AlexIndustrial/mesa
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
144,472 Commits 1 Branch 0 Tags
bb956464cbc3d5012527ea3d16dbd3f2c96cdae1
Commit Graph

812 Commits

Author SHA1 Message Date
Samuel Pitoiset 1402c17e4f radv: optimize VRS when no depth stencil attachment is bound 
This is allowed by the Vulkan spec and we have to handle this situation
internally. We used to create and bind a 4096x4096 image to copy the
VRS rates but this wasted too much VRAM (~33MiB). Now, the driver only
allocates a HTILE buffer (~1MiB) and bind it to the framebuffer.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12243>
 2021-09-02 19:39:04 +00:00
Samuel Pitoiset ab635b024b radv: pass the HTILE buffer to radv_copy_vrs_htile() 
Will be used to use a global HTILE buffer without an image.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12243>
 2021-09-02 19:39:04 +00:00
Samuel Pitoiset ad60354a92 radv: optimize copying VRS rates to the global HTILE buffer 
By skipping the read operation which is unnecessary.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12243>
 2021-09-02 19:39:04 +00:00
Samuel Pitoiset df688e6941 radv: do not load/store the clear value for comp-to-single images 
Images that are fast cleared with the comp-to-single mode clears DCC
to 0x10 which tells the hardware to get the clear value from the
main surface instead of the reg.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12565>
 2021-08-30 07:18:19 +00:00
Samuel Pitoiset 80e5e059fa radv: fix pre-computing viewport xform when setting new viewports 
viewportCount is the number of viewports in pViewports while
firstViewport is the index.

Fixes new CTS dEQP-VK.draw.depth_clamp.*_clamp_four_viewports

Fixes: a2ef92d7a5 ("radv: pre-calculate viewport transforms")
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12353>
 2021-08-17 08:20:02 +02:00
Samuel Pitoiset 7ae3881a4b radv: flush caches before performing separate depth/stencil aspect init 
It's a RMW operation, also note that DB doesn't use L2 on GFX6-8.

Fixes test_clear_depth_stencil_view() and test_discard_resource() tests
from vkd3d-proton.

Cc: 21.2 mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12223>
 2021-08-09 16:35:41 +00:00
Bas Nieuwenhuizen 02b6015945 radv: Allocate space for inline push constants. 
In the compute dispatch path we do not allocate a huge amount
of space to cover everything so the individual functions have to
allocate. This was missing here, causing a hang in Cyberpunk when
accessing the system menu at some locations with thread tracing
enabled.

Fixes: bd1186572f ("radv: add support for push constants inlining when possible")
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12271>
 2021-08-09 14:26:21 +00:00
Samuel Pitoiset 1db36422b9 radv: fix initializing the DS clear metadata value for separate aspects 
We shouldn't overwrite the clear value of the other aspect (in case
separate depth/stencil layouts are used).

Found by inspection.

Cc: 21.2 mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12222>
 2021-08-09 10:41:39 +00:00
Timur Kristóf 74181ffcc5 radv: Write RSRC2_GS for NGGC when pipeline is dirty but not emitted. 
The radv_emit_ngg_culling_state function won't write the
SPI_SHADER_PGM_RSRC2_GS register when it knows in advance that
radv_emit_graphics_pipeline will overwrite it anyway.

However, there is an unhandled case:

radv_emit_graphics_pipeline will not emit anything (including this
register) when the pipeline is already emitted. Hence, improve
the check in radv_emit_ngg_culling_state to consider this.

Fixes: 9a95f5487f
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12237>
 2021-08-06 12:18:37 +02:00
Timur Kristóf 61eb3debcf radv: Don't toggle PC oversubscription for NGG culling. 
We are going to add this directly to the pipeline.
If a pipeline has such a shader, NGG culling is turned on
most of the time, so it's not useful to toggle this setting.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11905>
 2021-08-04 15:37:05 +00:00
Samuel Pitoiset 87b0962fef radv: do not use radeon_set_context_reg_seq() for only one register 
radeon_set_context_reg() is a shortcut for that.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12099>
 2021-07-30 17:05:38 +02:00
Samuel Pitoiset f693b94d2f radv: reduce number of emitted DWORDS for contiguous context registers 
radeon_set_context_reg() emits 3 DWORDS, while
radeon_set_context_reg_seq() + 2 values only needs 4 DWORDS.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12099>
 2021-07-30 17:05:35 +02:00
Timur Kristóf 6e17931d21 radv: Use pre-computed viewport transform for NGG culling state. 
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11839>
 2021-07-16 15:22:46 +00:00
Mike Blumenkrantz c9a478f1cd radv: remove unused variable from radv_emit_viewport 
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11839>
 2021-07-16 15:22:46 +00:00
Mike Blumenkrantz a2ef92d7a5 radv: pre-calculate viewport transforms 
this requires more storage in the viewport struct, but it avoids
the need to repeatedly calculate the same transform if e.g., a meta
operation occurs, which can save about 5% cpu in some cases

