-
7fc153c9
by Eric Engestrom at 2025-09-03T13:55:58+02:00
docs: add sha sum for 25.2.2
-
f77b2d5c
by Eric Engestrom at 2025-09-17T13:59:13+02:00
.pick_status.json: Update to f99230cc67b0af4d403374a7c12529e566590a8b
-
10f74cca
by Eric Engestrom at 2025-09-17T13:59:13+02:00
.pick_status.json: Mark 83bf75e307b2171094811884dbd350399cf145ba as denominated
-
58f1b12b
by Eric Engestrom at 2025-09-17T13:59:13+02:00
.pick_status.json: Mark 22606815722fdc36ecb8afb4a21a825f1b459e1f as denominated
-
6dab5618
by Eric Engestrom at 2025-09-17T13:59:13+02:00
.pick_status.json: Mark a09d18a316efb339c0ace43c4f3192cac169df4a as denominated
-
6df08558
by Eric Engestrom at 2025-09-17T13:59:13+02:00
.pick_status.json: Mark d68344bffe893c1ecaa63ce5f2795c3479517da6 as denominated
-
12f8f6d6
by Eric Engestrom at 2025-09-17T13:59:13+02:00
.pick_status.json: Mark 87ce4c6ca4a13952a5b0658c94d638cf52afaed0 as denominated
-
46c12e80
by Eric Engestrom at 2025-09-17T13:59:13+02:00
.pick_status.json: Mark bd074a542e203255f1daa5657fd09b88d2e6679e as denominated
-
8cc2dd19
by Eric Engestrom at 2025-09-17T13:59:13+02:00
.pick_status.json: Mark aed7e7add135bbf33de689048a2766ea9e7f095f as denominated
-
63f47a97
by Erik Faye-Lund at 2025-09-17T13:59:13+02:00
panvk: fix up vk1.4 properties
These two properties reports how the interaction between MSAA coverage
and occlusion queries works. We need to report the correct value here,
otherwise applications might misbehave.
Fixes: 5ee3c10d1ed ("panvk: advertise vulkan 1.4 on v10+")
Reviewed-by: Christoph Pillmayer <christoph.pillmayer@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37136>
(cherry picked from commit 166d650c10ce28a2fa00d39d8591be98948a4b4c)
-
dda8f9f5
by Danylo Piliaiev at 2025-09-17T13:59:14+02:00
tu: Reset rp_trace on tu_reset_cmd_buffer
Secondary command buffers with RENDER_PASS_CONTINUE_BIT don't reset
rp_trace, and without reset we get garbage tracepoints.
Fixes garbage sysmem_clear_all tracepoints in some games running
though DXVK.
Fixes: 630380349bb ("tu: Give renderpass events a separate trace buffer")
Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37161>
(cherry picked from commit 482e0d0d1e965b3fa1f9faa6f63b11fa83e8fd82)
-
8dc2746a
by Yiwei Zhang at 2025-09-17T13:59:14+02:00
radv: bind aliased wsi image at memory offset zero
This aligns with common wsi, and also obeys dedicated alloc requirement.
Fixes: 825c05a7e8f ("radv: adopt wsi_common_get_memory")
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37152>
(cherry picked from commit 94d8a4a4656738c32457d19e0562c3cb54811fd6)
-
8f756344
by Yiwei Zhang at 2025-09-17T13:59:14+02:00
nvk: bind aliased wsi image at memory offset zero
This aligns with common wsi, and also obeys dedicated alloc requirement.
Fixes: 273df23a219 ("nvk: adopt wsi_common_get_memory")
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37152>
(cherry picked from commit 611749a7f93dfba0802b49f2b14e455daec35781)
-
b6e6285d
by Eric R. Smith at 2025-09-17T13:59:14+02:00
panfrost: fix typo in register allocation
The intention of the code was to allow PHI values to be propagated
if they were in registers (as opposed to in memory). As written though
values were never propagated. I think this typo was due to some
debug code that wasn't removed properly.
Fixes: 6c64ad934f7 ("panfrost: spill registers in SSA form")
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Ashley Smith <ashley.smith@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37092>
(cherry picked from commit d482b6ca68666a4905a2561a92ddedd16972c157)
-
3e543205
by Eric R. Smith at 2025-09-17T13:59:14+02:00
panfrost: fix debug print of spilled registers
We were testing some conditions in the wrong order, so spilled
registers were being printed as if they were uniforms. This is
incorrect, but only subtly so, and lead to confusion.
Fixes: 6c64ad934f7 ("panfrost: spill registers in SSA form")
Reviewed-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Ashley Smith <ashley.smith@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37092>
(cherry picked from commit e3552c427eae1a829ff20f3b977fb088687c0ca2)
-
677f1c45
by Mike Blumenkrantz at 2025-09-17T13:59:14+02:00
tc: don't unset resolve resource in set_framebuffer_state
this breaks the whole mechanism
Fixes: 2eb45daa9c8 ("gallium: de-pointerize pipe_surface")
Acked-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37166>
(cherry picked from commit 7984a16e27bb30cbe3f6a6361603557c24322e1d)
-
ec27b06c
by Yonggang Luo at 2025-09-17T13:59:14+02:00
tgsi: Fixes ntt_should_vectorize_io parameters
Fixes: 5f757bb95c2 ("nir: Make the load_store_vectorizer provide align_mul + align_offset.")
This is found when I am trying to narrow bit_size and num_components to uint8_t
Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Emma Anholt <emma@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37042>
(cherry picked from commit 949a05693415cd39ffadbe44867be0b71e3627f5)
-
46d835f8
by Yiwei Zhang at 2025-09-17T13:59:14+02:00
tu: bind aliased wsi image at memory offset zero
The vulkan spec says that we should ignore memoryOffset when
VkBindImageMemorySwapchainInfoKHR is present. wsi common assumes that we
bind the wsi image at offset 0, so set the offset to 0. This change
aligns with common wsi, and also obeys dedicated alloc requirement.
Fixes: f887116c499 ("turnip: adopt wsi_common_get_memory")
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37099>
(cherry picked from commit cef48af271b8bdb360f2a31e876ecdc05a820753)
-
a5e92689
by Mel Henning at 2025-09-17T13:59:14+02:00
nvk: Clear cond_render_gart_* in reset_cmd_buffer
nvk_cmd_pool_free_gart_mem_list frees this buffer, so we need to clear
the pointers to it in order to avoid a use after free.
