Samuel Pitoiset [Fri, 23 Sep 2022 06:37:46 +0000 (08:37 +0200)]
zink/ci: skip GLX tests with RADV
Can't work with GBM.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Martin Roukala (né Peres) <martin.roukala@mupuf.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18315>
Samuel Pitoiset [Thu, 22 Sep 2022 16:57:48 +0000 (18:57 +0200)]
zink/ci: skip arb_shader_clock with RADV
They are too random.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Martin Roukala (né Peres) <martin.roukala@mupuf.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18315>
Samuel Pitoiset [Mon, 22 Aug 2022 20:31:41 +0000 (22:31 +0200)]
zink/ci: skip ext_external_objects from Piglit with RADV
This subset randomly hangs.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Martin Roukala (né Peres) <martin.roukala@mupuf.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18315>
Lionel Landwerlin [Wed, 20 Jul 2022 14:01:21 +0000 (17:01 +0300)]
intel/nir/rt: fixup alignment of memcpy iterations
Not sure if fixes anything because it's always 16 at least, but this
is more correct.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17396>
Lionel Landwerlin [Wed, 20 Jul 2022 13:53:14 +0000 (16:53 +0300)]
intel/fs: fixup a64 messages
And run algebraic when either int64 for float64 are not supported so
those don't end up in the generated code.
Cc: mesa-stable
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17396>
Lionel Landwerlin [Tue, 19 Jul 2022 19:09:17 +0000 (22:09 +0300)]
intel/nir/rt: store ray query state in scratch
Initially I tried to store ray query state in the RT scratch space but
got the offset wrong. In the end putting this in the scratch surface
makes more sense, especially for non RT stages.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes:
c78be5da300ae3 ("intel/fs: lower ray query intrinsics")
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17396>
Lionel Landwerlin [Tue, 19 Jul 2022 19:04:35 +0000 (22:04 +0300)]
intel/nir/rt: change scratch check validation
It's very unfortunate that we have the RT scratch being conflated with
the usual scratch. In our implementation those are 2 different buffers.
The usual scratch access are done through the scratch surface state
(delivered through thread payload), while RT scratch (which outlives
thread dispatch with shader calls) is its own buffer.
So checking the NIR scratch size makes no sense as we can have normal
scratch accesses completely unrelated to RT scratch accesses.
This change switches the validation by looking at whether the scratch
base pointer intrinsic is being used (which is what we use/abuse to
implement RT scratch).
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes:
c78be5da300ae3 ("intel/fs: lower ray query intrinsics")
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17396>
Lionel Landwerlin [Tue, 19 Jul 2022 18:50:44 +0000 (21:50 +0300)]
intel/nir/rt: fix ray query proceed level
Initially the level is world (top level), then it's whatever level the
potential hit is.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes:
c78be5da300ae3 ("intel/fs: lower ray query intrinsics")
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17396>
Lionel Landwerlin [Wed, 20 Jul 2022 14:03:22 +0000 (17:03 +0300)]
intel/nir/rt: remove ray query mem hit writes at initialization
This will not even be read by HW.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17396>
Lionel Landwerlin [Tue, 19 Jul 2022 18:47:45 +0000 (21:47 +0300)]
intel/nir/rt: spill/fill the entire ray query data
We need the traversal stack to saved/restored along with mem hits.
Total spill/fill is 256bytes.
We can potentially optimize this but we have to be very careful about
what state the query is in.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes:
c78be5da300ae3 ("intel/fs: lower ray query intrinsics")
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17396>
Lionel Landwerlin [Thu, 7 Jul 2022 10:31:46 +0000 (13:31 +0300)]
intel/nir/rt: fixup generate hit
This function copies the potential hit from its memory location to the
committed hit location. A couple of fields got their bit offset wrong.
Fixes some CTS tests in dEQP-VK.ray_query.*
v2: Copy primitive/instance leaf pointers
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes:
046571479028 ("intel/nir/rt: add more helpers for ray queries")
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17396>
Xin LI [Fri, 23 Sep 2022 06:05:53 +0000 (23:05 -0700)]
meson: Improve detection of qsort_r().
Instead of trying to guess the interface style by solely checking
the OS name, check if a fake test program can be built and linked.
This will give more accurate result when FreeBSD and other systems
moved to the interface based on GNU qsort_r().
