AlexIndustrial/mesa
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
211,065 Commits 1 Branch 0 Tags
cd91d2f4abb2ea76534a57baee90a5ff184ed7df
Commit Graph

211065 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Samuel Pitoiset cd91d2f4ab radv: add radv_cmd_set_patch_control_points() 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36979>
 2025-08-27 09:22:16 +00:00
Samuel Pitoiset 97597f65bf radv: add radv_cmd_set_tessellation_domain_origin() 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36979>
 2025-08-27 09:22:15 +00:00
Samuel Pitoiset 091f8d53cb radv: add radv_cmd_set_line_width() 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36979>
 2025-08-27 09:22:15 +00:00
Georg Lehmann e270a7480b nir/lower_io: fix boolean output stores 
Stores don't have a definition, we have to check the bit size of the source.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13762
Fixes: c217ee8d35 ("nir: Insert b2b1s around booleans in nir_lower_to")
Reviewed-by: Konstantin Seurer <konstantin.seurer@gmail.com>
Reviewed-by: Mary Guillemard <mary@mary.zone>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36966>
 2025-08-27 08:46:34 +00:00
Georg Lehmann 13a9f27432 ac/nir: do not assume mesh cull flag is 1bit 
It will no longer be 1bit after a nir/lower_io bug is fixed.

Cc: mesa-stable
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36966>
 2025-08-27 08:46:33 +00:00
Georg Lehmann 8a5178755e radv: reorder cmat properties according to performance 
On GFX12, int8 is twice as fast as fp16/bf16.
On GFX11, they have the same throughput, but int8 at least still uses
less registers.

Also reorder 16bit accumulators before 32bit, because they use less
registers on GFX12.

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37002>
 2025-08-27 08:20:47 +00:00
Georg Lehmann 5a10142a9f radv/nir/lower_cmat: split up larger nested switches 
This has been annoying me for quite some while, the level of indention
makes reviewing code changes in Gitlab harder.

I think now is a good time to change this before more cmat lowering is added.

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37002>
 2025-08-27 08:20:47 +00:00
Georg Lehmann 047b95a8c3 nir/shrink_vec_array_vars: detect zero init shared memory using constant initializer 
More consistent.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36956>
 2025-08-27 06:37:41 +00:00
Georg Lehmann edc5bea61e nir/shrink_vec_array_vars: update constant initializer after shrinking 
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13751
Fixes: c7df3b4f64 ("nir/shrink_vec_array_vars: allow nir_var_mem_shared")
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36956>
 2025-08-27 06:37:41 +00:00
Samuel Pitoiset 47cca88f3c radv: check flush_bits before calling radv_emit_cache_flush() in the draw path 
This is also faster.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37005>
 2025-08-27 06:19:14 +00:00
Samuel Pitoiset 5aba130eab radv: check if SQTT is enabled before calling radv_describe_draw() 
This is faster.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37005>
 2025-08-27 06:19:14 +00:00
Samuel Pitoiset a3f9d85270 radv: add a new state for forced VRS rates 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37005>
 2025-08-27 06:19:14 +00:00
Samuel Pitoiset 6d1e7c2ddf radv: clear descriptors state dirty bit outside of the caller 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37005>
 2025-08-27 06:19:13 +00:00
Samuel Pitoiset 9b912f00c7 radv: handle fbfetch output after binding graphics shaders 
uses_fbfetch_output might be updated when a fragment shader is bound.
This would only affect ESO and I'm not sure it's possible though.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37005>
 2025-08-27 06:19:13 +00:00
Samuel Pitoiset 664c9697aa radv: clear RADV_CMD_DIRTY_xxx bits outside of the caller in more cases 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37005>
 2025-08-27 06:19:13 +00:00
Julia Zhang 20b809f1f0 pps: init driver in OnSetup 
Initialization of driver has been moved to register_data_source() from
OnSetup() by: a739889789 ("pps: Report available counters when
gpu.counters* data source is registered")

With above change, pps will destroy driver when collecting data stops
(pps may keep running) then the driver will become nullptr when user try
to collect data again. This will cause segmentation fault in OnSetup().

So this remove driver = nullptr in OnStop() and init driver in OnSetup()
to make sure driver exists when pps-producer run more than once.

Fixes: a739889789 ("pps: Report available counters when gpu.counters*
data source is registered")

Signed-off-by: Julia Zhang <Julia.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36548>
 2025-08-27 02:27:03 +00:00
Faith Ekstrand a1d5e8bfdb compiler/rust: Fix the DFS loop detection algorithm 
The previous algorithm just looked at the dominator's loop header.
However, if you have multiple consecutive loops like:

    function_impl {
        loop {
            // Stuff
        }
        loop {
            // Other stuff
        }
    }

then it will look like the second loop is contained in the first loop
because the first loop's header dominates the second loop.  This isn't
actually what we want.  Instead, we want a node N to be considered part
of a loop with header H if H dominates N and H is reachable from N.

