mesa/src/intel/compiler/meson.build

332 lines
8.7 KiB
Meson
Raw Normal View History

# Copyright © 2017 Intel Corporation
# SPDX-License-Identifier: MIT
intel_nir_files = files(
'intel_nir.h',
'intel_nir.c',
'intel_nir_blockify_uniform_loads.c',
'intel_nir_clamp_image_1d_2d_array_sizes.c',
'intel_nir_clamp_per_vertex_loads.c',
'intel_nir_lower_conversions.c',
'intel_nir_lower_non_uniform_barycentric_at_sample.c',
'intel_nir_lower_non_uniform_resource_intel.c',
'intel_nir_lower_printf.c',
'intel_nir_lower_shading_rate_output.c',
'intel_nir_lower_sparse.c',
'intel_nir_lower_texture.c',
'intel_nir_opt_peephole_ffma.c',
'intel_nir_opt_peephole_imul32x16.c',
'intel_nir_tcs_workarounds.c',
)
libintel_compiler_brw_files = files(
'brw_analysis.h',
'brw_analysis.cpp',
'brw_analysis_def.cpp',
'brw_analysis_liveness.cpp',
'brw_analysis_performance.cpp',
'brw_builder.cpp',
'brw_builder.h',
'brw_cfg.cpp',
'brw_cfg.h',
'brw_compile_bs.cpp',
'brw_compile_cs.cpp',
'brw_compile_fs.cpp',
'brw_compile_gs.cpp',
'brw_compile_mesh.cpp',
'brw_compile_tcs.cpp',
'brw_compile_tes.cpp',
'brw_compile_vs.cpp',
'brw_compiler.c',
'brw_compiler.h',
'brw_debug_recompile.c',
'brw_disasm.c',
'brw_disasm_info.cpp',
'brw_disasm_info.h',
'brw_eu.c',
'brw_eu_compact.c',
'brw_eu_defines.h',
'brw_eu_emit.c',
'brw_eu_inst.h',
'brw_eu.h',
'brw_eu_validate.c',
'brw_from_nir.cpp',
'brw_fs.cpp',
'brw_fs.h',
'brw_generator.cpp',
'brw_generator.h',
'brw_inst.cpp',
'brw_inst.h',
'brw_isa_info.h',
'brw_lower.cpp',
'brw_lower_dpas.cpp',
'brw_lower_integer_multiplication.cpp',
'brw_lower_logical_sends.cpp',
'brw_lower_pack.cpp',
'brw_lower_regioning.cpp',
'brw_lower_scoreboard.cpp',
'brw_lower_simd_width.cpp',
'brw_lower_subgroup_ops.cpp',
'brw_nir.h',
'brw_nir.c',
'brw_nir_analyze_ubo_ranges.c',
'brw_nir_lower_cooperative_matrix.c',
'brw_nir_lower_cs_intrinsics.c',
'brw_nir_lower_alpha_to_coverage.c',
'brw_nir_lower_intersection_shader.c',
'brw_nir_lower_ray_queries.c',
'brw_nir_lower_rt_intrinsics.c',
'brw_nir_lower_shader_calls.c',
'brw_nir_lower_storage_image.c',
intel/brw: Move fsat instructions closer to the source Intel GPUs have a saturate destination modifier, and brw_fs_opt_saturate_propagation tries to replace explicit saturate operations with this destination modifier. That pass is limited in several ways. If the source of the explicit saturate is in a different block or if the source of the explicit saturate is live after the explicit saturate, brw_fs_opt_saturate_propagation will be unable to make progress. This optimization exists to help brw_fs_opt_saturate_propagation make more progress. It tries to move NIR fsat instructions to the same block that contains the definition of its source. It does this only in cases where it will not create additional live values. It also attempts to do this only in cases where the explicit saturate will ultimiately be converted to a destination modifier. v2: Fix metadata_preserve when theres no progress and use nir_metadata_control_flow when there is progress. All suggested by Alyssa. v3: Fix a typo in the file header comment. Noticed by Ken. Don't require nir_metadata_instr_index. Use nir_def_rewrite_uses_after instead of