  That meant low would always be a
vec8 or less, so (8 - low->num_components) was a positive number.

Now that we see overlapping loads, we can see a vec16 low, vec4 high,
and also a negative hole size, giving us fun comparisons like:

   -16 > 4 * (8 - 16)   =>   -16 > -32   => true, don't vectorize

Which is absolutely the wrong thing to do, because the high load's data
is entirely included within the former load's data.

The idea here was to make sure the second load would be able to pack at
least one component into the first's V8 result.  But even this isn't the
best, because...even if it's simply adjacent, doing one V16 load is more
efficient than requesting two back to back V8 loads.

So, we just simplify down to a static check: if there's an entire V8 of
hole, don't vectorize.  This already won't happen because the core pass
has max_hole set to 28 bytes (7 32-bit components), but that could
change based on the needs of other drivers, so let's be defensive.

fossil-db results on Alchemist:

   Instrs: 161533978 -> 161295137 (-0.15%); split: -0.20%, +0.05%
   Subgroup size: 8092544 -> 8092568 (+0.00%)
   Send messages: 7915233 -> 7844503 (-0.89%); split: -0.94%, +0.05%
   Cycle count: 16577700697 -> 16702609256 (+0.75%); split: -0.59%, +1.35%
   Spill count: 72338 -> 67226 (-7.07%); split: -7.36%, +0.29%
   Fill count: 134058 -> 125980 (-6.03%); split: -6.83%, +0.80%
   Scratch Memory Size: 4092928 -> 3786752 (-7.48%); split: -7.53%, +0.05%
   Max live registers: 33031460 -> 32945994 (-0.26%); split: -0.27%, +0.01%
   Max dispatch width: 5778384 -> 5778536 (+0.00%); split: +0.26%, -0.26%
   Non SSA regs after NIR: 179809505 -> 152735471 (-15.06%); split: -15.08%, +0.03%

Fixes: c21bc65ba7 ("nir/opt_load_store_vectorize: make hole_size signed to indicate overlapping loads")
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32932>
 2025-01-08 00:19:54 +00:00
Sagar Ghuge
33d9a685a5 anv: Add pipelined coarse pixel state 
3DSTATE_CPS_POINTERS is deprecated on PTL, so let's switch to
3DSTATE_COARSE_PIXEL to deliver CPS state as pipelined state.

Signed-off-by: Sagar Ghuge <sagar.ghuge@intel.com>
Reviewed-by: Rohan Garg <rohan.garg@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32737>
 2025-01-07 23:53:44 +00:00