AlexIndustrial/mesa
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
158,132 Commits 1 Branch 0 Tags
5076bed2fbebb995eca9c17a2b14199b7a010cfd
Commit Graph

9779 Commits

Author SHA1 Message Date
Samuel Pitoiset 5076bed2fb radv: slightly change the color write enable mask 
For switching to the common graphics pipeline state.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18015>
 2022-08-16 08:41:36 +02:00
Samuel Pitoiset ccb3730200 radv: do not translate primitive topology during gfx info initialization 
For switching to the common graphics pipeline state.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18015>
 2022-08-16 08:41:33 +02:00
Samuel Pitoiset e34440a1bf radv: do not translate blend op/factor during gfx info initialization 
For switching to the common graphics pipeline state.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18015>
 2022-08-16 08:41:30 +02:00
Samuel Pitoiset 0f2ca61dcc radv: do not translate logic op during gfx info initialization 
For switching to the common graphics pipeline state.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18015>
 2022-08-16 08:41:27 +02:00
Samuel Pitoiset 1eb18154da radv: do not translate polygon mode during gfx info initialization 
For switching to the common graphics pipeline state.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18015>
 2022-08-16 08:41:24 +02:00
Samuel Pitoiset faaf8422f3 radv: simplify the depth clip enable/disable logic 
For switching to the common graphics pipeline state.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18015>
 2022-08-16 08:41:20 +02:00
Samuel Pitoiset e70baccf38 radv: store the provoking vertex mode in the graphics pipeline info 
For switching to the common graphics pipeline state.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18015>
 2022-08-16 08:41:18 +02:00
Georg Lehmann 7b9d3ebe42 aco: Use v_cmpx pre GFX10. 
Foz-DB Vega10:
Totals from 29508 (21.85% of 135041) affected shaders:
CodeSize: 184345656 -> 184345820 (+0.00%)
Instrs: 35906154 -> 35906195 (+0.00%)
Latency: 581696114 -> 581530021 (-0.03%); split: -0.03%, +0.00%
InvThroughput: 245625572 -> 245561351 (-0.03%); split: -0.03%, +0.00%
Copies: 3134925 -> 3278672 (+4.59%)

Signed-off-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18049>
 2022-08-15 13:25:38 +00:00
Georg Lehmann 393e577435 aco/assembler: Fix v_cmpx pre GFX10. 
The second destination is implicit exec.

Signed-off-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18049>
 2022-08-15 13:25:38 +00:00
Georg Lehmann 410eff4d2f aco: Fix optimizing branching sequence with s_and_saveexec. 
This optimization was broken for two reasons:
- s_and_saveexec has two operands, the copy value and exec
- s_and_saveexec has an exec read, so exec_write_used will always be true
  before we find branch_exec_val_idx

Foz-DB Navi21:
Totals from 31453 (23.31% of 134913) affected shaders:
CodeSize: 204831260 -> 204831156 (-0.00%)
Instrs: 38157117 -> 38157091 (-0.00%)
Latency: 533708882 -> 531211721 (-0.47%); split: -0.47%, +0.00%
InvThroughput: 107088408 -> 106719188 (-0.34%); split: -0.35%, +0.00%
Copies: 2326179 -> 2502490 (+7.58%)

Signed-off-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18049>
 2022-08-15 13:25:38 +00:00
Georg Lehmann 5ffc73896f aco/assembler: Fix v_cmpx with SDWA. 
We need to use the implicit destination.

Fixes: baab6f18c9 ("aco: Optimize branching sequence during SSA elimination.")
Signed-off-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18049>
 2022-08-15 13:25:38 +00:00
Georg Lehmann 9e5f311efe aco: Check that we don't override exec_val operands during branching sequence optimization. 
Fixes: baab6f18c9 ("aco: Optimize branching sequence during SSA elimination.")
Signed-off-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18049>
 2022-08-15 13:25:38 +00:00
Timur Kristóf baf314e2c0 aco: Check for instructions that inhibit the branching sequence optimization. 
Fixes: baab6f18c9 ("aco: Optimize branching sequence during SSA elimination.")
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18049>
 2022-08-15 13:25:38 +00:00
Timur Kristóf d88b2e4ab5 aco: Fix invalidated reference in branching sequence optimization. 
Inserting in the instructions vector may invalidate the exec_val reference,
so do that last.

