fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2026-05-20 02:38:07 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 4
186682 commits 66 branches 936 tags 981 MiB
Commit graph

2941 commits

Author SHA1 Message Date
Daniel Schürmann
09c5414760 aco/spill: don't prefer to spill phis at merge blocks 
Totals from 44 (0.06% of 79242) affected shaders: (GFX11)

Instrs: 165578 -> 165545 (-0.02%); split: -0.08%, +0.06%
CodeSize: 833528 -> 832988 (-0.06%); split: -0.11%, +0.04%
SpillSGPRs: 346 -> 329 (-4.91%)
Latency: 1176341 -> 1176231 (-0.01%); split: -0.02%, +0.01%
InvThroughput: 235771 -> 235761 (-0.00%); split: -0.01%, +0.01%
SClause: 3287 -> 3289 (+0.06%); split: -0.21%, +0.27%
Copies: 16290 -> 16299 (+0.06%); split: -0.62%, +0.68%
Branches: 3633 -> 3641 (+0.22%)
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27774>
 2024-03-25 09:21:20 +00:00
Daniel Schürmann
d234f789b5 aco/spill: keep live-out variables spilled at branch blocks 
There is no reason to reload early. Leave scheduling to the scheduler.

Totals from 98 (0.12% of 79242) affected shaders: (GFX11)

Instrs: 841944 -> 841995 (+0.01%); split: -0.02%, +0.03%
CodeSize: 4354332 -> 4354544 (+0.00%); split: -0.02%, +0.02%
SpillSGPRs: 2557 -> 2554 (-0.12%)
Latency: 5772080 -> 5772209 (+0.00%); split: -0.00%, +0.01%
InvThroughput: 1694834 -> 1694873 (+0.00%); split: -0.00%, +0.00%
SClause: 16036 -> 16042 (+0.04%); split: -0.07%, +0.11%
Copies: 85117 -> 85125 (+0.01%); split: -0.18%, +0.19%
Branches: 25222 -> 25219 (-0.01%); split: -0.02%, +0.00%
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27774>
 2024-03-25 09:21:20 +00:00
Daniel Schürmann
32882d647e aco/spill: refactor adding spilled vars into separate function add_to_spills() 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27774>
 2024-03-25 09:21:20 +00:00
Daniel Schürmann
a5e8f9f879 aco/spill: add spills_entry interferences only when necessary 
For variables which are already spilled in previous blocks,
the interferences already exist.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27774>
 2024-03-25 09:21:20 +00:00
Daniel Schürmann
1a1d6aef56 aco/spill: don't allocate extra spill_id for phi operands in add_coupling_code() 
We can re-use the phi definition's spill_id.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27774>
 2024-03-25 09:21:20 +00:00
Daniel Schürmann
b1468fc952 aco/spill: refactor SSA repairing 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27774>
 2024-03-25 09:21:19 +00:00
Marek Olšák
7c25243f86 aco: add a helper printing shader asm by disassembling via LLVM 
ACO uses the LLVM disassembler. It was used to verify that manually-written
inline assembly code is correct.

Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27952>
 2024-03-22 21:58:02 +00:00
Marek Olšák
df6fe90926 aco: implement aco_is_gpu_supported using switch statement 
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27952>
 2024-03-22 21:58:02 +00:00
Marek Olšák
6773595ed0 nir: rename AMD XFB intrinsics to *_gfx11_amd 
to indicate it's only for gfx11.

Acked-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27952>
 2024-03-22 21:58:02 +00:00
Rhys Perry
588b1ce533 aco: split instruction assembly into functions 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27913>
 2024-03-22 10:29:43 +00:00
Rhys Perry
5651aa7644 aco/gfx11: fix scratch ST mode assembly 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27913>
 2024-03-22 10:29:43 +00:00
Rhys Perry
cdcc856cbf aco: unify MIMG opcode lists 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27913>
 2024-03-22 10:29:43 +00:00
Rhys Perry
20b14381c4 aco: move dot/wmma instructions into VOP3P list 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27913>
 2024-03-22 10:29:43 +00:00
Rhys Perry
6bfd3dc39f aco: use op() 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27913>
 2024-03-22 10:29:43 +00:00
Rhys Perry
a806e48982 aco: use single tuple for all opcode numbers 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27913>
 2024-03-22 10:29:43 +00:00
Rhys Perry
5864c22fb7 aco: refactor VOPC opcode list 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27913>
 2024-03-22 10:29:43 +00:00
Rhys Perry
d23f8b0dcf aco: rename opcode->instruction 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27913>
 2024-03-22 10:29:43 +00:00
Rhys Perry
b723a15bd0 aco: fix GFX6 buffer_load_dwordx4 opcode number 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27913>
 2024-03-22 10:29:43 +00:00
Rhys Perry
76e089ea48 aco/cssa: update comments 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28248>
 2024-03-20 17:50:27 +00:00
Rhys Perry
0c0819f0da aco/cssa: reset equal_anc_out if merging fails 
try_merge_merge_set() expects equal_anc_out to be Temp() in the beginning,
so we need to reset it in case it's used again.