Fixes: 07c70c77de ("nvk: add cond render upload buffer.")
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Reviewed-by: Mary Guillemard <mary@mary.zone>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37153>
(cherry picked from commit eaa547f6f2193e50ffd5564e1f143ca42defaef8)
-
8534c245
by Mel Henning at 2025-09-17T13:59:14+02:00
nvk: Clear second SET_RENDER_ENABLE operand
The hardware actually compares a pair of 64-bit values, rather than
comparing a single value against zero like we previously assumed.
This wasn't an issue in most cases before because if the buffer is
zero-initialized the previous code happens to work. If we get a
buffer with garbage in it though we would run into issues.
Fixes: 80eac1337d ("nvk: Always copy conditional rendering value before compare")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13821
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Reviewed-by: Mary Guillemard <mary@mary.zone>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37153>
(cherry picked from commit 90ac7d13dce89d236b2273255fc291edf49bc204)
-
4623e861
by Samuel Pitoiset at 2025-09-17T13:59:14+02:00
radv: fix vk_error in radv_update_preambles()
This needs to be a vk object, otherwise it asserts or crashes.
Cc: mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37179>
(cherry picked from commit 57deff5658cc0fa7dcd2676c552afc4a50e5048d)
-
36ef404a
by Timur Kristóf at 2025-09-17T13:59:14+02:00
radv/amdgpu: Fix crash with RADV_DEBUG=noibs
After a refactor last year, the noibs option stopped working
because it hits an assertion when empty IBs are submitted.
Emit a single large NOP packet to avoid submitting empty IBs.
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37121>
(cherry picked from commit 132a61c6b72d8e96ae7e0022a394ac470446b7ce)
-
15d67c30
by Timur Kristóf at 2025-09-17T13:59:14+02:00
radv/amdgpu: Use correct NOP packets when unchaining a CS
GFX6 doesn't support single-dword PKT3 NOP packets,
so they shouldn't be used when unchaining a CS.
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37121>
(cherry picked from commit 3056279d096981a119a3e8a020367df1736f80e8)
-
79a84be7
by Timur Kristóf at 2025-09-17T13:59:14+02:00
radv/amdgpu: Don't use IB2 on GFX6 (for now)
GFX6 actually supports IB2, but doesn't support chaining between
chunks inside the IB2. See WaCpIb2ChainingUnsupported in PAL.
Disable IB2 on GFX6 for now.
The proper fix will be to disable use_ib in just secondary
command buffers on GFX6 and emit multiple IB2 packets in the
main command buffer. This will be implemented later.
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37121>
(cherry picked from commit e8d1e935fb41bfc593f473d99f4711073307d013)
-
56d84fb0
by Timur Kristóf at 2025-09-17T13:59:14+02:00
radv: Don't use EVENT_WRITE_EOS on GFX7
EOS events are buggy on GFX7 and can cause hangs when used
together in the same IB with CP DMA packets that use L2.
While we don't use the L2 for CP DMA copies, we still use it
with CP DMA prefetches, so the issue needs to be mitigated.
As a mitigation, avoid using EVENT_WRITE_EOS and prefer to use
the BOTTOM_OF_PIPE event instead of PS_DONE/CS_DONE, which should
be close enough.
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37121>
(cherry picked from commit c56c746b718d518ddaea687a229d2815d04d1314)
-
3403c49b
by Timur Kristóf at 2025-09-17T13:59:15+02:00
radeonsi: Flush L2 for render condition when CP can't use L2
If CP can't use L2 then it also can't read the render condition
through L2, so we need a flush, just like on GFX6-8.
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37121>
(cherry picked from commit 637f618ac5c46c2813645fc5279e89bfc768c280)
-
390cf756
by Valentine Burley at 2025-09-17T13:59:15+02:00
tu: Enable robustBufferAccessUpdateAfterBind
This is supported and must be enabled when
descriptorBinding*UpdateAfterBind is active.
Fixes the following VVL error:
Validation Error: [ VUID-VkDeviceCreateInfo-robustBufferAccess-10247 ]
vkCreateDevice(): robustBufferAccessUpdateAfterBind is false, but both
robustBufferAccess and a descriptorBinding*UpdateAfterBind feature are
enabled.
Fixes: d9fcf5de55a ("turnip: Enable nonuniform descriptor indexing")
Signed-off-by: Valentine Burley <valentine.burley@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36787>
(cherry picked from commit 31f6235126b8d6d1f3c8afce86e61af0d92be2f0)
-
770debce
by Mike Blumenkrantz at 2025-09-17T13:59:15+02:00
zink: add another flag to determine whether linked program compile is done
it's otherwise possible for this to race and hit the draw before
precompile finishes without ever waiting on the fence
I guess this just worked coincidentally before?
cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37197>
(cherry picked from commit 6596bf69c636ccf6c3abee1ef8f2e266c8007718)
-
50f967cd
by Georg Lehmann at 2025-09-17T13:59:15+02:00
mesa: clamp fog scale to -FLT_MAX instead of FLT_MIN
FLT_MIN is the smallest positive float, not the smallest negative float.
Fixes: 35ae5dce39c ("mesa: don't pass Infs to the shader via gl_Fog.scale")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/11412
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37204>
(cherry picked from commit f47e4fee4cd292f0c85f0c61495369e9066b8f6c)
-
4a2273eb
by Caio Oliveira at 2025-09-17T13:59:15+02:00
util: Avoid invalid access in ralloc_print_info()
Check if allocation is large enough to hold the
linear and gc contexts before probing for them.
Fixes: 7b5b1642815 ("util: Add function print information about a ralloc tree")
Acked-by: Eric R. Smith <eric.smith@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37017>
(cherry picked from commit 62815cc91fee45ca11a0d36a6425fd88f975fa90)
-
64fd494f
by Christoph Neuhauser at 2025-09-17T13:59:15+02:00
iris: Increase max_shader_buffer_size to max_buffer_size
This commit increases max_shader_buffer_size to max_buffer_size for Iris.
