AlexIndustrial/mesa
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
166,536 Commits 1 Branch 0 Tags
d7f9e2db598b60fbb83e8d05834bc3a23a4ecd07
Commit Graph

11235 Commits

Author SHA1 Message Date
Samuel Pitoiset bd30f3619d radv: implement graphics shaders relocation for a RGP workaround 
RGP requires shaders to be uploaded consecutively inside the same
buffer object. Otherwise, either it makes the driver generating
huge traces (ie. in GiB) or it fails to load traces at all. Hopefully,
this will be improved soon when AMDGPU drivers will have GPL support.

To workaround this, the driver relocates graphics shaders in the same
buffer object when a pipeline is created. Then at draw time, it
overwrites SPI_SHADER_PGM_xxx registers to make sure SQTT can match
between emitted and exported shaders. It's a bit suboptimal because
graphics shaders are uploaded twice but it's the best solution I found.

This will allow to implement GPL caching without breaking capturing
shaders with RGP.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21078>
 2023-02-10 13:42:14 +00:00
Samuel Pitoiset 69bd1c0c40 radv: restore uploading shaders individually instead of consecutively 
The shaders were uploaded consecutively to fit a RGP constraint but
this was more like a workaround. This upload path doesn't work well for
graphics pipeline library and it was the main blocker for GPL caching.

This commit breaks capturing shaders with RGP if the offset between
shaders is too big. Next commit should fix it by using shaders reloc.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21078>
 2023-02-10 13:42:14 +00:00
Georg Lehmann 533d0008c7 aco: remove stale TODOs about v_interp opsel 
These are already handled correctly according to the ISA docs.

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21096>
 2023-02-10 12:01:56 +00:00
Samuel Pitoiset 75ccf7c003 radv/ci: bump the number of runners to 3 for vkcts-navi21-valve 
RADV_PERFTEST=gpl increased execution time, so let's try with a 3d
runner.

dEQP-VK.api.device_init.create_instance_device_intentional_alloc_fail.basic
seems reliably fixed now for some reasons.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21214>
 2023-02-10 11:06:54 +01:00
Samuel Pitoiset 6533709d10 radv/ci: set RADV_PERFTEST=GPL for all VKCTS jobs 
The Vulkan CTS version in Mesa CI is so old that a bunch of tests
are broken, but it's expected.

This runs +283939 tests and the overall VKCTS execution time increased
from ~23 minutes to ~26 minutes (+~13%) on my Threadripper 1950X.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21214>
 2023-02-10 11:06:54 +01:00
Yonggang Luo 09061e2fe9 vulkan: Use static_assert for check HWVULKAN_DISPATCH_MAGIC == ICD_LOADER_MAGIC 
Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Eric Engestrom <eric@igalia.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21210>
 2023-02-10 07:21:31 +00:00
Emma Anholt b01c3553f7 ci/radv: Update navi21 llvm xfails. 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21205>
 2023-02-10 02:53:09 +00:00
Tatsuyuki Ishi 9fdc145ad6 radv: Assert the hardware support rbplus when emitting rbplus state. 
If someone forget to check for rbplus before setting dirty bits, it's going
to cause really mysterious bugs.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21189>
 2023-02-09 08:32:20 +00:00
Tatsuyuki Ishi a36efe3e19 radv: Fix missing rbplus_allowed check for dynamic PS epilogs. 
This created really mysterious bugs on gfx10.

Fixes flickering in a bunch of DXVK games, most visibly Overwatch 2.

Fixes: eb07a11b8f ("radv: add support for compiling PS epilogs on-demand")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/8258
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21189>
 2023-02-09 08:32:20 +00:00
Samuel Pitoiset 46aee750c6 radv: stop skipping the cache for monolithic graphics pipelines with GPL 
Only libraries and optimized (LTO) pipelines are still unsupported,
but there is no reason to skip the cache for monolithic pipelines.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21198>
 2023-02-09 08:11:45 +00:00
Samuel Pitoiset 26896616c1 radv: stop skipping the cache for compute/raytracing pipelines with GPL 
This was a hard solution somewhat.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21198>
 2023-02-09 08:11:45 +00:00
Samuel Pitoiset 53cea404f7 radv: simplify creating a FS epilog from a library 
It's now compiled in radv_graphics_pipeline_compile().

