fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2026-05-05 13:58:04 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 2
196697 commits 66 branches 933 tags 988 MiB
Commit graph

196697 commits

This branch
This branch
All branches
Author SHA1 Message Date
Adam Jackson
514ba16d95 rusticl: Enable out-of-order execution 
Ought to work, let's find out.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/12029
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31672>
 2024-10-22 00:03:48 +00:00
Karol Herbst
1798597637 radeonsi: move si_compute::global_buffers to si_context 
si_set_global_binding is a context function, but it touches the bound
compute program. As radeonsi also advertizes PIPE_CAP_SHAREABLE_SHADERS
this function is supposed to be safe when the same compute state object is
bound to multiple contexts at once.

In order to fix this data race global_buffers is moved to si_context so it
becomes context private data instead.

Cc: mesa-stable
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31672>
 2024-10-22 00:03:48 +00:00
Christian Gmeiner
fad599a619 etnaviv: Enable ARB_framebuffer_no_attachments 
There is not much needed to pass the supported tests.

Signed-off-by: Christian Gmeiner <cgmeiner@igalia.com>
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31578>
 2024-10-21 23:27:30 +00:00
David Heidelberg
3dbbc24cb3 ci/build: debian-build-testing should follow same limit as other jobs 
This was temporarily measure to prevent long LTO builds get stuck.

Signed-off-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31758>
 2024-10-21 22:30:09 +00:00
Eric Engestrom
5a65c478dd ci: fix shader-db rules 
All the post-merge pipelines are failing to be created with this error:

  'shader-db' job needs 'debian-build-testing' job, but 'debian-build-testing' is not in any previous stage

Like all the other test jobs, shader-db should never run in post-merge
pipelines.

Fixes: 0724b00dad ("ci: run shader-db only when one of the drivers it tests is modified")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31758>
 2024-10-21 22:30:08 +00:00
Amber
9ace01edbe tu, ir3: Implement VK_KHR_shader_atomic_int64 for a7xx. 
Passes dEQP-VK.glsl.atomic_operations.*64bit*

Signed-off-by: Amber Harmonia <amber@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27776>
 2024-10-21 21:47:44 +00:00
Amber
a3afe22dc9 nir: add pass to lower atomic arithmetic to a loop with cmpxchg. 
Signed-off-by: Amber Harmonia <amber@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27776>
 2024-10-21 21:47:44 +00:00
Amber
7d0870e5d5 ir3: add support for 64 bit atomics 
Signed-off-by: Amber Harmonia <amber@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27776>
 2024-10-21 21:47:44 +00:00
Amber
5628a01523 ir3: add encoding support for 64-bit atomics introduced in a7xx. 
Signed-off-by: Amber Harmonia <amber@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27776>
 2024-10-21 21:47:44 +00:00
Mary Guillemard
84d57e1fb1 nir: Move atomic_op_to_alu to common code 
Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27776>
 2024-10-21 21:47:44 +00:00
quic_lkondred
334af37697 freedreno: Add support for Adreno 663 GPU 
Add support to enable basic functionality of Adreno 663 GPU.

Signed-off-by: quic_lkondred <quic_lkondred@quicinc.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31211>
 2024-10-21 21:05:57 +00:00
Eric Engestrom
ce02c9b5ef ci: simplify setting lava job priority 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31755>
 2024-10-21 20:18:23 +00:00
Eric Engestrom
e70db26d18 ci: simplify ci-tron priority variable 
Set it to low, and only override in merge pipelines.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31755>
 2024-10-21 20:18:23 +00:00
Rebecca Mckeever
c2299b6642 panvk/csf: Implement vkCmdExecuteCommands 
Signed-off-by: Rebecca Mckeever <rebecca.mckeever@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31159>
 2024-10-21 19:39:07 +00:00
Rebecca Mckeever
3513960fe6 panvk: Move rendering info to panvk_rendering_state struct 
Signed-off-by: Rebecca Mckeever <rebecca.mckeever@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31159>
 2024-10-21 19:39:07 +00:00
Rebecca Mckeever
ddb2f6f1fa panvk: Track VkFormats for depth and stencil attachments 
These can be used directly in vk_meta_rendering_info.

Signed-off-by: Rebecca Mckeever <rebecca.mckeever@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31159>
 2024-10-21 19:39:07 +00:00
Rebecca Mckeever
26a03222ef panvk/csf: Split tiler flags initialization into two steps 
The first step conditionally updates the tiler flags based on dirty bits,
and the second step is the override flags, which are unconditionally
updated at draw time.

Use pan_pack_nodefaults() to avoid default initialized fields.