Signed-off-by: Christoph Neuhauser <christoph.neuhauser@intel.com>
Co-authored-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37182>
(cherry picked from commit 2f8b8649f015720fd1ac08f779accbfb4701c074)
-
2b14506b
by David Rosca at 2025-09-17T13:59:15+02:00
radv/video: Set rate control to default on reset
If we never get rate control info, we would treat it as rate control
disabled and use QP value from slice info which is always 0 for default
rate control.
Cc: mesa-stable
Reviewed-by: Autumn Ashton <misyl@froggi.es>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37156>
(cherry picked from commit a03e055cd344942024b18ef09879c1ece5b785c6)
-
2cd143b0
by Patrick Lerda at 2025-09-17T13:59:15+02:00
r600: fix r600_resource_copy_region behavior for some formats
The function resource_copy_region is expected to have a memcpy()
like behavior. Some formats, like r11g11b10_float, do not preserve
the original raw values. This was the problem. This change
updates r600_resource_copy_region() to use the generic copy path
for those formats.
This change takes into account how 8235d3aa1950 "radeonsi: preserve
NaNs in draw-based resource_copy_region" proceeds.
This change fixes the remaining khr-gl and deqp-gles31 copy_image tests. This
change was tested on palm, barts and cayman. Here are the tests fixed:
khr-gl4[2-6]/texture_view/view_classes: fail pass
khr-gl4[3-6]/copy_image/functional_src_target_texture_2d_array_src_format_rgb9_e5_dst_target_texture_2d_array_dst_format_r11f_g11f_b10f: fail pass
khr-gl4[3-6]/copy_image/functional_src_target_texture_2d_array_src_format_rgb9_e5_dst_target_texture_3d_dst_format_r11f_g11f_b10f: fail pass
khr-gl4[3-6]/copy_image/functional_src_target_texture_2d_array_src_format_rgb9_e5_dst_target_texture_rectangle_dst_format_r11f_g11f_b10f: fail pass
khr-gl4[3-6]/copy_image/functional_src_target_texture_3d_src_format_rgb9_e5_dst_target_texture_2d_array_dst_format_r11f_g11f_b10f: fail pass
khr-gl4[3-6]/copy_image/functional_src_target_texture_3d_src_format_rgb9_e5_dst_target_texture_3d_dst_format_r11f_g11f_b10f: fail pass
khr-gl4[3-6]/copy_image/functional_src_target_texture_3d_src_format_rgb9_e5_dst_target_texture_rectangle_dst_format_r11f_g11f_b10f: fail pass
khr-gl4[3-6]/copy_image/functional_src_target_texture_rectangle_src_format_rgb9_e5_dst_target_texture_2d_array_dst_format_r11f_g11f_b10f: fail pass
khr-gl4[3-6]/copy_image/functional_src_target_texture_rectangle_src_format_rgb9_e5_dst_target_texture_3d_dst_format_r11f_g11f_b10f: fail pass
khr-gl4[3-6]/copy_image/functional_src_target_texture_rectangle_src_format_rgb9_e5_dst_target_texture_rectangle_dst_format_r11f_g11f_b10f: fail pass
deqp-gles31/functional/copy_image/non_compressed/viewclass_16_bits/.*: fail pass
deqp-gles31/functional/copy_image/non_compressed/viewclass_32_bits/.*: fail pass
Cc: mesa-stable
Signed-off-by: Patrick Lerda <patrick9876@free.fr>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37117>
(cherry picked from commit df5e8e1252f2703fd3d14ef08b531f85b508ad67)
-
985bbb27
by Patrick Lerda at 2025-09-17T13:59:15+02:00
r600: update multi_draw_indirect_params drm version requirement
The definitive linux kernel counterpart update is now merged.
The linux kernel commit dc5c742f41c0a1c2b14e4357c752851e015f3bcd
(v6.17-rc1) set the new r600 drm version requirement. This change
updates the drm version accordingly to 2.51.0.
Fixes: 9e1180b33523 ("r600: implement ARB_indirect_parameters")
Signed-off-by: Patrick Lerda <patrick9876@free.fr>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37118>
(cherry picked from commit ef01013b71098cadaef583ba48df79c27f8d92a8)
-
81b4015c
by Lionel Landwerlin at 2025-09-17T13:59:15+02:00
anv: fix partial queries
Partial results should be computed for all types of queries.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: mesa-stable
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36916>
(cherry picked from commit 262baafe27457dfbe3840ed957afb3b2ac076ff7)
-
4d60b13d
by Connor Abbott at 2025-09-17T13:59:15+02:00
tu: Lower ViewIndex to 0 when multiview is disabled
This is an optimization, but it also seems to be required because the HW
sometimes fails to set ViewIndex to 0. This fixes flakes with
dEQP-VK.renderpass2.fragment_density_map.*multiviewport where the VS for
the main renderpass is reused for the copy renderpass afterwards and it
copies ViewIndex to ViewportIndex expecting it to be 0 since multiview
is disabled for the copy renderpass.
Closes: #13534
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37206>
(cherry picked from commit 7527ad001aabb4f2983f6fd8b7f3591e3c191633)
-
f88d6cf7
by Caio Oliveira at 2025-09-17T13:59:15+02:00
brw: Run validation as soon as we have the CFG around
Fixes: affa7567c29 ("intel/brw: Add phases to backend")
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37148>
(cherry picked from commit 4e253184de021f12b077461723dcbe2ff389af4e)
-
30d1a6db
by Caio Oliveira at 2025-09-17T13:59:15+02:00
brw: Fix printing of blocks in disassembly when BRW is available
When disassembling and BRW IR is available (which happens in the
generator), there will be pointers to the BRW's basic block structures
that are used to print the block numbers and predecessor/successors
in the output.
There are two challenges:
- Because DO and FLOW instructions are not real instructions, they are
not emitted in the output but would still cause the output to contain
empty blocks. Previous code accounted for DO but still had problems.
- DO blocks have special physical links that don't make sense when the
DO is not emitted at the end, but they would be shown even if that
block was omitted.
