mesa.git
17 months agovulkan/wsi: Add callback to set ownership of buffer.
Bas Nieuwenhuizen [Tue, 24 Mar 2020 16:59:07 +0000 (17:59 +0100)]
vulkan/wsi: Add callback to set ownership of buffer.

For radv BO list pruning.

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Acked-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4306>

17 months agoac,radeonsi: fix compilations issues with LLVM 11
Samuel Pitoiset [Mon, 27 Apr 2020 09:01:24 +0000 (11:01 +0200)]
ac,radeonsi: fix compilations issues with LLVM 11

Latest LLVM replaced LLVMVectorTypeKind.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/2826
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Michel Dänzer <mdaenzer@redhat.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4755>

17 months agogallium/gallivm: remove unused header include for newer LLVM
Jan Zielinski [Mon, 27 Apr 2020 11:34:59 +0000 (13:34 +0200)]
gallium/gallivm: remove unused header include for newer LLVM

In the top of the trunk LLVM (11) llvm/IR/CallSite.h header
has been removed. The file compiles without this include also
for LLVM 8, but I'm not sure about 9, 10, and older versions
so I disable it only for the latest LLVM

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4748>

17 months agogallium/gallivm: fix compilation issues with llvm 11
Jan Zielinski [Sat, 25 Apr 2020 17:27:07 +0000 (19:27 +0200)]
gallium/gallivm: fix compilation issues with llvm 11

Top of the trunk LLVM removes old vector type
and introduces two new ones - one for fixed-width
and one for scalable vectors. This commit fixes
compilation issues by switching from LLVMVectorTypeKind
to LLVMFixedVectorTypeKind for new LLVM.

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4748>

17 months agoir3: Use shared mediump output lowering
Alyssa Rosenzweig [Thu, 23 Apr 2020 23:12:43 +0000 (19:12 -0400)]
ir3: Use shared mediump output lowering

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Acked-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4716>

17 months agonir: Move nir_lower_mediump_outputs from ir3
Alyssa Rosenzweig [Thu, 23 Apr 2020 23:10:43 +0000 (19:10 -0400)]
nir: Move nir_lower_mediump_outputs from ir3

(Original code from ir3)

Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Acked-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4716>

17 months agonir/algebraic: Detect some kinds of malformed variable names
Ian Romanick [Fri, 24 Apr 2020 00:39:07 +0000 (17:39 -0700)]
nir/algebraic: Detect some kinds of malformed variable names

I spent over an hour trying to debug a problem if a condition on a
variable not being applied.  The problem turned out to be
"a(is_not_negative" instead of "a(is_not_negative)".  This commit would
have detected that problem and failed to build.

v2: Just add $ to the end of the existing regex, and it will fail to
match a malformed string.  Suggested by Jason.

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com> [v1]
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4720>

17 months agopan/bi: Implement 16-bit COMBINE lowering
Alyssa Rosenzweig [Mon, 27 Apr 2020 14:37:20 +0000 (10:37 -0400)]
pan/bi: Implement 16-bit COMBINE lowering

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bi: Fix RA wrt 16-bit swizzles
Alyssa Rosenzweig [Mon, 27 Apr 2020 14:38:00 +0000 (10:38 -0400)]
pan/bi: Fix RA wrt 16-bit swizzles

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bit: Add SELECT tests
Alyssa Rosenzweig [Sat, 25 Apr 2020 00:52:40 +0000 (20:52 -0400)]
pan/bit: Add SELECT tests

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bit: Interpret BI_SELECT
Alyssa Rosenzweig [Fri, 24 Apr 2020 23:14:21 +0000 (19:14 -0400)]
pan/bit: Interpret BI_SELECT

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bi: Force BI_SELECT arguments scalar
Alyssa Rosenzweig [Sat, 25 Apr 2020 00:52:28 +0000 (20:52 -0400)]
pan/bi: Force BI_SELECT arguments scalar

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bi: Pack ADD SEL16
Alyssa Rosenzweig [Sat, 25 Apr 2020 00:52:17 +0000 (20:52 -0400)]
pan/bi: Pack ADD SEL16

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bi: Pack FMA SEL8
Alyssa Rosenzweig [Sat, 25 Apr 2020 00:48:59 +0000 (20:48 -0400)]
pan/bi: Pack FMA SEL8

