platform/upstream/mesa.git
2 years agofreedreno: Document rest of GRAS_LRZ_CNTL, clarify UNK_25 event
Danylo Piliaiev [Fri, 13 May 2022 14:10:36 +0000 (17:10 +0300)]
freedreno: Document rest of GRAS_LRZ_CNTL, clarify UNK_25 event

Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/16251>

2 years agomeson: drop dladdr check on win32
Michel Zou [Mon, 27 Jun 2022 19:08:40 +0000 (21:08 +0200)]
meson: drop dladdr check on win32

since !17208 there are 2 paths for disk_cache_get_function_identifier
on mingw: DETECT_OS_WINDOWS or HAVE_DLADDR (if dlfcn shims is present)

../src/util/disk_cache_os.c:47:1: error: redefinition of 'disk_cache_get_function_identifier'
   47 | disk_cache_get_function_identifier(void *ptr, struct mesa_sha1 *ctx)
      | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from ../src/util/disk_cache_os.c:36:
../src/util/disk_cache.h:121:1: note: previous definition of 'disk_cache_get_function_identifier' with type '_Bool(void *, struct _SHA1_CTX *)'
  121 | disk_cache_get_function_identifier(void *ptr, struct mesa_sha1 *ctx)

here we disable the dladdr path from meson for consistency with msvc

fixes: 2dcbe8727

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Yonggang Luo <luoyonggang@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17261>

2 years agoci/dzn: Add pipeline cache tests to the include list
Boris Brezillon [Wed, 22 Jun 2022 09:01:19 +0000 (11:01 +0200)]
ci/dzn: Add pipeline cache tests to the include list

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Cache pipeline info
Boris Brezillon [Wed, 22 Jun 2022 08:43:04 +0000 (10:43 +0200)]
dzn: Cache pipeline info

We are already caching DXIL shaders individually, but that forces us
to retrieve the NIR shader, do the linking and binding translation
steps, to finally query the cache for each DXIL shader. This pipeline
caching is about skipping those steps when we can.

Note that a graphics/compute pipeline cache entry doesn't contain the
DXIL shaders, but hashes to retrieve those shaders from the same cache.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Cache DXIL shaders
Boris Brezillon [Wed, 22 Jun 2022 08:32:41 +0000 (10:32 +0200)]
dzn: Cache DXIL shaders

This way we avoid the NIR -> DXIL translation when we have a cache hit.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Caculate a binding translation hash
Boris Brezillon [Wed, 22 Jun 2022 07:46:49 +0000 (09:46 +0200)]
dzn: Caculate a binding translation hash

Binding translation has an impact on the final DXIL shader, and this
binding translation depends on the pipeline layout. Let's extend the
adjust_var_bindings() pass to caculate a hash we can then use in the
DXIL shader hash caculation.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Cache NIR shaders
Boris Brezillon [Wed, 22 Jun 2022 07:12:51 +0000 (09:12 +0200)]
dzn: Cache NIR shaders

Saves us the SPIRV -> NIR translation, and all the lowering passes run
in dxil_spirv_nir_passes().

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Move the compute shader compilation logic to a sub-function
Boris Brezillon [Wed, 22 Jun 2022 07:19:04 +0000 (09:19 +0200)]
dzn: Move the compute shader compilation logic to a sub-function

Will make things easier when we introduce shader caching.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Drop unneeded goto statement in dzn_compute_pipeline_create()
Boris Brezillon [Wed, 22 Jun 2022 07:33:12 +0000 (09:33 +0200)]
dzn: Drop unneeded goto statement in dzn_compute_pipeline_create()

The 'out' label is placed just after the if () block, we can thus
remove the 'goto out;' statement.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Add a generic cached blob wrapper
Boris Brezillon [Tue, 21 Jun 2022 15:12:17 +0000 (17:12 +0200)]
dzn: Add a generic cached blob wrapper

Basically what vk_pipeline_cache's raw_data_object abstraction provides,
but I'm not sure it makes sense to make it generic so I copied it here.
Might be more appropriate to make raw_data_object public.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Hash pipeline layout stages
Boris Brezillon [Tue, 21 Jun 2022 15:09:27 +0000 (17:09 +0200)]
dzn: Hash pipeline layout stages

DXIL shaders depend on the vulkan -> d3d12 binding translation done in
adjust_var_bindings(). In order to maximize our chances to re-use those
shaders, we need to hash the binding translation information and take
this hash into account when computing the DXIL shader hash.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Don't delegate binding translation to dxil_spirv_nir_passes()
Boris Brezillon [Tue, 21 Jun 2022 09:53:05 +0000 (02:53 -0700)]
dzn: Don't delegate binding translation to dxil_spirv_nir_passes()

We will need to hash var bindings if we want to cache DXIL shaders.
Let's move this pass to dzn_pipeline.c to prepare this transition.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Save a few indentation levels in graphics_pipeline_compile_shaders()
Boris Brezillon [Mon, 20 Jun 2022 10:43:46 +0000 (03:43 -0700)]
dzn: Save a few indentation levels in graphics_pipeline_compile_shaders()