These issues can be seen here (edited to remove non-essential bits)
```
START B0 (2 cycles)
mov(8) g126<1>UD 0x3f800000UD
END B0 ->B1
START B2 <-B1 <-B4 (0 cycles)
END B2 ->B3
START B3 <-B2 (260 cycles)
LABEL1:
mov(8) g1<1>D 0D
cmp.ge.f0.0(8) null<1>D g2<0,1,0>D 10D
sync nop(1) null<0,1,0>UB
send(1) g0UD g1UD nullUD
(+f0.0) break(8) JIP: LABEL0 UIP: LABEL0
END B3 ->B1 ->B5 ->B4
START B4 <-B3 (1000 cycles)
sync nop(1) null<0,1,0>UB
mov(8) g126<1>UD g0<0,1,0>UD
LABEL0:
while(8) JIP: LABEL1
END B4 ->B2
START B5 <-B1 <-B3 (20 cycles)
```
For example:
- Block 1 is missing (a skipped DO block)
- Block 2 is empty (it was a FLOW block)
- Block 3 ends with a link to Block 1 (the special links involving DO
blocks).
Two key changes were made to fix this. First, skip the DO and FLOW
blocks completely. The use_tail ensures that the instruction group is
reused to avoid empty blocks. Second, when printing, the successors and
predecessors, walk through the skipped blocks. And finally, don't print
the special blocks.
With the fix, here's the output. Note the blocks retain their original
BRW IR number.
```
START B0 (2 cycles)
mov(8) g127<1>UD 0x3f800000UD
END B0 ->B3
START B3 <-B0 <-B4 (260 cycles)
LABEL1:
mov(8) g1<1>D 0D
cmp.ge.f0.0(8) null<1>D g2<0,1,0>D 10D
sync nop(1) null<0,1,0>UB
send(1) g0UD g1UD nullUD
(+f0.0) break(8) JIP: LABEL0 UIP: LABEL0
END B3 ->B5 ->B4
START B4 <-B3 (1000 cycles)
sync nop(1) null<0,1,0>UB
mov(8) g127<1>UD g0<0,1,0>UD
LABEL0:
while(8) JIP: LABEL1
END B4 ->B3
START B5 <-B3 (20 cycles)
```
Issue was spotted by Ken.
Fixes: d2c39b17793 ("intel/brw: Always have a (non-DO) block after a DO in the CFG")
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36226>
(cherry picked from commit f37c9c873cc048920758af41e76afc2dc619201c)
-
e22ca94e
by Karol Herbst at 2025-09-17T13:59:15+02:00
nak/qmd: import size_of for rust-1.78 compatability
-
e5c8ed73
by Mike Blumenkrantz at 2025-09-17T13:59:16+02:00
zink: toggle ctx->has_swapchain when flushing clears
this otherwise breaks resolve handling
Fixes: b122c3eaa9e ("zink: unify/fix clear flushing")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13841
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13847
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37234>
(cherry picked from commit 504a522a1ea0dc638373db970a3d9979b92f79d7)
-
88a8e4fe
by Jose Maria Casanova Crespo at 2025-09-17T13:59:16+02:00
v3dv: Fix stencil clear values for only stencil clears
The attachment loadOp for stencil buffers is not stored at
attachment->desc.loadOp. It is stored at attachment->desc.stencilLoadOp
so v3dv was ignoring the clear values for stencil if the depth buffer
wasn't setup with the VK_ATTACHMENT_LOAD_OP_CLEAR.
Fixes: 4d0e497a3e6 ("v3dv: implement support for depth testing")
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37228>
(cherry picked from commit 914b60f0dd412e1c64c74cd98d4c605926892de8)
-
8f81950c
by Rob Clark at 2025-09-17T13:59:16+02:00
freedreno/registers: Re-enable validation for gen_header.py
Commit 84e93daa2656 ("freedreno/registers: allow skipping the
validation") synced a change that made validation optional for
kernel builds, to avoid a lxml dependency for kernel builds.
But this inadvertantly also disabled schema validation on the
mesa side. CI (and meson "test" target) still validates the
xml against the schema, but it is easier if this is also done
as part of the normal build to avoid suprises from Marge.
Fixes: 84e93daa2656 ("freedreno/registers: allow skipping the validation")
Signed-off-by: Rob Clark <rob.clark@oss.qualcomm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37216>
(cherry picked from commit a70279adc26d825d285f84bc8a0150ef67f084e7)
-
462f8c59
by Mike Blumenkrantz at 2025-09-17T13:59:16+02:00
zink: flag pipeline_changed when updating shader modules
this otherwise breaks everything when switching from pipeline -> shobj draw
cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37239>
(cherry picked from commit fe82d156bf32facceb344408a197c4db8e5dc21c)
-
af7f3d1d
by Mike Blumenkrantz at 2025-09-17T13:59:16+02:00
zink: clamp subgroup op return types to required int/uint types
required by spec
cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37239>
(cherry picked from commit a11a02d6f60d3904889938d963a9de715e586b94)
-
f8a88724
by Mike Blumenkrantz at 2025-09-17T13:59:16+02:00
zink: fix edgeflags check on program creation
can't check the shaders before the shaders are set
Fixes: a93685ef557 ("zink: add `has_edgeflags` flag to zink_shader and zink_gfx_program")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37239>
(cherry picked from commit 683e3ea0af5dd0eddc767fba6e90e583137e53b7)
-
113db9ac
by Mike Blumenkrantz at 2025-09-17T13:59:16+02:00
zink: correctly handle batch_id==0 in check_last_finished()
in some places this assert is used to hang check, but here we
can assume that the batch_id is always valid (i.e., not a hang) and
check it accordingly to avoid spurious crashes
cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37239>
(cherry picked from commit 28342d1f11317168f7c0017b0a3c3420c86d8344)
-
91f3af81
by Sviatoslav Peleshko at 2025-09-17T13:59:16+02:00
anv: Always disable Color Blending for unused Render Targets
Commit d2f7b6d5 changed the BLEND_STATE update process so that only
the used render targets will be updated. This mostly works fine, but
in cases when the Dual Source Blending was used previously, we still
must turn it off to avoid the undefined behavior that leads to hangs.