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bi: Pack FMA SEL16
Alyssa Rosenzweig [Sat, 25 Apr 2020 00:13:17 +0000 (20:13 -0400)]
pan/bi: Pack FMA SEL16

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bi: Rename BI_SWIZZLE to BI_SELECT
Alyssa Rosenzweig [Fri, 24 Apr 2020 23:10:44 +0000 (19:10 -0400)]
pan/bi: Rename BI_SWIZZLE to BI_SELECT

The select version is more general.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bi: Eliminate writemasks in the IR
Alyssa Rosenzweig [Fri, 24 Apr 2020 21:20:28 +0000 (17:20 -0400)]
pan/bi: Eliminate writemasks in the IR

Since the hardware doesn't support them, they're a burden to deal with,
so let's ensure we never get to a place where we would need to at all.

Disables COMBINE lowering for now.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bi: Fix ADD.v4i8 opcode
Alyssa Rosenzweig [Fri, 24 Apr 2020 21:20:15 +0000 (17:20 -0400)]
pan/bi: Fix ADD.v4i8 opcode

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bi: Add missing BI_VECTOR
Alyssa Rosenzweig [Fri, 24 Apr 2020 21:20:07 +0000 (17:20 -0400)]
pan/bi: Add missing BI_VECTOR

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bi: Assign blend descriptor for BLEND op
Alyssa Rosenzweig [Fri, 10 Apr 2020 04:25:50 +0000 (00:25 -0400)]
pan/bi: Assign blend descriptor for BLEND op

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bi: Passthrough blend types
Alyssa Rosenzweig [Thu, 23 Apr 2020 23:26:01 +0000 (19:26 -0400)]
pan/bi: Passthrough blend types

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bi: Passthrough type for ATEST
Alyssa Rosenzweig [Thu, 23 Apr 2020 23:03:44 +0000 (19:03 -0400)]
pan/bi: Passthrough type for ATEST

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agopan/bi: Pack fp16 ATEST
Alyssa Rosenzweig [Thu, 23 Apr 2020 23:03:35 +0000 (19:03 -0400)]
pan/bi: Pack fp16 ATEST

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4766>

17 months agomesa: Skip 3-byte array formats in _mesa_array_format_flip_channels
Michel Dänzer [Fri, 24 Apr 2020 09:40:14 +0000 (11:40 +0200)]
mesa: Skip 3-byte array formats in _mesa_array_format_flip_channels

Byte swapping makes no sense for 3-byte formats: Swapping the order of 2
or 4 bytes at a time would inevitably result in bytes getting mixed up
between neighbouring pixels.

Fixes crash with a debugging build on a big endian machine due hitting
the unreachable() at the end of the function.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/2665
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4735>

17 months agomesa: replace GLenum target with gl_shader_stage in NewProgram
Marek Olšák [Tue, 31 Mar 2020 14:03:55 +0000 (10:03 -0400)]
mesa: replace GLenum target with gl_shader_stage in NewProgram

So that the GLSL compiler doesn't have to use the GLenum conversion
functions.

Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4756>

17 months agodrm-uapi,radv,radeonsi: Add amdgpu_drm.h header.
Bas Nieuwenhuizen [Sat, 25 Apr 2020 18:03:15 +0000 (20:03 +0200)]
drm-uapi,radv,radeonsi: Add amdgpu_drm.h header.

Use it instead of the libdrm provided amdgpu_drm.h header. I used
the kernel revision from the README to get the header so the
header versions should be consistent.

Tested by removing /usr/include/libdrm/amdgpu_drm.h from my dev-machine.

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/4749>

17 months agomesa: extend _mesa_bind_vertex_buffer to take ownership of the buffer reference
Marek Olšák [Sun, 22 Mar 2020 22:13:45 +0000 (18:13 -0400)]
mesa: extend _mesa_bind_vertex_buffer to take ownership of the buffer reference

This reduces overhead of _mesa_reference_buffer_object_ from 6% to 4%
with glthread when profiling the game "torcs" with non-VBO data uploaded
by glthread.

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4758>

17 months agomesa: add offset_is_int32 param into _mesa_bind_vertex_buffer for glthread
Marek Olšák [Sat, 7 Mar 2020 02:23:11 +0000 (21:23 -0500)]
mesa: add offset_is_int32 param into _mesa_bind_vertex_buffer for glthread

glthread will pass signed integer offsets, so don't reset negative offsets
to 0 there.