We can compute the yz_flip_mode and force_sample_rate_shading outside
of the foreach_shader loop.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Use vk_to_mesa_shader_stage()
Boris Brezillon [Mon, 20 Jun 2022 10:33:32 +0000 (03:33 -0700)]
dzn: Use vk_to_mesa_shader_stage()

Use vk_to_mesa_shader_stage() to convert a VkShaderStageFlagBits into
its gl_shader_stage counterpart.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Fix potential nir_shader leak
Boris Brezillon [Mon, 20 Jun 2022 10:08:52 +0000 (03:08 -0700)]
dzn: Fix potential nir_shader leak

We leak the nir_shader object f a failure happens between the NIR
shader creation and the DXIL compilation. Let's drop the local
nir_shader pointer and use the one in the graphics_pipeline object
to avoid that.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Let dzn_pipeline_init() initialize the root signature in the stream
Boris Brezillon [Fri, 17 Jun 2022 17:51:59 +0000 (10:51 -0700)]
dzn: Let dzn_pipeline_init() initialize the root signature in the stream

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Drop unused allocator passed to dzn_graphics_pipeline_create()
Boris Brezillon [Fri, 17 Jun 2022 17:45:17 +0000 (19:45 +0200)]
dzn: Drop unused allocator passed to dzn_graphics_pipeline_create()

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Drop useless while(link_mask != 0)
Boris Brezillon [Fri, 17 Jun 2022 17:43:53 +0000 (19:43 +0200)]
dzn: Drop useless while(link_mask != 0)

That's not a `do {} while();`, just a simple `while() {}`, the second
while statement is a NOP.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Pass the maximum stream size to d3d12_pipeline_state_stream_new_desc()
Boris Brezillon [Fri, 17 Jun 2022 17:34:19 +0000 (10:34 -0700)]
dzn: Pass the maximum stream size to d3d12_pipeline_state_stream_new_desc()

This way we can use d3d12_pipeline_state_stream_new_desc() directly
without doing

   if (type == GRAPHICS)
      d3d12_gfx_pipeline_state_stream_new_desc()
   else
      d3d12_compute_pipeline_state_stream_new_desc()

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Use d3d12_compute_pipeline_state_stream_new_desc() in the compute path
Boris Brezillon [Fri, 17 Jun 2022 17:28:14 +0000 (10:28 -0700)]
dzn: Use d3d12_compute_pipeline_state_stream_new_desc() in the compute path

Fixes: 9feda65a83c ("dzn: Use CreatePipelineState()")
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Fix assertion in d3d12_pipeline_state_stream_new_desc()
Boris Brezillon [Fri, 17 Jun 2022 17:26:46 +0000 (10:26 -0700)]
dzn: Fix assertion in d3d12_pipeline_state_stream_new_desc()

Fixes: 9feda65a83c ("dzn: Use CreatePipelineState()")
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Add missing D3D12_CACHED_PIPELINE_STATE to MAX_COMPUTE_PIPELINE_STATE_STREAM_SIZE
Boris Brezillon [Fri, 17 Jun 2022 17:08:36 +0000 (19:08 +0200)]
dzn: Add missing D3D12_CACHED_PIPELINE_STATE to MAX_COMPUTE_PIPELINE_STATE_STREAM_SIZE

Fixes: 9feda65a83c ("dzn: Use CreatePipelineState()")
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Drop dzn_pipeline_cache.c
Boris Brezillon [Tue, 14 Jun 2022 09:39:24 +0000 (02:39 -0700)]
dzn: Drop dzn_pipeline_cache.c

The core provides a conformant pipeline cache implementation, let's
use it.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Initialize UUIDs
Boris Brezillon [Thu, 23 Jun 2022 11:06:54 +0000 (04:06 -0700)]
dzn: Initialize UUIDs

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agodzn: Fix indentation
Boris Brezillon [Mon, 20 Jun 2022 11:52:14 +0000 (13:52 +0200)]
dzn: Fix indentation

Replace tabs by spaces

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17140>

2 years agopytracediff: implement pager ('less') invocation internally
Matti Hamalainen [Mon, 20 Jun 2022 11:20:09 +0000 (14:20 +0300)]
pytracediff: implement pager ('less') invocation internally

In order to get rid of the ntracediff.sh wrapper script, implement
invocation of 'less' internally, if the stdout is determined to
be a tty. Otherwise just print out normally.

Signed-off-by: Matti Hamalainen <ccr@tnsp.org>
Acked-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17135>

2 years agopytracediff: change how 'junk' calls are handled
Matti Hamalainen [Sun, 19 Jun 2022 18:28:42 +0000 (21:28 +0300)]
pytracediff: change how 'junk' calls are handled

Instead of discarding them at parsing phase, let the difflib
SequenceMatcher always ignore them, and optionally suppress
them from output if -I/--ignore-junk option is given.