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18527>
Gert Wollny [Fri, 16 Sep 2022 14:32:33 +0000 (16:32 +0200)]
r600: Make NIR the default shader IR
By default sb is disabled, to enable it one can
set
R600_DEBUG=nirsb
TGSI can be enabled by using
R600_DEBUG=use_tgsi
In this case "sb" will be enabled, but one can use
the TGSI backend without sb by setting
R600_DEBUG=use_tgsi,sb
Closes: #6858
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18753>
Gert Wollny [Fri, 16 Sep 2022 14:43:12 +0000 (16:43 +0200)]
r600: Add slot to ALU disassambly
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18753>
Dave Airlie [Mon, 19 Sep 2022 20:40:21 +0000 (06:40 +1000)]
lavapipe: use slow clear path if conditional rendering.
the ->clear path checks render condition which isn't necessary.
Fixes dEQP-VK.conditional_rendering.draw.condition_host_memory_expect_noop_rp_clear.draw*
Reviewed-by: Brian Paul <brianp@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18686>
Samuel Pitoiset [Wed, 21 Sep 2022 11:35:42 +0000 (13:35 +0200)]
radv: rework and rename radv_make_buffer_descriptor()
Pass the VA directly and rename to radv_make_texel_buffer_descriptor().
This helper will be re-used for future work.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18750>
Samuel Pitoiset [Wed, 21 Sep 2022 11:32:27 +0000 (13:32 +0200)]
radv: remove unnecessary radv_buffer_view::vk_format
It's not really useful.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18750>
Samuel Pitoiset [Tue, 20 Sep 2022 14:03:53 +0000 (16:03 +0200)]
radv: pass a VkSampler to write_sampler_descriptor()
For future work, and remove useless radv_device parameter.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18750>
Iago Toral Quiroga [Thu, 22 Sep 2022 06:41:50 +0000 (08:41 +0200)]
broadcom/compiler: don't apply robust buffer access to shared variables
This feature is only concerned with buffers bound through a descriptor
set. We are still keeping the code for this (disabled by default) since
it may be useful for debugging some scenarios.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18744>
Iago Toral Quiroga [Thu, 22 Sep 2022 07:20:46 +0000 (09:20 +0200)]
broadcom/compiler: handle shared stores with robust buffer access
For some reason we supported all shared intrinsics but this one.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18744>
Iago Toral Quiroga [Thu, 22 Sep 2022 07:18:18 +0000 (09:18 +0200)]
broadcom/compiler: fix robust buffer access
Our implemention was bogus, it was only putting a cap on the offset
based on the aligned buffer size and this doesn't ensure the access
to the buffer happens within its valid range.
I think the only reason we have been passing the tests is that we
align all buffers sizes to 256B and the tests create buffers with a
size that is smaller than that (like 64B). When get the size of the
buffer from the shader, we get the actual bound range (so 64B in this
case) and by capping to that we don't ensure the access will stay
within that range, but we ensure it will stay within the underlying
memory bound to the buffer (256B), and this is fine by the spec,
however, I think if the actual buffer range was the same as the
underlying allocation we would fail the tests.
A valid behavior for robust buffer access on an out-of-bounds access
is to return any valid bytes within the buffer, so we can just
make that offset 0.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18744>
Paul Gofman [Tue, 28 Jun 2022 01:39:04 +0000 (20:39 -0500)]
vbo/dlist: keep buffers used in loopback_vertex_list() mapped.
When display list loopback path is hit the major performance
drop during glCallList() happens due to constantly mapping and
unmapping the buffer.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17285>
Marek Olšák [Wed, 21 Sep 2022 20:12:25 +0000 (16:12 -0400)]
gallium/u_threaded: add an option for unsychronized create_fence_fd
And enable it for radeonsi. This may be needed by Android.
Reviewed-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18731>
Mike Blumenkrantz [Thu, 22 Sep 2022 19:33:48 +0000 (15:33 -0400)]
zink: handle culldistance xfb outputs like clipdistance
fixes:
spec@ext_transform_feedback@builtin-varyings gl_culldistance
cc: mesa-stable
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18724>
Mike Blumenkrantz [Wed, 21 Sep 2022 14:49:33 +0000 (10:49 -0400)]
glsl: fix gl_CullDistance xfb linking
this matches the behavior for gl_ClipDistance
ref #7152
Fixes:
0c8492cd3b2 ("glsl: fix location for array subscript")
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18724>
Caio Oliveira [Wed, 21 Sep 2022 20:43:56 +0000 (13:43 -0700)]
spirv: Make a helper function to check glslang/shaderc
When deciding on workarounds that apply to GLSLang we should also check
for ShaderC (when using glslang). Add a helper to do that and reuse in
the three places we check for it.