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11839>
 2021-07-16 15:22:46 +00:00
Mike Blumenkrantz 1e13cb1965 radv: merge si_write_viewport into radv_emit_viewport 
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11839>
 2021-07-16 15:22:46 +00:00
Timur Kristóf d07e5bde75 radv: Remove num_viewports from radv_skip_ngg_culling. 
NGG culling is not compiled into shaders that can use multiple
viewports, so it's not necessary to check it here.

Fixes: 9a95f5487f
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11910>
 2021-07-16 13:00:36 +00:00
Samuel Pitoiset 7a1cc56e40 radv: fix bounds checking for zero vertex stride on GFX6-7 
GFX6 and GFX10+ have similar logic.

This fixes test_zero_vertex_stride from vkd3d-proton on
Pitcairn (GFX6) and on Bonaire (GFX7).

Cc: 21.2 mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11904>
 2021-07-16 08:30:07 +02:00
Samuel Pitoiset 1ea156f44c radv: only init the TC-compat ZRANGE metadata for the depth aspect 
With separate depth/stencil layouts, if the depth aspect is first
initialized and then cleared, the ZRANGE_PRECISION metadata might
be different than 0. Initializing it again for the stencil aspect
will overwrite the value.

Fixes rendering glitches with Scarlet Nexus on GFX8-9.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5052
Cc: 21.1 21.2 mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11883>
 2021-07-15 07:21:50 +00:00
Timur Kristóf 9a95f5487f radv: New shader args for NGG culling settings and viewport. 
Add new shader arguments in RADV for:
- NGG culling settings
- Viewport transform

These will be used by NGG culling shaders.

Additionally, some tweaks are made to some config registers
in order to make culling shaders more efficient.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10525>
 2021-07-13 23:56:33 +00:00
Samuel Pitoiset 29f264f258 ac,radv: implement the cs_regalloc_hang HW bug workaround 
Might fix spurious failures on GFX6 and some GFX7 chips.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11675>
 2021-07-09 13:37:37 +00:00
Mike Blumenkrantz a4713ddb56 radv: implement VK_EXT_multi_draw 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11532>
 2021-07-06 13:16:56 +00:00
Mike Blumenkrantz ba30adce07 radv: add a gfx10 bug workaround for NOT_EOP 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11532>
 2021-07-06 13:16:56 +00:00
Mike Blumenkrantz d9132fcb50 radv: split indexed draw cases based on whether drawid is used 
fewer conditionals in the draw loop = less overhead

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11532>
 2021-07-06 13:16:56 +00:00
Mike Blumenkrantz 5aaf45f5e1 radv: determine if hardware can emit NOT_EOP before emitting 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11532>
 2021-07-06 13:16:56 +00:00
Mike Blumenkrantz d28329ed02 radv: emit drawid for multidraws 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11532>
 2021-07-06 13:16:56 +00:00
Mike Blumenkrantz 00206b1664 radv: emit NOT_EOP for multi indexed draws 
this is never emitted currently because drawCount is always 1

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11532>
 2021-07-06 13:16:56 +00:00
Mike Blumenkrantz 72e75092a9 radv: use multidraw iteration for direct draws 
there's still no arrays of draw info being passed, but this takes care of
all the work for it

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11532>
 2021-07-06 13:16:56 +00:00
Rhys Perry 307a4bf4ab radv: use null vertex descriptor if num_records=0 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11642>
 2021-07-06 13:00:35 +00:00
Rhys Perry a2ac660eb7 radv: adjust num_records when offset>stride 
If an attribute's offset is larger than the stride, the compiler will
increase the vertex index and use offset%stride instead as the offset.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5007
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11642>
 2021-07-06 13:00:35 +00:00
Rhys Perry e4fbb200fc radv: don't ever convert num_records to bytes if it's zero 
If num_records==0, this conversion doesn't work. We should just keep it at
zero.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Gitlab: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5007
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11642>
 2021-07-06 13:00:35 +00:00
Bas Nieuwenhuizen f9cc94af7b radv/winsys: Add support for a fixed VA address for replay. 
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10570>
 2021-07-05 17:27:51 +00:00
Bas Nieuwenhuizen 8025b4120f radv/winsys: Return vulkan errors for buffer creation. 
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10570>
 2021-07-05 17:27:51 +00:00
Samuel Pitoiset 9a95aba377 radv: implement VK_EXT_color_write_enable 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11530>
 2021-07-02 10:02:02 +00:00
Mike Blumenkrantz c0397a42ce radv: pre-calc vertex buffer descriptor size on pipeline object 
util_bitcount has a nonzero cost, and calling it like this in a hotpath
generates unnecessary overhead