Fixes compilation of metro_exodus/163b3b895730d37b with
VK_PIPELINE_CREATE_2_DISABLE_OPTIMIZATION_BIT_KHR.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Fixes: 18ba93e673 ("aco/cssa: rewrite lower_to_cssa pass")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28248>
 2024-03-20 17:50:27 +00:00
Timur Kristóf
58e3b1f930 aco: Allow passing constant operand to is_overwritten_since. 
This is to make it more intuitive and also consistent
with last_writer_idx which does allow constant operands.

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/28046>
 2024-03-19 20:50:12 +00:00
Daniel Schürmann
9bbb9f1104 aco: use small_vec as Block::edge_vec for predecessors and successors 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27984>
 2024-03-19 13:06:58 +00:00
Daniel Schürmann
3e58a736e4 aco/util: small_vec few additions 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27984>
 2024-03-19 13:06:58 +00:00
Rhys Perry
5cbd7689be aco/util: add small_vec 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27984>
 2024-03-19 13:06:58 +00:00
Daniel Schürmann
4564ca313b aco: reorder code and use namespaces in aco_interface.cpp 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27984>
 2024-03-19 13:06:58 +00:00
Georg Lehmann
67997fd735 aco: avoid full 32bit imul for uniform reduce/scan 
Foz-DB Navi31:
Totals from 24 (0.03% of 79395) affected shaders:
Instrs: 1172275 -> 1172078 (-0.02%)
CodeSize: 5974424 -> 5973860 (-0.01%)
Latency: 5896216 -> 5895923 (-0.00%)
InvThroughput: 1167928 -> 1167869 (-0.01%)
VALU: 625756 -> 625636 (-0.02%)

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28223>
 2024-03-18 13:37:28 +00:00
Georg Lehmann
96ff511b75 aco/optimizer: combine v_mul_i32_i24 and add to mad 
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28223>
 2024-03-18 13:37:28 +00:00
Georg Lehmann
0c57340c23 aco/builder: use 24bit mul if low bits of imm are zero 
Foz-DB Navi31:
Totals from 39 (0.05% of 79395) affected shaders:
Instrs: 62712 -> 62696 (-0.03%)
CodeSize: 330096 -> 329896 (-0.06%)
Latency: 192747 -> 192561 (-0.10%)
InvThroughput: 34078 -> 33889 (-0.55%)
VALU: 38979 -> 38963 (-0.04%)

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28223>
 2024-03-18 13:37:28 +00:00
Georg Lehmann
b48a101d8f aco/builder: improve v_mul_imm for negative imm 
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28223>
 2024-03-18 13:37:27 +00:00
Rhys Perry
ec892c4d2b aco: don't reuse misaligned attribute destination VGPRs in VS prologs 
Since we split misaligned attributes, we could overwrite one of these
VGPRs in the middle of loading the attribute.

For example:
   v_add_u32_e32 v4, vcc, s7, v1
   s_waitcnt lgkmcnt(0)
   buffer_load_dword v4, v4, s[32:35], 0 idxen
   buffer_load_dword v5, v4, s[32:35], 0 idxen offset:4
can overwrite the vertex index in the load of the first component.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27920>
 2024-03-14 15:30:12 +00:00
Rhys Perry
df7024bcdd radv,aco: allow VS prologs to increase VGPR usage 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27920>
 2024-03-14 15:30:12 +00:00
Samuel Pitoiset
0ed6aba85a radv,aco: stop duplicating PS/TCS epilog fields 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28042>
 2024-03-08 07:28:47 +00:00
Konstantin Seurer
eb249bb18e aco: Only fix used variables to registers 
Reviewed-by: Friedrich Vock <friedrich.vock@gmx.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27137>
 2024-03-07 20:07:07 +00:00
Rhys Perry
dea8b02b03 aco: don't pass constant to is_overwritten_since() 
Fixes usage of uninitialised value in dead_space/065515347dca4851 and
other dead_space shaders.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Fixes: def0c275c4 ("aco: Eliminate SCC copies when possible.")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28011>
 2024-03-07 12:17:39 +00:00
Daniel Schürmann
61854009f3 aco: rematerialize constants in every basic block during optimizer 
Totals from 16837 (21.25% of 79242) affected shaders: (GFX11)