This patch will now consider shaderc for wa_glslang_cs_barrier workaround.
Suggested by Timur Kristóf.
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18442>
Caio Oliveira [Mon, 19 Sep 2022 22:19:36 +0000 (15:19 -0700)]
spirv: Call vtn_emit_ret_store() only when handling return branch
No need to call it at the end of every block, since we already
know and identified the blocks terminating with returns.
Suggested by Jason Ekstrand.
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18442>
Caio Oliveira [Mon, 19 Sep 2022 20:50:55 +0000 (13:50 -0700)]
spirv: Add workaround to ignore OpReturn after OpEmitMeshTasksEXT
Fixes:
7d1bcf1f55b ("spirv, nir: Handle EmitMeshTasksEXT opcode.")
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18442>
Timur Kristóf [Tue, 6 Sep 2022 14:12:02 +0000 (16:12 +0200)]
spirv: Treat EmitMeshTasksEXT as a terminating instruction.
glslang had a bug and didn't actually emit it as a terminating
instruction, therefore this went unnoticed thus far. See
https://github.com/KhronosGroup/glslang/issues/3020.
Modified by Caio Oliveira to emit the corresponding NIR intrinsic
right before the halt.
Fixes:
7d1bcf1f55b ("spirv, nir: Handle EmitMeshTasksEXT opcode.")
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18442>
Caio Oliveira [Mon, 19 Sep 2022 20:35:13 +0000 (13:35 -0700)]
spirv: Pass vtn_block into vtn_emit_branch()
This will be used later to handle branches instructions that have
parameters in them, e.g. SpvOpEmitMeshTasksEXT. No behavior change in
this commit.
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18442>
Caio Oliveira [Wed, 21 Sep 2022 14:43:52 +0000 (07:43 -0700)]
nir/lower_task_shader: Don't fail adding a launch when last instruction is a jump
Fixes:
8aff8d3dd42 ("nir: Add common task shader lowering to make the backend's job easier.")
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18442>
Mihai Preda [Tue, 6 Sep 2022 16:00:30 +0000 (19:00 +0300)]
gallivm: LLVM opaque pointers: use lp_build_array_get_ptr2() in lp_bld_nir_soa.c
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18721>
Mihai Preda [Tue, 6 Sep 2022 15:57:42 +0000 (18:57 +0300)]
gallivm: LLVM opaque pointers: add lp_build_array_get[_ptr]2 in lp_bld_struct.c
i.e. variants taking an explicit LLVM type
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18721>
Mihai Preda [Wed, 31 Aug 2022 12:38:52 +0000 (15:38 +0300)]
gallivm: LLVM opaque pointer fixes in lp_bld_sample.c
load_mip()
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18721>
Mihai Preda [Wed, 14 Sep 2022 10:01:49 +0000 (13:01 +0300)]
gallium/draw: fix LLVM opaque pointer warnings in store_aos()
Acked-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18720>
Mihai Preda [Wed, 21 Sep 2022 11:00:02 +0000 (14:00 +0300)]
gallium/draw: fix LLVM opaque pointer warnings in draw_tcs_llvm_emit_store_output()
Acked-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18720>
Mihai Preda [Wed, 21 Sep 2022 10:26:00 +0000 (13:26 +0300)]
gallium/draw: fix LLVM opaque pointer warnings in draw_tcs_llvm_emit_fetch_output()
Acked-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18720>
Mihai Preda [Wed, 21 Sep 2022 10:08:59 +0000 (13:08 +0300)]
gallium/draw: fix LLVM opaque pointer warnings in draw_tcs_llvm_emit_fetch_input()
Acked-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18720>
Mihai Preda [Wed, 21 Sep 2022 09:59:37 +0000 (12:59 +0300)]
gallium/draw: fix LLVM opaque pointer warnings in draw_gs_llvm_epilogue()
Acked-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18720>
Mihai Preda [Wed, 21 Sep 2022 09:43:47 +0000 (12:43 +0300)]
gallium/draw: fix LLVM opaque pointer warnings in draw_gs_llvm_end_primitive()
Acked-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18720>
Mihai Preda [Wed, 14 Sep 2022 11:55:59 +0000 (14:55 +0300)]
gallium/draw: fix LLVM opaque pointer warnings in draw_gs_llvm_fetch_input()
Acked-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18720>
Mihai Preda [Wed, 14 Sep 2022 10:30:51 +0000 (13:30 +0300)]
gallium/draw: fix LLVM opaque pointer warnings in generate_clipmask()
