03d79bc09d
nouveau/headers: Import the video class headers from NVIDIA
...
From open-gpu-doc commit c8607fe576b5 ("Add hopper and blackwell
dma-copy class header files").
Signed-off-by: Daniel Almeida <daniel.almeida@collabora.com >
Signed-off-by: Faith Ekstrand <faith.ekstrand@collabora.com >
Reviewed-by: Mel Henning <mhenning@darkrefraction.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36641 >
2025-08-07 20:09:10 +00:00
31a6e80bb3
nouveau/class_parser: Strip unnecessary parens
...
Some of the headers (video is a particularly egregious culprit) wrap
defined values in parens, even when they don't need to be. This changes
the parser to up-front strip all outer pairs of parens to get us down to
the original value.
Reviewed-by: Mel Henning <mhenning@darkrefraction.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36641 >
2025-08-07 20:09:10 +00:00
ba92104b56
wsi/display: make HDR_OUTPUT_METADATA, Colorspace properties optional
...
Fixes some dEQP-VK.wsi.display.* issues on platforms that don't
support said properties following their addition in !35461 .
Fixes: b4176393a0simon.perretta@imgtec.com >
Reviewed-by: David Rosca <david.rosca@amd.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36607 >
2025-08-07 20:32:02 +01:00
2bd5639152
nvk: Disable non-graphics timestamp queries
...
We can support this in the future, but we don't implement it just yet.
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36651 >
2025-08-07 18:47:47 +00:00
9c6d6daf1b
nvk: Check subchannels are valid in nv_push
...
The hardware already checks that any subchannel that we use is valid,
but those error messages don't provide much context and can be hard to
debug. Add an assert in nv_push that the subchannel we're using has
been allocated. This way, we can pinpoint exactly where the incorrect
method was used by looking at the stack trace.
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36651 >
2025-08-07 18:47:47 +00:00
521537503c
nvk: Factor out nvk_queue_engines_from_queue_flags
...
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36651 >
2025-08-07 18:47:47 +00:00
87d6840110
nvk: Call cmd_buffer_begin_* based on queue flags
...
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36651 >
2025-08-07 18:47:47 +00:00
66017ae4a6
nvk: Handle empty pushes in nvk_queue_push()
...
The only reason why this did not crash until now is because NVKMD_ENGINE_3D
and NVKMD_ENGINE_COMPUTE are always set in the only queue we returned.
With the addition of a video queue, where no state must be submitted, the logic
would try to push an empty push_buffer. This is wrong and the first place where
it fails is when it tries to allocate a BO of size 0 for the push data.
Reviewed-by: Mel Henning <mhenning@darkrefraction.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36651 >
2025-08-07 18:47:46 +00:00
9a579673dd
meson,nvk: Require rustc-hash 2.0 or later
...
rustc-hash 1.1 does not have FxBuildHasher, and 1.2 has been yanked from
crates.io. I've verified that version 2.0 builds and seems to work, so
set the required version as 2.0.
Fixes: c1f979d84ahttps://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36650 >
2025-08-07 18:31:08 +00:00
900e56fc44
ac/nir: clarify the behavior of ac_nir_lower_ngg_options::can_cull
...
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36578 >
2025-08-07 18:12:53 +00:00
1ddbe7e818
radeonsi/gfx12: print swizzle modes for AMD_TEST=imagecopy
...
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36578 >
2025-08-07 18:12:53 +00:00
9244618e5f
ac/nir/meta: allow compute blits with R5G6B5 & R5G5B5A1 formats on GFX9+
...
v2: add a workaround for incorrect hw rounding
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36578 >
2025-08-07 18:12:52 +00:00
82a7085898
freedreno/ci: Move a660-gl-cl job to nightly
...
The sm8350-hdk devices in the Collabora lab are experiencing power
issues: most are unable to boot via the barrel jack and only receive
power over the USB-C port, which is also used for fastboot.
Out of 9 devices, only 4 are currently operational. To reduce the load
on the remaining devices, move the 3 a660-gl-cl jobs to nightly.
Signed-off-by: Valentine Burley <valentine.burley@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36623 >
2025-08-07 17:16:29 +00:00
a8ab696033
meson: Allow configuring with Android-internal perfetto
...