MaxWaves: 441634 -> 444546 (+0.66%); split: +0.66%, -0.00%
Instrs: 25908303 -> 25838469 (-0.27%); split: -0.36%, +0.09%
CodeSize: 133943168 -> 135446948 (+1.12%); split: -0.04%, +1.16%
VGPRs: 985332 -> 977440 (-0.80%); split: -0.83%, +0.03%
SpillSGPRs: 9133 -> 7535 (-17.50%); split: -17.74%, +0.24%
SpillVGPRs: 1418 -> 1359 (-4.16%); split: -4.58%, +0.42%
Scratch: 5047552 -> 5040640 (-0.14%)
Latency: 204330340 -> 204179212 (-0.07%); split: -0.32%, +0.25%
InvThroughput: 36584220 -> 36508856 (-0.21%); split: -0.40%, +0.19%
VClause: 437847 -> 437344 (-0.11%); split: -0.34%, +0.22%
SClause: 771311 -> 771013 (-0.04%); split: -0.42%, +0.38%
Copies: 1774950 -> 1712070 (-3.54%); split: -4.46%, +0.91%
Branches: 580595 -> 580478 (-0.02%); split: -0.03%, +0.01%
PreSGPRs: 877017 -> 817549 (-6.78%)
PreVGPRs: 852747 -> 846966 (-0.68%); split: -0.68%, +0.00%
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/26875>
 2024-03-06 15:02:21 +00:00
Rhys Perry
3b28ba8239 aco: optimize for purely linear VGPR copies 
fossil-db:
Totals from 2 (0.00% of 79242) affected shaders:
Instrs: 1344 -> 1340 (-0.30%)
CodeSize: 6968 -> 6952 (-0.23%)
Latency: 4414 -> 4410 (-0.09%)
InvThroughput: 1018 -> 1020 (+0.20%)
Copies: 60 -> 56 (-6.67%)
SALU: 40 -> 36 (-10.00%)

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27697>
 2024-03-06 12:55:46 +00:00
Rhys Perry
8cd3a3a520 aco/tests: add tests for linear VGPR register allocation 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27697>
 2024-03-06 12:55:46 +00:00
Rhys Perry
f9b37723d0 aco/ra: emit linear VGPR parallel copy separately 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27697>
 2024-03-06 12:55:46 +00:00
Rhys Perry
d9b69a7cbf aco/ra: disable live range splitting of linear vgprs 
These shouldn't happen anymore.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27697>
 2024-03-06 12:55:46 +00:00
Rhys Perry
b7738de4f9 aco/ra: rework linear VGPR allocation 
We allocate them at the end of the register file and keep them separate
from normal VGPRs. This is for two reasons:
- Because we only ever move linear VGPRs into an empty space or a space
  previously occupied by a linear one, we never have to swap a normal VGPR
  and a linear one. This simplifies copy lowering.
- As linear VGPR's live ranges only start and end on top-level blocks, we
  never have to move a linear VGPR in control flow.

fossil-db (navi31):
Totals from 5493 (6.93% of 79242) affected shaders:
MaxWaves: 150365 -> 150343 (-0.01%)
Instrs: 7974740 -> 7976073 (+0.02%); split: -0.06%, +0.08%
CodeSize: 41296024 -> 41299024 (+0.01%); split: -0.06%, +0.06%
VGPRs: 283192 -> 329560 (+16.37%)
Latency: 64267936 -> 64268414 (+0.00%); split: -0.17%, +0.17%
InvThroughput: 10954037 -> 10951735 (-0.02%); split: -0.09%, +0.07%
VClause: 132792 -> 132956 (+0.12%); split: -0.06%, +0.18%
SClause: 223854 -> 223841 (-0.01%); split: -0.01%, +0.01%
Copies: 559574 -> 561395 (+0.33%); split: -0.24%, +0.56%
Branches: 179630 -> 179636 (+0.00%); split: -0.02%, +0.02%
VALU: 4572683 -> 4574487 (+0.04%); split: -0.03%, +0.07%
SALU: 772076 -> 772111 (+0.00%); split: -0.01%, +0.01%
VOPD: 1095 -> 1099 (+0.37%); split: +0.73%, -0.37%

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27697>
 2024-03-06 12:55:46 +00:00
Rhys Perry
2d49c79c7e aco/ra: change get_reg_bounds() helper 
We will have a separate bounds for linear VGPRs.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27697>
 2024-03-06 12:55:46 +00:00
Rhys Perry
a38bc9e165 aco/ra: move parallelcopy creation into helper 
This is almost a direct copy+paste into it's own function.

