Commit graph

3 commits

Author SHA1 Message Date
Ian Romanick
97e3c6a12a intel/brw: Use range analysis to optimize fsign
shader-db:

Meteor Lake, DG2, and Tiger Lake had similar results. (Meteor Lake shown)
total instructions in shared programs: 19674784 -> 19665960 (-0.04%)
instructions in affected programs: 933425 -> 924601 (-0.95%)
helped: 3656 / HURT: 0

total cycles in shared programs: 810343919 -> 810241030 (-0.01%)
cycles in affected programs: 56752034 -> 56649145 (-0.18%)
helped: 3032 / HURT: 434

LOST:   11
GAINED: 0

Ice Lake and Skylake had similar results. (Ice Lake shown)
total instructions in shared programs: 20315795 -> 20305856 (-0.05%)
instructions in affected programs: 979698 -> 969759 (-1.01%)
helped: 3845 / HURT: 0

total cycles in shared programs: 830600281 -> 830534694 (<.01%)
cycles in affected programs: 45675615 -> 45610028 (-0.14%)
helped: 3250 / HURT: 325

total spills in shared programs: 4583 -> 4565 (-0.39%)
spills in affected programs: 180 -> 162 (-10.00%)
helped: 3 / HURT: 0

total fills in shared programs: 5245 -> 5219 (-0.50%)
fills in affected programs: 379 -> 353 (-6.86%)
helped: 3 / HURT: 0

LOST:   14
GAINED: 8

fossil-db:

All Intel platforms except Tiger Lake had similar results. (Meteor Lake shown)
Totals:
Instrs: 154024263 -> 154023814 (-0.00%)
Cycle count: 17463341602 -> 17461726239 (-0.01%); split: -0.01%, +0.00%

Totals from 322 (0.05% of 631440) affected shaders:
Instrs: 199933 -> 199484 (-0.22%)
Cycle count: 168492537 -> 166877174 (-0.96%); split: -0.96%, +0.00%

Tiger Lake
Instrs: 149984723 -> 149984287 (-0.00%)
Cycle count: 15238596937 -> 15239260415 (+0.00%); split: -0.00%, +0.01%
Max dispatch width: 5553408 -> 5553424 (+0.00%)

Totals from 318 (0.05% of 631414) affected shaders:
Instrs: 179624 -> 179188 (-0.24%)
Cycle count: 160724533 -> 161388011 (+0.41%); split: -0.06%, +0.48%
Max dispatch width: 3296 -> 3312 (+0.49%)

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29095>
2024-05-14 01:28:21 +00:00
Ian Romanick
e578657313 intel/brw: Implement more strictly correct fsign lowering
The huge amount of helped shaders is due to the "~" versions of the
patterns.

shader-db:

Meteor Lake and DG2 had similar results. (Meteor Lake shown)
total instructions in shared programs: 19672345 -> 19662605 (-0.05%)
instructions in affected programs: 1147766 -> 1138026 (-0.85%)
helped: 2691 / HURT: 1650

total cycles in shared programs: 810323688 -> 810145191 (-0.02%)
cycles in affected programs: 68918312 -> 68739815 (-0.26%)
helped: 3651 / HURT: 1832

LOST:   29
GAINED: 38

Tiger Lake
total instructions in shared programs: 19489619 -> 19479909 (-0.05%)
instructions in affected programs: 1124564 -> 1114854 (-0.86%)
helped: 2682 / HURT: 1643

total cycles in shared programs: 811468406 -> 811706747 (0.03%)
cycles in affected programs: 66397690 -> 66636031 (0.36%)
helped: 3692 / HURT: 1775

total spills in shared programs: 3906 -> 3907 (0.03%)
spills in affected programs: 16 -> 17 (6.25%)
helped: 0 / HURT: 1

total fills in shared programs: 3220 -> 3222 (0.06%)
fills in affected programs: 50 -> 52 (4.00%)
helped: 0 / HURT: 1