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4758>

17 months agomesa: add Const.BufferCreateMapUnsynchronizedThreadSafe & MESA_MAP_THREAD_SAFE
Marek Olšák [Sun, 8 Mar 2020 21:43:26 +0000 (17:43 -0400)]
mesa: add Const.BufferCreateMapUnsynchronizedThreadSafe & MESA_MAP_THREAD_SAFE

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4758>

17 months agogallium: add PIPE_CAP_MAP_UNSYNCHRONIZED_THREAD_SAFE for glthread
Marek Olšák [Fri, 6 Mar 2020 02:49:30 +0000 (21:49 -0500)]
gallium: add PIPE_CAP_MAP_UNSYNCHRONIZED_THREAD_SAFE for glthread

and add radeonsi support.

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4758>

17 months agoglthread: sort variables in marshal structures to pack them optimally
Marek Olšák [Mon, 23 Mar 2020 23:39:40 +0000 (19:39 -0400)]
glthread: sort variables in marshal structures to pack them optimally

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4758>

17 months agoglthread: use GLenum16 in batch buffers to save space
Marek Olšák [Mon, 23 Mar 2020 01:07:38 +0000 (21:07 -0400)]
glthread: use GLenum16 in batch buffers to save space

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4758>

17 months agoglthread: reduce dereferences of the next batch
Marek Olšák [Fri, 27 Mar 2020 10:06:31 +0000 (06:06 -0400)]
glthread: reduce dereferences of the next batch

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4758>

17 months agoglthread: use 32-bit align instead of 64-bit ALIGN
Marek Olšák [Fri, 27 Mar 2020 09:57:56 +0000 (05:57 -0400)]
glthread: use 32-bit align instead of 64-bit ALIGN

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4758>

17 months agomesa: remove exec="dynamic" from Draw functions that are not really dynamic
Marek Olšák [Sun, 22 Mar 2020 23:35:28 +0000 (19:35 -0400)]
mesa: remove exec="dynamic" from Draw functions that are not really dynamic

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4758>

17 months agomesa: reset primitive restart state in glClientAttribDefaultEXT
Marek Olšák [Sun, 22 Mar 2020 00:01:37 +0000 (20:01 -0400)]
mesa: reset primitive restart state in glClientAttribDefaultEXT

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4758>

17 months agomesa: replace _NEW_EVAL with vbo_exec_update_eval_maps
Marek Olšák [Sat, 21 Mar 2020 05:40:30 +0000 (01:40 -0400)]
mesa: replace _NEW_EVAL with vbo_exec_update_eval_maps

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4758>

17 months agoac: reassociate FP expressions for inexact instructions for radeonsi
Marek Olšák [Thu, 23 Apr 2020 03:01:28 +0000 (23:01 -0400)]
ac: reassociate FP expressions for inexact instructions for radeonsi

Totals:
SGPRS: 2591784 -> 2590696 (-0.04 %)
VGPRS: 1666888 -> 1666736 (-0.01 %)
Spilled SGPRs: 4131 -> 4107 (-0.58 %)
Spilled VGPRs: 38 -> 38 (0.00 %)
Private memory VGPRs: 2176 -> 2176 (0.00 %)
Scratch size: 2228 -> 2228 (0.00 %) dwords per thread
Code Size: 52715468 -> 52693584 (-0.04 %) bytes
LDS: 92 -> 92 (0.00 %) blocks
Max Waves: 479897 -> 479892 (-0.00 %)
Wait states: 0 -> 0 (0.00 %)

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4696>

17 months agoac: generate FMA for inexact instructions for radeonsi
Marek Olšák [Thu, 23 Apr 2020 02:45:45 +0000 (22:45 -0400)]
ac: generate FMA for inexact instructions for radeonsi

NIR mostly does this already.