Signed-off-by: Matti Hamalainen <ccr@tnsp.org>
Acked-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17135>

2 years agopytracediff: make -M ("method only") option print arguments for differing calls
Matti Hamalainen [Tue, 25 Jan 2022 13:27:34 +0000 (15:27 +0200)]
pytracediff: make -M ("method only") option print arguments for differing calls

Basically implement the last item on the original feature request list of #4609.

Example: ./pytracediff.py good.xml bad.xml -NM

Or suppress common calls completely via -C, e.g. -NC etc.

Signed-off-by: Matti Hamalainen <ccr@tnsp.org>
Acked-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17135>

2 years agopytracediff: add per-line difference highlighting for blocks
Matti Hamalainen [Mon, 24 Jan 2022 21:02:48 +0000 (23:02 +0200)]
pytracediff: add per-line difference highlighting for blocks

Highlight differing _lines_ in the differing blocks, with somewhat
different ANSI colors.

Signed-off-by: Matti Hamalainen <ccr@tnsp.org>
Acked-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17135>

2 years agogallium/tools: reimplement tracediff completely in Python
Matti Hamalainen [Fri, 21 Jan 2022 21:43:25 +0000 (23:43 +0200)]
gallium/tools: reimplement tracediff completely in Python

The limitations of current approach for Gallium trace diffing via
'tracediff.sh' are becoming all too apparent, as we are first dumping
both trace to text and performing plain line-based sdiff on them.

This obviously loses the context of calls and sometimes results in
confusing diffs when subsequent calls are similar enough. It also
prevents us from formatting the diff output in ways that would
benefit readability.

In attempt to rectify the situation, reimplement the diffing completely
in Python, using difflib and adding the necessary plumbing into the trace
model objects etc.

Signed-off-by: Matti Hamalainen <ccr@tnsp.org>
Acked-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17135>

2 years agogallium/tools: add option for ignoring junk calls in trace dumper
Matti Hamalainen [Fri, 21 Jan 2022 21:35:33 +0000 (23:35 +0200)]
gallium/tools: add option for ignoring junk calls in trace dumper

Previously tracediff.sh used postprocessing sed-script to remove unwanted
calls from the dump output. Instead of that, add option to parse.py to
ignore a list of calls at parsing phase. Currently this list is hardcoded
in parse.py.

Also clean up the trace model code and pointer tracking a bit to avoid
static state in Pointer class.

Signed-off-by: Matti Hamalainen <ccr@tnsp.org>
Acked-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17135>

2 years agotrace: Allow to control nir dumping via an environment variable.
Jose Fonseca [Mon, 27 Jun 2022 18:38:19 +0000 (19:38 +0100)]
trace: Allow to control nir dumping via an environment variable.

As requested by Mike Blumenkrantz.

Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17107>

2 years agotrace: Bring state dump up to speed.
Jose Fonseca [Thu, 16 Jun 2022 10:06:31 +0000 (11:06 +0100)]
trace: Bring state dump up to speed.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17107>

2 years agotrace: Parse character data more efficiently.
Jose Fonseca [Thu, 16 Jun 2022 10:06:12 +0000 (11:06 +0100)]
trace: Parse character data more efficiently.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17107>

2 years agotrace: Dump NIR.
Jose Fonseca [Thu, 16 Jun 2022 10:05:48 +0000 (11:05 +0100)]
trace: Dump NIR.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17107>

2 years agolavapipe: Prevent mapping buffers beyond their size.
Jose Fonseca [Thu, 16 Jun 2022 08:28:34 +0000 (09:28 +0100)]
lavapipe: Prevent mapping buffers beyond their size.

This was breaking trace driver serialization, since it relies upon the
transfer box to know what to serialize.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17107>

2 years agotrace: Fix framebuffer state serialization.
Jose Fonseca [Thu, 16 Jun 2022 08:27:33 +0000 (09:27 +0100)]
trace: Fix framebuffer state serialization.

Ensure the argument name is serialized "state" not, "&tr_ctx->unwrapped_state".

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17107>

2 years agodzn: Hook-up device-lost detection
Boris Brezillon [Mon, 27 Jun 2022 08:23:10 +0000 (10:23 +0200)]
dzn: Hook-up device-lost detection

Provide a vk_device::check_status() implementation so the code can
call it at key moments to detect when a device-lost event (device-removed
in D3D12) has been received.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17249>

2 years agoradv: Use vk_pipeline_hash_shader_stage()
Boris Brezillon [Wed, 22 Jun 2022 17:02:44 +0000 (19:02 +0200)]
radv: Use vk_pipeline_hash_shader_stage()

Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17186>

2 years agoradv: Kill unused fs_m local var in radv_create_shaders()
Boris Brezillon [Wed, 22 Jun 2022 21:27:43 +0000 (23:27 +0200)]
radv: Kill unused fs_m local var in radv_create_shaders()

Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17186>

2 years agotu: Use vk_pipeline_hash_shader_stage()
Boris Brezillon [Wed, 22 Jun 2022 16:10:12 +0000 (18:10 +0200)]
tu: Use vk_pipeline_hash_shader_stage()

