OpenHarmony/third_party_mesa3d
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
194,142 Commits 9 Branches 1 Tag
c60a421f0c2357fe13636f0d51f21e7be25b625e
Commit Graph

194142 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Faith Ekstrand
c60a421f0c vtn: Add a debug flag to dump SPIR-V assembly 
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30875>
 2024-08-28 21:52:59 +00:00
Faith Ekstrand
9520fb8ecc vtn: Allow SPIR-V debug options in release builds 
As long as no one sets any flags, they cost us virtually nothing.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30875>
 2024-08-28 21:52:58 +00:00
Faith Ekstrand
ee7b596abc vtn: Move initialization of mesa_spirv_debug to a helper 
Importantly, this gives us a single once_flag instead of two so we
actually only initialize it once.  This race will probably never matter
in practice because it's just a debug var but this is cleaner.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30875>
 2024-08-28 21:52:58 +00:00
Faith Ekstrand
d8644a533e v3dv: Update a debug message 
v3dv switched to vk_pipeline_shader_stage_to_nir a while ago.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30875>
 2024-08-28 21:52:58 +00:00
Kenneth Graunke
da395e6985 intel/brw: Fix extract_imm for subregion reads of 64-bit immediates 
We could be trying to extract a D/UD from a Q/UQ, for example.  We were
ignoring the top 32-bits, which is incorrect.

Fixes: 580e1c592d ("intel/brw: Introduce a new SSA-based copy propagation pass")
Reviewed-by: Caio Oliveira <caio.oliveira@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30884>
 2024-08-28 12:33:26 -07:00
Kenneth Graunke
51c85e0363 intel/brw: Drop misguided sign extension attempts in extract_imm() 
This function never expands a type - it only narrows it.  As such, we
don't need to ever sign extend to fill additional new bits.  I think
this code was left over from earlier versions of my optimization pass
that was buggy and trying to handle cases it should not have.

Fixes: 580e1c592d ("intel/brw: Introduce a new SSA-based copy propagation pass")
Reviewed-by: Caio Oliveira <caio.oliveira@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30884>
 2024-08-28 12:33:26 -07:00
Karol Herbst
53869ae45b rusticl/mesa: Return CStr for device and vendor names. 
This allows us to skip converting between Rust and C strings when fetching
those.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30888>
 2024-08-28 18:24:04 +00:00
Karol Herbst
6b43c5c552 zink: cache generated vendor and device name inside zink_screen 
get_name and get_device_vendor are supposed to return immutable strings.

This prevents the CL_DEVICE_NAME and CL_DEVICE_VENDOR queries from
randomly failing.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30888>
 2024-08-28 18:24:04 +00:00
Karol Herbst
73c8e2c3cd rusticl: reuse PipeContext 
Gallium drivers are likely to leak CPU and GPU resources as with OpenGL
they rarely have to create more than a single pipe_context.

This also reduces the cost of creating CL queues.

In order to debug drivers leaking memory a new debug option is added to
disable the reuse of PipeContexts

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30888>
 2024-08-28 18:24:04 +00:00
David Rosca
64ca0fd2f2 frontends/va: Allocate surface buffers on demand 
This saves memory by not allocating encode DPB surface buffers which
are currently not used by drivers.

Reviewed-By: Sil Vilerino <sivileri@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30843>
 2024-08-28 15:37:10 +00:00
David Rosca
de61cb72fe pipe: Use correct struct for h264/h265 dpb entry 
Fixes: cc14724d73 ("frontends/va: Implement DPB management for H264/5 encode")
Reviewed-By: Sil Vilerino <sivileri@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30843>
 2024-08-28 15:37:10 +00:00
David Heidelberg
83b74bfc6d bin/gen_release_notes: adjust checksums section 
We currently provide SHA256 and SHA512, differenciated by length.