This is useful both for future work and the make the caller smaller.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27697>
 2024-03-06 12:55:46 +00:00
Rhys Perry
a8b72082cf aco/ra: constify various RegisterFile 
This makes it more obvious that these functions don't change it.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27697>
 2024-03-06 12:55:45 +00:00
Rhys Perry
6a195bb521 aco: only allow linear vgpr kills in top-level blocks 
This is already the case, and requiring it will be useful in the future.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27697>
 2024-03-06 12:55:45 +00:00
Rhys Perry
07eab4ab40 aco: require linear vgpr uses to be late kill 
This also removes some linear VGPR tests which will be replaced soon
anyway.

fossil-db (navi31):
Totals from 107 (0.14% of 79242) affected shaders:
Instrs: 66203 -> 66211 (+0.01%); split: -0.09%, +0.10%
CodeSize: 354644 -> 354588 (-0.02%); split: -0.08%, +0.07%
VGPRs: 4476 -> 4452 (-0.54%); split: -0.80%, +0.27%
Latency: 513863 -> 513877 (+0.00%); split: -0.08%, +0.08%
InvThroughput: 68871 -> 68870 (-0.00%); split: -0.02%, +0.02%
SClause: 1589 -> 1590 (+0.06%)
PreVGPRs: 3404 -> 3415 (+0.32%)

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27697>
 2024-03-06 12:55:45 +00:00
Rhys Perry
5e17a39b15 aco: allow p_start_linear_vgpr to use multiple operands 
Merging the p_create_vector into the p_start_linear_vgpr is useful since
we stopped attempting to place the p_start_linear_vgpr definition in the
same registers as the operand.

fossil-db (navi31):
Totals from 927 (1.17% of 79242) affected shaders:
MaxWaves: 26412 -> 26442 (+0.11%)
Instrs: 938328 -> 938181 (-0.02%); split: -0.14%, +0.13%
CodeSize: 4891448 -> 4890820 (-0.01%); split: -0.11%, +0.10%
VGPRs: 47016 -> 47004 (-0.03%); split: -0.13%, +0.10%
SpillSGPRs: 222 -> 226 (+1.80%)
Latency: 5076065 -> 5075191 (-0.02%); split: -0.12%, +0.10%
InvThroughput: 712316 -> 712421 (+0.01%); split: -0.09%, +0.10%
SClause: 27992 -> 27972 (-0.07%); split: -0.09%, +0.02%
Copies: 38042 -> 38104 (+0.16%); split: -1.95%, +2.12%
PreVGPRs: 39448 -> 39369 (-0.20%)
VALU: 570157 -> 570224 (+0.01%); split: -0.13%, +0.14%
SALU: 51672 -> 51678 (+0.01%); split: -0.01%, +0.02%

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27697>
 2024-03-06 12:55:45 +00:00
Rhys Perry
f764f6848a aco/ra: disable p_start_linear_vgpr allocation hint 
As this is, this will become useless soon.

fossil-db (navi31):
Totals from 176 (0.22% of 79242) affected shaders:
Instrs: 101932 -> 102413 (+0.47%); split: -0.01%, +0.49%
CodeSize: 541352 -> 543256 (+0.35%); split: -0.01%, +0.36%
VGPRs: 7884 -> 7896 (+0.15%)
Latency: 588129 -> 588559 (+0.07%); split: -0.07%, +0.15%
InvThroughput: 83349 -> 83689 (+0.41%); split: -0.01%, +0.42%
Copies: 4324 -> 4691 (+8.49%)
VALU: 61431 -> 61798 (+0.60%)

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27697>
 2024-03-06 12:55:45 +00:00
Rhys Perry
f99443a68b aco: don't combine linear and normal VGPR copies 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27697>
 2024-03-06 12:55:45 +00:00
Timur Kristóf
def0c275c4 aco: Eliminate SCC copies when possible. 
Foz-DB Navi31:
Totals from 2517 (3.22% of 78112) affected shaders:
Instrs: 5992126 -> 5972611 (-0.33%); split: -0.33%, +0.00%
CodeSize: 30986404 -> 30914536 (-0.23%); split: -0.23%, +0.00%
Latency: 43221112 -> 43217422 (-0.01%); split: -0.02%, +0.01%
InvThroughput: 6675983 -> 6674598 (-0.02%); split: -0.02%, +0.00%
SClause: 181987 -> 181976 (-0.01%); split: -0.01%, +0.00%
Copies: 538852 -> 519419 (-3.61%)

Co-authored-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27072>
 2024-03-05 22:51:01 +00:00
Chia-I Wu
3d4dfae7eb aco: fix nir_op_pack_32_4x8 handling 
I started seeing

  ACO ERROR:
      In file ../src/amd/compiler/aco_validate.cpp:98
      Operand and Definition types do not match:  s1: %44 = p_parallelcopy %158
  test_basic: ../src/amd/compiler/aco_interface.cpp:85: void validate(aco::Program*):
      Assertion `is_valid' failed.

since commit 52ee4cf229 ("nir/builder: Teach nir_pack_bits and
nir_unpack_bits about 32_4x8").

Fixes: e0d232c2fc ("aco: implement nir_op_pack_32_4x8").  I
Suggested-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27972>
 2024-03-05 20:38:34 +00:00