Acked-by: Emma Anholt <emma@anholt.net>
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17186>

2 years agov3dv: Use vk_pipeline_hash_shader_stage()
Boris Brezillon [Wed, 22 Jun 2022 16:08:03 +0000 (18:08 +0200)]
v3dv: Use vk_pipeline_hash_shader_stage()

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17186>

2 years agov3dv: Fix nir_shader leaks in v3dv_meta_{clear,copy}()
Boris Brezillon [Wed, 22 Jun 2022 23:45:03 +0000 (01:45 +0200)]
v3dv: Fix nir_shader leaks in v3dv_meta_{clear,copy}()

Reported-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17186>

2 years agoanv: Use vk_pipeline_hash_shader_stage()
Boris Brezillon [Wed, 22 Jun 2022 16:07:40 +0000 (18:07 +0200)]
anv: Use vk_pipeline_hash_shader_stage()

Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17186>

2 years agovulkan: Add a vk_pipeline_hash_shader_stage() helper
Boris Brezillon [Wed, 22 Jun 2022 11:14:38 +0000 (04:14 -0700)]
vulkan: Add a vk_pipeline_hash_shader_stage() helper

All drivers implement some sort of shader hashing, but each of
them does it slightly differently. Let's provide a generic helper
to avoid new copies of the same logic and encourage new drivers
to use one of the already implemented function.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17186>

2 years agonir/serialize: Silence integer-overflow false positive
Boris Brezillon [Thu, 23 Jun 2022 12:39:34 +0000 (14:39 +0200)]
nir/serialize: Silence integer-overflow false positive

Use util_sign_extend() to silence the following integer-overflow
error.

src/compiler/nir/nir_serialize.c:1333:40: runtime error: left shift of 1000165000 by 13 places cannot be represented in type 'int'

Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17186>

2 years agonir/serialize: Support texop >= 16
Boris Brezillon [Thu, 23 Jun 2022 14:32:12 +0000 (16:32 +0200)]
nir/serialize: Support texop >= 16

Extend the packed_instr struct to support texops above
nir_texop_fragment_fetch_amd.

Fixes: 603e6ba972da ("nir: add two new texture ops for multisample fragment color/mask fetches")
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17186>

2 years agobroadcom/compiler: don't predicate postponed spills
Iago Toral Quiroga [Wed, 22 Jun 2022 07:43:30 +0000 (09:43 +0200)]
broadcom/compiler: don't predicate postponed spills

The postponed spill is predicated using the condition from the
last write, but this is only correct if the register was only
written once in the TMU sequence, or if it is always written with
the same predication.

While we could try to track whether this is the case or not, it
would make the postponed spill path even more complex than it
already is, so let's just avoid predicating these. We are already
discouraging TMU spilling of registers in the middle of TMU
sequences, so this should not be a very common case.

Cc: mesa-stable
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17201>

2 years agobroadcom/compiler: fix postponed TMU spills with multiple writes
Iago Toral Quiroga [Wed, 22 Jun 2022 06:29:42 +0000 (08:29 +0200)]
broadcom/compiler: fix postponed TMU spills with multiple writes

If we are spilling a register that is used in the middle of a TMU
sequence, we postpone the spill until the TMU sequence finishes,
at which point we inject the spill and rewrite the original
instruction to write to the new temp.

However, this doesn't work if the register is written multiple
times during the TMU sequence. In that scenario, we need to ensure
that all writes are rewritten to use the new temp, not just the last
one.

Cc: mesa-stable
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17201>

2 years agov3dv: fix leak
Iago Toral Quiroga [Wed, 22 Jun 2022 09:44:12 +0000 (11:44 +0200)]
v3dv: fix leak

Cc: mesa-stable
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17201>

2 years agozink: delete zink_resource_object::dt_has_data
Mike Blumenkrantz [Wed, 22 Jun 2022 17:24:33 +0000 (13:24 -0400)]
zink: delete zink_resource_object::dt_has_data

this should be on the swapchain image

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agozink: use kopper_displaytarget type directly
Mike Blumenkrantz [Wed, 22 Jun 2022 16:02:14 +0000 (12:02 -0400)]
zink: use kopper_displaytarget type directly

casting this all over is super annoying and unnecessary

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agozink: acquire swapchain images on image map
Mike Blumenkrantz [Tue, 21 Jun 2022 20:35:36 +0000 (16:35 -0400)]
zink: acquire swapchain images on image map

this is a weird one

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agozink: no-op pixmap frontbuffer flush calls
Mike Blumenkrantz [Tue, 21 Jun 2022 20:32:17 +0000 (16:32 -0400)]
zink: no-op pixmap frontbuffer flush calls

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agozink: delete zink_resource_object::acquire
Mike Blumenkrantz [Tue, 21 Jun 2022 20:25:32 +0000 (16:25 -0400)]
zink: delete zink_resource_object::acquire

this was duplicated from the swapchain object for convenience, but really
it just leads to desync if a swapchain is shared, so use the swapchain
one instead