Cc: mesa-stable
Signed-off-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30897>
 2024-08-28 15:25:27 +00:00
David Rosca
2aa373aee5 radeonsi/vcn: Implement get_feedback_fence 
Reviewed-by: Ruijing Dong <ruijing.dong@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30780>
 2024-08-28 13:38:52 +00:00
David Rosca
49596f70ee frontends/va: Check for errors from end_frame in vlVaEndPicture 
Reviewed-By: Sil Vilerino <sivileri@microsoft.com>
Reviewed-by: Ruijing Dong <ruijing.dong@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30780>
 2024-08-28 13:38:52 +00:00
David Rosca
db45d16dfe pipe: Add int return value to video end_frame 
Make it possible to report errors to frontend.

Reviewed-By: Sil Vilerino <sivileri@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30780>
 2024-08-28 13:38:52 +00:00
David Rosca
96fe9fde3f frontends/va: Implement sync buffer/surface timeout for encode feedback 
Same as processing/decode, the driver will assign fence to
pipe_picture_desc::fence and frontend will wait on it with
get_feedback_fence.

Reviewed-By: Sil Vilerino <sivileri@microsoft.com>
Reviewed-by: Ruijing Dong <ruijing.dong@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30780>
 2024-08-28 13:38:52 +00:00
David Rosca
6d69748542 d3d12: Remove get_feedback_fence implementation 
The function signature and semantics will be changed in next commit.

Reviewed-By: Sil Vilerino <sivileri@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30780>
 2024-08-28 13:38:52 +00:00
Maíra Canal
ef4c41d7fd v3d, vc4: remove unused functions 
Signed-off-by: Maíra Canal <mcanal@igalia.com>
Reviewed-by: Juan A. Suarez <jasuarez@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30889>
 2024-08-28 13:15:45 +00:00
Faith Ekstrand
518b2d548f nir: Preserve fp_fast_math in nir_opt_vectorize() 
Fixes the following CTS tests on NVK:

dEQP-VK.spirv_assembly.instruction.*.float_controls.fp16.generated_args.signed_zero_sub_var_preserve*

Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30859>
 2024-08-28 12:29:06 +00:00
Samuel Pitoiset
7392e3306e radv: remove useless check about non-indexed draws and DGC 
The index buffer is only emitted inside the indexed draw path.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30852>
 2024-08-28 11:03:36 +00:00
Samuel Pitoiset
0e1e5264b5 radv: specialize indirect command layout stride for DGC 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30852>
 2024-08-28 11:03:36 +00:00
Samuel Pitoiset
8edbfbfe68 radv: specialize push constant DGC token 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30852>
 2024-08-28 11:03:36 +00:00
Samuel Pitoiset
7d0972711c radv: simplify allocating push constants with DGC 
Using a condition will allow to specialize it.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30852>
 2024-08-28 11:03:36 +00:00
Samuel Pitoiset
545949d12f radv: specialize VBO DGC token 
Can't really specialize more without rewriting VBO completely.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30852>
 2024-08-28 11:03:36 +00:00
Samuel Pitoiset
64076c652c radv: specialize pipeline DGC token 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30852>
 2024-08-28 11:03:36 +00:00
Samuel Pitoiset
7270bf7aa3 radv: specialize index buffer DGC token 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30852>
 2024-08-28 11:03:36 +00:00
Samuel Pitoiset
3128eca2d0 radv: specialize draw DGC token 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30852>
 2024-08-28 11:03:36 +00:00
Samuel Pitoiset
ccd55b55da radv: specialize dispatch DGC token 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30852>
 2024-08-28 11:03:36 +00:00
Samuel Pitoiset
b4793400f3 radv: add a pointer to the DGC layout in dgc_cmdbuf 
Will be useful.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30852>
 2024-08-28 11:03:36 +00:00
Samuel Pitoiset
c7540d3fd6 radv: prepare for specialized DGC shaders 
The DGC prepare shader is getting crazy and it takes a non-trivial
amount of time. Using specialized DGC shaders is cleaner and it's
faster than a pile of conditional SALU instructions.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30852>
 2024-08-28 11:03:36 +00:00
David Heidelberg
0dc791ed57 freedreno: use unicode © instead of DOS-like (C) copyright sign 
Acked-by: Rob Clark <robclark@freedesktop.org>
Signed-off-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30721>
 2024-08-28 08:54:00 +00:00
David Heidelberg
89366ff523 freedreno: Convert to SPDX-License-Identifier instead of pasting whole license 
SPDX is ISO standard now, let's leverage it to cleanup our code.