This enables ninja-to-soong to generate an Android.bp that builds Mesa
against Android's `libperfetto_client_experimental` library.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36561 >
2025-08-07 16:22:37 +00:00
eb6f6c1976
meson: Allow controlling perfetto fallback
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36561 >
2025-08-07 16:22:37 +00:00
d476a8c07b
zink: stop always syncing threaded flushes
...
async flushes are meant to be async here, and the other ones are manually
synced later in the function
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36643 >
2025-08-07 16:00:54 +00:00
f5f7a5f801
zink: don't access ctx in submit_queue
...
this can be executed from a thread
Fixes: 63e17ccc0ahttps://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36643 >
2025-08-07 16:00:54 +00:00
7dc4f28507
pan/bi: schedule simple iterators to avoid extra move
...
Try to move iterator as close to the end of the block as possible. The
goal is to avoid the iterator being used after being updated, to
prevent the need for an extra move instruction.
shader-db report on `Mali-G725`:
```
total instrs in shared programs: 720530 -> 716482 (-0.56%)
instrs in affected programs: 231842 -> 227794 (-1.75%)
helped: 3804
HURT: 1
helped stats (abs) min: 1.0 max: 8.0 x̄: 1.06 x̃: 1
helped stats (rel) min: 0.14% max: 6.25% x̄: 2.75% x̃: 2.86%
HURT stats (abs) min: 1.0 max: 1.0 x̄: 1.00 x̃: 1
HURT stats (rel) min: 0.05% max: 0.05% x̄: 0.05% x̃: 0.05%
95% mean confidence interval for instrs value: -1.08 -1.04
95% mean confidence interval for instrs %-change: -2.79% -2.71%
Instrs are helped.
total cycles in shared programs: 35295.80 -> 35295.55 (<.01%)
cycles in affected programs: 3.50 -> 3.25 (-7.14%)
helped: 8
HURT: 0
helped stats (abs) min: 0.03125 max: 0.03125 x̄: 0.03 x̃: 0
helped stats (rel) min: 6.90% max: 7.41% x̄: 7.15% x̃: 7.15%
95% mean confidence interval for cycles value: -0.03 -0.03
95% mean confidence interval for cycles %-change: -7.38% -6.92%
Cycles are helped.
total fma in shared programs: 5054.34 -> 5054.34 (0.00%)
fma in affected programs: 0 -> 0
helped: 0
HURT: 0
total cvt in shared programs: 4707.69 -> 4644.44 (-1.34%)
cvt in affected programs: 1471.28 -> 1408.03 (-4.30%)
helped: 3804
HURT: 1
helped stats (abs) min: 0.015625 max: 0.125 x̄: 0.02 x̃: 0
helped stats (rel) min: 0.37% max: 12.50% x̄: 6.01% x̃: 6.67%
HURT stats (abs) min: 0.015625 max: 0.015625 x̄: 0.02 x̃: 0
HURT stats (rel) min: 0.13% max: 0.13% x̄: 0.13% x̃: 0.13%
95% mean confidence interval for cvt value: -0.02 -0.02
95% mean confidence interval for cvt %-change: -6.07% -5.94%
Cvt are helped.
total sfu in shared programs: 1878.25 -> 1878.25 (0.00%)
sfu in affected programs: 0 -> 0
helped: 0
HURT: 0
total v in shared programs: 2353 -> 2353 (0.00%)
v in affected programs: 0 -> 0
helped: 0
HURT: 0
total t in shared programs: 5530 -> 5530 (0.00%)
t in affected programs: 0 -> 0
helped: 0
HURT: 0
total ls in shared programs: 27975 -> 27975 (0.00%)
ls in affected programs: 0 -> 0
helped: 0
HURT: 0
total code size in shared programs: 6386560 -> 6289664 (-1.52%)
code size in affected programs: 508544 -> 411648 (-19.05%)
helped: 757
HURT: 0
helped stats (abs) min: 128.0 max: 128.0 x̄: 128.00 x̃: 128
helped stats (rel) min: 0.83% max: 33.33% x̄: 31.09% x̃: 33.33%
95% mean confidence interval for code size value: -128.00 -128.00
95% mean confidence interval for code size %-change: -31.57% -30.60%
Code size are helped.
total threads in shared programs: 14698 -> 14698 (0.00%)
threads in affected programs: 0 -> 0
helped: 0
HURT: 0
total loops in shared programs: 166 -> 166 (0.00%)
loops in affected programs: 0 -> 0
helped: 0
HURT: 0
total spills in shared programs: 37 -> 37 (0.00%)
spills in affected programs: 0 -> 0
helped: 0
HURT: 0
total fills in shared programs: 111 -> 111 (0.00%)
fills in affected programs: 0 -> 0
helped: 0
HURT: 0
```
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com >
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com >
Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36021 >
2025-08-07 15:09:56 +00:00
4a53eca97d
pan/bi: add pass to simplify control flow
...