Fixes: d2f7b6d5 ("anv: implement VK_KHR_dynamic_rendering_local_read")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13675
Signed-off-by: Sviatoslav Peleshko <sviatoslav.peleshko@globallogic.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37246>
(cherry picked from commit b148d47c3eea366087094be48253fe3aec0fc46b)
-
c99a6f84
by Mike Blumenkrantz at 2025-09-17T13:59:16+02:00
zink: only set compute module info on dispatch (after compile fence)
this otherwise can attempt to access thread data
cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37139>
(cherry picked from commit 3fecf687841b787aad5319aed01bf80165e1f14d)
-
d9094cf0
by Mike Blumenkrantz at 2025-09-17T13:59:16+02:00
zink: set current compute prog after comparing against current compute prog
not sure how this was never caught before now?
cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37139>
(cherry picked from commit e21438192a7d3ddff99af5f6fd94e682b7a32d0c)
-
95171205
by Mike Blumenkrantz at 2025-09-17T13:59:16+02:00
zink: ensure transient surface is created when doing msaa expand
forgetting this can lead to res->transient being NULL
Fixes: ef3f7989574 ("zink: prune zink_surface down to the imageview and create/fetch on demand")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37257>
(cherry picked from commit 7c1c2f8fce13f73ce3e022ad3092f64cc55cc43d)
-
693b1211
by Ella Stanforth at 2025-09-17T13:59:16+02:00
util/list: Fix next instruction removal usecase for non safe iterators
Introducing this iterator debug information breaks the usecase of removing
elements in the list other than the current element.
Fixes: 372e83b95f9
Reviewed-by: Rob Clark <rob.clark@oss.qualcomm.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37061>
(cherry picked from commit 68632230336d79434a04a93e894a81dff105118b)
-
09043aba
by Mike Blumenkrantz at 2025-09-17T13:59:16+02:00
zink: flag vertex element state for rebind after vstate draws
vstate draws bind their own elements unrelated to the bound
gallium elements, so any draw occurring after a vstate draw must
rebind to ensure the correct ones are bound
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13570
cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37274>
(cherry picked from commit 5fefb9e795dafe195399fac4d4a66c1fcda1b33e)
-
c6aed295
by David Rosca at 2025-09-17T13:59:16+02:00
radeonsi/vcn: Fix calculating QP map region dimensions
It needs to be aligned to block size otherwise it would skip last
row/column on resolutions like 1080p.
Cc: mesa-stable
Reviewed-by: Ruijing Dong <ruijing.dong@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37143>
(cherry picked from commit 8eb84f8854219e95e2795d36aee99418a7bcda79)
-
283734b5
by Dylan Baker at 2025-09-17T13:59:16+02:00
intel/brw: Fix implementaiton of |= operator for enum
The current implementation does nothing, since it has no side effects,
only a return value. By passing `x` as a reference we can mutate the
value before returning.
Fixes: df37c7ca743 ("brw: fix analysis dirtying with pulled constants")
CID: 1665293
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37263>
(cherry picked from commit f18aca868993576822e41175588e9bd683bb81ac)
-
3df08d26
by Karol Herbst at 2025-09-17T13:59:17+02:00
rusticl/mem: fix Image::read for 1Darray images
Some drivers do not return identical strides when mapping 1Darray images
leading to data being written in the wrong place.
Cc: mesa-stable
Tested-by: Rob Clark <rob.clark@oss.qualcomm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37284>
(cherry picked from commit b584b47d0184f4b960962141f604d6a956c948d1)
-
5fb8e4af
by Erik Faye-Lund at 2025-09-17T13:59:17+02:00
aux/pp: fixup sampler-view release
Without this, the prost-processing mechanism crashes badly, whoops.
Fixes: 73da0dcddcb ("gallium: eliminate frontend refcounting from samplerviews")
Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37290>
(cherry picked from commit fc0418a69fd7c5d1b6c426314649e7449477c578)
-
fd1c62be
by Danylo Piliaiev at 2025-09-17T13:59:17+02:00
egl: Bring back util_cpu_trace_init
Partially reverts "perfetto: unify init". EGL lib uses perfetto,
but after the change, util_cpu_trace_init doesn't happen inside
libEGL.so, so when it tries to use perfetto, it crashes.
Perfetto initialization happens inside other Mesa's libs, but
each lib that uses perfetto must initialize it.
Fixes: c7045e3e633 ("perfetto: unify init")
Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37303>
(cherry picked from commit 715b8ba539dd04a6735411d0aae644272cf28ebf)
-
8dd3853f
by Sagar Ghuge at 2025-09-17T13:59:17+02:00
isl: Respect driconf option for EnableSamplerRoutetoLSC
For EnableSamplerRoutetoLSC, we do check driconf option. Buffer state
setup is just missing that option so add check for that too.
Fixes: 7934b70f ("isl/iris/anv: provide drirc toggle intel_sampler_route_to_lsc")
Cc: mesa-stable
Signed-off-by: Sagar Ghuge <sagar.ghuge@intel.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Nanley Chery <nanley.g.chery@intel.com>
Reviewed-by: Caleb Callaway <caleb.callaway@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37190>
(cherry picked from commit 99cd6ffd1ffce18325b7b491b90b2b87c6b27e7f)
-
7337e571
by Mike Blumenkrantz at 2025-09-17T13:59:17+02:00
zink: use implicit offsets for function temp variables in ntv
this is also apparently illegal
Fixes: 0b7611824a1 ("zink: use implicit stride in ntv for temp vars"
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37312>
(cherry picked from commit f419a96d769148474784f49848cbbdf2e5c87722)
-
b0f0a809
by Samuel Pitoiset at 2025-09-17T13:59:17+02:00
radv: fix capture/replay with sampler border color
The border color index must be captured and returned to the app,
otherwise it's broken if samplers aren't replayed in-order.
Cc: mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37291>
(cherry picked from commit a52483d9e7214ded4401edc7e6e9c8bc3e768603)
-
fb70b9ba
by Samuel Pitoiset at 2025-09-17T13:59:17+02:00
spirv: add missing non-uniform access for SSBO atomics
This fixes new VKCTS coverage
dEQP-VK.descriptor_indexing.non_uniform_atomics.
Found this while implementing a new extension.
Cc: mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37295>
(cherry picked from commit 9fbf29fc6d40a19edc76ffa18dae51bcdf35f142)
-
50394e4f
by Christian Meissl at 2025-09-17T13:59:17+02:00
panfrost: take reference from pool used for allocation
Fixes bo leaks caused by wrong ref counting when
using temporary sampler views.