Acked-by: Rob Clark <robclark@freedesktop.org>
Signed-off-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30721>
 2024-08-28 08:54:00 +00:00
Georg Lehmann
ef970c5a9d nir: optimize pack_uint_2x16 of pack_half(a, 0) 
Foz-DB Navi31:
Totals from 31 (0.04% of 79395) affected shaders:
Instrs: 6157 -> 6065 (-1.49%)
CodeSize: 35676 -> 34936 (-2.07%)
Latency: 23979 -> 23805 (-0.73%); split: -0.79%, +0.07%
InvThroughput: 5248 -> 5124 (-2.36%)
VALU: 3224 -> 3162 (-1.92%)

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30855>
 2024-08-28 07:16:55 +00:00
David Rosca
79637d5942 radeonsi/vcn: Support H264 constrained intra prediction 
We now get this flag from PPS, so it can be enabled when requested.

Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Reviewed-by: Ruijing Dong <ruijing.dong@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30842>
 2024-08-28 06:48:32 +00:00
David Rosca
c53641f810 radeonsi: Implement buf_ofst_info in si_video_resize_buffer GPU copy path 
This is used when resizing VP9 Tier1 DPB.

Fixes: 322240fcff ("radeonsi: Add GPU copy path to si_video_resize_buffer")
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Reviewed-by: Ruijing Dong <ruijing.dong@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30842>
 2024-08-28 06:48:32 +00:00
David Rosca
134e95cdb1 radeonsi: Enable transform_skip HEVC feature support 
This used to be hardcoded in PPS, but now it can be selected by apps.
However, we were not reporting this feature as supported in HEVC feature
flags and apps would always disable it.

Fixes: af849516f0 ("radeonsi/vcn: Use pipe header params in HEVC header encoder")
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Reviewed-by: Ruijing Dong <ruijing.dong@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30842>
 2024-08-28 06:48:32 +00:00
Deborah Brouwer
18f15da94d ci/intel: add i915/MTL firmware to rootfs 
Add Meteor Lake firmware directly to rootfs since it is not available
from debian package.

Signed-off-by: Deborah Brouwer <deborah.brouwer@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30770>
 2024-08-28 04:31:10 +00:00
Deborah Brouwer
0441202d6b ci: add firmware files to rootfs 
Currently only package versions of firmware files are available in the
rootfs.

This commit allows firmware files to be pulled directly from a specific
git hash of a remote repository.

Signed-off-by: Deborah Brouwer <deborah.brouwer@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30770>
 2024-08-28 04:31:10 +00:00
Caio Oliveira
695f5314d6 intel/brw: Simplify fs_inst annotation 
When INTEL_DEBUG=ann is also set, the disassembler would annotate the
output with either a string or the string verison of a NIR instruction.
This was done by keeping two pointers (but only using one at a time).

Change the code to print the instruction into a string instead of
keeping it pointer around (peg the string to the shader).  That way,
only one pointer is needed for annotations.  Because that serialization
is not free, only do that when the environment variable is set.

Since we are here, move the annotation string field to the end, moving
it to the least commonly used cacheline.  Further packing might allow
the entire fs_inst to fit in two cachelines.

For release builds, don't even add the debug annotation to the struct.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30822>
 2024-08-28 03:59:50 +00:00
Caio Oliveira
ec15cdfa2a intel/brw: Pack brw_reg struct 
The alignment required for the second union (has 64-bit size) causes
a hole between the first and second union.  Move the remaining data
there.