Totals:
SGPRS: 2588520 -> 2591784 (0.13 %)
VGPRS: 1666984 -> 1666888 (-0.01 %)
Spilled SGPRs: 4074 -> 4131 (1.40 %)
Spilled VGPRs: 38 -> 38 (0.00 %)
Private memory VGPRs: 2176 -> 2176 (0.00 %)
Scratch size: 2228 -> 2228 (0.00 %) dwords per thread
Code Size: 52726872 -> 52715468 (-0.02 %) bytes
LDS: 92 -> 92 (0.00 %) blocks
Max Waves: 479872 -> 479897 (0.01 %)
Wait states: 0 -> 0 (0.00 %)

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4696>

17 months agoac: update and document fast math flags used by radeonsi
Marek Olšák [Thu, 23 Apr 2020 02:38:14 +0000 (22:38 -0400)]
ac: update and document fast math flags used by radeonsi

This should have no effect, because we never use FP division, but
it's safer for the future.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4696>

17 months agoac: force enable -structurizecfg-skip-uniform-regions for LLVM 11
Marek Olšák [Thu, 23 Apr 2020 01:05:36 +0000 (21:05 -0400)]
ac: force enable -structurizecfg-skip-uniform-regions for LLVM 11

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4696>

17 months agost/mesa: Treat vertex inputs absent in inputMapping as zero in mesa_to_tgsi
Danylo Piliaiev [Thu, 23 Apr 2020 09:41:34 +0000 (12:41 +0300)]
st/mesa: Treat vertex inputs absent in inputMapping as zero in mesa_to_tgsi

After updating vertex inputs being read based on optimized NIR, they may go out
of sync with inputs in mesa IR. Which is translated to TGSI and used together
with NIR if draw doesn't have llvm.

It's much easier to treat such inputs as zero because there is no pass to
entirely get rid of them and they don't contribute to shader's output.

Fixes: d684fb37bfbc47d098158cb03c0672119a4469fe
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/2815
Signed-off-by: Danylo Piliaiev <danylo.piliaiev@globallogic.com>
Reviewed-by: Jose Maria Casanova Crespo <jmcasanova@igalia.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4705>

17 months agogitlab-ci: add lists of expected failures for RADV CI
Samuel Pitoiset [Mon, 20 Apr 2020 07:02:18 +0000 (09:02 +0200)]
gitlab-ci: add lists of expected failures for RADV CI

Currently only supports PITCAIRN, POLARIS10, VEGA10 and NAVI10
with ACO only, but it's a start.

Unfortunately, we have to duplicate and we will have to try to
keep these lists up-to-date, but it's better than nothing.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4689>

17 months agoradv: fix robust_buffer_access if enabled via VkPhysicalDeviceFeatures2
Samuel Pitoiset [Thu, 23 Apr 2020 12:05:07 +0000 (14:05 +0200)]
radv: fix robust_buffer_access if enabled via VkPhysicalDeviceFeatures2

It can be enabled via pEnabledFeatures or via vkPhysicalDeviceFeatures2.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4706>

17 months agogallivm: fix stencil border
Dave Airlie [Fri, 24 Apr 2020 01:23:19 +0000 (11:23 +1000)]
gallivm: fix stencil border

Fixes:
dEQP-GLES31.functional.texture.border_clamp.unused_channels.depth32f_stencil8_sample_stencil
dEQP-GLES31.functional.texture.border_clamp.sampler.uint_stencil

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4574>

17 months agollvmpipe: clamp color storage for integer types.
Dave Airlie [Tue, 21 Apr 2020 05:28:38 +0000 (15:28 +1000)]
llvmpipe: clamp color storage for integer types.

If storing to an integer for lower bit size (i.e. 16-bit uint to
10-bit uint), we need to clamp to the maximum value not truncate.

Fixes:
dEQP-VK.api.copy_and_blit.core.blit_image.all_formats.color.r16_uint.a2b10g10r10_uint_pack32.optimal_optimal_nearest

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4574>

17 months agollvmpipe: enable stencil only formats. (v2)
Dave Airlie [Tue, 3 Dec 2019 06:01:37 +0000 (16:01 +1000)]
llvmpipe: enable stencil only formats. (v2)

This fixes two bugs, one in clearing and one in sign extensions
for S8 only types, and enables it for use.

These are useful for vulkan support later.

v2: move casting to same place as Z casting.

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4574>

17 months agollvmpipe/setup: add point size clamping
Dave Airlie [Thu, 16 Apr 2020 07:15:28 +0000 (17:15 +1000)]
llvmpipe/setup: add point size clamping

Fixes
dEQP-GLES2.functional.rasterization.limits.points
dEQP-VK.rasterization.primitive_size.points.point_size*

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4574>

17 months agollvmpipe: fix d32 unorm depth conversions.
Dave Airlie [Wed, 22 Apr 2020 00:05:59 +0000 (10:05 +1000)]
llvmpipe: fix d32 unorm depth conversions.