That pass tries to remove blocks that are not needed:
blocks with only 1 predecessor and 1 successor containing no
instruction or only 1 branch instruction.
Also remove unnecessary branch at the end of block with only 1
successor which is the next block
Run this pass at the end of the compiler flow once all optimisations
have been applied.
Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com >
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36021 >
2025-08-07 15:09:56 +00:00
fa2bd65171
r600/sfn: Move RA helper class declaration into implementation file
...
There is no need that these types are exposed.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36586 >
2025-08-07 14:23:25 +00:00
2bb3e5c265
r60/sfn: Update .clang-format
...
* Short inline methods in headers may stay on one line.
* Include system headers last to avoid hidden dependencies.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36542 >
2025-08-07 14:08:46 +00:00
6b693e281a
pan/va: improve lowering of SWZ_V4I8
...
Use bi_make_vec_to to allow to use only 1 MKVEC.v2i8 when possible.
Also add support for all swizzles instead of only mono-byte ones,
using bi_swizzle_to_byte_channels.
Update assert in bi_byte.
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35643 >
2025-08-07 13:48:33 +00:00
857f29d67b
pan/bi: use only 1 MKVEC.v2i8 to generate v4i8 when possible
...
When making a vector of 4 elements, try to make it with only 1
instruction instead of 2 at the moment, if the last 2 elements respect
the pattern supported by MKVEC.v2i8
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35643 >
2025-08-07 13:48:32 +00:00
8b96f66da6
agx: gate scratch opt on internal shaders
...
fixes a bunch of OpenCL CTS including test_basic vload_private due to failing
to relower the derefs but also lol.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io >
Backport-to: 25.1
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36631 >
2025-08-07 13:34:22 +00:00
45c28a20a4
rusticl/gl: flush and wait on gl objects inside clEnqueueAcquireGLObjects
...
This fixes flakes in the gl/gles OpenCL CTS tests on radeonsi and zink.
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36249 >
2025-08-07 12:58:33 +00:00
f6b844e239
rusticl/gl: only flush objects on import if we get a valid fd
...
If no context is current we might not get anything to wait on.
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36249 >
2025-08-07 12:58:33 +00:00
4d8b08e7ec
st/interup: flushing objects is a no-op when no context is bound
...
Luckily this is backwards compatible, because without this change users
would simply run into crashes within _mesa_glthread_finish.
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36249 >
2025-08-07 12:58:32 +00:00
2e1fff8e5d
rusticl/gl: store the mesa_glinterop_export_in
...
We need it later for proper flushing and waiting.
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36249 >
2025-08-07 12:58:32 +00:00
8617cf46d6
rusticl/mesa: wire up fence_server
...
We need this to wait on imported fences.
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36249 >
2025-08-07 12:58:32 +00:00
24c692c981
panvk: fix a NULL pointer dereference in occlusion queries
...
If a meta operation (like a blit or clear) happens while occlusion
queries are active, we temporarily disable the query. Unfortunately
the code for this did not clear out the `syncobj` field. In rare
combinations of circumstances this could cause an attempt to issue
a write back of the occlusion query values, and since we've zeroed
the `ptr` field it writes to a NULL value, causing a bus fault and
device lost error.
Fixes: 61534faf4eerik.faye-lund@collabora.com >
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36525 >
2025-08-07 11:58:53 +00:00
2d691d7dd3
meson: remove intel-clc options
...
This tool is gone.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com >
Fixes: f0e18c475balyssa@rosenzweig.io >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36625 >
2025-08-07 11:25:41 +00:00
3e19c0f129
Revert "ci: mark igalia farm as offline"
...
This reverts commit 09272e5da9https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36627 >
2025-08-07 11:02:05 +00:00
94a2740cab
panvk: rewrite pan_nir_lower_static_noperspective
...
Really this should be using the common code I added in
https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36501 but one thing at
a time.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io >
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com >
Reviewed-by: Olivia Lee <olivia.lee@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36600 >
2025-08-07 10:33:36 +00:00
dff1d91c64
panvk: Enable VK_ARM_shader_core_builtins
...
While the shaderCoreBuiltins feature is only supported on v9+, the
extension itself provides some useful physical device properties and is
supported on all GPUs.