The leak was observed when playing videos using the
DRM_FORMAT_MOD_VENDOR_MTK tiling format
through gstreamer.
Fixes: 0795f3d7e40 ("panfrost: Add a pool to sampler_view")
Signed-off-by: Christian Meissl <meissl.christian@gmail.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37325>
(cherry picked from commit 31a627bdc2a6d66f1a27b182a69d7ca4da6b6ece)
-
8ec6872f
by Mike Blumenkrantz at 2025-09-17T13:59:17+02:00
cso: unbind vertex buffers when unbinding context
cc: mesa-stable
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37276>
(cherry picked from commit 9f764486cb746259d42408c419b299b099d6dc51)
-
838d3c13
by Ian Romanick at 2025-09-17T13:59:17+02:00
iris: Limit max_shader_buffer_size to INT32_MAX
GLSL arrays are sized by signed integers. Other places in the Mesa treat
this value is int32_t. Having the value larger than (uint32_t)INT32_MAX
causes assertion failures in many piglit tests.
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Closes: #13873
Fixes: 2f8b8649f01 ("iris: Increase max_shader_buffer_size to max_buffer_size")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37282>
(cherry picked from commit 1eb4a2f5cd5a5bc557be57b16028cf372baad8b4)
-
bb2c8704
by Corentin Noël at 2025-09-17T13:59:17+02:00
virgl: Skip resource destruction only when there are actually needed references
Suggested by Jan Alexander Steffens in https://gitlab.freedesktop.org/mesa/mesa/-/issues/13844#note_3087461
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13844
Fixes: 6c1c116a0fb66f2d74976260775ec73025c4a697
cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37346>
(cherry picked from commit 63cd410c29f1a052672fda0316810c613ae76571)
-
e93b3052
by Timur Kristóf at 2025-09-17T13:59:17+02:00
radv: Flush L2 before CP DMA copy/fill when CP DMA doesn't use L2
In case the source or destination were previously written
through L2, we need to writeback L2 to avoid the CP DMA accessing
stale data.
However, as the CP DMA doesn't write L2 either, an invalidation
is also needed to make sure other clients don't access stale data
when they read it through L2 after the CP DMA is complete.
Doing an invalidation before the CP DMA operation should take
care of both.
Additionally, radv_src_access_flush also invalidates L2 before
the copied data can be read.
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36820>
(cherry picked from commit c183eb5bc8747e401a38df427429570c4ef074ca)
-
f70ac948
by Qiang Yu at 2025-09-17T13:59:17+02:00
panfrost: fix image plane array copy
Reference the original code, each plane resource have
the same image plane array.
Fixes: 53e5e07c4b7 ("pan: Add the concept of modifier handler")
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Eric R. Smith <eric.smith@collabora.com>
Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37247>
(cherry picked from commit 95579b7af6a38267f8234bbb63580106c6a0eb72)
-
652ce02b
by Jose Maria Casanova Crespo at 2025-09-17T13:59:17+02:00
v3d: Don't enable Early-z with discards when stencil updates are enabled
This fixes font rendering artifacts when Chromium/Firefox use Office365
PowerPoint set into fullscreen presentation mode.
Fixes: 5b951bcdd7f ("v3d: Enable Early-Z with discards when depth updates are disabled")
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37349>
(cherry picked from commit 72d7864c6b587c481d549f56ec12f5727f38edab)
-
9174a024
by Boris Brezillon at 2025-09-17T13:59:17+02:00
panvk/jm: Preload the FB even if we have no draws queued
There seems to be cases where we end up with a depth/stencil preload,
with no draw associated to it, so let's make sure we preload in that
case, otherwise we will write invalid data back.
Fixes: 0bc3502ca36b ("panvk: Implement a custom FB preload logic")
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Christoph Pillmayer <christoph.pillmayer@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37158>
(cherry picked from commit 023972827af13f9dbc91a334fc5b59bb1d5474d7)
-
9eaf7fe7
by Boris Brezillon at 2025-09-17T13:59:18+02:00
panfrost: Fix panfrost_batch_to_fb_info() for stencil-only attachments
Stencil-only attachments should go in the stencil view, not the
depth/stencil one.
Cc: mesa-stable
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Eric R. Smith <eric.smith@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37158>
(cherry picked from commit 2668316917134adc7c2345e323e0590bee134523)
-
fefa020a
by Quentin Schulz at 2025-09-17T13:59:18+02:00
nvk: remove unused relative_dir variable
In commit f35172b6a465 ("meson: make CL args common"), the
fmacro-prefix-map using relative_dir variable was moved from child
meson.build files to the main meson.build file. However, the
relative_dir variable assignment from src/nouveau/vulkan/meson.build
wasn't removed, while src/asahi/libagx's and src/panfrost/libpan's were.
This seems to be an oversight, so let's remove this seemingly now unused
relative_dir meson variable from nvk's meson.build.
This also allows us to remove the use of discouraged global_build_root()
[1] and global_source_root()[2].
[1] https://mesonbuild.com/Reference-manual_builtin_meson.html#mesonglobal_build_root
[2] https://mesonbuild.com/Reference-manual_builtin_meson.html#mesonglobal_source_root
Fixes: f35172b6a465 ("meson: make CL args common")
Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36927>
(cherry picked from commit e1765a2914c0e19a9749177924fc12cd3fd23222)
-
e1ddd68a
by Quentin Schulz at 2025-09-17T13:59:18+02:00
meson: replace global_source_root/global_build_root with project_*
global_source_root() and global_build_root() typically shouldn't be used
in case the current project can be used as a subproject as it'll return
the path to the main project[1][2].
Instead, let's do as suggested by the documentation and use
project_source_root() and project_build_root() instead.
[1] https://mesonbuild.com/Reference-manual_builtin_meson.html#mesonglobal_source_root
[2] https://mesonbuild.com/Reference-manual_builtin_meson.html#mesonglobal_build_root
Suggested-by: Eric Engestrom <eric@igalia.com>
Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36927>
(cherry picked from commit 6afe429b7c935335bd9382842ae5bc51095b550a)
-
3445cb55
by Quentin Schulz at 2025-09-17T13:59:18+02:00
meson: fix libcl assert() reproducibility
The current -fmacro-prefix-map only handles removal of relative paths
but we still need to handle absolute paths.