Signed-off-by: Rebecca Mckeever <rebecca.mckeever@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31159>
 2024-10-21 19:39:07 +00:00
Rebecca Mckeever
07346ea1e6 panvk/csf: Set and clear vb.dirty flag 
Fixes: 5544d39f44 ("panvk: Add a CSF backend for panvk_queue/cmd_buffer")

Signed-off-by: Rebecca Mckeever <rebecca.mckeever@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31159>
 2024-10-21 19:39:07 +00:00
Marek Olšák
fb6184f89c nir: add shader_info::tess::tcs_same_invocation_inputs_read(_indirect) 
We need both the same-invocation usage mask and cross-invocation usage
mask. The AMD reason is below.

Cross-invocation TCS input access doesn't prevent the same-invocation
fast path in AMD hw because it's just a different way to load the same
data, and we want to use both paths for the same TCS input based on
the load instruction. The fast path can't be used for indirect access,
which is gathered separately for same-invocation access.

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31645>
 2024-10-21 18:53:51 +00:00
David Heidelberg
9ef6ff1702 build: Fix LTO with Android NDK 
`no-emulated-tls` is passed, but also link arguments are required for LTO.

Tested-by: Mauro Rossi <issor.oruam@gmail.com>
Signed-off-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31615>
 2024-10-21 17:39:59 +00:00
Rhys Perry
8221367fba radv: use explicitly sized types for some radv_shader_info members 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31593>
 2024-10-21 15:52:53 +00:00
Rhys Perry
4383a917e6 radv: optimize VS input load components to constants earlier 
This helps some linking optimizations.

fossil-db (navi21):
Totals from 2262 (2.85% of 79395) affected shaders:
MaxWaves: 57680 -> 57738 (+0.10%); split: +0.11%, -0.01%
Instrs: 1061526 -> 1053937 (-0.71%); split: -0.79%, +0.07%
CodeSize: 5766352 -> 5736784 (-0.51%); split: -0.60%, +0.08%
VGPRs: 89376 -> 89000 (-0.42%); split: -0.43%, +0.01%
Latency: 4102938 -> 4059773 (-1.05%); split: -1.14%, +0.08%
InvThroughput: 1105885 -> 1092291 (-1.23%); split: -1.24%, +0.01%
VClause: 18917 -> 18972 (+0.29%); split: -0.12%, +0.41%
SClause: 28839 -> 28115 (-2.51%); split: -3.32%, +0.81%
Copies: 73396 -> 72671 (-0.99%); split: -1.63%, +0.65%
PreSGPRs: 65866 -> 65838 (-0.04%); split: -0.22%, +0.17%
PreVGPRs: 69752 -> 69278 (-0.68%)
VALU: 680351 -> 673489 (-1.01%); split: -1.03%, +0.02%
SALU: 121459 -> 121515 (+0.05%); split: -0.00%, +0.05%
VMEM: 29632 -> 30021 (+1.31%); split: -0.02%, +1.33%
SMEM: 73744 -> 73836 (+0.12%); split: -0.01%, +0.14%

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31593>
 2024-10-21 15:52:53 +00:00
Rhys Perry
9784165de5 radv: fix output statistic for fragment shaders 
This is a per-component bit mask (0xf for each output).

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Fixes: 0e0c2574d1 ("radv: Add shader stats for inputs and outputs.")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31593>
 2024-10-21 15:52:52 +00:00
Marek Olšák
45d8cd037a ac/nir: rewrite ac_nir_lower_ps epilog to fix dual src blending with mono PS 
Unigine Heaven with AMD_DEBUG=mono has incorrect rendering on gfx11
because it doesn't set nir_io_semantics::dual_source_blend_index for
the second output, resulting in garbage asm.

Instead of trying to find out what's wrong, I decided to rewrite this
to make it the same as the LLVM IR path. It simplifies the code and fixes
Unigine Heaven with AMD_DEBUG=mono.

Reviewed-by: Qiang Yu <yuq825@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31669>
 2024-10-21 12:06:14 +00:00
Eric Engestrom
d117411309 {freedreno,intel}/ci: add missing tracking of merge-skips.txt files 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31739>
 2024-10-21 10:22:57 +00:00
Eric Engestrom
ee0d782229 {freedreno,intel}/ci: rename "premerge-skips.txt" to "merge-skips.txt" to accurately reflect what they are 
This also means the infrastructure added by @gallo in 1dc64d0613
("ci: Use merge-skips files during merge pipelines") can be used and all
the manual adding of these files can be dropped, reducing the likeliness
of bugs.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31739>
 2024-10-21 10:22:57 +00:00
Eric Engestrom
bb98949134 ci: rename "freedreno" farm to "google-freedreno" 
There are 3 freedreno farms: google, collabora, and valve.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31709>
 2024-10-21 09:36:05 +00:00
Eric Engestrom
18489da224 freedreno/ci: use {freedreno,turnip}{,-manual}-rules 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31709>
 2024-10-21 09:36:05 +00:00
Eric Engestrom
51125f8b39 freedreno/ci: add {freedreno,turnip}{,-manual}-rules to simplify the rest of the code 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31709>
 2024-10-21 09:36:05 +00:00
Eric Engestrom
85de24ebc5 ci/prepare-artifacts: use more than one thread when compressing the install tarball 
Also, add `--quiet` to avoid partial progress output.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30416>
 2024-10-21 08:52:23 +00:00
Eric Engestrom
e9998fdef5 ci/{deqp,piglit}-runner: make zstd respect FDO_CI_CONCURRENT when set, and fallback to the current "all the threads" 
Also, expand `-q` to `--quiet` to be more explicit, and drop a stray
`-c` which did nothing except cancel out `--rm`.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30416>
 2024-10-21 08:52:22 +00:00
Pavel Ondračka
9672f9732f r300/ci: add new RV410 flakes 
Signed-off-by: Pavel Ondračka <pavel.ondracka@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31753>
 2024-10-21 08:22:32 +00:00
Georg Lehmann
10951bb11a aco: fix 64bit extract_i8/extract_i16 
The old code only sign extended to 32bit, with a zero hi half.