Fixes: baab6f18c9 ("aco: Optimize branching sequence during SSA elimination.")
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18049>
 2022-08-15 13:25:38 +00:00
Timur Kristóf 5ead973824 aco: Add faster code path to store_lds for consecutive write mask. 
This makes it more likely to hit the fast path for count == 1
in the split_store_data function.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17923>
 2022-08-14 15:09:07 +00:00
Konstantin Seurer 2d39227a90 radv: Fix stack size calculation with stage ids 
In create_rt_shader, we were setting group_idx to the stage index before.

Fixes the following tests:

dEQP-VK.ray_query.builtin.instancecustomindex.miss.aabbs
dEQP-VK.ray_query.builtin.objectrayorigin.miss.triangles

Fixes: c39ccce ("radv/rt: use stage ID as handle for general and closestHit shaders")
Signed-off-by: Konstantin Seurer <konstantin.seurer@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17936>
 2022-08-11 17:59:47 +00:00
Samuel Pitoiset 58bf84044b radv: copy the sample locations only if the state isn't dynamic 
If user sample locations is enabled and the state is dynamic we don't
have to copy them to the dynamic state struct.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18007>
 2022-08-11 14:53:32 +00:00
Samuel Pitoiset 0ca33a6c0d radv: copy line stipple factor/pattern only if the state isn't dynamic 
If line stipple is enabled and the state is dynamic we don't have to
copy factor/pattern to the dynamic state struct.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18007>
 2022-08-11 14:53:32 +00:00
Samuel Pitoiset 1d92ab921b radv: remove unecessary radv_pipeline::uses_dynamic_stride 
We can just rely on the dynamic states.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18007>
 2022-08-11 14:53:32 +00:00
Mike Blumenkrantz bc7edc1c7a radv: force inlining on another draw function 
Reviewed-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/17993>
 2022-08-11 12:27:23 +00:00
Mike Blumenkrantz 64395f0d85 radv: remove stupid conditional from draw path 
compilers are now provably smart enough to handle this

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17993>
 2022-08-11 12:27:23 +00:00
Mike Blumenkrantz 632e1b66f5 radv: fix return type for meta resolve shaders 
this should match the image type

cc: mesa-stable

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17874>
 2022-08-10 19:44:58 +00:00
Samuel Pitoiset 524f173c03 radv: explain why radv_invariant_geom is still needed for SotTR DX 
It's disabled by DXVK and vkd3d because it breaks NV.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Martin Roukala <martin.roukala@mupuf.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17951>
 2022-08-10 19:21:12 +00:00
Samuel Pitoiset 95232bf261 radv: remove radv_invariant_geom workarounds for DX games 
DXVK (v1.8+, Feb 19, 2021) enabled invariant position with:
    "[d3d11] Enable d3d11.invariantPosition by default"
    (8de9dc93781b61a902a2dc15cc4192514a821237)

vkd3d-proton (v2.5+, Oct 18, 2021) enabled invariant position with:
    "vkd3d-shader: Add INVARIANT_POSITION quirk."
    (32c5abf4963e07b3d7cbf1c95a9249dce16473d7)

It's recommanded to use Proton 7 or newer to ensure DXVK/vkd3d-proton
contain the invariant position fixes. Otherwise this change might
introduce regressions with older Proton versions.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Martin Roukala <martin.roukala@mupuf.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17951>
 2022-08-10 19:21:12 +00:00
Samuel Pitoiset bea050345e radv: do not set registers for PSO states that are dynamic 
Dynamic states are emitted from the cmdbuf, setting them from the
pipeline has no effects because they should be ignored anyways.

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/17677>
 2022-08-10 17:26:26 +00:00
Samuel Pitoiset 42803c70ef radv: simplify dynamic states uses when determining NGG culling settings 
Looking at the dynamic states is enough.