Fixes: 741f7067f1 ("nak: Add loop detection to the CFG")
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36524>
 2025-08-27 01:20:05 +00:00
Faith Ekstrand ab97fea323 nak: Increase Imma latencies on Blackwell by 4 
We already do +8 on HMMA.  This gives us a +4 on IMMA which seems to fix
the test failures I'm seeing.  This fixes the IMMA CTS tests on my RTX
5090 with NAK_DEBUG=spill.

Fixes: 477533ee00 ("nvk: add sm120 latencies via csv files.")
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36524>
 2025-08-27 01:20:05 +00:00
Mel Henning 606fa193af nvk: Reduce subc switches in cmd_invalidate_deps 
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37018>
 2025-08-27 01:04:51 +00:00
Mel Henning cb22409d72 nvK: Add nvk_cmd_buffer_last_subchannel 
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37018>
 2025-08-27 01:04:51 +00:00
Mel Henning abd374628b nvk: Split out NVC0_FIFO_SUBC_FROM_PKHDR helper 
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37018>
 2025-08-27 01:04:50 +00:00
Mohamed Ahmed 7692d3c0e1 nvk: Advertise VK_KHR_maintenance9 
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35804>
 2025-08-27 00:50:27 +00:00
Mohamed Ahmed b7be6e73c9 nvk: Add support for VK_QUERY_POOL_CREATE_RESET_BIT_KHR 
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35804>
 2025-08-27 00:50:27 +00:00
Mohamed Ahmed bfc6ec0981 nvk: Skip creating a nvkmd device if we don't have to 
KHR_maintenance9 allows for the creation of VkDevices with no queues for the
purpose of offline compilation and other similar usages. Notably, devices with
0 queues aren't allowed to allocate memory, create command buffers, or create
fences/semaphores; this allows us to completely skip creating a nvkmd device
and going through the kernel for these objects.

Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35804>
 2025-08-27 00:50:27 +00:00
Mohamed Ahmed 969e639344 nvk: Ensure we have nvkmd before sampler descriptor upload 
We need to be able to create samplers without necessarily having to have to
upload them. This is to allow us later to create a VkDevice without having to
touch the kernel for VkDevice objects without VkQueues used for things like
shader compilation.

Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35804>
 2025-08-27 00:50:27 +00:00
Mohamed Ahmed 5945d42d3b nvk: Ensure we have nvkmd before shader upload 
This is to allow us later to create a VkDevice without having to touch the
kernel for VkDevice objects without VkQueues used for things like shader
compilation.

Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35804>
 2025-08-27 00:50:27 +00:00
Mohamed Ahmed 49db51e5ff nak: Enable lowering for bitfield manipulation at <32bit sizes 
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35804>
 2025-08-27 00:50:27 +00:00
Mohamed Ahmed 16f3b9d984 nak: Fix 64-bit bit_count, ufind_msb, ifind_msb, find_lsb 
Previously, these were going through nir_lower_bit_size, which handles
upscaling from lower bit sizes to 32-bit only and is incorrect for higher sizes.
This change uses nir_lower_int64 instead, which handles them properly and
correctly.

Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35804>
 2025-08-27 00:50:27 +00:00
Ritesh Raj Sarraf 3dc9f2f70f ci: Uprev kernel to pick up virtio fix and use it for venus 
The regressed changes have been reverted for Linux 6.16 in commit 9d85a32f0e6bf6878676a9ec7ce3c8032c500c7c

This should now allow us to use Linux 6.16 for the rest of the testing

Signed-off-by: Ritesh Raj Sarraf <ritesh.sarraf@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36909>
 2025-08-27 00:15:57 +00:00
Faith Ekstrand 0a14ce7f50 nak: NAK_MAX_QMD_SIZE_B should be 384 
Also add a static assert so we don't miss this again.

Fixes: 00a845a698 ("nak/qmd: QMD versions 4.0 and 5.0 are both 384B")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37015>
 2025-08-27 00:00:54 +00:00
Mel Henning 265afd9bfd loader: Don't fall back to nouveau GL without zink 
Long term, we don't want to support nouveau gl on new cards. Remove
the fallback so users without zink will get software rendering
instead of nouveau gl.

For now, NOUVEAU_USE_ZINK will still select nouveau gl on cards where
that is possible, but that isn't really supported and will likely be
removed for a lot of cards in the future.

Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36971>
 2025-08-26 23:36:46 +00:00
Autumn Ashton 8e1734d2f2 radv: Allow MUTABLE_FORMAT and EXTENDED_USAGE for qp_map images 
Signed-off-by: Autumn Ashton <misyl@froggi.es>

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36943>
 2025-08-26 23:19:58 +00:00
Autumn Ashton f4597930ef radv: Support VK_IMAGE_TILING_OPTIMAL for quantization maps 
Saw this when implementing something else, that I could just add the handling here to radv_choose_tiling in order to expose OPTIMAL.