Acked-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18720>
Mihai Preda [Wed, 14 Sep 2022 10:14:04 +0000 (13:14 +0300)]
gallium/draw: fix LLVM opaque pointer warnings in store_clip()
Acked-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18720>
Lionel Landwerlin [Thu, 22 Sep 2022 05:13:22 +0000 (08:13 +0300)]
anv: add missing wokraround for texture cache invalidate
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Nanley Chery <nanley.g.chery@intel.com>
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18743>
Connor Abbott [Thu, 22 Sep 2022 15:32:53 +0000 (17:32 +0200)]
ir3/analyze_ubo_ranges: Account for reserved consts
We weren't accounting for the reserved consts when calculating how much
we can upload. This led to assertion failures later if we pushed too
much.
Fixes:
d3b7681df28 ("tu: ir3: Emit push constants directly")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18757>
Connor Abbott [Thu, 22 Sep 2022 13:19:09 +0000 (15:19 +0200)]
tu: Set textures_used for input attachments correctly
We were accidentally multiplying by 2 twice. Noticed by inspection.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18757>
Connor Abbott [Thu, 22 Sep 2022 13:07:21 +0000 (15:07 +0200)]
tu: Fix maxPerStageDescriptorUpdateAfterBindInputAttachments
We need this to be the same as maxPerStageDescriptorInputAttachments.
Fixes:
d9fcf5de55a ("turnip: Enable nonuniform descriptor indexing")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18757>
Sil Vilerino [Thu, 22 Sep 2022 17:41:20 +0000 (13:41 -0400)]
d3d12: Fix texture array DPB encode path
Mark D3D12_VIDEO_ENCODER_PICTURE_CONTROL_FLAG_USED_AS_REFERENCE_PICTURE when applicable
Reviewed-by: Giancarlo Devich <gdevich@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18763>
Sil Vilerino [Thu, 22 Sep 2022 17:34:28 +0000 (13:34 -0400)]
d3d12: Add debug verbosity for output reconstructed picture
Reviewed-by: Giancarlo Devich <gdevich@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18763>
Sil Vilerino [Thu, 22 Sep 2022 17:07:59 +0000 (13:07 -0400)]
d3d12: Fix out of bounds index check for texture array DPB
Reviewed-by: Giancarlo Devich <gdevich@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18763>
Sil Vilerino [Thu, 22 Sep 2022 11:58:48 +0000 (07:58 -0400)]
d3d12: Relax slice caps requirements for video encode
The check in d3d12_has_video_encode_support was returning no encode support unless
the driver supported D3D12_VIDEO_ENCODER_FRAME_SUBREGION_LAYOUT_MODE_UNIFORM_PARTITIONING_SUBREGIONS_PER_FRAME
when it should also return support when the driver reports only full frame support
(ie. D3D12_VIDEO_ENCODER_FRAME_SUBREGION_LAYOUT_MODE_FULL_FRAME)
Reviewed-by: Giancarlo Devich <gdevich@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18763>
Mike Blumenkrantz [Thu, 1 Sep 2022 14:23:35 +0000 (10:23 -0400)]
zink: delete some descriptor caching cruft
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18664>
Mike Blumenkrantz [Thu, 1 Sep 2022 14:22:51 +0000 (10:22 -0400)]
zink: delete more sw conditional render handling
I forgot to delete this when I deleted the draw path handling
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18664>
Mike Blumenkrantz [Thu, 1 Sep 2022 13:37:06 +0000 (09:37 -0400)]
zink: remove batch usage from sampler states
this was useless
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18664>
Mike Blumenkrantz [Thu, 1 Sep 2022 13:34:41 +0000 (09:34 -0400)]
zink: remove batch tracking/usage from view types
instead of incurring all the overhead of tracking lifetimes for these,
it makes more sense to just let them delete whenever and then store
the vk object onto its parent image/buffer to be deleted when that gets
freed, as the parent object's lifetime will always be >= the view's lifetime
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18664>
Mike Blumenkrantz [Thu, 1 Sep 2022 13:15:30 +0000 (09:15 -0400)]
zink: split out bind_sampler_states hook for drivers without EXT_non_seamless_cube_map
this incurs some overhead and is seldom used, so avoid having it in the codepath