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/17677>
 2022-08-10 17:26:26 +00:00
Samuel Pitoiset 2a96521bb4 radv: remove one useless check about line width and dynamic state 
It's now checked when the rasterization state is initialized.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17677>
 2022-08-10 17:26:26 +00:00
Samuel Pitoiset 101f9dd8cf radv: properly ignore PSO states that are dynamic 
They should be ignored because it's allowed to have garbage in there.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17677>
 2022-08-10 17:26:26 +00:00
Pierre-Eric Pelloux-Prayer 70891edd97 nir: add a nir_opt_if_options enum 
And don't enable nir_opt_if_optimize_phi_true_false on radeonsi with
LLVM 14 because it crashes Blender.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/6976
Cc: mesa-stable
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Acked-by: Emma Anholt <emma@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17949>
 2022-08-10 12:55:39 +00:00
Rhys Perry bd40e1b012 aco: fix hash statistic 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Fixes: 897561b7b9 ("aco: add aco_postprocess_shader() helper")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17954>
 2022-08-10 11:56:29 +00:00
Friedrich Vock 82fd379d9e amd/common: move ac_memory_ops_per_clock into ac_gpu_info.h 
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Tatsuyuki Ishi <ishitatsuyuki@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17973>
 2022-08-10 10:58:54 +00:00
Samuel Pitoiset 37dfa4e3f3 radv: fix cleaning the meta query state if an error occured 
It's already correctly cleaned in radv_device_init_meta().

This fixes a recent regression with
dEQP-VK.api.device_init.create_instance_device_intentional_alloc_fail.

Fixes: 1a95d43e55 ("radv: Simplify the meta init fail path")
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Konstantin Seurer <konstantin.seurer@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17952>
 2022-08-10 07:21:46 +00:00
Samuel Pitoiset 5b5d5554f6 radv: respect the render area for depth/stencil resolves 
Subpass resolves don't necessarily start from 0,0.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17950>
 2022-08-10 06:17:06 +00:00
Samuel Pitoiset e95531e101 radv: fix gathering XFB info if there is dead outputs 
The driver should still gather XFB info even if all XFB outputs are
dead, otherwise the pipeline can't find the streamout shader.

RADV should use vk_spirv_to_nir() at some point to reduce code
duplication during SPIRV->NIR compilation.

This fixes new dEQP-VK.transform_feedback.simple.*.

Cc: mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17939>
 2022-08-09 16:04:25 +00:00
Tomeu Vizoso a35742c434 radv/ci: Put one more board to run the CTS on Stoney Ridge 
These jobs were taking more than 20 minutes to run, slowing the pipeline
down.

Reviewed-by: David Heidelberg <david.heidelberg@collabora.com>
Reviewed-by: Emma Anholt <emma@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17935>
 2022-08-09 15:39:41 +00:00
Martin Roukala (né Peres) 8d00c55e9e radv/ci: bump the console activity timeout of VanGogh to 3 minutes 
It seems like VKCTS's startup time is close to the current console
activity timeout (2 minutes...) on the Steam Deck, leading to spurious
timeouts...

Let's bump the timeout by 50%!

Signed-off-by: Martin Roukala (né Peres) <martin.roukala@mupuf.org>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17829>
 2022-08-09 09:04:09 +03:00
Martin Roukala (né Peres) 1db0b56609 radv/ci: skip host_write_transfer_src.1048576 on renoir 
The test can take a little longer than 60s, which leads to spurious
timeouts and thus unstable CI pipelines.

Signed-off-by: Martin Roukala (né Peres) <martin.roukala@mupuf.org>
Reviewed-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/17829>
 2022-08-09 09:04:09 +03:00
Martin Roukala (né Peres) e091e56fc1 radv/ci: add more tests to the flake list of vega10 
Signed-off-by: Martin Roukala (né Peres) <martin.roukala@mupuf.org>
Reviewed-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/17829>
 2022-08-09 09:04:09 +03:00
Martin Roukala (né Peres) 6bbe3c6d36 radv/ci: move some tests from the vega10 fail to its flake list 
v2:
 - Move the other 2 `sample_texture` tests to the flake list (Ishi)

Signed-off-by: Martin Roukala (né Peres) <martin.roukala@mupuf.org>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Tatsuyuki Ishi <ishitatsuyuki@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17829>
 2022-08-09 09:04:09 +03:00
Martin Roukala (né Peres) 3f8a356bd3 radv/ci: document a recent regression 
This regression may have been introduced by mesa/mesa!17652 or
mesa/mesa!17625.

