51a2e1eb94
aco/ra: don't use kill-flags as indicator in get_reg_create_vector()
...
We are about to re-use this function for vector-aligned operands.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34359 >
2025-05-28 09:24:17 +00:00
3d8b355f22
aco/assembler: support vector-aligned operands on MIMG instructions
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34359 >
2025-05-28 09:24:17 +00:00
8cb1700c74
aco/print_ir: print parenthesis around vector-aligned operands
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34359 >
2025-05-28 09:24:16 +00:00
6aabcb02a1
aco/print_ir: only print 'lateKill' if requested via print_kill flag
...
Also only print lateKill for actually killed operands.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34359 >
2025-05-28 09:24:16 +00:00
a9645fdd89
aco: introduce concept of vector-aligned Operands
...
Operand::isVectorAligned indicates that the Operand is part of a vector
consisting of multiple operands. Therefore, it must reside in a register
aligned with the next Operand.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34359 >
2025-05-28 09:24:16 +00:00
a4fa3935fd
aco/live_var_analysis: set same lateKill flags for same operands
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34359 >
2025-05-28 09:24:16 +00:00
ee0ee282b9
aco: simplify Operand() constructor
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34359 >
2025-05-28 09:24:16 +00:00
19531a0162
pan/csf: Fix typo in cs_trace_run_idvs2
...
This should have been a multiply not an add.
Fix an assertion when running in tracing mode on panvk.
Fixes: 79a1d98e1emary.guillemard@collabora.com >
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35176 >
2025-05-28 10:12:19 +02:00
2ebfa64be7
radv: add radv_disable_hiz_his_gfx12 and enable for Mafia Definitive Edition
...
This is a workaround for random GPU hangs with HiZ/HiS on GFX12
because the correct fix is complex and it will take time to be
implemented properly.
Mafia Definitive Edition is the first known game affected by this.
Cc: mesa-stable
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13222
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35182 >
2025-05-28 07:20:26 +00:00
35c76bc7f7
nir/tcs_info: use range analysis to determine the range of tess levels
...
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35195 >
2025-05-28 06:46:56 +00:00
24c3f30e4a
nir/tcs_info: gather which patch outputs are only read/written by invoc 0
...
Tested thoroughly by a shader test.
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35195 >
2025-05-28 06:46:56 +00:00
a3632d7d88
nir/tcs_info: gather for all patch outputs whether they're written by all invocs
...
This substantially rewrites the pass. It also makes it easier to read.
Tested thoroughly by a shader test.
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35195 >
2025-05-28 06:46:56 +00:00
6b265d9d7d
nvk: Use NAK by default on Kepler
...
At this point it's way more correct than codegen.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:19 +00:00
a37c50e985
nak/sm20: Encode OpSuLdGa and OpSuStGa
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:19 +00:00
2a5ac439f5
nak/sm20: Encode imadsp
...
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:19 +00:00
2c0957d086
nak/sm20: Encode surface address ops
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:19 +00:00
4d86f95cff
nak/sm32: Encode surface address ops
...
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:19 +00:00
2c0d0bad01
nak: Remove unused intrinsic image_load_raw_nv
...
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:19 +00:00
ac3fd5768b
nak: Add surface address ops
...
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:19 +00:00
82d789d22a
nvk: Lower images to addresses on Kepler
...
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:19 +00:00
bb3f4b86bf
nvk: Use nil_su_info for image descriptors on Kepler
...
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:19 +00:00
e5d4c4934d
nak: Add an image address lowering pass
...
On Kepler, we don't have real surface opcodes. Instead, we have a bunch
of helper ops which make image calculations in the shader reasonably
efficient and we have to lower to that and an address-based load/store.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:19 +00:00
0b06feee92
nouveau: Move bitview to a common rust util folder
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:19 +00:00
5fbcdd6e32
nir,nak: Add NV-specific image intrinsics
...
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:19 +00:00
a27a711c3a
nil: Add helpers for filling out Fermi/Kepler surface info
...
Kepler needs different descriptors for Image storage, these descriptors
are not directly used by the hardware but by lowering instructions
(added in later commits)
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:19 +00:00
e388f25a2a
nil: Align linear image strides to the image alignment
...
This is a function no-op right. In the case where we don't have an
explicit_row_stride_B, we use an alignment of 128B which is what we were
aligning the row stride to before.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:18 +00:00
7a577d2e3a
nvk: Drop the union from nvk_buffer_view
...
The space saved is tiny and not having the union at all makes me way
more comfortable about the modifications we're about to do.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34975 >
2025-05-28 01:47:18 +00:00
4b1c824b67
vtn/opencl: Handle OpenCLstd_F{Min,Max}_common
...
Normal fmin doesn't make any promises about NaN, common additionally
doesn't make any promises about infinities. Would be nice to hook that
up to codegen but lowering them to normal works for now.
Cc: mesa-stable
Reviewed-by: Karol Herbst <kherbst@redhat.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34941 >
2025-05-28 01:08:20 +00:00
92f07860a4
vtn: (Silently) handle FunctionParameterAttributeNo{Capture,Write}
...
Silences a few thousand warnings in sycl/test-e2e
Cc: mesa-stable
Reviewed-by: Karol Herbst <kherbst@redhat.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34941 >
2025-05-28 01:08:20 +00:00
fd7589d775
vtn: Handle SPV_KHR_uniform_group_instructions
...
Reviewed-by: Karol Herbst <kherbst@redhat.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34941 >
2025-05-28 01:08:20 +00:00
52db0e1480
intel/compiler: fix SHA generation for shader replace
...
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35140 >
2025-05-27 22:57:19 +00:00
8ffe0098be
nvk: Reserve a sampler for TXF on Kepler
...
The SPIR-V spec says texelFetch and friends don't take a sampler.