open-coding something slightly more specific. Both suggested by Ken. shader-db: All Intel platforms had similar results. (Meteor Lake shown) total instructions in shared programs: 19733645 -> 19733028 (<.01%) instructions in affected programs: 193300 -> 192683 (-0.32%) helped: 246 HURT: 1 helped stats (abs) min: 2 max: 48 x̄: 2.51 x̃: 2 helped stats (rel) min: 0.18% max: 0.39% x̄: 0.33% x̃: 0.34% HURT stats (abs) min: 1 max: 1 x̄: 1.00 x̃: 1 HURT stats (rel) min: 0.31% max: 0.31% x̄: 0.31% x̃: 0.31% 95% mean confidence interval for instructions value: -2.87 -2.13 95% mean confidence interval for instructions %-change: -0.34% -0.32% Instructions are helped. total cycles in shared programs: 916180971 -> 916264656 (<.01%) cycles in affected programs: 30197180 -> 30280865 (0.28%) helped: 194 HURT: 142 helped stats (abs) min: 1 max: 21251 x̄: 872.75 x̃: 19 helped stats (rel) min: <.01% max: 23.17% x̄: 2.59% x̃: 0.23% HURT stats (abs) min: 1 max: 28058 x̄: 1781.68 x̃: 399 HURT stats (rel) min: <.01% max: 37.21% x̄: 4.85% x̃: 1.63% 95% mean confidence interval for cycles value: -196.84 694.97 95% mean confidence interval for cycles %-change: -0.17% 1.27% Inconclusive result (value mean confidence interval includes 0). fossil-db: Meteor Lake, DG2, and Tiger Lake had similar results. (Meteor Lake shown) Totals: Instrs: 151512021 -> 151511351 (-0.00%); split: -0.00%, +0.00% Cycle count: 17209013596 -> 17209840995 (+0.00%); split: -0.02%, +0.02% Max live registers: 32013312 -> 32013549 (+0.00%) Max dispatch width: 5512304 -> 5512136 (-0.00%) Totals from 774 (0.12% of 630172) affected shaders: Instrs: 1559285 -> 1558615 (-0.04%); split: -0.05%, +0.01% Cycle count: 1312656268 -> 1313483667 (+0.06%); split: -0.24%, +0.30% Max live registers: 82195 -> 82432 (+0.29%) Max dispatch width: 6664 -> 6496 (-2.52%) Ice Lake Totals: Instrs: 151416791 -> 151416137 (-0.00%); split: -0.00%, +0.00% Cycle count: 15162468885 -> 15163298824 (+0.01%); split: -0.00%, +0.01% Max live registers: 32471367 -> 32471603 (+0.00%) Max dispatch width: 5623752 -> 5623712 (-0.00%) Totals from 733 (0.12% of 635598) affected shaders: Instrs: 877965 -> 877311 (-0.07%); split: -0.09%, +0.01% Cycle count: 190763628 -> 191593567 (+0.44%); split: -0.21%, +0.64% Max live registers: 72067 -> 72303 (+0.33%) Max dispatch width: 6216 -> 6176 (-0.64%) Skylake Totals: Instrs: 140794845 -> 140794075 (-0.00%); split: -0.00%, +0.00% Cycle count: 14665159301 -> 14665320514 (+0.00%); split: -0.00%, +0.01% Max live registers: 31783341 -> 31783662 (+0.00%); split: -0.00%, +0.00% Totals from 659 (0.11% of 625670) affected shaders: Instrs: 829061 -> 828291 (-0.09%); split: -0.09%, +0.00% Cycle count: 185478478 -> 185639691 (+0.09%); split: -0.33%, +0.41% Max live registers: 67491 -> 67812 (+0.48%); split: -0.01%, +0.48% Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29774>
2024-06-07 13:08:06 -07:00
'brw_nir_opt_fsat.c',
'brw_nir_rt.h',
'brw_nir_rt.c',
'brw_nir_rt_builder.h',
'brw_opt.cpp',