Fixes: 1f2518ef9f ("aco: implement nir_op_extract/nir_op_insert")
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31734>
 2024-10-21 07:13:57 +00:00
David Heidelberg
4375133abb ci/alpine: stay trendy, Alpine flies on LLVM 19 now 
Signed-off-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31656>
 2024-10-20 19:39:32 -07:00
Georg Lehmann
3120fec0c0 ci: remove selinux from containers 
Reviewed-by: Michel Dänzer <mdaenzer@redhat.com>
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31656>
 2024-10-21 01:14:35 +00:00
Georg Lehmann
894c4f0c78 meson: remove selinux option 
Reviewed-by: Michel Dänzer <mdaenzer@redhat.com>
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31656>
 2024-10-21 01:14:35 +00:00
Daniel Stone
ec95f07a1f ci: Quieten test execution 
Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31602>
 2024-10-20 11:32:43 +01:00
Daniel Stone
ef175ea594 ci: Make section emission really quiet 
If you're emitting a section header under set -x, you will see:
+ section_start foo "foo"
+ x_off
[the section header]

This is kind of annoying. Instead of trying to squash it everywhere by
dancing around local set +x management, play some extremely stupid
tricks with shells to make sure we never emit it.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31602>
 2024-10-20 11:32:43 +01:00
Daniel Stone
8316654dac ci: Quieten post-test cleanup 
Ordinary people are not interested in reading this.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31602>
 2024-10-20 11:32:43 +01:00
Daniel Stone
dd415e45bb ci/traces: Use sections to quiet execution 
Only print the good bits by default.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31602>
 2024-10-20 11:32:43 +01:00
Daniel Stone
5dcd5bfc3e ci/crosvm: Replace SCRIPT_DIR with SCRIPTS_DIR 
This is what gets used everywhere else, so do that.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31602>
 2024-10-20 11:32:43 +01:00
Daniel Stone
5edb11e9ea ci/crosvm: Make crosvm-runner safe for set +u 
This is a really useful thing to have; the next commit cost me an
embarrassing amount of time for not having it.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31602>
 2024-10-20 11:32:43 +01:00
Daniel Stone
da292b019c ci/bare-metal: Use section for boot messages from poe-powered 
Very few people ever want to see the exact details of how we attempt to
boot a device.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31602>
 2024-10-20 11:32:43 +01:00
Daniel Stone
d9e1b29504 ci/bare-metal: Truncate printed times and prefixes 
The date isn't interesting, since you can tell if it wraps around
anyway. Just print the time down to the millisecond, which should be
plenty enough. We also don't need to print 'R SERIAL_CPU> ' when almost
every line is reading from the CPU serial buffer, so make that the
default.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31602>
 2024-10-20 11:32:42 +01:00
Daniel Stone
3ed9fd3140 ci/bare-metal: Use section for boot messages from cros-servo 
Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31602>
 2024-10-20 11:32:42 +01:00
Daniel Stone
a9b03a8fc4 ci/bare-metal: Move preparation steps into section 
Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31602>
 2024-10-20 11:32:42 +01:00
Daniel Stone
d171f47f44 ci/lava: Coalesce post-processed job information 
We can combine a section _and_ a print.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31602>
 2024-10-20 11:32:42 +01:00
Daniel Stone
9be46b29f0 ci/lava: Print relative timestamps in sections 
Follow what the shell executor does and print the time since the job
started in the section header.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31602>
 2024-10-20 11:32:42 +01:00
Daniel Stone
8ee6241a8c ci/hw: Wrap pre-test setup in collapsed section 
Most people don't care about environment variables and starting Weston.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31602>
 2024-10-20 11:32:42 +01:00
Daniel Stone
970f37be09 ci/lava: Change default section colour to cyan 
This matches the sections from the shell prints, which are quite nice.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31602>
 2024-10-20 11:32:42 +01:00