When the depth value was 1.0 and was being converted to Z32_UNORM
the conversion would scale it up to INT32_MAX + 1 which would
cause FPToSI to give incorrect results, changing it to use
FPToUI for the unsigned 32-bit case only fixes it.

Fixes:
GTF-GL45.gtf30.GL3Tests.depth_texture.depth_texture_fbo_clear

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4574>

17 months agodraw/tess: fix TES patch vertices in.
Dave Airlie [Mon, 6 Apr 2020 06:40:04 +0000 (16:40 +1000)]
draw/tess: fix TES patch vertices in.

Fixes CTS KHR-GL45.tessellation_shader.single.max_patch_vertices

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4574>

17 months agollvmpipe: fix ssbo alignment
Dave Airlie [Mon, 6 Apr 2020 07:00:19 +0000 (17:00 +1000)]
llvmpipe: fix ssbo alignment

KHR-GL45.geometry_shader.api.max_shader_storage_blocks

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4574>

17 months agollvmpipe: bump max images to 16
Dave Airlie [Tue, 7 Apr 2020 02:08:20 +0000 (12:08 +1000)]
llvmpipe: bump max images to 16

This is needed to make some tests run, and helps for vulkan later.

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4574>

17 months agoutil/indirect: handle stride less than number of parameters.
Dave Airlie [Mon, 20 Apr 2020 05:39:50 +0000 (15:39 +1000)]
util/indirect: handle stride less than number of parameters.

It's legal to have a stride less than the num of parameters,
in this case no need to try and over map the buffer which asserts

Fixes:
GTF-GL45.gtf43.GL3Tests.multi_draw_indirect.multi_draw_indirect_stride

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4574>

17 months agogallivm/nir: add helper invocation support
Dave Airlie [Fri, 27 Mar 2020 05:27:26 +0000 (15:27 +1000)]
gallivm/nir: add helper invocation support

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4574>

17 months agogallivm/nir: fix image store conversions
Dave Airlie [Tue, 7 Apr 2020 05:06:26 +0000 (15:06 +1000)]
gallivm/nir: fix image store conversions

This fixes a few of the image store paths, to do the
correct clamping of unsigned/signed values

Fixes: KHR-GLES31.core.layout_binding.block_layout_binding_block_ComputeShader
KHR-GL45.shader_image_load_store.basic-allFormats-store
KHR-GL46.shader_image_load_store.multiple-uniforms

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4574>

17 months agotu: Don't invert point coords
Connor Abbott [Fri, 24 Apr 2020 13:53:19 +0000 (15:53 +0200)]
tu: Don't invert point coords

We shouldn't need to invert them, and the Vulkan blob doesn't either.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4733>

17 months agoir3: Remove VARYING_SLOT_PNTC remapping hack
Connor Abbott [Thu, 23 Apr 2020 14:23:18 +0000 (16:23 +0200)]
ir3: Remove VARYING_SLOT_PNTC remapping hack

The st now does this for us.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4732>

17 months agost/nir: Fix assigning PointCoord location with !PIPE_CAP_TEXCOORD
Connor Abbott [Thu, 23 Apr 2020 14:08:21 +0000 (16:08 +0200)]
st/nir: Fix assigning PointCoord location with !PIPE_CAP_TEXCOORD

This was trying to emulate the effect of mapping GL -> TGSI -> NIR,
but failed to handle VARYING_SLOT_PNTC which led to a kludgy workaround
in freedreno.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4732>

17 months agofreedreno/a6xx: Implement PrimID passthrough
Connor Abbott [Thu, 23 Apr 2020 09:56:07 +0000 (11:56 +0200)]
freedreno/a6xx: Implement PrimID passthrough

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4704>

17 months agotu: Implement PrimID passthrough
Connor Abbott [Wed, 22 Apr 2020 13:27:24 +0000 (15:27 +0200)]
tu: Implement PrimID passthrough

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4704>

17 months agoir3: Skip missing VS outputs in VS out map when linking
Connor Abbott [Wed, 22 Apr 2020 15:54:41 +0000 (17:54 +0200)]
ir3: Skip missing VS outputs in VS out map when linking