However, on Kepler and earlier hardware, the sampler is read even for
tld. In particular, the hardware reads the sRGB conversion bit in the
sampler and this can be in an inconsistent state if we haven't
initialized samplers properly. On Kepler, we should just reserve a
sampler at device creation time and always use that for tld.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35190 >
2025-05-27 22:23:37 +00:00
bfdc95b109
nak: Scalarize non-constant ald/ast on Kepler
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35190 >
2025-05-27 22:23:36 +00:00
f10c42fcc1
bitview: Implement SetFieldU64 for all BitViewMutable
...
There's no reason why we need to keep overriding this. There's also
probably not a good reason why this trait exists at all but that's a
problem for another day.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35190 >
2025-05-27 22:23:36 +00:00
6ec69e5a07
gallium: drop dynamic pipe-loader leftovers
...
Fixes: 30c5c7d811https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35185 >
2025-05-27 21:44:31 +00:00
0a9b871b9e
rusticl: support cl_khr_kernel_clock
...
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35179 >
2025-05-27 19:57:28 +00:00
8144c7ee86
rusticl: generate bindings for llvm version macros
...
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35179 >
2025-05-27 19:57:28 +00:00
f5a9a80f29
clc: wire up cl_khr_kernel_clock
...
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35179 >
2025-05-27 19:57:28 +00:00
63758bc093
radv: fix capture/replay with sparse images and descriptor buffer
...
The sparse image VA needs to be returned to the application for replay.
Reported by Baldur.
VKCTS has coverage but it doesn't verify this yet.
Cc: mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35162 >
2025-05-27 19:30:18 +00:00
a628190e0d
ci/vkd3d: collapse section by default
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35180 >
2025-05-27 18:52:50 +00:00
f12f2bcae1
ci: uprev vkd3d
...
A new group of descriptor aliasing tests was added; let's run these in CI :)
https://github.com/HansKristian-Work/vkd3d-proton/compare/7e829e88364f9429596b14d0c0ea0205e1337edc...041cfa75580090c6b30af2f2721015d15493f14b
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35180 >
2025-05-27 18:52:50 +00:00
738b95a35e
panfrost: add max_4x_msaa quirk
...
Turns out, some Midgard GPUs don't support more than 4x MSAA. Add a
quirk for those GPUs, so don't expose it when it doesn't work.
Reviewed-by: Eric R. Smith <eric.smith@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35184 >
2025-05-27 17:41:47 +00:00
3a78d9213b
panfrost: pass quirks through __VA_ARGS__
...
We can't easily pass multiple initializers here, because a comma in the
initializer list would be treated as a preprocessor argument separator
and not a separator in the initializer lst.
We could also have fixed this with some nested macro ugliness, but let's
instead do what nir_builder does for intrinsic indices and use __VA_ARGS__
to keep this neat.
Reviewed-by: Eric R. Smith <eric.smith@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35184 >
2025-05-27 17:41:47 +00:00
e77e4d5c17
panfrost: change tie-breaking rule for 16x MSAA
...
When using 16x MSAA, we have two sample-positions on the negative
boundary of the unit-square covering the pixel. This causes problems
when using the default tie-breaking rule, where we miss some
sample-positions when rasterizing primitives covering the entire
viewport.
This works fine on Bifrost and later, but this setting is ignored on
those GPUs, and they assume the default (e.g MINUS_180_OUT_0_IN).
Because we'd prefer for rasterization to match between Midgard and
Bifrost when we can, we only apply this when we have 16x MSAA.
As an added bonus, this behavior matches what the DDK does.
Fixes these tests when 16x MSAA is enabled:
- dEQP-GLES31.functional.texture.multisample.samples_16.use_texture_*
- dEQP-GLES3.functional.multisample.fbo_max_samples.proportionality_alpha_to_coverage
Reviewed-by: Eric R. Smith <eric.smith@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35184 >
2025-05-27 17:41:47 +00:00
130bbf20c7
panfrost: properly compute tile-buffer requirements
...
This should be helful when we want to enable higher sample counts.
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com >
Reviewed-by: Eric R. Smith <eric.smith@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35184 >
2025-05-27 17:41:47 +00:00
64becaf142
panfrost: re-align table
...
The formatting here has grown some warts over time, let's clean it up a
bit.
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com >
Reviewed-by: Eric R. Smith <eric.smith@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35184 >
2025-05-27 17:41:47 +00:00
cf28edd8bd
panfrost: do not calculate max-msaa on v4
...
The V4 GPUs doesn't have the dynamic allocation logic that V5 and later
has. There's nothing to calculate here; the GPU either supports 8x MSAA,
or 4x MSAA.
Since 8x MSAA is the architectural max, let's have this function report
that. We deal with the 4x limit separately as a quirk, because this
applies to some V5 GPUs as well.
Reviewed-by: Eric R. Smith <eric.smith@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35184 >
2025-05-27 17:41:47 +00:00
483ce5a1dc
panfrost: do not try to use 4x4 tiles on v4 gpus
...
Mali V4 GPUs only ever use 16x16 tiles, so we need to set the minimum
tile-size to match.
Fixes: 329568b5eberic.smith@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35184 >
2025-05-27 17:41:47 +00:00
7de12da5a2
panfrost: v4 does not support 16x msaa
...
This value isn't valid on V4, so let's make sure we don't try to use it.
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com >
Reviewed-by: Eric R. Smith <eric.smith@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35184 >
2025-05-27 17:41:47 +00:00
439b88c619
mesa/main: remove non-existing function prototype
...
This function was removed about a decade ago, let's get rid of the
prototype as well!
Fixes: a347a0f53feric.smith@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35184 >
2025-05-27 17:41:47 +00:00
Daniel Schürmann