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11525>
 2021-06-30 13:42:42 +00:00
Samuel Pitoiset 8efa897168 radv: implement dynamic logic op 
This is part of VK_EXT_extended_dynamic_state2.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10880>
 2021-06-22 15:32:31 +00:00
Samuel Pitoiset 977355c6e5 radv: fix dynamic culling and depth/stencil related dynamic states 
To avoid overwriting previous dynamic state with default state from
the pipeline.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4926
Cc: 21.1 mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11375>
 2021-06-18 16:27:57 +00:00
Mike Blumenkrantz 651c6b16ff radv: move pipe_misaligned and l2_coherent image checks to flags set on init 
this should save 4-5% cpu in some cases

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11462>
 2021-06-18 16:02:26 +00:00
Bas Nieuwenhuizen f7c622307d radv: Don't skip barriers that only change queues. 
We depend on the queue mask for some decisions ...

CC: mesa-stable
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10802>
 2021-06-14 11:20:59 +00:00
Samuel Pitoiset a234840e60 radv: do not launch an IB2 for secondary cmdbuf with INDIRECT_MULTI on GFX7 
It's illegal to emit DRAW_{INDEX}_INDIRECT_MULTI from an IB2 on GFX7.

PAL applies this workaround for indirect dispatches and also on
GFX8-9 but it doesn't seem needed.

This fixes various GPU hangs on Bonaire (GFX7).

Cc: 21.1 mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11214>
 2021-06-10 08:31:08 +02:00
Mike Blumenkrantz 79742d41c0 radv: declare index_va in a single call for indexed draw packet emit 
this is an extreme hotpath, so having a single calculation in a const
variable is slightly better for compiler microoptimizing

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11124>
 2021-06-02 13:31:43 +00:00
Rhys Perry 0f8fef1261 radv: make attrib_end variable in radv_flush_vertex_descriptors 32-bit 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Fixes: 1e9dc0474e ("radv: make radv_pipeline::attrib_ends 32bit")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11089>
 2021-05-31 18:28:12 +00:00
Samuel Pitoiset 69ae02151d radv: remove DFSM 
DFSM has never been enabled by default because it was slower.
RadeonSI is also dropping support for this because they discovered
that's actually not efficient in practice.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10968>
 2021-05-26 17:22:14 +00:00
Samuel Pitoiset 9984ebf173 radv: use radv_dcc_enabled() for the FB mip flush workaround 
This has no effects because radv_image_has_CB_metadata() still
accounts for DCC which is incorrect. This should be changed.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10667>
 2021-05-26 06:59:35 +00:00
Samuel Pitoiset ff38e3aadd radv: only mark DCC as compressed when drawing if layout allows it 
Just having DCC enabled on the base level doesn't mean we are
using compressed rendering.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10667>
 2021-05-26 06:59:35 +00:00
Samuel Pitoiset bdb9634151 radv: pass an image range to radv_layout_dcc_compressed() 
With DCC and mipmaps, some mips can't be compressed and it makes
sense to check this here.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10667>
 2021-05-26 06:59:35 +00:00
Samuel Pitoiset a00be79d80 radv: remove small overhead of radv_pipeline_has_ngg() 
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4784
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10908>
 2021-05-24 08:50:47 +00:00
Samuel Pitoiset 726cb2d6f6 ac: ac_gpu_info::has_vgt_flush_ngg_legacy_bug 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10911>
 2021-05-21 19:46:56 +00:00
Timur Kristóf 2bb1bf78f0 radv/cmd_buffer: Fix warning by initializing instance count. 
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10806>
 2021-05-20 17:11:22 +00:00
Samuel Pitoiset b1171e5ffe Revert "radv: Do not access set layout during vkCmdBindDescriptorSets." 
This was a temporary workaround for a Baldur's Gate 3 bug which
should now be fixed.

This reverts commit eb104e949e.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/3701
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10839>
 2021-05-20 13:38:33 +00:00