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21140>
 2023-02-09 07:48:03 +00:00
Samuel Pitoiset 06d17fccfb radv: make sure to disable MRT compaction when compiling a PS epilog with GPL 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21140>
 2023-02-09 07:48:03 +00:00
Samuel Pitoiset 16e4dd3e37 radv: fix disabling MRT compaction for on-demand PS epilogs 
Some dynamic states require to compile PS epilogs on-demand. In this
case, MRT compaction should be disabled because we don't know the CB
state when compiling the fragment shader.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21140>
 2023-02-09 07:48:03 +00:00
Samuel Pitoiset ea5c893289 radv: regroup PS epilog info when generating the graphics pipeline key 
No logical change.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21140>
 2023-02-09 07:48:03 +00:00
Samuel Pitoiset 84d006ef11 radv: simplify determining when the fragment shader needs an epilog 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21140>
 2023-02-09 07:48:03 +00:00
Samuel Pitoiset b41c6c3a50 radv: cleanup graphics pipeline library flags uses 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21140>
 2023-02-09 07:48:03 +00:00
Samuel Pitoiset 7976316f3e radv: fix skipping graphics pipeline compilation when the FS is NULL 
Fixes: 3eb97b9d33 ("radv: skip compilation when possible with GPL fast-linking")
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21192>
 2023-02-09 07:27:53 +00:00
Rhys Perry b4383821e7 aco: don't modify exec in p_interp_gfx11 
The RDNA3 ISA docs say that lds_param_load write the entire quad
regardless of exec, so this isn't needed.

fossil-db (gfx1100):
Totals from 5291 (3.93% of 134574) affected shaders:
Instrs: 4891396 -> 4789628 (-2.08%)
CodeSize: 25519032 -> 25111960 (-1.60%)
Latency: 36122982 -> 36074300 (-0.13%); split: -0.14%, +0.00%
InvThroughput: 4162436 -> 4161424 (-0.02%); split: -0.02%, +0.00%
Copies: 263862 -> 263838 (-0.01%)
PreSGPRs: 225012 -> 224179 (-0.37%)

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21171>
 2023-02-08 19:35:54 +00:00
Georg Lehmann 6e4598f7b9 aco: support omod/imod for v_fmac_f16 
Only matters for post-RA DPP16.

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21174>
 2023-02-08 18:52:28 +00:00
Georg Lehmann 2deda5c0be aco: don't list imod/omod support v_fmaak_f32/v_fmamk_f32 
We can never use them anyway because these opcodes don't support VOP3/DPP16/SDWA

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21174>
 2023-02-08 18:52:28 +00:00
Georg Lehmann 4c9ac73064 aco: allow output modifiers for ldexp_f16 
It also supports imod for the first operand, but we cannot express that at
moment.

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21174>
 2023-02-08 18:52:28 +00:00
Georg Lehmann b63aa2bb8e aco: don't allow output modifiers for v_cvt_pkrtz_f16_f32 
Cc: mesa-stable
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21174>
 2023-02-08 18:52:28 +00:00
Georg Lehmann 5038a049f1 aco: add mov/cndmask opcodes to does_fp_op_flush_denorms 
For completeness sake also add v_mov_b32, even if we don't use imod for it
because it's only supported since gfx10.

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21170>
 2023-02-08 13:07:46 +00:00
Georg Lehmann c8adf16278 aco: fix imod/omod for gfx11 VOP3 opcodes 
Fixes: d8d99c3c4f ("aco: add GFX11 opcode numbers")
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21170>
 2023-02-08 13:07:46 +00:00
Tatsuyuki Ishi c6ebd2de71 radv: Fix noop FS not getting constructed for GPL pipelines. 
The condition was inverted, causing compilation to be actually skipped when
a noop FS is used and straight emitting the pipeline from the default
initialized struct.

Fixes: 3eb97b9d33 ("radv: skip compilation when possible with GPL fast-linking")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21166>
 2023-02-08 10:52:54 +00:00
Daniel Schürmann e5600d5257 radv: CSE ray_launch_{size|id} 
Stats for Q2RTX
Totals from 7 (0.01% of 134913) affected shaders: (GFX10.3)

VGPRs: 736 -> 704 (-4.35%)
CodeSize: 204424 -> 204400 (-0.01%); split: -0.03%, +0.02%
MaxWaves: 67 -> 69 (+2.99%)
Instrs: 37540 -> 37549 (+0.02%); split: -0.02%, +0.04%
Latency: 973556 -> 973267 (-0.03%)
InvThroughput: 209068 -> 200902 (-3.91%)
VClause: 920 -> 921 (+0.11%); split: -0.11%, +0.22%
SClause: 1045 -> 1043 (-0.19%)
Copies: 4853 -> 4865 (+0.25%); split: -0.08%, +0.33%
Branches: 1571 -> 1578 (+0.45%)
PreSGPRs: 421 -> 407 (-3.33%)
PreVGPRs: 638 -> 632 (-0.94%)
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21156>
 2023-02-08 10:10:27 +00:00
Timur Kristóf 8ebb34ee57 radv: Refactor radv_emit_ngg_culling_state so it's based on dirty flags. 
This significantly lowers the CPU overhead of this function.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20980>
 2023-02-08 09:28:20 +00:00
Timur Kristóf 22947ef0cc radv: Remove NGG culling skip from command buffer. 
This code used to runtime-disable NGG culling for small draw calls.
However, this had too much CPU overhead, let's remove it.
It will be solved by the shaders instead.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20980>
 2023-02-08 09:28:19 +00:00
Timur Kristóf 88a4e17018 radv: Use shader code to skip NGG culling in small workgroups. 
Extend RADV lowering of the load_cull_any_enabled_amd intrinsic to
take into account the number of primitives in the current workgroup.