The following path can be found in libvulkan_panfrost.so when building
with Yocto for example:
/work/build/tmp/work/cortexa76-cortexa55-crypto-poky-linux/mesa/25.2.0/sources/mesa-25.2.0/src/panfrost/libpan/../../util/bitpack_helpers.h
These paths currently seem to only appear in the binary in strings that
start with "Shader assertion fail at " which seems to indicate there are
calls to assert() with absolute paths as well as relative paths (which
are already patched with fmacro-prefix-map).
By stripping the project source and build root directories from the
paths with -fmacro-prefix-map, we get rid of this reproducibility issue
(which incidentally makes the build fail due to buildpaths QA issue).
Signed-off-by: Quentin Schulz <quentin.schulz@cherry.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36927>
(cherry picked from commit b1a28b1165bd3d749064402d08c665dd7770aa9f)
-
d4353d5d
by Patrick Lerda at 2025-09-17T13:59:18+02:00
r600: fix emit_ssbo_atomic_op when ssbo_image_offset is non-zero
The value read was incorrect.
This change was tested on palm, barts and cayman, it fixes all the variants
of the advanced-sso-simple test:
khr-gl4[56]/es_31_compatibility/shader_image_load_store/advanced-sso-simple: fail pass
khr-gles31/core/shader_image_load_store/advanced-sso-simple: fail pass
Cc: mesa-stable
Signed-off-by: Patrick Lerda <patrick9876@free.fr>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37225>
(cherry picked from commit 06da0e7310ce9e73f30091e733f3eb96406abbfb)
-
8e7efd98
by Mike Blumenkrantz at 2025-09-17T13:59:18+02:00
zink: stop unsetting zink_gfx_pipeline::modules on shader unbind
this conflicts with the optimal_key
cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37370>
(cherry picked from commit f7480771c68e7b1590e53166c1e7e0ea8a3d2f0a)
-
222fe87c
by Mike Blumenkrantz at 2025-09-17T13:59:18+02:00
zink: don't use screen ralloc context for screen::pipeline_libs
set_rehash will realloc this memory, which triggers a UAF on screen destroy
cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37370>
(cherry picked from commit dfcdae378acb6b6372581aac7ed36c7b059502cb)
-
c98c42af
by Job Noorman at 2025-09-17T13:59:18+02:00
ir3: don't create merge sets for subreg moves
There are multiple places where RA assumes merge sets are either
all-full or all-half registers. Creating merge sets for subreg moves
mixes full and half registers which may lead to RA failures.
Fix this by not creating merge sets for subreg moves anymore. Instead,
we manually try to allocate a subreg move's src for its dst when
selecting a register during RA, similar to how ALU/SFU instructions try
to reuse their srcs.
Totals:
Instrs: 363174291 -> 363175216 (+0.00%); split: -0.00%, +0.00%
CodeSize: 922975364 -> 922977230 (+0.00%); split: -0.00%, +0.00%
NOPs: 47652421 -> 47652444 (+0.00%); split: -0.00%, +0.00%
MOVs: 15652959 -> 15653065 (+0.00%); split: -0.00%, +0.00%
COVs: 4097203 -> 4097052 (-0.00%); split: -0.01%, +0.00%
(ss): 7806025 -> 7806183 (+0.00%); split: -0.00%, +0.00%
(sy): 3981862 -> 3981855 (-0.00%); split: -0.00%, +0.00%
(ss)-stall: 26612057 -> 26612789 (+0.00%); split: -0.00%, +0.00%
(sy)-stall: 111568786 -> 111568721 (-0.00%); split: -0.00%, +0.00%
STPs: 345796 -> 345792 (-0.00%)
LDPs: 191118 -> 191111 (-0.00%)
Preamble Instrs: 160491915 -> 160492355 (+0.00%); split: -0.00%, +0.00%
Last helper: 116587870 -> 116588273 (+0.00%); split: -0.00%, +0.00%
Cat0: 53288367 -> 53288384 (+0.00%); split: -0.00%, +0.00%
Cat1: 20954383 -> 20954336 (-0.00%); split: -0.00%, +0.00%
Cat2: 155294307 -> 155295252 (+0.00%); split: -0.00%, +0.00%
Cat6: 4623070 -> 4623059 (-0.00%)
Cat7: 9302363 -> 9302384 (+0.00%); split: -0.00%, +0.00%
Totals from 979 (0.07% of 1352016) affected shaders:
Instrs: 1324850 -> 1325775 (+0.07%); split: -0.07%, +0.14%
CodeSize: 2596114 -> 2597980 (+0.07%); split: -0.04%, +0.11%
NOPs: 330197 -> 330220 (+0.01%); split: -0.23%, +0.24%
MOVs: 62592 -> 62698 (+0.17%); split: -0.35%, +0.52%
COVs: 49011 -> 48860 (-0.31%); split: -0.62%, +0.31%
(ss): 35671 -> 35829 (+0.44%); split: -0.28%, +0.73%
(sy): 18936 -> 18929 (-0.04%); split: -0.13%, +0.09%
(ss)-stall: 157929 -> 158661 (+0.46%); split: -0.36%, +0.82%
(sy)-stall: 543371 -> 543306 (-0.01%); split: -0.20%, +0.19%
STPs: 2741 -> 2737 (-0.15%)
LDPs: 3022 -> 3015 (-0.23%)
Preamble Instrs: 322588 -> 323028 (+0.14%); split: -0.01%, +0.14%
Last helper: 298996 -> 299399 (+0.13%); split: -0.05%, +0.19%
Cat0: 361575 -> 361592 (+0.00%); split: -0.21%, +0.22%
Cat1: 111733 -> 111686 (-0.04%); split: -0.45%, +0.41%
Cat2: 487366 -> 488311 (+0.19%); split: -0.04%, +0.23%
Cat6: 21239 -> 21228 (-0.05%)
Cat7: 37170 -> 37191 (+0.06%); split: -0.06%, +0.12%
Signed-off-by: Job Noorman <jnoorman@igalia.com>
Fixes: c757b22c5fa ("ir3: add subreg move optimization")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37368>
(cherry picked from commit 563b0b347a50366f4a75f2638d966e42addda7af)
-
bf21cd87
by Nanley Chery at 2025-09-17T13:59:18+02:00
anv,hasvk: Take trace submission ID out of lock
The Vulkan spec requires that access to the queue parameter be
externally synchronized for vkQueueSubmit(). So, each submit call to a
specific queue will have a unique ID.