at all if it won't be used
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18664>
Mike Blumenkrantz [Thu, 1 Sep 2022 13:15:06 +0000 (09:15 -0400)]
zink bump batch state limit a bit
this should yield better throughput
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18664>
Mike Blumenkrantz [Wed, 31 Aug 2022 19:38:29 +0000 (15:38 -0400)]
zink: rework batch tracking for resources
this uses a revised version of radeonsi's buffer list array to track
resources for lifetime/usage management, massively improving performance
by avoiding hash table lookups
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18664>
Mike Blumenkrantz [Mon, 29 Aug 2022 15:40:50 +0000 (11:40 -0400)]
zink: defer more semaphore destruction
these have noticeable overhead, so handle them in the submit thread
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18664>
Mike Blumenkrantz [Wed, 21 Sep 2022 13:07:35 +0000 (09:07 -0400)]
zink: rework flush_frontbuffer to always flush
previously this would no-op unacquired swapchains and then try to
execute a copy from the last-presented swapchain image to the current
image to avoid presenting garbage data
instead, just present garbage like the app requested
fixes #7165
Tested-By: Matti Hamalainen <ccr@tnsp.org>
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18730>
Mike Blumenkrantz [Tue, 6 Sep 2022 16:14:01 +0000 (12:14 -0400)]
zink: check format compatibility up front when creating framebuffer surface
this avoids an assert in create_ivci while enforcing correct behavior
cc: mesa-stable
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18446>
Adam Jackson [Tue, 20 Sep 2022 18:42:23 +0000 (14:42 -0400)]
egl: Rewrite eglSetBlobCacheFuncsANDROID to use the standard macros
As the deleted comment hints, you just need the macro to expand to
return /* void */;
for things to work.
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18713>
Adam Jackson [Tue, 20 Sep 2022 18:14:51 +0000 (14:14 -0400)]
egl: Factor out the eglTerminate write-lock pattern
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18713>
Adam Jackson [Tue, 20 Sep 2022 17:51:00 +0000 (13:51 -0400)]
egl: Untangle some spaghetti around _eglLookupDisplay
Only eglapi needs any of this.
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18713>
Adam Jackson [Tue, 20 Sep 2022 17:38:04 +0000 (13:38 -0400)]
egl: Promote _eglLock/UnlockDisplay to internal API
This reads a bit more cleanly than egl_lock/egl_unlock I think.
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18713>
Adam Jackson [Fri, 16 Sep 2022 17:06:18 +0000 (13:06 -0400)]
egl: Fix a bunch of maybe-uninitialized warnings
Fedora's gcc 12 is apparently not smart enough to deduce that these are
always initialized, so set some clear default values.
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18713>
Yonggang Luo [Wed, 21 Sep 2022 18:26:44 +0000 (18:26 +0000)]
Update 22.2.0.rst
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18726>
Dylan Baker [Wed, 21 Sep 2022 16:24:29 +0000 (09:24 -0700)]
docs: Add calendar entries for 22.2 release.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18726>
Dylan Baker [Wed, 21 Sep 2022 16:23:31 +0000 (09:23 -0700)]
docs: update calendar and link releases notes for 22.2.0
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18726>
Dylan Baker [Wed, 21 Sep 2022 16:17:10 +0000 (09:17 -0700)]
docs: Add sha256 sum for 22.2.0
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18726>
Dylan Baker [Wed, 21 Sep 2022 16:16:34 +0000 (09:16 -0700)]
docs: add release notes for 22.2.0
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18726>
Yiwei Zhang [Thu, 22 Sep 2022 05:43:56 +0000 (05:43 +0000)]
venus: ensure to clear VkPipelineCreationFeedback::flags if needed
Fixes:
d0cb99e96a2 ("venus: Enable VK_EXT_pipeline_creation_feedback")
Signed-off-by: Yiwei Zhang <zzyiwei@chromium.org>
Reviewed-by: Ryan Neph <ryanneph@google.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18741>
Emma Anholt [Wed, 21 Sep 2022 23:42:34 +0000 (16:42 -0700)]
turnip: Fix busy-waiting on syncobjs with OS_TIMEOUT_INFINITE.