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agozink: ensure pending present flushes are handled during frontbuffer flush
Mike Blumenkrantz [Tue, 21 Jun 2022 20:24:25 +0000 (16:24 -0400)]
zink: ensure pending present flushes are handled during frontbuffer flush

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agozink: flag batch as having work during frontbuffer flush
Mike Blumenkrantz [Tue, 21 Jun 2022 19:16:14 +0000 (15:16 -0400)]
zink: flag batch as having work during frontbuffer flush

avoids having the flush deferred

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agozink: improve no-op acquire detection
Mike Blumenkrantz [Tue, 21 Jun 2022 15:22:41 +0000 (11:22 -0400)]
zink: improve no-op acquire detection

the swapchain may have successfully completed the acquire in a previous
batch, in which case a new acquire is not needed

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agozink: remove unnecessary code in kopper readback acquire
Mike Blumenkrantz [Tue, 21 Jun 2022 01:46:21 +0000 (21:46 -0400)]
zink: remove unnecessary code in kopper readback acquire

this used to be important but no longer is, so modify the loop
code to handle this case implicitly

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agozink: remove broken kopper assert
Mike Blumenkrantz [Tue, 21 Jun 2022 01:36:58 +0000 (21:36 -0400)]
zink: remove broken kopper assert

in a single-buffered scenario this can and does happen

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agozink: use better determination for handling swapchain acquire submits
Mike Blumenkrantz [Mon, 20 Jun 2022 16:33:27 +0000 (12:33 -0400)]
zink: use better determination for handling swapchain acquire submits

if the swapchain has data, it has already passed through this function and
thus does not need to attempt to synchronize another swapchain acquire
for the same image

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agozink: move 'acquired' flag onto swapchain struct
Mike Blumenkrantz [Tue, 21 Jun 2022 14:59:55 +0000 (10:59 -0400)]
zink: move 'acquired' flag onto swapchain struct

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agozink: move to AoS for storing swapchain struct data
Mike Blumenkrantz [Wed, 22 Jun 2022 15:46:35 +0000 (11:46 -0400)]
zink: move to AoS for storing swapchain struct data

this is a bit less cumbersome

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agozink: track whether current swapchain has data
Mike Blumenkrantz [Mon, 20 Jun 2022 16:30:23 +0000 (12:30 -0400)]
zink: track whether current swapchain has data

duplicated from resource object since swapchains can be shared

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agozink: fix kopper_acquire return value
Mike Blumenkrantz [Mon, 20 Jun 2022 16:29:19 +0000 (12:29 -0400)]
zink: fix kopper_acquire return value

this should be a VkResult

Fixes: ab1941fc0e9 ("zink: handle zombie swapchains")

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17149>

2 years agolima: wire up MSAA 4x support
Vasily Khoruzhick [Fri, 26 Nov 2021 07:56:41 +0000 (23:56 -0800)]
lima: wire up MSAA 4x support

Utgard supports MSAA 4x, so wire it up.

RSW bits were already REd by Luc, the only remaining part was storing
non-resolved buffers, reloading them (including for depth/stencil) and
doing MSAA resolve.

To store non-resolved buffer we need to set mrt_pitch and mrt_bits
registers in WB, and to resolve non-resolved buffer we need to reload
it into individual samples and then write out with mrt_bits = 0, it's
now done by lima blitter.

We also need to do resolve on transfer_map() of multi-sampled buffers,
so utilize u_transfer_helper for that.

As a side fix, it turns out that our wb_reg definition wasn't correct,
'zero' isn't always zero, it's set if we need to swap channels, and
it goes before mrt_bits. mrt_bits actually enables multiple MRTs,
so this commit renames 'zero' to 'flags' and changes its position.

If mrt_bits == 0 and MSAA is enabled, GPU does resolve
in place, to expose this functionality we set PIPE_CAP_SURFACE_SAMPLE_COUNT.

Fixes dEQP-GLES2.functional.multisample.*

Reviewed-by: Erico Nunes <nunes.erico@gmail.com>
Signed-off-by: Vasily Khoruzhick <anarsoul@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13963>

2 years agoci/turnip: Trim the a630 VK run a bit.
Emma Anholt [Tue, 21 Jun 2022 21:37:48 +0000 (14:37 -0700)]
ci/turnip: Trim the a630 VK run a bit.

We have a lot of spilling coverage in a618 pre-merge, don't do it all (~2
minutes) here.  Also, force-gmem touch testing should probably test less than
the default run does!

This should help make up for having added the tu-zink run.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17125>

2 years agoci/turnip: Test traces on turnip using zink.
Emma Anholt [Tue, 13 Jul 2021 23:41:07 +0000 (16:41 -0700)]
ci/turnip: Test traces on turnip using zink.

Reviewed-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17125>

2 years agoci/freedreno: Add vulkan+gl integration testing in piglit.
Emma Anholt [Mon, 20 Jun 2022 16:48:11 +0000 (09:48 -0700)]
ci/freedreno: Add vulkan+gl integration testing in piglit.