Backport-to: 25.2
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37307>
(cherry picked from commit 27167fdcb553c6ab18de522cfcabc63679b833cc)
-
63548eec
by Nanley Chery at 2025-09-17T13:59:18+02:00
anv: Rework locking for sparse binding with TR-TT
When sparse binding functions submit batches, they may modify the
exec_obj_index field of anv_bo structs. This field is used to ensure a
unique list of buffers is sent to the kernel (i915). Add a lock in these
functions to prevent multiple threads from modifying this field during
the batch submission process. To avoid creating a deadlock, also rework
the locking done in anv_queue_submit().
When playing the Monster Hunter Wilds Benchmark on a mesa build which
enables slab allocation of batch buffers (6f7a32ec928), this avoids a
sporadic assert failure:
nsterHunterWilds.exe:
../../src/intel/vulkan/i915/anv_batch_chain.c:489:
setup_execbuf_for_cmd_buffers:
Assertion `execbuf->bos[idx] == first_batch_bo_real' failed.
This issue was seemingly first introduced in 04bfe828db8
("anv/sparse: allow sparse resouces to use TR-TT as its backend")
Backport-to: 25.2
Ref: https://gitlab.freedesktop.org/mesa/mesa/-/issues/12582
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37307>
(cherry picked from commit 7c8e38ac6727c3846be9db6174442fcccbc026ee)
-
a98c46f9
by Samuel Pitoiset at 2025-09-17T13:59:18+02:00
radv/meta: fix saving push constants for depth/stensil resolves on compute
Found by inspection.
Cc: mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37300>
(cherry picked from commit 2207d1e732fac36417c20e2b7f3b9147afa3b086)
-
fa2492c1
by Samuel Pitoiset at 2025-09-17T13:59:18+02:00
radv: set DRLR mapping info from inheritance info when present
These two structs are allowed to be in pNext and they should match
the primary command buffer info.
Found while implementing a new extension.
Cc: mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37278>
(cherry picked from commit 0bc0ead674565087c1af5dee5de7292ddead80a1)
-
04d7a945
by Tapani Pälli at 2025-09-17T13:59:18+02:00
egl: allocate device info lazily only when queried
Cc: mesa-stable
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13746
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Acked-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36998>
(cherry picked from commit 72f2565fc99ea80b7f7ddc4392ef2313edf96692)
-
0d94a27b
by Georg Lehmann at 2025-09-17T13:59:18+02:00
aco/isel: fix output args init stack buffer overflow
BITSET range functions include the end of the range.
Fixes: eb249bb18e2 ("aco: Only fix used variables to registers")
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37361>
(cherry picked from commit d029686e2029aab746559de88b3dff082c7c3dc3)
-
de0e1a9e
by Mary Guillemard at 2025-09-17T13:59:18+02:00
pan/decode: Fix SYNC_SET32 double dots
Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Fixes: 3778df877898 ("pan/decode: Untangle CS disassembling and interpretation")
Reviewed-by: Christoph Pillmayer <christoph.pillmayer@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37396>
(cherry picked from commit 664a83e4136b87647a89782c6b95eb479aefc0e8)
-
a750d144
by Mary Guillemard at 2025-09-17T13:59:19+02:00
panvk: Fix wrong type for sb_mask in CmdSetEvent2
Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Fixes: f8598e599e1f ("panvk: Add VkEvent support to the CSF backend")
Reviewed-by: Christoph Pillmayer <christoph.pillmayer@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37396>
(cherry picked from commit 9dcc7c0d0a7a19f0f39d5d747860bf2cbcdc1070)
-
59b5f7c9
by Rhys Perry at 2025-09-17T13:59:19+02:00
aco: avoid wraparound for smem global loads with both offsets
fossil-db (gfx1201):
Totals from 296 (0.37% of 79839) affected shaders:
Instrs: 382593 -> 380149 (-0.64%)
CodeSize: 1981452 -> 1970988 (-0.53%); split: -0.53%, +0.00%
Latency: 1575286 -> 1574252 (-0.07%)
InvThroughput: 215839 -> 215818 (-0.01%)
SClause: 8679 -> 8677 (-0.02%); split: -0.03%, +0.01%
Copies: 19642 -> 19641 (-0.01%); split: -0.03%, +0.02%
PreSGPRs: 14521 -> 14515 (-0.04%)
SALU: 57097 -> 55718 (-2.42%)
fossil-db (polaris10):
Totals from 30 (0.05% of 62201) affected shaders:
Instrs: 23341 -> 23379 (+0.16%); split: -0.01%, +0.18%
CodeSize: 121316 -> 121516 (+0.16%); split: -0.01%, +0.17%
SGPRs: 2368 -> 2384 (+0.68%)
Latency: 235153 -> 235374 (+0.09%); split: -0.01%, +0.11%
InvThroughput: 92582 -> 92566 (-0.02%)
SClause: 616 -> 619 (+0.49%)
Copies: 2717 -> 2720 (+0.11%)
PreSGPRs: 1204 -> 1213 (+0.75%)
SALU: 3654 -> 3692 (+1.04%); split: -0.08%, +1.12%
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Backport-to: 25.2
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37301>
(cherry picked from commit 6d71521ecd81aaa4479a8d79f901916c80ab1815)
-
0b384915
by Eric Engestrom at 2025-09-17T16:18:36+02:00
docs: add release notes for 25.2.3
-
62f9be9a
by Eric Engestrom at 2025-09-17T16:18:36+02:00
VERSION: bump for 25.2.3
-
bc1d8f96
by Timo Aaltonen at 2025-09-17T17:51:42+03:00
Merge branch 'upstream-unstable' into debian-unstable
-
59d4beac
by Timo Aaltonen at 2025-09-17T17:55:16+03:00
version bump
-
1e64f53f
by Timo Aaltonen at 2025-09-17T18:03:40+03:00
patches: Fix crash on intel xe with 16kB page size. (Closes: #1115467)
-
4222697d
by Timo Aaltonen at 2025-09-18T11:08:24+03:00
releasing package mesa version 25.2.3-1