LOST:   33
GAINED: 36

Ice Lake and Skylake had similar results. (Ice Lake shown)
total instructions in shared programs: 20317882 -> 20307495 (-0.05%)
instructions in affected programs: 1199651 -> 1189264 (-0.87%)
helped: 2863 / HURT: 1680

total cycles in shared programs: 830880024 -> 830457927 (-0.05%)
cycles in affected programs: 63347102 -> 62925005 (-0.67%)
helped: 4118 / HURT: 1622

total spills in shared programs: 4593 -> 4583 (-0.22%)
spills in affected programs: 205 -> 195 (-4.88%)
helped: 4 / HURT: 0

total fills in shared programs: 5284 -> 5245 (-0.74%)
fills in affected programs: 464 -> 425 (-8.41%)
helped: 4 / HURT: 0

LOST:   70
GAINED: 33

fossil-db:

Meteor Lake and DG2 had similar results. (Meteor Lake shown)
Totals:
Instrs: 154025275 -> 154022035 (-0.00%); split: -0.00%, +0.00%
Cycle count: 17472869499 -> 17463289530 (-0.05%); split: -0.06%, +0.00%
Spill count: 141269 -> 141246 (-0.02%); split: -0.02%, +0.00%
Fill count: 265342 -> 265159 (-0.07%); split: -0.11%, +0.04%
Max live registers: 32597829 -> 32597986 (+0.00%); split: -0.00%, +0.00%
Max dispatch width: 5536776 -> 5537048 (+0.00%)

Totals from 1590 (0.25% of 631423) affected shaders:
Instrs: 1146532 -> 1143292 (-0.28%); split: -0.44%, +0.16%
Cycle count: 1230843330 -> 1221263361 (-0.78%); split: -0.83%, +0.05%
Spill count: 15832 -> 15809 (-0.15%); split: -0.19%, +0.04%
Fill count: 36071 -> 35888 (-0.51%); split: -0.79%, +0.29%
Max live registers: 93529 -> 93686 (+0.17%); split: -0.00%, +0.17%
Max dispatch width: 15168 -> 15440 (+1.79%)

Tiger Lake, Ice Lake, and Skylake had similar results. (Tiger Lake shown)
Totals:
Instrs: 149564084 -> 149562467 (-0.00%); split: -0.00%, +0.00%
Cycle count: 15151701515 -> 15158290114 (+0.04%); split: -0.00%, +0.04%
Max live registers: 32249443 -> 32249620 (+0.00%); split: -0.00%, +0.00%
Max dispatch width: 5540536 -> 5540488 (-0.00%)

Totals from 1605 (0.25% of 630303) affected shaders:
Instrs: 584950 -> 583333 (-0.28%); split: -0.49%, +0.21%
Cycle count: 160926321 -> 167514920 (+4.09%); split: -0.05%, +4.14%
Max live registers: 90851 -> 91028 (+0.19%); split: -0.00%, +0.20%
Max dispatch width: 15440 -> 15392 (-0.31%)

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29095>
2024-05-14 01:28:20 +00:00
Ian Romanick
3f151c03af intel/brw: Handle fsign optimization in a NIR algebraic pass
This is a lot less code, and it makes it easier to experiment with other
pattern-based optimizations in the future.

The results here are nearly identical to the results I got from Ken's
"intel/brw: Make fsign (for 16/32-bit) in SSA form"... which are not
particularly good.

In this commit and in Ken's, all of the shader-db shaders hurt for
spills and fills are from Deus Ex Mankind Divided. Each shader has a
bunch of texture instructions with a single fsign between the
blocks. With the dependency on the flag removed, the scheduler puts all
of the texture instructions at the start... and there are a LOT of them.

shader-db:

All Intel platforms had similar results. (Meteor Lake shown)
total instructions in shared programs: 19647060 -> 19650207 (0.02%)
instructions in affected programs: 734718 -> 737865 (0.43%)
helped: 382 / HURT: 1984

total cycles in shared programs: 823238442 -> 822785913 (-0.05%)
cycles in affected programs: 426901157 -> 426448628 (-0.11%)
helped: 3408 / HURT: 3671

total spills in shared programs: 3887 -> 3891 (0.10%)
spills in affected programs: 256 -> 260 (1.56%)
helped: 0 / HURT: 4

total fills in shared programs: 3236 -> 3306 (2.16%)
fills in affected programs: 882 -> 952 (7.94%)
helped: 0 / HURT: 12

LOST:   37
GAINED: 34

fossil-db:

DG2 and Meteor Lake had similar results. (Meteor Lake shown)
Totals:
Instrs: 154005469 -> 154008294 (+0.00%); split: -0.00%, +0.00%
Cycle count: 17551859277 -> 17554293955 (+0.01%); split: -0.02%, +0.04%
Spill count: 142078 -> 142090 (+0.01%)
Fill count: 266761 -> 266729 (-0.01%); split: -0.02%, +0.01%
Max live registers: 32593578 -> 32593858 (+0.00%)
Max dispatch width: 5535944 -> 5536816 (+0.02%); split: +0.02%, -0.01%

Totals from 5867 (0.93% of 631350) affected shaders:
Instrs: 5475544 -> 5478369 (+0.05%); split: -0.04%, +0.09%
Cycle count: 1649032029 -> 1651466707 (+0.15%); split: -0.24%, +0.39%
Spill count: 26411 -> 26423 (+0.05%)
Fill count: 57364 -> 57332 (-0.06%); split: -0.10%, +0.04%
Max live registers: 431561 -> 431841 (+0.06%)
Max dispatch width: 49784 -> 50656 (+1.75%); split: +2.38%, -0.63%

Tiger Lake
Totals:
Instrs: 149530671 -> 149533588 (+0.00%); split: -0.00%, +0.00%
Cycle count: 15261418953 -> 15264764921 (+0.02%); split: -0.00%, +0.03%
Spill count: 60317 -> 60316 (-0.00%); split: -0.02%, +0.01%
Max live registers: 32249201 -> 32249464 (+0.00%)
Max dispatch width: 5540608 -> 5540584 (-0.00%)

Totals from 5862 (0.93% of 630309) affected shaders:
Instrs: 4740800 -> 4743717 (+0.06%); split: -0.04%, +0.10%
Cycle count: 566531248 -> 569877216 (+0.59%); split: -0.13%, +0.72%
Spill count: 11709 -> 11708 (-0.01%); split: -0.09%, +0.08%
Max live registers: 424560 -> 424823 (+0.06%)
Max dispatch width: 50304 -> 50280 (-0.05%)

Ice Lake
Totals:
Instrs: 150499705 -> 150502608 (+0.00%); split: -0.00%, +0.00%
Cycle count: 15105629116 -> 15105425880 (-0.00%); split: -0.00%, +0.00%
Spill count: 60087 -> 60090 (+0.00%)
Fill count: 100542 -> 100541 (-0.00%); split: -0.00%, +0.00%
Max live registers: 32605215 -> 32605495 (+0.00%)
Max dispatch width: 5617752 -> 5617792 (+0.00%); split: +0.00%, -0.00%

Totals from 5882 (0.93% of 634934) affected shaders:
Instrs: 4737206 -> 4740109 (+0.06%); split: -0.04%, +0.10%
Cycle count: 598882104 -> 598678868 (-0.03%); split: -0.08%, +0.05%
Spill count: 10278 -> 10281 (+0.03%)
Fill count: 22504 -> 22503 (-0.00%); split: -0.01%, +0.01%
Max live registers: 424184 -> 424464 (+0.07%)
Max dispatch width: 50216 -> 50256 (+0.08%); split: +0.25%, -0.18%

Skylake
Totals:
Instrs: 139092612 -> 139095257 (+0.00%); split: -0.00%, +0.00%
Cycle count: 14533550285 -> 14533544716 (-0.00%); split: -0.00%, +0.00%
Spill count: 58176 -> 58172 (-0.01%)
Fill count: 95877 -> 95796 (-0.08%)
Max live registers: 31924594 -> 31924874 (+0.00%)
Max dispatch width: 5484568 -> 5484552 (-0.00%); split: +0.00%, -0.00%

Totals from 5789 (0.93% of 625512) affected shaders:
Instrs: 4481987 -> 4484632 (+0.06%); split: -0.04%, +0.10%
Cycle count: 578310124 -> 578304555 (-0.00%); split: -0.05%, +0.05%
Spill count: 9248 -> 9244 (-0.04%)
Fill count: 19677 -> 19596 (-0.41%)
Max live registers: 415340 -> 415620 (+0.07%)
Max dispatch width: 49720 -> 49704 (-0.03%); split: +0.10%, -0.13%

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29095>
2024-05-14 01:28:20 +00:00