AlexIndustrial/mesa
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
184,955 Commits 1 Branch 0 Tags
0589dfe4e2a12f463df79f48b9b2ca5e1eb57eba
Commit Graph

184955 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Marek Olšák
0589dfe4e2 util/idalloc: optimize foreach by tracking the greatest non-zero element 
also add util_idalloc_foreach_no_zero_safe.

Reviewed-by: Adam Jackson <ajax@redhat.com>
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27494>
 2024-02-21 16:04:51 +00:00
Chris Rankin
e1e84b0721 vdpau: Declare texture object as immutable using helper function. 
Cc: mesa-stable

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/10618
Signed-off-by: Chris Rankin <rankincj@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27681>
 2024-02-21 14:54:48 +00:00
Rhys Perry
d0595e4805 aco/ra: fix GFX9- writelane 
061b8bfd29 moved handling of fixed operands earlier, but it should have
moved the fixing of writelane operands earlier too.

This fixes Crucible's func.uniform-subgroup.exclusive.imin64 on GFX8.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Fixes: 061b8bfd29 ("aco/ra: rework fixed operands")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27583>
 2024-02-21 12:40:14 +00:00
Rhys Perry
53800191a3 aco/ra: don't initialize assigned in initializer list 
According to Valgrind, vcc/m0 are uninitialized and this fixes it.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27583>
 2024-02-21 12:40:14 +00:00
Mike Blumenkrantz
30b7771aa9 ci: make clang-format job warn on failure instead of killing the pipeline 
it's insane to fail a pipeline over formatting errors.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27702>
 2024-02-21 12:04:08 +00:00
Lionel Landwerlin
137b7e874f anv: disable Wa_16013994831 
We've implemented another workaround completely disabling high
priority preemption.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: e6e320fc79 ("anv: make Wa_16013994831 to use intel_needs_workaround")
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27716>
 2024-02-21 11:39:50 +00:00
Samuel Pitoiset
78ea304a06 spirv: only consider IO variables when adjusting patch locations for TES 
With TES, the primitive ID is an input variable but it's considered a
sysval by SPIRV->NIR. Though, its value is greater than
VARYING_SLOT_VAR0 which means its location was adjusted by mistake.

This fixes compiling a tessellation evaluation shader in debug build
with Enshrouded.

Fixes: dfbc03fa88 ("spirv: Fix locations for per-patch varyings")
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27413>
 2024-02-21 10:36:07 +00:00
Kenneth Graunke
c12300844d intel/fs: Don't rely on CSE for VARYING_PULL_CONSTANT_LOAD 
In the past, we didn't have a good solution for combining scalar loads
with a variable index plus a constant offset.  To handle that, we took
our load offset and rounded it down to the nearest vec4, loaded an
entire vec4, and trusted in the backend CSE pass to detect loads from
the same address and remove redundant ones.

These days, nir_opt_load_store_vectorize() does a good job of taking
those scalar loads and combining them into vector loads for us, so we
no longer need to do this trick.  In fact, it can be better not to:
our offset need only be 4 byte (scalar) aligned, but we were making it
16 byte (vec4) aligned.  So if you wanted to load an unaligned vec2,
we might actually load two vec4's (___X | Y___) instead of doing a
single load at the starting offset.

This should also reduce the work the backend CSE pass has to do,
since we just emit a single VARYING_PULL_CONSTANT_LOAD instead of 4.

shader-db results on Alchemist:
- No changes in SEND count or spills/fills
- Instructions: helped 95, hurt 100, +/- 1-3 instructions
- Cycles: helped 3411 hurt 1868, -0.01% (-0.28% in affected)
- SIMD32: gained 5, lost 3

fossil-db results on Alchemist:
- Instrs: 161381427 -> 161384130 (+0.00%); split: -0.00%, +0.00%
- Cycles: 14258305873 -> 14145884365 (-0.79%); split: -0.95%, +0.16%
- SIMD32: Gained 42, lost 26

- Totals from 56285 (8.63% of 652236) affected shaders:
- Instrs: 13318308 -> 13321011 (+0.02%); split: -0.01%, +0.03%
- Cycles: 7464985282 -> 7352563774 (-1.51%); split: -1.82%, +0.31%

From this we can see that we aren't doing more loads than before
and the change is pretty inconsequential, but it requires less
optimizing to produce similar results.

Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27568>
 2024-02-20 23:16:27 -08:00
Iago Toral Quiroga
1880e7cfed v3d,v3dv: fix BO allocation for shared vars 
We need to allocate "shared size" bytes for each workgroup but
we were incorrectly multiplying by the number of workgroups in
each supergroup instead, which would typically cause us to allocate
less memory than actually required.

The reason this issue was not visible until now is that the kernel
driver is using a large page alignment on all BO allocations and
this causes us to "waste" a lot of memory after each allocation.
Incidentally, this wasted memory ensured that out of bounds
accesses would not cause issues since they would typically land
in unused memory regions in between aligned allocations, however,
experimenting with reduced memory aligments raised the issue,
which manifested with the UE4 Shooter demo as a GPU hang caused
by corrupted state from out of bounds memory writes to CS
shared memory.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27675>
 2024-02-21 06:17:55 +00:00
Dave Airlie
f47858b837 nvk: fix dri options leak. 
Noticed in valgrind.

Fixes: edb5229538 ("nvk: Hook up driconf for nvk_instance")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27712>
 2024-02-21 05:23:09 +00:00
Dave Airlie
fd04c14306 nouveau/winsys: fix bda heap leak. 
found with valgrind.

Fixes: b4cfac64c8 ("nvk: Add a separate VMA heap for BDA capture/replay")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27712>
 2024-02-21 05:23:09 +00:00
Mike Blumenkrantz
2ad0146179 zink: use KHR_dynamic_rendering_local_read 
no more renderpasses!

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27299>
 2024-02-21 04:24:36 +00:00
Mike Blumenkrantz
19789aeeb1 zink: hook up KHR_dynamic_rendering_local_read 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27299>
 2024-02-21 04:24:36 +00:00
Faith Ekstrand
3b658028dd nvk: Invalidate the texture cache before MSAA resolves 
Fixes: 4bd2ba31fc ("nvk: Use meta for MSAA resolves")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27710>
 2024-02-21 04:07:18 +00:00
Jesse Natalie
705e646e38 microsoft/compiler: Fix SM6.6 non-bindless handle annotation for UAV counter 
Counter bool is 8, not 6.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27708>
 2024-02-21 03:39:22 +00:00
Mike Blumenkrantz
ccbf9b0ea7 zink: force host-visible allocations for MAP_COHERENT resources 
this fixes persistent maps for systems without any BAR

cc: mesa-stable

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27707>
 2024-02-21 02:57:06 +00:00
Mike Blumenkrantz
fb2ae7736f zink: add a second fence disambiguation case 
this is the case where:
* a batch A is submitted
* a no-op flush occurs
* the frontend gets the fence from already-flushed batch A
* zink recycles batch A
* the frontend waits on fence A

fixes #10598

cc: mesa-stable

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27623>
 2024-02-21 02:28:48 +00:00
Dave Airlie
6a688e65a0 loader: handle picking zink for nouveau for certain GPUs. 
This adds NOUVEAU_USE_ZINK env var, but also has commented out
code to pick it for turing+ if mesa is built with nvk and zink.

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27628>
 2024-02-21 11:16:25 +10:00
Dave Airlie
caf99133ec glx/dri3: handle zink fallback if loader picks it. 
if we get a zink driver selected for the fd then don't bind it to dri3
return a special error and this should stop dri2 binding it.

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27628>
 2024-02-21 11:16:22 +10:00
Dave Airlie
b907eb4750 egl: don't bind zink under dri2/3 
If we are in dri2/3 paths and the loader picks zink for this fd
don't bind it here, let the fallbacks work.