The hardware is capable of automatically filling in certain values in
the VPC without writing them from the last geometry stage, like
gl_PointCoord or gl_PrimitiveID when there is no GS. However, we *do*
have to enable these outputs (i.e. set the VPC_VAR_DISABLE bit to 0) as
VPC_VAR_DISABLE is really about FS inputs rather than VS outputs. To do
this, we move the computation of the enable bits to ir3_link_add(),
which is also a nice refactor anyway. In addition we detect the PrimID
case specifically so that the driver can program the location.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4704>

17 months agofreedreno/a6xx: Document PrimID passthrough registers
Connor Abbott [Wed, 22 Apr 2020 13:04:25 +0000 (15:04 +0200)]
freedreno/a6xx: Document PrimID passthrough registers

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4704>

17 months agoradv: Pass logical device to si_emit_graphics
Joshua Ashton [Fri, 24 Apr 2020 10:13:51 +0000 (11:13 +0100)]
radv: Pass logical device to si_emit_graphics

We'll need this in order to retrieve the va of a bo for a future ext.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4728>

17 months agofreedreno/ir3: Print @tex write mask using 0x%x
Kristian H. Kristensen [Thu, 23 Apr 2020 18:34:31 +0000 (11:34 -0700)]
freedreno/ir3: Print @tex write mask using 0x%x

That way we can parse it again with the assembler.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4741>

17 months agofreedreno/ir3: Reset lex line number when we start parsing
Kristian H. Kristensen [Thu, 23 Apr 2020 18:33:57 +0000 (11:33 -0700)]
freedreno/ir3: Reset lex line number when we start parsing

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4741>

17 months agofreedreno/ir3: Parse, but ignore @in, @out and @tex headers
Kristian H. Kristensen [Thu, 23 Apr 2020 18:33:04 +0000 (11:33 -0700)]
freedreno/ir3: Parse, but ignore @in, @out and @tex headers

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4741>

17 months agofreedreno/ir3: Move ir3 assembler to backend compiler
Kristian H. Kristensen [Wed, 22 Apr 2020 23:57:52 +0000 (16:57 -0700)]
freedreno/ir3: Move ir3 assembler to backend compiler

For easier reuse.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4741>

17 months agofreedreno/computerator: Decouple ir3 assembler
Kristian H. Kristensen [Wed, 22 Apr 2020 23:52:46 +0000 (16:52 -0700)]
freedreno/computerator: Decouple ir3 assembler

Specifically, don't include ir3_asm.h in the parser as that's
computerator specific.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4741>

17 months agoRevert "meson,ci: Disable sparse_array tests on windows"
Andres Gomez [Wed, 22 Apr 2020 13:36:33 +0000 (16:36 +0300)]
Revert "meson,ci: Disable sparse_array tests on windows"

The Wine version in the build image has been upgraded.

This reverts commit 6be65b077743fc80efe061b1e05cb13b2ff1a6b1.

Reviewed-by: Michel Dänzer <mdaenzer@redhat.com>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4678>

17 months agogitlab-ci: install winehq-stable to get 5.0 instead of 4.0
Andres Gomez [Wed, 22 Apr 2020 13:51:48 +0000 (16:51 +0300)]
gitlab-ci: install winehq-stable to get 5.0 instead of 4.0

Additionally, purge the winehq-stable package and its dependencies to
avoid crashing when building for s390x.

v2:
  - Remove winehq-stable and dependencies for s390x.

Closes: https://gitlab.freedesktop.org/mesa/mesa/issues/2657
Signed-off-by: Andres Gomez <agomez@igalia.com>
Reviewed-by: Daniel Stone <daniels@collabora.com> [v1]
Reviewed-by: Michel Dänzer <mdaenzer@redhat.com>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4678>

17 months agoetnaviv: Fix depth stencil ops on GC880/GC2000
Marek Vasut [Mon, 9 Mar 2020 03:14:26 +0000 (04:14 +0100)]
etnaviv: Fix depth stencil ops on GC880/GC2000

This patch fixes depth stencil ops on MX6S GC880 and MX6Q GC2000.
The following dEQPs now pass:
  dEQP-GLES2.functional.fragment_ops.depth_stencil.stencil_depth_funcs.*
  dEQP-GLES2.functional.fragment_ops.depth_stencil.stencil_ops.*
which is roughly 600 fixed dEQP tests.