I noticed that glmark2's glFinish()es in its offscreen rendering tests
under zink were spinning. When we passed -1 as the timeout for
drmSyncobjWait(), the kernel would immediately return ETIME.
Fixes:
0a82a26a1854 ("turnip: Porting to common implementation for timeline semaphore")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18739>
Leo Liu [Thu, 22 Sep 2022 16:53:25 +0000 (12:53 -0400)]
frontends/va: fix build error for vaSyncBuffer with older VA
return VA_STATUS_ERROR_TIMEDOUT;
| ^~~~~~~~~~~~~~~~~~~~~~~~
| VA_STATUS_ERROR_UNKNOWN
../src/gallium/frontends/va/buffer.c:434:14: note: each undeclared identifier is reported only once for each function it appears in
../src/gallium/frontends/va/buffer.c:436:22: error: ‘VA_TIMEOUT_INFINITE’ undeclared (first use in this function); did you mean ‘PIPE_TIMEOUT_INFINITE’?
436 | if (timeout_ns != VA_TIMEOUT_INFINITE)
| ^~~~~~~~~~~~~~~~~~~
| PIPE_TIMEOUT_INFINITE
Fixes:
229c6f79a ("frontends/va: Implement vaSyncBuffer")
Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Ruijing Dong <ruijing.dong@amd.com>
Reviewed-by: Sil Vilerino <sivileri@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18760>
Rob Clark [Tue, 20 Sep 2022 21:15:43 +0000 (14:15 -0700)]
freedreno: Handle xfb invalidation
So using the same buffer for CPU writes and GPU writes might be kinda
unusual. But I noticed we were missing handling for the case while
debugging something unrelated.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18732>
Yonggang Luo [Thu, 22 Sep 2022 08:41:55 +0000 (16:41 +0800)]
lavapipe/meson: rename module_dir to icd_lib_path and swap two lines for consistence
Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18747>
Yonggang Luo [Thu, 22 Sep 2022 08:42:32 +0000 (16:42 +0800)]
dzn: rename module_dir to icd_lib_path for consistence
Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18747>
Yonggang Luo [Thu, 22 Sep 2022 03:58:57 +0000 (11:58 +0800)]
radv: Getting radeon_icd to be generated properly on win32
Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18747>
Tapani Pälli [Thu, 22 Sep 2022 08:17:21 +0000 (11:17 +0300)]
mesa/st: fix a set_sampler_views call parameter order
Fixes:
e7b95619596 ("gallium: implement compute pbo download")
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18745>
Rajnesh Kanwal [Thu, 1 Sep 2022 12:13:12 +0000 (13:13 +0100)]
pvr: Implement vkCmdNextSubpass API.
Signed-off-by: Rajnesh Kanwal <rajnesh.kanwal@imgtec.com>
Reviewed-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18707>
Rajnesh Kanwal [Tue, 30 Aug 2022 18:58:56 +0000 (19:58 +0100)]
pvr: Pass load-op struct instead of idx to support subpass load-ops.
Signed-off-by: Rajnesh Kanwal <rajnesh.kanwal@imgtec.com>
Reviewed-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18707>
Rajnesh Kanwal [Tue, 30 Aug 2022 14:50:15 +0000 (15:50 +0100)]
pvr: Add support to create subpass load ops.
Signed-off-by: Rajnesh Kanwal <rajnesh.kanwal@imgtec.com>
Reviewed-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18707>
Rajnesh Kanwal [Fri, 2 Sep 2022 14:15:42 +0000 (15:15 +0100)]
pvr: Add attachment resolve support.
Signed-off-by: Rajnesh Kanwal <rajnesh.kanwal@imgtec.com>
Reviewed-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18707>
Rajnesh Kanwal [Fri, 2 Sep 2022 14:12:06 +0000 (15:12 +0100)]
pvr: Add const to structs that are not supposed to be modified.