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36019 >
2025-08-07 11:46:33 +02:00
c46407de88
pan/va: Add support for SPV_ARM_core_builtins
...
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36019 >
2025-08-07 11:46:33 +02:00
000bd3046d
nir,spirv: Add support for SPV_ARM_core_builtins
...
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36019 >
2025-08-07 11:46:33 +02:00
a68a825aad
nir,agx: unvendor core_id_agx
...
core_id will be used by SPV_ARM_core_builtins
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com >
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36019 >
2025-08-07 11:46:33 +02:00
dad0c7e6d4
zink/ci: make zink-radv-navi31-valve a pre-merge job
...
Otherwise, it's regularly broken and a waste of time for everyone.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13501
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36594 >
2025-08-07 09:14:00 +00:00
06338ccf9e
zink/ci: reduce timeout of zink-radv-navi31-valve
...
It usually runs in ~5 minutes, and it's important to reduce it for
pre-merge testing.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36594 >
2025-08-07 09:14:00 +00:00
b6ab6139c1
ci/container: add comment to bump image tag
...
Add a comment to update ALPINE_X86_64_LAVA_SSH_TAG when this
script is modified.
Signed-off-by: Vignesh Raman <vignesh.raman@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36620 >
2025-08-07 08:44:53 +00:00
cfd8b9d3b3
rusticl: use pipe_sampler_view_release
...
Fixes: a245ed462aleftmostcat@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36613 >
2025-08-07 08:05:55 +00:00
8e39bab014
addrlib: __debugbreak only present on Windows and from intrin.h
...
Both MSVC/MINGW support this.
For fixes compiling with MINGW/GCC
Signed-off-by: Yonggang Luo <luoyonggang@gmail.com >
Reviewed-by: Marek Olšák <marek.olsak@amd.com >
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36598 >
2025-08-07 07:47:42 +00:00
d83234c8f3
radv: Move the amdgpu.h defines for Win32 to ac_linux_drm.h
...
This is for fixes compiling with MINGW/GCC
Signed-off-by: Yonggang Luo <luoyonggang@gmail.com >
Reviewed-by: Marek Olšák <marek.olsak@amd.com >
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36598 >
2025-08-07 07:47:42 +00:00
dc2c3cf06b
intel: Disable 16x MSAA support on Xe3
...
16x MSAA isn't supported at all on certain Xe3 variants, and on its way
out on the rest. Most vendors choose not to support it, and many apps
offer more modern multisampling and upscaling techniques these days.
Only 2/4/8x are supported going forward.
Cc: mesa-stable
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com >
Reviewed-by: Sagar Ghuge <sagar.ghuge@intel.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36592 >
2025-08-07 07:30:35 +00:00
f7def1ac1e
crocus: Fix a comment about supporting 16x MSAA
...
The mask here was already 0xff (8-bits) because only 8x MSAA is
supported, but the comment was copy-pasted from iris which did 16x.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com >
Reviewed-by: Sagar Ghuge <sagar.ghuge@intel.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36592 >
2025-08-07 07:30:35 +00:00
d098b3b973
crocus: Drop 16X MSAA code remnants
...
No platforms supported by crocus actually support 16X MSAA, so this is
just copy-pasted from iris without actually offering any benefit.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com >
Reviewed-by: Sagar Ghuge <sagar.ghuge@intel.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36592 >
2025-08-07 07:30:35 +00:00
9765624ed8
radv/ci: fix GPU hang detection regex with recent kernels
...
Since July 2024, *ERROR* is no longer printed in dmesg and this caused
GPU hangs to not be detected with recent kernels. Hopefully, this won't
uncover more problems in CI...
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36611 >
2025-08-07 07:08:14 +00:00
418a335626
zink/ci: skip spec@arb_fragment_program@fog-modes on RADV
...
This test seems to hang.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36611 >
2025-08-07 07:08:14 +00:00
b987ec42ac
radv: Avoid calls to strlen when parsing umr output to speed up hang progressing
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36451 >
2025-08-07 06:44:41 +00:00
4094706828
radv: invalidate compute/rt descriptors at dispatch time
...
Invalidating descriptors at bind time is wrong in a scenario like
trace rays -> dispatch -> trace rays because RT uses compute shaders
and the user SGPRs need to be re-emitted.
Noticed by Konstantin.
Fixes: 40ceece75fsamuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36605 >
2025-08-07 06:26:43 +00:00
Daniel Almeida