The problem is that if the front-facing stencil has a value mask 0x00 and
the back-facing stencil has some non-zero value mask, then the stencil part
of the depth stencil buffer is written with 0x00 unconditionally. The blob
replicates the value mask of the back-facing stencil to the value mask of
the front-facing stencil to achieve correct rendering, replicate the same
behavior here.

Signed-off-by: Marek Vasut <marex@denx.de>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4275>

17 months agoradv/aco: enable 8/16-bit storage and int8/int16 on GFX8+
Rhys Perry [Mon, 6 Apr 2020 10:22:55 +0000 (11:22 +0100)]
radv/aco: enable 8/16-bit storage and int8/int16 on GFX8+

With this, Doom Eternal should now run with ACO on GFX8+.

The generated 8/16-bit storage code is okay but the generated int8/int16
code is currently pretty bad but it works and apparently Doom Eternal
doesn't actually use it (even though it requires it).

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4707>

17 months agoaco: lower 8/16-bit integer arithmetic
Rhys Perry [Mon, 6 Apr 2020 10:15:00 +0000 (11:15 +0100)]
aco: lower 8/16-bit integer arithmetic

dEQP-VK.spirv_assembly.type.* passes with the features and extensions
enabled.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4707>

17 months agoaco: improve sub-dword emit_split_vector() with sgprs
Rhys Perry [Thu, 23 Apr 2020 16:23:21 +0000 (17:23 +0100)]
aco: improve sub-dword emit_split_vector() with sgprs

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: clobber scc in s_bfe_u32 in get_alu_src()
Rhys Perry [Thu, 23 Apr 2020 16:17:49 +0000 (17:17 +0100)]
aco: clobber scc in s_bfe_u32 in get_alu_src()

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: handle undef p_create_vector operands in the optimizer
Rhys Perry [Thu, 16 Apr 2020 19:18:23 +0000 (20:18 +0100)]
aco: handle undef p_create_vector operands in the optimizer

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: vectorize global loads/stores
Rhys Perry [Thu, 16 Apr 2020 19:15:00 +0000 (20:15 +0100)]
aco: vectorize global loads/stores

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: allow 8/16-bit shared loads
Rhys Perry [Thu, 16 Apr 2020 15:01:31 +0000 (16:01 +0100)]
aco: allow 8/16-bit shared loads

These should work now

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: add and use get_buffer_store_op() helper
Rhys Perry [Thu, 16 Apr 2020 18:46:09 +0000 (19:46 +0100)]
aco: add and use get_buffer_store_op() helper

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: refactor visit_store_scratch() to use new helpers
Rhys Perry [Thu, 16 Apr 2020 18:27:13 +0000 (19:27 +0100)]
aco: refactor visit_store_scratch() to use new helpers

Should support 8/16-bit stores now

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: refactor visit_store_global() to use new helpers
Rhys Perry [Thu, 16 Apr 2020 18:25:06 +0000 (19:25 +0100)]
aco: refactor visit_store_global() to use new helpers

Should support 8/16-bit stores now

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: refactor visit_store_ssbo() to use new helpers
Rhys Perry [Thu, 16 Apr 2020 18:22:23 +0000 (19:22 +0100)]
aco: refactor visit_store_ssbo() to use new helpers

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: refactor store_vmem_mubuf() to use new helpers
Rhys Perry [Thu, 16 Apr 2020 18:20:26 +0000 (19:20 +0100)]
aco: refactor store_vmem_mubuf() to use new helpers

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: refactor store_lds() to use new helpers
Rhys Perry [Thu, 16 Apr 2020 18:12:19 +0000 (19:12 +0100)]
aco: refactor store_lds() to use new helpers

It should also work correctly for 8/16-bit stores

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: add helpers for splitting stores
Rhys Perry [Thu, 16 Apr 2020 18:07:06 +0000 (19:07 +0100)]
aco: add helpers for splitting stores

split_store_data() splits a vector and p_as_uniforms it if needed.

scan_write_mask()/advance_write_mask() are similar to
u_bit_scan_consecutive_range(), but makes it easier to only clear part of
the range and will also give ranges for zero'd bits.

split_buffer_store() is a helper for splitting VMEM/SMEM stores.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: use emit_load helper for VMEM/SMEM loads
Rhys Perry [Wed, 15 Apr 2020 14:39:44 +0000 (15:39 +0100)]
aco: use emit_load helper for VMEM/SMEM loads