In 64-bit build, shrinks brw_reg from 24 bytes to 16 bytes.  And by
consequence, shirnks fs_inst from 200 bytes to 160 bytes, making it
use one less cacheline.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30822>
 2024-08-28 03:59:50 +00:00
Iván Briano
2261b298d1 anv: fix adding to wa_addr 
Fixes: 6336e0fe7f ("anv: order data in wa_bo to leave wa_addr last")

Reviewed-by: Caio Oliveira <caio.oliveira@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30881>
 2024-08-27 18:10:58 -07:00
Yiwei Zhang
7941d705c3 venus: workaround cacheline overflush issue on Intel JSL 
We observed that Venus on ANV on JSL platform has some cacheline flush
issue. The overflush shows up as:
1. There're 2 threads venus bliting the feedback buffers suballocated
   from the same backing device memory, back to back.
2. On thread A, flushing the feedback buffer for cpu read is placed
   behind flushing a shader storage buffer for cpu read.
3. On thread B, flushing a different feedback buffer with the same
   backing device memory (different offset bound to) can kick the
   feedback buffer flush in (2) earlier than it should be flushed.
4. As a result, CPU polling thread for thread B results would see venus
   feedback buffer update earlier than shader storage buffer results
   being updated, breaking Venus sync primitives optimization.

During investigation, a solid workaround for JSL platform is to force
Venus to align up to 128 bytes for feedback buffer suballocation while
the default is at 64 bytes.

Cc: mesa-stable
Signed-off-by: Yiwei Zhang <zzyiwei@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30879>
 2024-08-28 00:43:59 +00:00
Eric Engestrom
c0762e88f3 ci/build: fix ppc64le and s390x jobs rules 
I think these were written with the idea of making it "(build rules) &&
(any relevant driver), but instead the driver rules are bypassing the
build rules because
1) it's not an AND, it's an OR; any line that matches applies, and
2) the driver rules are `when: on_success` when these need to be `when:
   manual` like the rest of the build jobs.

Let's stop trying to be special and simply behave like all the other
build jobs.

We can always try making complex rules later, but once we're on a base
that works.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30871>
 2024-08-28 00:11:26 +00:00
Eric Engestrom
f6eeb3c6d1 ci/image-tags: re-generate all the images building deqp-runner 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30866>
 2024-08-27 22:26:17 +00:00
Eric Engestrom
8a95129aee ci/deqp-runner: add infra to apply patches 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30866>
 2024-08-27 22:26:17 +00:00
Eric Engestrom
83d9cfa58d ci/deqp-runner: build from git checkout even on linux 
This allows things like patching deqp-runner, and unifies linux and
android.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30866>
 2024-08-27 22:26:17 +00:00
Eric Engestrom
03e50318ff ci/deqp-runner: be less verbose in the loop printing the deqp builds info 
The bash stuff printed in the middle makes it unnecessarily hard to read
the useful part of the output.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30866>
 2024-08-27 22:26:17 +00:00
Eric Engestrom
3b0c527b56 ci/deqp: simplify command to list local deqp patches 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30866>
 2024-08-27 22:26:17 +00:00
Eric Engestrom
ca7fde8761 ci/deqp-runner: restore CC after temporarily overriding it 
Fixes: 6edfb09dda ("ci/deqp-runner: unset CC for arm32 cross-compilation")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30866>
 2024-08-27 22:26:17 +00:00
Sagar Ghuge
17f97a69c1 iris: Reduce clear color state alignment to 64B 
Closes https://gitlab.freedesktop.org/mesa/mesa/-/issues/10067

Signed-off-by: Sagar Ghuge <sagar.ghuge@intel.com>
Reviewed-by: Nanley Chery <nanley.g.chery@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/26793>
 2024-08-27 21:13:30 +00:00