The libvulkan-dev was needed for building zink, which ended up turning on
the vulkan tests in piglit.  Split them out here.

Reviewed-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17125>

2 years agoci/traces: Drop ZINK_USE_LAVAPIPE ICD override.
Emma Anholt [Thu, 2 Jun 2022 22:24:03 +0000 (15:24 -0700)]
ci/traces: Drop ZINK_USE_LAVAPIPE ICD override.

Not set in the tree any more.

Reviewed-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17125>

2 years agoci/traces: GC unused code for DXVK trace replay.
Emma Anholt [Thu, 2 Jun 2022 22:23:11 +0000 (15:23 -0700)]
ci/traces: GC unused code for DXVK trace replay.

I haven't found any use of it since it was introduced, and it got in the
way of zink trace testing.

Reviewed-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17125>

2 years agolavapipe: always set point_tri_clip
Mike Blumenkrantz [Mon, 27 Jun 2022 17:56:23 +0000 (13:56 -0400)]
lavapipe: always set point_tri_clip

this invokes GLES-compatible point clipping, which is more consistent
with vulkan expectations and fixes a number of zink tests

Acked-by: Emma Anholt <emma@anholt.net>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17259>

2 years agolima,panfrost: Do not ralloc contexts against the screen
Alyssa Rosenzweig [Fri, 24 Jun 2022 14:42:00 +0000 (10:42 -0400)]
lima,panfrost: Do not ralloc contexts against the screen

ralloc is not thread-safe. While a given context can only be accessed from a
single thread at once, multiple contexts can be created against the same screen
at once. The ralloc allocations against the shared screens will race. Depending
on the result of the race, the same block of memory can be returned as the two
new contexts in two different threads, causing a use-after-free when the context
is freed later.

We free the context explicitly when it's destroyed anyway. If screens are
getting destroyed without the contexts getting destroyed first, that's a state
tracker bug, not a Panfrost one.

This matches what Iris does.

Fixes crash in test_integer_ops.int_math on Panfrost.

Fixes: 0fcf73bc2d4 ("panfrost: Move to use ralloc for some allocations")
Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Vasily Khoruzhick <anarsoul@gmail.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17234>

2 years agozink: remove swizzle from fbfetch lowering
Mike Blumenkrantz [Mon, 27 Jun 2022 15:00:49 +0000 (11:00 -0400)]
zink: remove swizzle from fbfetch lowering

I had this in at one point to fix something, but now it somehow just
breaks fbfetch instead of fixing anything

cc: mesa-stable

fixes:
dEQP-GLES31.functional.blend_equation_advanced*

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17254>

2 years agoglx/dri: Fix DRI drawable release at MakeCurrent time
Adam Jackson [Tue, 21 Jun 2022 20:07:19 +0000 (16:07 -0400)]
glx/dri: Fix DRI drawable release at MakeCurrent time

We want to release the drawables of the context we're coming from, but
we were releasing them from the context we're switching to. This is
probably not a big deal normally because both contexts are likely to be
on the same display, which is all that driReleaseDrawables is really
sensitive to. But if the contexts are on different Displays this would
go quite wrong.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17176>

2 years agomicrosoft/clc: Add a unit test for unused image kernel args
Jesse Natalie [Mon, 27 Jun 2022 04:09:07 +0000 (21:09 -0700)]
microsoft/clc: Add a unit test for unused image kernel args

Reviewed-by: Bill Kristiansen <billkris@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17246>

2 years agomicrosoft/clc: Remove dead image vars
Jesse Natalie [Mon, 27 Jun 2022 04:08:56 +0000 (21:08 -0700)]
microsoft/clc: Remove dead image vars

Reviewed-by: Bill Kristiansen <billkris@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17246>

2 years agomicrosoft/clc: Fix test double free in the case of compilation failure
Jesse Natalie [Mon, 27 Jun 2022 03:58:24 +0000 (20:58 -0700)]
microsoft/clc: Fix test double free in the case of compilation failure

Reviewed-by: Bill Kristiansen <billkris@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17246>

2 years agomicrosoft/clc: Enable tests that pass on server 2022
Jesse Natalie [Mon, 27 Jun 2022 03:58:02 +0000 (20:58 -0700)]
microsoft/clc: Enable tests that pass on server 2022

Reviewed-by: Bill Kristiansen <billkris@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17246>

2 years agoutil/disk_cache: Implement disk_cache_get_function_identifier for Windows
Jesse Natalie [Thu, 23 Jun 2022 14:31:10 +0000 (07:31 -0700)]
util/disk_cache: Implement disk_cache_get_function_identifier for Windows

Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17208>

2 years agointel/clc: enable fp16 & subgroups for GRL
Lionel Landwerlin [Mon, 27 Jun 2022 10:13:14 +0000 (13:13 +0300)]
intel/clc: enable fp16 & subgroups for GRL

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17253>

2 years agoclc: add new feature options for intel_clc
Lionel Landwerlin [Mon, 27 Jun 2022 10:12:46 +0000 (13:12 +0300)]
clc: add new feature options for intel_clc