Also implements 8/16-bit loads for scratch/global.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: refactor load_lds to use new helpers
Rhys Perry [Thu, 16 Apr 2020 19:02:31 +0000 (20:02 +0100)]
aco: refactor load_lds to use new helpers

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: add emit_load helper
Rhys Perry [Thu, 16 Apr 2020 18:51:02 +0000 (19:51 +0100)]
aco: add emit_load helper

This helper is used for recombining split loads, passing the result to
p_as_uniform, aligning the offset down and shifting it right if needed and
handling large constant offsets.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: add and use RegClass::get() helper
Rhys Perry [Tue, 14 Apr 2020 19:32:39 +0000 (20:32 +0100)]
aco: add and use RegClass::get() helper

Eventually, we'll probably want to replace the current
RegClass(type, size) constructor with this.

This has a functional change in that get_reg_class() now creates v1/v2
instead of v4b/v8b.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: be more careful about using SMEM for load_global
Rhys Perry [Mon, 6 Apr 2020 19:15:36 +0000 (20:15 +0100)]
aco: be more careful about using SMEM for load_global

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoradv: allocate larger shader memory slabs if needed
Rhys Perry [Tue, 14 Apr 2020 19:15:46 +0000 (20:15 +0100)]
radv: allocate larger shader memory slabs if needed

Fixes dEQP-VK.ssbo.phys.layout.random.16bit.scalar.13 hang with ACO
(features needed for the test are implemented in a later commit)

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoradv: align buffer descriptor sizes to dword
Rhys Perry [Fri, 10 Apr 2020 13:25:46 +0000 (14:25 +0100)]
radv: align buffer descriptor sizes to dword

This is needed to prevent bounds checking issues when load 8/16-bit values
with dword loads.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4639>

17 months agoaco: Move s_setprio to correct place after the gs_alloc_req.
Timur Kristóf [Thu, 23 Apr 2020 13:53:09 +0000 (15:53 +0200)]
aco: Move s_setprio to correct place after the gs_alloc_req.

Previously the setprio was inside the branch, so it would only reset
the priority on the first wave, but not the others.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4536>

17 months agoaco: Use 24-bit multiplication for NGG wave id and thread id.
Timur Kristóf [Thu, 23 Apr 2020 13:50:56 +0000 (15:50 +0200)]
aco: Use 24-bit multiplication for NGG wave id and thread id.

Both of them should always fit 24 bits anyway.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4536>

17 months agoaco: Use 24-bit multiplication in TCS I/O
Timur Kristóf [Thu, 23 Apr 2020 13:39:56 +0000 (15:39 +0200)]
aco: Use 24-bit multiplication in TCS I/O

The TCS inputs and outputs must always fit into the LDS,
which implies that their addresses also always fit 24 bits.

On AMD GPUs, 24-bit multiplication is much faster than 32-bit
multiplication, so we can take the opportunity to use that
for TCS I/O instead.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4536>

17 months agoaco: Const correctness for aco_print_ir.
Timur Kristóf [Thu, 23 Apr 2020 13:24:47 +0000 (15:24 +0200)]
aco: Const correctness for aco_print_ir.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4536>

17 months agoaco: Const correctness for get_barrier_interaction.
Timur Kristóf [Wed, 18 Mar 2020 09:40:06 +0000 (10:40 +0100)]
aco: Const correctness for get_barrier_interaction.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4536>

17 months agoaco: Abort when RA can't find a register.
Timur Kristóf [Mon, 10 Feb 2020 15:34:56 +0000 (16:34 +0100)]
aco: Abort when RA can't find a register.

Previously, it was just unreachable, which means it will generate
invalid shaders when it encounters a situation when it can't allocate
registers for eg. a large load.

This commit makes it slightly easier to notice such problems without
triggering a GPU hang.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4536>

17 months agoaco: Increase barrier_count to 7 to include barrier_barrier.
Timur Kristóf [Thu, 23 Apr 2020 13:17:11 +0000 (15:17 +0200)]
aco: Increase barrier_count to 7 to include barrier_barrier.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4536>

17 months agoaco: Only store TCS outputs to VMEM when they are read by TES.
Timur Kristóf [Mon, 13 Apr 2020 17:03:55 +0000 (19:03 +0200)]
aco: Only store TCS outputs to VMEM when they are read by TES.

Totals from affected shaders (GFX10):
Code Size: 10832 -> 10736 (-0.89 %) bytes

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4536>