Signed-off-by: Martin Roukala (né Peres) <martin.roukala@mupuf.org>
Reviewed-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/17829>
 2022-08-09 09:04:09 +03:00
Samuel Pitoiset d4b8abe511 radv: simplify radv_bind_dynamic_state() slightly 
This adds RADV_CMP_COPY to compact copies. Based on ANV.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17649>
 2022-08-08 14:59:47 +00:00
Samuel Pitoiset 18e9ba3e3b radv: remove unused states parameter from some radv_emit_XXX() helpers 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17649>
 2022-08-08 14:59:47 +00:00
Samuel Pitoiset 1f6e32ff7c radv: simplify saving/restoring all dynamic states 
Instead of copying every field individually, just use a whole memcpy.
This could be optimized but that's not the point here.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17649>
 2022-08-08 14:59:47 +00:00
Samuel Pitoiset 1d82ec1b3f radv: remove RADV_META_SAVE_SAMPLE_LOCATIONS 
We already save/restore all other dynamic states unconditionally, it's
not really useful to make an exception for sample locations.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17649>
 2022-08-08 14:59:47 +00:00
Rhys Perry bf0af80045 aco: improve VcmpxPermlaneHazard workaround 
According to LLVM, we only need to care about VOPC which writes exec.

No fossil-db changes.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17697>
 2022-08-08 13:59:17 +00:00
Rhys Perry 5912c7d3fa aco: only add vscnt wait when visiting VMEM/DS 
This prevents issues where we insert a s_waitcnt_vscnt(0) at the start of
a block or very end of the shader because we're joining two blocks (for
example, one with has_VMEM=true and the other with
has_branch_after_DS=true).

fossil-db (navi10):
Totals from 2441 (1.51% of 161220) affected shaders:
Instrs: 1383964 -> 1384094 (+0.01%); split: -0.07%, +0.08%
CodeSize: 7438212 -> 7438760 (+0.01%); split: -0.05%, +0.06%
Latency: 13780665 -> 13679664 (-0.73%); split: -1.53%, +0.80%
InvThroughput: 2950835 -> 2921511 (-0.99%); split: -1.06%, +0.07%

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17697>
 2022-08-08 13:59:17 +00:00
Rhys Perry 52156d6b26 aco: set has_VMEM,has_DS=false after a branch 
fossil-db (navi10):
Totals from 161 (0.10% of 161220) affected shaders:
Instrs: 206726 -> 207179 (+0.22%); split: -0.02%, +0.24%
CodeSize: 1114152 -> 1116032 (+0.17%); split: -0.01%, +0.18%
Latency: 2119380 -> 2147403 (+1.32%); split: -0.16%, +1.48%
InvThroughput: 462960 -> 461922 (-0.22%); split: -0.42%, +0.19%

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17697>
 2022-08-08 13:59:17 +00:00
Rhys Perry b17e59a03b aco: fix LdsBranchVmemWARHazard with 2+ branch chains 
For example, "DS -> branch -> VMEM -> branch -> DS".

fossil-db (navi10):
Totals from 639 (0.40% of 161220) affected shaders:
Instrs: 629090 -> 628254 (-0.13%); split: -0.19%, +0.06%
CodeSize: 3410164 -> 3406748 (-0.10%); split: -0.14%, +0.04%
Latency: 7834755 -> 7821011 (-0.18%); split: -0.70%, +0.52%
InvThroughput: 1369698 -> 1374495 (+0.35%); split: -0.12%, +0.47%

A lot of the fossil-db changes are noise.
threekingdoms.8db138826c386a62.1.foz/0b222ed175eebad0 is an example of a
shader that actually has this issue.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Fixes: c037ba1bb7 ("aco/gfx10: Mitigate LdsBranchVmemWARHazard.")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17697>
 2022-08-08 13:59:17 +00:00
Timur Kristóf dccd6f495a ac/nir/cull: Fix typo in bounding box culling. 
Bounding box culling is only viable when the W of all
vertices are positive. Always accept triangles whose any
W is negative.

Fixes: 0d527bb1aa
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/7018
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Qiang Yu <yuq825@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17929>
 2022-08-08 11:16:04 +00:00
Samuel Pitoiset 1fb12d2cce radv: use ref counting for VS prologs and PS epilogs 
With GPL, it will be possible to create VS prologs and PS epilogs
from libraries, so reference counting is useful here too.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-By: Tatsuyuki Ishi <ishitatsuyuki@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17628>
 2022-08-08 10:52:31 +00:00