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17253>

2 years agoradv: vectorize nir_op_fabs
Daniel Schürmann [Thu, 7 Oct 2021 18:20:23 +0000 (20:20 +0200)]
radv: vectorize nir_op_fabs

Totals from 4 (0.00% of 134913) affected shaders: (GFX10.3)
CodeSize: 37868 -> 36576 (-3.41%)
Instrs: 5332 -> 5169 (-3.06%)
Latency: 24452 -> 24174 (-1.14%)
InvThroughput: 9784 -> 9462 (-3.29%)
VClause: 54 -> 50 (-7.41%)
Copies: 520 -> 519 (-0.19%)
PreVGPRs: 266 -> 264 (-0.75%)

Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/15176>

2 years agoradv: vectorize nir_op_fdiv & nir_op_flrp & nir_op_ineg
Daniel Schürmann [Tue, 10 Aug 2021 16:06:58 +0000 (18:06 +0200)]
radv: vectorize nir_op_fdiv & nir_op_flrp & nir_op_ineg

These instructions are lowered to fmul/isub.
So, prevent scalarization.

Totals from 49 (0.04% of 134913) affected shaders: (GFX10.3)
VGPRs: 2576 -> 2568 (-0.31%)
SpillVGPRs: 1145 -> 1132 (-1.14%); split: -2.10%, +0.96%
CodeSize: 663968 -> 659376 (-0.69%); split: -1.08%, +0.38%
Scratch: 113664 -> 112640 (-0.90%)
Instrs: 110274 -> 109683 (-0.54%); split: -0.81%, +0.27%
Latency: 2904434 -> 2869588 (-1.20%); split: -1.64%, +0.44%
InvThroughput: 1414237 -> 1396600 (-1.25%); split: -1.69%, +0.44%
VClause: 2899 -> 2891 (-0.28%); split: -0.93%, +0.66%
SClause: 1520 -> 1537 (+1.12%); split: -0.07%, +1.18%
Copies: 28829 -> 28662 (-0.58%); split: -1.90%, +1.32%
Branches: 3560 -> 3564 (+0.11%)
PreVGPRs: 2550 -> 2427 (-4.82%)

Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/15176>

2 years agoradv: use callback for nir_lower_to_scalar
Daniel Schürmann [Tue, 6 Jul 2021 17:09:51 +0000 (19:09 +0200)]
radv: use callback for nir_lower_to_scalar

Now uses nir_lower_alu_width.
This avoids scalarization and re-vectorization of 16bit instructions.

Totals from 289 (0.21% of 134913) affected shaders: (GFX10.3)
VGPRs: 12864 -> 13072 (+1.62%); split: -0.50%, +2.11%
SpillSGPRs: 609 -> 505 (-17.08%)
SpillVGPRs: 946 -> 1145 (+21.04%)
CodeSize: 2537024 -> 2576976 (+1.57%); split: -0.10%, +1.67%
Scratch: 89088 -> 113664 (+27.59%)
MaxWaves: 7150 -> 7134 (-0.22%)
Instrs: 458352 -> 460830 (+0.54%); split: -0.45%, +0.99%
Latency: 6615279 -> 6844092 (+3.46%); split: -0.08%, +3.54%
InvThroughput: 1929504 -> 2044989 (+5.99%); split: -0.22%, +6.21%
VClause: 7186 -> 7338 (+2.12%); split: -0.08%, +2.20%
SClause: 13144 -> 13116 (-0.21%)
Copies: 46152 -> 50127 (+8.61%); split: -0.11%, +8.73%
Branches: 16530 -> 16572 (+0.25%); split: -0.02%, +0.27%
PreSGPRs: 14903 -> 14905 (+0.01%); split: -0.01%, +0.03%
PreVGPRs: 11806 -> 11730 (-0.64%); split: -1.83%, +1.19%

Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/15176>

2 years agoradv: don't lower vectorized instructions to 32bit
Daniel Schürmann [Mon, 27 Jun 2022 12:49:49 +0000 (14:49 +0200)]
radv: don't lower vectorized instructions to 32bit

Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/15176>

2 years agoaco: correctly validate v_fma_mixhi_f16 register assignment
Daniel Schürmann [Wed, 1 Jun 2022 15:16:55 +0000 (17:16 +0200)]
aco: correctly validate v_fma_mixhi_f16 register assignment

Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/15176>

2 years agoanv: disable injection of primitive shading rate for mesh
Marcin Ślusarz [Fri, 6 May 2022 12:38:43 +0000 (14:38 +0200)]
anv: disable injection of primitive shading rate for mesh

It's not needed and causes issues for mesh code (it doesn't
mark the output as per-primitive, which confuses brw_compute_mue_map)

Fixes many tests matching:
dEQP-VK.fragment_shading_rate.dynamic_rendering.*.ms

Fixes: 1542ab70eb4 ("anv: handle primitive shading rate for mesh")
Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/16196>

2 years agointel/compiler: adjust task payload offsets as late as possible
Marcin Ślusarz [Mon, 23 May 2022 15:09:33 +0000 (17:09 +0200)]
intel/compiler: adjust task payload offsets as late as possible

Otherwise passes which expect offsets to be in bytes (like
brw_nir_lower_mem_access_bit_sizes, called from brw_postprocess_nir)
may produce incorrect results.