Signed-off-by: Autumn Ashton <misyl@froggi.es>

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36943>
 2025-08-26 23:19:58 +00:00
Guilherme Gallo 70a2eec2ab ci/baremetal: Use curl-with-retry in build scripts 
Signed-off-by: Guilherme Gallo <guilherme.gallo@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36946>
 2025-08-26 21:58:02 +00:00
Guilherme Gallo 5e745f653f ci/android: Use curl-with-retry in build scripts 
Let's standardize llvm android build script, by calling the
`curl-with-retry` instead of a specific curl arg set.

Signed-off-by: Guilherme Gallo <guilherme.gallo@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36946>
 2025-08-26 21:58:02 +00:00
Guilherme Gallo a51b12bbf6 ci/android: Use find_s3_project_artifact in build script 
So we can reuse artifact finding functionality and also can make the
`gallo/mesa` workaround to work.

Signed-off-by: Guilherme Gallo <guilherme.gallo@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36946>
 2025-08-26 21:58:02 +00:00
Guilherme Gallo de5f96b4de ci/baremetal: Use find_s3_project_artifact on baremetal_build.sh 
So we can reuse artifact finding functionality and also can make the
`gallo/mesa` workaround to work.

Signed-off-by: Guilherme Gallo <guilherme.gallo@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36946>
 2025-08-26 21:58:02 +00:00
Faith Ekstrand 50f5ade796 nak: Assert that QMDs are big enough 
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36995>
 2025-08-26 21:42:20 +00:00
Faith Ekstrand 0c71f5f1bd nouveau/headers: Add a MAX_BIT for structs 
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36995>
 2025-08-26 21:42:20 +00:00
Faith Ekstrand 00a845a698 nak/qmd: QMD versions 4.0 and 5.0 are both 384B 
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Backport-to: 25.2
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36995>
 2025-08-26 21:42:20 +00:00
Faith Ekstrand 0e268dad00 nvk: Allow for larger QMDs 
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Backport-to: 25.2
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36995>
 2025-08-26 21:42:20 +00:00
Faith Ekstrand 238534e069 nak/hw_runner: Allow for variable sized QMDs 
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Backport-to: 25.2
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36995>
 2025-08-26 21:42:19 +00:00
Faith Ekstrand 02ef6a5158 nak: Add a nak_qmd_size_B() query 
Also add a #define NAK_QMD_ALIGN_B for alignments.  The alignment is
always 256B and there's no evidence of that changing.

Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Backport-to: 25.2
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36995>
 2025-08-26 21:42:19 +00:00
Faith Ekstrand 8e93a763a3 nouveau/push: Handle more recent versions of 6F 
This also requires us to rework SET_OBJECT handling a bit.

Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36995>
 2025-08-26 21:42:19 +00:00
Faith Ekstrand 38108b2c4b nvk,nvkmd: Move push dumping to NVKMD 
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36995>
 2025-08-26 21:42:19 +00:00
Faith Ekstrand 7a50d7004b nvk/nvkmd: Track all memory objects by default 
A single lock on allocate/free is nothing compared to the ioctls we're
already taking.  This ensures that we always have all our memory
objects.

Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36995>
 2025-08-26 21:42:19 +00:00
Faith Ekstrand 7f95c07080 nvk/nvkmd: Stop setting WAIT_FOR_SUBMIT for sync 
We add the syncobj and exec right before the wait so we know it's been
submitted. WAIT_FOR_SUBMIT does nothing except fail to catch potential
errors.

Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36995>
 2025-08-26 21:42:18 +00:00
Faith Ekstrand 63263baeca nak: Lower away ldcx when NAK_DEBUG=no_ugpr is set 
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36995>
 2025-08-26 21:42:18 +00:00
Mary Guillemard becccebd21 nvk: Use MEM_LOCAL for nvk_cmd_mem_create 
It is safe to allocate command buffer memory as MEM_LOCAL (VRAM or GART)
when we can map GPU memory (we trust NVKMD to force GART).

This reduces latency on pyrowave for compute dispatch (~1.90ms -> ~1.48ms),
likely as QMD and the root tables should not have been in GART in the first place.

Signed-off-by: Mary Guillemard <mary@mary.zone>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36970>
 2025-08-26 21:25:08 +00:00
Mary Guillemard 5d77057ff6 nvk: Force GART for command buffers 
We are already always allocating on GART at the moment, let's force it
for now as we are going to change the default.

Signed-off-by: Mary Guillemard <mary@mary.zone>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36970>
 2025-08-26 21:25:08 +00:00