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27628>
 2024-02-21 11:15:38 +10:00
Caio Oliveira
ccbccfff91 iris: Include brw_compiler.h only when needed 
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27646>
 2024-02-21 00:38:35 +00:00
Caio Oliveira
376296a8c9 iris: Move iris_get_compiler_options to iris_program.c 
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27646>
 2024-02-21 00:38:35 +00:00
Caio Oliveira
880e452abf iris: Add helper for indirect_ubos_use_sampler 
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27646>
 2024-02-21 00:38:35 +00:00
Caio Oliveira
947ef049df iris: Add helper to access use_tcs_multi_patch 
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27646>
 2024-02-21 00:38:35 +00:00
Caio Oliveira
634a8ece67 iris: Add IRIS_MAX_* constants to replace BRW_MAX_* usage 
They are still the same, but we don't rely on the BRW compiler
specific symbols.  STATIC_ASSERT catches at compile time if they
change independently.  At some point we might revisit the need
for them to match.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27646>
 2024-02-21 00:38:35 +00:00
Caio Oliveira
875aa527b8 iris: Move compiler creation to iris_program.c 
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27646>
 2024-02-21 00:38:35 +00:00
Caio Oliveira
19574a8156 iris: Use uint32_t instead of brw_param_builtin 
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27646>
 2024-02-21 00:38:35 +00:00
Caio Oliveira
0b135c9f80 iris: Take ownership of prog_data when applying it 
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27646>
 2024-02-21 00:38:35 +00:00
Caio Oliveira
04364768f2 iris: Reduce dependency on brw_*_prog_data structs 
Once the brw_*_prog_data are available, copy down all the relevant
fields to iris_compiled_shader (and iris_*_data corresponding structs)
so that most of Iris code will be independent of brw types.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27646>
 2024-02-21 00:38:35 +00:00
Caio Oliveira
9fd4b7460f iris: Don't use prog_data to guard 3DSTATE_CONSTANT_* code 
At this point in the code, the prog_data is always non-NULL (and was
already used before by setup_constant_buffers() to fill push_bos.
Suggested by Ken.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27646>
 2024-02-21 00:38:35 +00:00
Caio Oliveira
be13c3ef9f iris: Add stage to iris_compiled_shader 
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27646>
 2024-02-21 00:38:35 +00:00
Caio Oliveira
8ae528331c intel/compiler: Use "intel" prefix for walk_order enum 
Will be used later in non-brw specific code in Iris.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27646>
 2024-02-21 00:38:35 +00:00
Timothy Arceri
74534397ac glsl: split var copies before lowering named interfaces 
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/10593

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27669>
 2024-02-20 23:29:17 +00:00
Timothy Arceri
4c11119825 glsl: support array wildcards in lower named interface blocks 
Will be required with the change in the following patch.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27669>
 2024-02-20 23:29:17 +00:00
Timothy Arceri
ec240e2cd8 nir: allow gather info to handle nir_deref_type_array_wildcard 
Needed for some changes to the glsl nir linker in the following
patches.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27669>
 2024-02-20 23:29:17 +00:00
Kenneth Graunke
1497f4e0c2 intel/fs: Don't include sync.nop in instruction count statistics 
With the advent of software scoreboarding, we emit sync instructions
in various places to synchronize the execution pipelines.  This results
in assembly being littered with a bunch of sync.nop instructions.  That
means that when you reorder anything in the program, the scoreboarding
changes, and the number of sync.nops can vary wildly - even if the code
isn't really materially better or worse.  This makes it hard to use
tools like shader-db or fossil-db on post-Icelake platforms.

For now, exclude sync.nops from the instruction count statistic.  One
day we may want to consider improving the software scoreboarding pass
to emit fewer redundant sync.nop instructions, at which point tracking
this as a separate stat might be useful.  For now though, it's simply
cluttering and confusing our results.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27701>
 2024-02-20 22:26:09 +00:00
Mike Blumenkrantz
83d1241cf5 zink: add checks/compat for low-spec descriptor buffer implementations 
for implementations that can only support 1 descriptor buffer:
* prefer templates
* allow db if explicitly requested, but disable bindless (for testing)

fix #10640

Fixes: b8b51d96b0 ("zink: delete maxDescriptorBufferBindings checks")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27696>
 2024-02-20 21:43:15 +00:00
Caio Oliveira
60cf1ede4f iris: Remove no-ops from storage image lowering 
The atomics lowering only applies to Gfx <= 7.5.
The get_size lowering only applies to Gfx <= 8.