Signed-off-by: Rajnesh Kanwal <rajnesh.kanwal@imgtec.com>
Reviewed-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18707>
Rajnesh Kanwal [Tue, 23 Aug 2022 10:35:25 +0000 (11:35 +0100)]
pvr: Rename userpass_spawn to isp_userpass.
Signed-off-by: Rajnesh Kanwal <rajnesh.kanwal@imgtec.com>
Reviewed-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18707>
Rajnesh Kanwal [Wed, 3 Aug 2022 16:53:32 +0000 (17:53 +0100)]
pvr: Add missing output register write in pvr_CreateRenderPass2.
Signed-off-by: Rajnesh Kanwal <rajnesh.kanwal@imgtec.com>
Reviewed-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18707>
Rajnesh Kanwal [Thu, 4 Aug 2022 13:05:44 +0000 (14:05 +0100)]
pvr: Use ~0 for masks instead of ~0U.
This makes code future proof in case someone changes the type
and also avoids half masks (0x00000000FFFFFFFF) in case of
64bit variables.
Signed-off-by: Rajnesh Kanwal <rajnesh.kanwal@imgtec.com>
Reported-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Reviewed-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18707>
Rajnesh Kanwal [Tue, 5 Jul 2022 11:26:37 +0000 (12:26 +0100)]
pvr: Add support for generating render pass hw setup data.
Signed-off-by: Rajnesh Kanwal <rajnesh.kanwal@imgtec.com>
Reviewed-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18707>
Rajnesh Kanwal [Tue, 2 Aug 2022 12:52:03 +0000 (13:52 +0100)]
pvr: Replace pvr_renderpass_surface_initop with VkAttachmentLoadOp.
Signed-off-by: Rajnesh Kanwal <rajnesh.kanwal@imgtec.com>
Reviewed-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18707>
Rajnesh Kanwal [Tue, 2 Aug 2022 12:27:35 +0000 (13:27 +0100)]
pvr: General reformatting and renaming some fields.
Mainly renaming driver_id and similar fields to index to
better represent what they contain.
Signed-off-by: Rajnesh Kanwal <rajnesh.kanwal@imgtec.com>
Reviewed-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18707>
Eric Engestrom [Thu, 22 Sep 2022 12:33:29 +0000 (14:33 +0200)]
docs: update calendar and link releases notes for 22.1.7
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18752>
Eric Engestrom [Thu, 22 Sep 2022 12:58:07 +0000 (14:58 +0200)]
docs: add release notes for 22.1.7
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18752>
Eric Engestrom [Thu, 22 Sep 2022 12:33:27 +0000 (14:33 +0200)]
docs: update calendar and link releases notes for 22.1.6
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18752>
Eric Engestrom [Thu, 22 Sep 2022 12:55:57 +0000 (14:55 +0200)]
docs: add release notes for 22.1.6
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18752>
Daniel Schürmann [Wed, 24 Aug 2022 17:13:52 +0000 (19:13 +0200)]
aco/optimizer: change inverse_comparison in-place
This avoids creating a second comparison which is more expensive
than just using the existing s_not.
Totals from 1089 (0.81% of 134913) affected shaders: (GFX10.3)
VGPRs: 50472 -> 50184 (-0.57%)
CodeSize: 4724692 -> 4760824 (+0.76%); split: -0.03%, +0.79%
MaxWaves: 23964 -> 24012 (+0.20%)
Instrs: 859588 -> 859687 (+0.01%); split: -0.11%, +0.12%
Latency:
10674653 ->
10650353 (-0.23%); split: -0.41%, +0.18%
InvThroughput: 1752987 -> 1750238 (-0.16%); split: -0.20%, +0.04%
VClause: 20921 -> 20872 (-0.23%); split: -0.68%, +0.45%
SClause: 31417 -> 31550 (+0.42%)
Copies: 69428 -> 68738 (-0.99%); split: -1.52%, +0.53%
PreSGPRs: 48033 -> 49649 (+3.36%)
PreVGPRs: 44490 -> 43699 (-1.78%)
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18253>
Timur Kristóf [Sat, 20 Aug 2022 20:55:45 +0000 (22:55 +0200)]
aco: Change inverse-comparison optimization to work with s_not
Some time ago we stopped using s_andn2 with exec for boolean NOT.