Workgroups that have less than 16 triangles are considered "small"
and will disable shader culling. Note that LLPC does the same,
but it checks the number of vertices not primitives.

The primary intention of this change is to eliminate the need to
check the draw size in radv_cmd_buffer, but this is actually
beneficial to larger draw calls too, specifically this may improve
the performance of the last workgroup of larger draws too.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20980>
 2023-02-08 09:28:19 +00:00
Timur Kristóf def53a085c radv: Move checking primitive topology to radv_get_ngg_culling_settings. 
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20980>
 2023-02-08 09:28:19 +00:00
Timur Kristóf 8457dfec47 radv: Don't change LDS_SIZE for NGG culling shaders. 
There was no measurable perf benefit from this optimization,
and it made the code messy and difficult to refactor.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20980>
 2023-02-08 09:28:19 +00:00
Rhys Perry fad1f716dd aco: fix out-of-bounds access when moving s_mem(real)time across SMEM 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/8224
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21138>
 2023-02-07 14:50:43 +00:00
Qiang Yu 4888dd7391 radv: use amd common force_vrs option 
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21010>
 2023-02-07 08:09:39 +00:00
Qiang Yu 1e3198c766 radv: move radv_consider_force_vrs above radv_fill_shader_info 
It will be used by radv_fill_shader_info, no function change.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21010>
 2023-02-07 08:09:39 +00:00
Dave Airlie 3e2c768aa8 radv/vcn: enable dynamic dpb tier 2 for h264/h265 on navi21+ 
navi21 can do separate image support, using tier 2 DPB messages.

This enables support for doing that in the vulkan video driver.

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20388>
 2023-02-07 12:48:28 +10:00
Dave Airlie 6c3c242361 radv/video: add h265 decode UVD support 
Add support for h265 decode on older UVD GPUs

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20388>
 2023-02-07 12:48:26 +10:00
Dave Airlie db62c38091 radv: add vcn h265 decode. 
This adds support for HEVC decode in VCN

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20388>
 2023-02-07 12:48:24 +10:00
Dave Airlie 8a29291dbe radv/video: add h264 support for uvd 
This adds support for the older UVD h264 decoder.

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20388>
 2023-02-07 12:48:22 +10:00
Dave Airlie 1693c03a39 radv/video: add initial h264 decoder for VCN 
This adds support for H264 decode on VCN hardware.

It uses the full DPB method, and relies on the application
to allocate an arrayed texture for the DPB to be stored into.

RADV_PERFTEST=video_decode is required to enable this.

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20388>
 2023-02-07 12:47:53 +10:00
Dave Airlie 9477f117f4 radv/video: add initial frameworking. 
This just adds the basic commands and objects, and hooks up some
of the queues and extensions.

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20388>
 2023-02-07 12:47:50 +10:00
Dave Airlie 3253340916 radv: add video decoder register setup. 
This just assigns the correct registers depending on the gpu family.

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20388>
 2023-02-07 12:47:48 +10:00
Dave Airlie 85eead4198 radv: adding video decode queue support 
This adds the video queue interactions to radv and builds
on the winsys code previously added.

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20388>
 2023-02-07 12:47:46 +10:00
Dave Airlie 30b6e9797d ac: add name to codec info struct 
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20388>
 2023-02-07 12:47:43 +10:00
Dave Airlie ee7837aae2 radv: add new upload alloc aligned api 
The video buffers need to have objects aligned at certain ranges,
this enhances the uploader to allow an alignment to be specified.

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20388>
 2023-02-07 12:47:39 +10:00
Dave Airlie 855ad612e1 radv: remove the status query mark it unsupported. 
The current firmware can't support the status query requirements.

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20388>
 2023-02-07 12:47:25 +10:00
Alejandro Piñeiro a12a71e6c0 radv: use shader_info->var_copies_lowered 
Instead of passing allow_copies as a parameter for radv_optimize_nir
(so manually doing that tracking).

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19338>
 2023-02-06 22:11:34 +00:00
Konstantin Seurer 569517d7ad radv: Use common ycbcr conversion lowering 
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20731>
 2023-02-06 18:36:29 +00:00
Konstantin Seurer dae893cb96 radv: Remove radv_indirect_unaligned_dispatch 
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21071>
 2023-02-06 17:50:53 +00:00
Konstantin Seurer b9d7306edc radv: Use an ordered dispatch for BVH encoding 
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21071>
 2023-02-06 17:50:53 +00:00