Note: the lower_store still applies for Gfx9+, to perform color
conversion.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27638>
 2024-02-20 19:47:07 +00:00
Caio Oliveira
078dbc4d33 anv: Remove lower_atomics from storage image lowering opts 
The code only applies to Gfx <= 7.5, so it is a no-op.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27638>
 2024-02-20 19:47:07 +00:00
Mike Blumenkrantz
1ab94ca17b mesa: add more driver support checks for more format queries 
fixes #10641

cc: mesa-stable

Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27698>
 2024-02-20 18:27:10 +00:00
Mike Blumenkrantz
1e849b12f5 vk/wsi/x11/sw: use swapchain depth for putimage 
this is otherwise broken for 32bit depth windows

cc: mesa-stable

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27605>
 2024-02-20 15:47:45 +00:00
Mike Blumenkrantz
198719e80e drisw/egl: delete unused buffer age handling 
age is never used here

Acked-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27554>
 2024-02-20 15:12:38 +00:00
Mike Blumenkrantz
81ecb21d50 drisw: hook up EXT_buffer_age 
drisw only has a single drawable, which means the buffer age is always 1

Acked-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27554>
 2024-02-20 15:12:38 +00:00
Mike Blumenkrantz
9eb4b86e76 egl/x11/swrast: deduplicate ANGLE_sync_control_rate enablement 
no functional changes

Acked-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27554>
 2024-02-20 15:12:37 +00:00
Lionel Landwerlin
0eb3c850c6 intel/clc: workaround LLVM17 opaque pointers 
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Caio Oliveira <caio.oliveira@intel.com>
Fixes: b52e25d3a8 ("anv: rewrite internal shaders using OpenCL")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27637>
 2024-02-20 14:41:43 +00:00
Lionel Landwerlin
62baa4df5f intel/clc: lower temp function/shader variables together 
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Caio Oliveira <caio.oliveira@intel.com>
Fixes: 4fd7495c69 ("intel/clc: add ability to output NIR")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27637>
 2024-02-20 14:41:43 +00:00
Lionel Landwerlin
19aeb274e6 genxml: generate opencl temporary variables with private qualifier 
To avoid generic pointers, makes the NIR prints a bit more readable.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Caio Oliveira <caio.oliveira@intel.com>
Fixes: 41b2ed65 ("genxml: generate opencl packing headers")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27637>
 2024-02-20 14:41:43 +00:00
Gert Wollny
a95a2e4d3f zink/nir_to_spirv: emit ViewportIndex cap also for inputs 
This fixes validation error VUID-VkShaderModuleCreateInfo-pCode-08737 for
piglits:

  spec@oes_viewport_array@viewport-gs-write-simple
  spec@oes_viewport_array@viewport-gs-writes-in-range
  spec@arb_fragment_layer_viewport@viewport-gs-no-write
  spec@arb_fragment_layer_viewport@viewport-gs-write-simple
  spec@arb_fragment_layer_viewport@viewport-gs-writes-in-range
  spec@arb_fragment_layer_viewport@viewport-no-gs
  spec@arb_fragment_layer_viewport@viewport-vs-write-gs-no-write-simple
  spec@arb_fragment_layer_viewport@viewport-vs-write-simple

Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27694>
 2024-02-20 13:39:34 +00:00
Lionel Landwerlin
96ab1d65cc anv: fix emission of Wa_14015055625 
When moving the static part, I missed that the
pipeline->primitive_id_override field isn't set yet when we check it
to emit 3DSTATE_TE.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 1e081bd680 ("anv: split 3DSTATE_TE packing between static & dynamic parts")
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27692>
 2024-02-20 09:49:00 +00:00
Oskar Viljasaar
68f80f9b0f vulkan/properties: Start looping from the next member in GPDP2 
We don't need to start iterating from `pProperties`, as the first member
is already handled in vk_common_GetPhysicalDeviceProperties2. Eliminate
this iteration by starting from pProperties->pNext.

Reviewed-by: Konstantin Seurer <konstantin.seurer@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27671>
 2024-02-20 09:16:23 +00:00