The reasoning behind that change was that those booleans will be
always ANDed with exec when necessary.
This inhibited the inverse-comparison optimization in most cases
which is fixed by this patch.
Fossil DB stats on Navi 21:
Totals from 12251 (9.08% of 134913) affected shaders:
VGPRs: 801744 -> 802016 (+0.03%); split: -0.00%, +0.04%
SpillSGPRs: 8863 -> 8893 (+0.34%)
CodeSize:
100593244 ->
100370684 (-0.22%); split: -0.22%, +0.00%
MaxWaves: 204994 -> 204948 (-0.02%); split: +0.00%, -0.02%
Instrs:
18717001 ->
18668965 (-0.26%); split: -0.26%, +0.00%
Latency:
263255046 ->
262874896 (-0.14%); split: -0.16%, +0.02%
InvThroughput:
52760249 ->
52721736 (-0.07%); split: -0.08%, +0.01%
VClause: 329631 -> 329680 (+0.01%); split: -0.03%, +0.04%
SClause: 681563 -> 681435 (-0.02%); split: -0.02%, +0.00%
Copies: 1331612 -> 1372446 (+3.07%); split: -0.03%, +3.10%
Branches: 548325 -> 548301 (-0.00%)
PreSGPRs: 911317 -> 909700 (-0.18%)
PreVGPRs: 766279 -> 767070 (+0.10%)
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18253>
Daniel Schürmann [Wed, 24 Aug 2022 13:12:40 +0000 (15:12 +0200)]
aco/optimizer: optimize s_and(exec, s_and(x, y)) more aggressively
It is enough if either of the two operands is respecting the
current exec mask.
Totals from 1680 (1.25% of 134913) affected shaders: (GFX10.3)
CodeSize: 3929436 -> 3922372 (-0.18%); split: -0.18%, +0.00%
Instrs: 730305 -> 728536 (-0.24%); split: -0.24%, +0.00%
Latency: 6839314 -> 6835154 (-0.06%); split: -0.07%, +0.01%
InvThroughput: 1371351 -> 1371267 (-0.01%); split: -0.01%, +0.00%
SClause: 32819 -> 32802 (-0.05%); split: -0.09%, +0.04%
Copies: 33264 -> 33271 (+0.02%); split: -0.01%, +0.03%
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18253>
Daniel Schürmann [Wed, 24 Aug 2022 10:25:23 +0000 (12:25 +0200)]
aco/optimizer: do can_eliminate_and_exec() optimization later
This will allow to optimize s_not(v_cmp()) safely.
Totals from 1024 (0.76% of 134913) affected shaders: (GFX10.3)
CodeSize: 5695424 -> 5701860 (+0.11%); split: -0.00%, +0.11%
Scratch: 242688 -> 241664 (-0.42%)
Instrs: 1040656 -> 1041635 (+0.09%); split: -0.00%, +0.09%
Latency:
16842282 ->
16922790 (+0.48%); split: -0.06%, +0.54%
InvThroughput: 4772728 -> 4810868 (+0.80%); split: -0.10%, +0.90%
VClause: 20013 -> 20000 (-0.06%); split: -0.12%, +0.05%
Copies: 115057 -> 114384 (-0.58%); split: -1.22%, +0.63%
Branches: 34531 -> 34532 (+0.00%); split: -0.00%, +0.01%
PreSGPRs: 46263 -> 46267 (+0.01%)
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18253>
Erik Faye-Lund [Fri, 22 Jul 2022 09:35:10 +0000 (11:35 +0200)]
zink: assert on unnormalized texcoords
We do not support PIPE_CAP_TEXRECT, so let's assert in case something
tries to use it behind our back.
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17708>
Erik Faye-Lund [Tue, 20 Sep 2022 08:12:50 +0000 (10:12 +0200)]
mesa/st: always use normalized coords for samplers
Normalized samplers is the norm, and non-normalized samplers might cause
some drivers and hardware to have to bend over backwards a bit.
We're using TXF here anyway, so it doesn't really matter what we set
this state to. So let's always use normalized samplers instead of
always using non-normalized samplers. That makes things easier for
everyone.
Fixes:
e7b95619596 ("gallium: implement compute pbo download")
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17708>