Fixes 64-bit load/stores in task/mesh shaders.

Fixes: c36ae42e4cc ("intel/compiler: Use nir_var_mem_task_payload")
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/16196>

2 years agointel/common: allocate space for at least one task urb
Marcin Ślusarz [Fri, 6 May 2022 14:37:27 +0000 (16:37 +0200)]
intel/common: allocate space for at least one task urb

Fixes: c93cbc77f78 ("intel/common: Add helper for URB allocation in Mesh pipeline")
Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/16196>

2 years agoc11: reinstate the original license and authorship
Emil Velikov [Fri, 24 Jun 2022 12:36:44 +0000 (13:36 +0100)]
c11: reinstate the original license and authorship

The original code that was copied in was Boost licensed, so keep that
in. Since Yonggang Luo has code quite some work, keep their copyright
alongside the original one.

Fixes: b2ddec4e98f ("c11: Implement c11/time.h with c11/impl/time.c")
Fixes: e6392fcf3d8 ("c11: Move the implementation of threads.h into c source code")
Acked-by: Yonggang Luo <luoyonggang@gmail.com>
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17232>

2 years agointel/blorp: Dirty depth bounds dynamic state bits after blorp
Sviatoslav Peleshko [Fri, 24 Jun 2022 07:43:09 +0000 (10:43 +0300)]
intel/blorp: Dirty depth bounds dynamic state bits after blorp

Blorp emits its own 3DSTATE_DEPTH_BOUNDS, so we'll have to re-emit the
expected state after that.

Signed-off-by: Sviatoslav Peleshko <sviatoslav.peleshko@globallogic.com>
Fixes: 56ef501e3aa4 ("blorp: disable depth bounds")
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17205>

2 years agoanv: Dirty all dynamic state bits when creating command buffer state
Sviatoslav Peleshko [Thu, 23 Jun 2022 11:20:24 +0000 (14:20 +0300)]
anv: Dirty all dynamic state bits when creating command buffer state

This makes sure that we'll handle situations when the new state has
the same value as the default one, so we won't dirty some bits, and
consequently will not emit necessary commands (e.g. 3DSTATE_DEPTH_BOUNDS).

Signed-off-by: Sviatoslav Peleshko <sviatoslav.peleshko@globallogic.com>
Fixes: 48229d11 ("anv: don't emit 3DSTATE_DEPTH_BOUNDS in pipeline batch")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/6722
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17205>

2 years agointel: fix typos found by codespell
Marcin Ślusarz [Wed, 22 Jun 2022 16:31:08 +0000 (18:31 +0200)]
intel: fix typos found by codespell

Acked-by: David Heidelberg <david.heidelberg@collabora.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17191>

2 years agodzn: Enable the depthClamp feature
Boris Brezillon [Fri, 24 Jun 2022 15:42:34 +0000 (08:42 -0700)]
dzn: Enable the depthClamp feature

depthClampEnable is actually the case we support properly.

!depthClampEnable requires extra work to make sure the
depth clamping that's forced by D3D12 is inactive (setting the
viewport depth range to [0,1] and dealing with the actual range
at the shader level), and clamp the depth value read by the
fragment shader in that case. This will be addressed separately.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17231>

2 years agodzn: Enable shader{Clip,Cull}Distance
Boris Brezillon [Fri, 24 Jun 2022 11:28:09 +0000 (04:28 -0700)]
dzn: Enable shader{Clip,Cull}Distance

DXIL has clip/cull distance builtins too.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17231>

2 years agodzn: Enable dynamic indexing on all kind of descriptors
Boris Brezillon [Fri, 24 Jun 2022 11:24:42 +0000 (04:24 -0700)]
dzn: Enable dynamic indexing on all kind of descriptors

nir_to_dxil() supports it.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17231>

2 years agodzn: Advertise shaderImageGatherExtended support
Boris Brezillon [Fri, 24 Jun 2022 11:04:46 +0000 (04:04 -0700)]
dzn: Advertise shaderImageGatherExtended support

nir_to_dxil() takes tg4 offsets into account.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17231>

2 years agodzn: Advertise anisotropic filtering support
Boris Brezillon [Fri, 24 Jun 2022 10:51:01 +0000 (03:51 -0700)]
dzn: Advertise anisotropic filtering support

We support it already, let's toggle the switch to expose it.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17231>

2 years agonir/serialize: Put dest last in packed_instr::tex
Boris Brezillon [Thu, 23 Jun 2022 16:53:31 +0000 (18:53 +0200)]
nir/serialize: Put dest last in packed_instr::tex

packed_instr::tex::dest must be last to match the packed_instr::any::dest
position.

Fixes: 35655865cbde ("nir/serialize: pack instructions better")
Cc: stable
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17212>