brw: move final send lowering up into the IR Because we do emit the final send message form in code generation, a lot of emissions look like this : add(8) vgrf0, u0, 0x100 mov(1) a0.1, vgrf0 # emitted by the generator send(8) ..., a0.1 By moving address register manipulation in the IR, we can get this down to : add(1) a0.1, u0, 0x100 send(8) ..., a0.1 This reduce register pressure around some send messages by 1 vgrf. All lost shaders in the below results are fragment SIMD32, due to the throughput estimator. If turned off, we loose no SIMD32 shaders with this change. DG2 results: Assassin's Creed Valhalla: Totals from 2044 (96.87% of 2110) affected shaders: Instrs: 852879 -> 832044 (-2.44%); split: -2.45%, +0.00% Subgroup size: 23832 -> 23824 (-0.03%) Cycle count: 53345742 -> 52144277 (-2.25%); split: -5.08%, +2.82% Spill count: 729 -> 554 (-24.01%); split: -28.40%, +4.39% Fill count: 2005 -> 1256 (-37.36%) Scratch Memory Size: 25600 -> 19456 (-24.00%); split: -32.00%, +8.00% Max live registers: 116765 -> 115058 (-1.46%) Max dispatch width: 19152 -> 18872 (-1.46%); split: +0.21%, -1.67% Cyberpunk 2077: Totals from 1181 (93.43% of 1264) affected shaders: Instrs: 667192 -> 663615 (-0.54%); split: -0.55%, +0.01% Subgroup size: 13016 -> 13032 (+0.12%) Cycle count: 17383539 -> 17986073 (+3.47%); split: -0.93%, +4.39% Spill count: 12 -> 8 (-33.33%) Fill count: 9 -> 6 (-33.33%) Dota2: Totals from 173 (11.59% of 1493) affected shaders: Cycle count: 274403 -> 280817 (+2.34%); split: -0.01%, +2.34% Max live registers: 5787 -> 5779 (-0.14%) Max dispatch width: 1344 -> 1152 (-14.29%) Hitman3: Totals from 5072 (95.39% of 5317) affected shaders: Instrs: 2879952 -> 2841804 (-1.32%); split: -1.32%, +0.00% Cycle count: 153208505 -> 165860401 (+8.26%); split: -2.22%, +10.48% Spill count: 3942 -> 3200 (-18.82%) Fill count: 10158 -> 8846 (-12.92%) Scratch Memory Size: 257024 -> 223232 (-13.15%) Max live registers: 328467 -> 324631 (-1.17%) Max dispatch width: 43928 -> 42768 (-2.64%); split: +0.09%, -2.73% Fortnite: Totals from 360 (4.82% of 7472) affected shaders: Instrs: 778068 -> 777925 (-0.02%) Subgroup size: 3128 -> 3136 (+0.26%) Cycle count: 38684183 -> 38734579 (+0.13%); split: -0.06%, +0.19% Max live registers: 50689 -> 50658 (-0.06%) Hogwarts Legacy: Totals from 1376 (84.00% of 1638) affected shaders: Instrs: 758810 -> 749727 (-1.20%); split: -1.23%, +0.03% Cycle count: 27778983 -> 28805469 (+3.70%); split: -1.42%, +5.12% Spill count: 2475 -> 2299 (-7.11%); split: -7.47%, +0.36% Fill count: 2677 -> 2445 (-8.67%); split: -9.90%, +1.23% Scratch Memory Size: 99328 -> 89088 (-10.31%) Max live registers: 84969 -> 84671 (-0.35%); split: -0.58%, +0.23% Max dispatch width: 11848 -> 11920 (+0.61%) Metro Exodus: Totals from 92 (0.21% of 43072) affected shaders: Instrs: 262995 -> 262968 (-0.01%) Cycle count: 13818007 -> 13851266 (+0.24%); split: -0.01%, +0.25% Max live registers: 11152 -> 11140 (-0.11%) Red Dead Redemption 2 : Totals from 451 (7.71% of 5847) affected shaders: Instrs: 754178 -> 753811 (-0.05%); split: -0.05%, +0.00% Cycle count: 3484078523 -> 3484111965 (+0.00%); split: -0.00%, +0.00% Max live registers: 42294 -> 42185 (-0.26%) Spiderman Remastered: Totals from 6820 (98.02% of 6958) affected shaders: Instrs: 6921500 -> 6747933 (-2.51%); split: -4.16%, +1.65% Cycle count: 234400692460 -> 236846720707 (+1.04%); split: -0.20%, +1.25% Spill count: 72971 -> 72622 (-0.48%); split: -8.08%, +7.61% Fill count: 212921 -> 198483 (-6.78%); split: -12.37%, +5.58% Scratch Memory Size: 3491840 -> 3410944 (-2.32%); split: -12.05%, +9.74% Max live registers: 493149 -> 487458 (-1.15%) Max dispatch width: 56936 -> 56856 (-0.14%); split: +0.06%, -0.20% Strange Brigade: Totals from 3769 (91.21% of 4132) affected shaders: Instrs: 1354476 -> 1321474 (-2.44%) Cycle count: 25351530 -> 25339190 (-0.05%); split: -1.64%, +1.59% Max live registers: 199057 -> 193656 (-2.71%) Max dispatch width: 30272 -> 30240 (-0.11%) Witcher 3: Totals from 25 (2.40% of 1041) affected shaders: Instrs: 24621 -> 24606 (-0.06%) Cycle count: 2218793 -> 2217503 (-0.06%); split: -0.11%, +0.05% Max live registers: 1963 -> 1955 (-0.41%) LNL results: Assassin's Creed Valhalla: Totals from 1928 (98.02% of 1967) affected shaders: Instrs: 856107 -> 835756 (-2.38%); split: -2.48%, +0.11% Subgroup size: 41264 -> 41280 (+0.04%) Cycle count: 64606590 -> 62371700 (-3.46%); split: -5.57%, +2.11% Spill count: 915 -> 669 (-26.89%); split: -32.79%, +5.90% Fill count: 2414 -> 1617 (-33.02%); split: -36.62%, +3.60% Scratch Memory Size: 62464 -> 44032 (-29.51%); split: -36.07%, +6.56% Max live registers: 205483 -> 202192 (-1.60%) Cyberpunk 2077: Totals from 1177 (96.40% of 1221) affected shaders: Instrs: 682237 -> 678931 (-0.48%); split: -0.51%, +0.03% Subgroup size: 24912 -> 24944 (+0.13%) Cycle count: 24355928 -> 25089292 (+3.01%); split: -0.80%, +3.81% Spill count: 8 -> 3 (-62.50%) Fill count: 6 -> 3 (-50.00%) Max live registers: 126922 -> 125472 (-1.14%) Dota2: Totals from 428 (32.47% of 1318) affected shaders: Instrs: 89355 -> 89740 (+0.43%) Cycle count: 1152412 -> 1152706 (+0.03%); split: -0.52%, +0.55% Max live registers: 32863 -> 32847 (-0.05%) Fortnite: Totals from 5354 (81.72% of 6552) affected shaders: Instrs: 4135059 -> 4239015 (+2.51%); split: -0.01%, +2.53% Cycle count: 132557506 -> 132427302 (-0.10%); split: -0.75%, +0.65% Spill count: 7144 -> 7234 (+1.26%); split: -0.46%, +1.72% Fill count: 12086 -> 12403 (+2.62%); split: -0.73%, +3.35% Scratch Memory Size: 600064 -> 604160 (+0.68%); split: -1.02%, +1.71% Hitman3: Totals from 4912 (97.09% of 5059) affected shaders: Instrs: 2952124 -> 2916824 (-1.20%); split: -1.20%, +0.00% Cycle count: 179985656 -> 189175250 (+5.11%); split: -2.44%, +7.55% Spill count: 3739 -> 3136 (-16.13%) Fill count: 10657 -> 9564 (-10.26%) Scratch Memory Size: 373760 -> 318464 (-14.79%) Max live registers: 597566 -> 589460 (-1.36%) Hogwarts Legacy: Totals from 1471 (96.33% of 1527) affected shaders: Instrs: 748749 -> 766214 (+2.33%); split: -0.71%, +3.05% Cycle count: 33301528 -> 34426308 (+3.38%); split: -1.30%, +4.68% Spill count: 3278 -> 3070 (-6.35%); split: -8.30%, +1.95% Fill count: 4553 -> 4097 (-10.02%); split: -10.85%, +0.83% Scratch Memory Size: 251904 -> 217088 (-13.82%) Max live registers: 168911 -> 168106 (-0.48%); split: -0.59%, +0.12% Metro Exodus: Totals from 18356 (49.81% of 36854) affected shaders: Instrs: 7559386 -> 7621591 (+0.82%); split: -0.01%, +0.83% Cycle count: 195240612 -> 196455186 (+0.62%); split: -1.22%, +1.84% Spill count: 595 -> 546 (-8.24%) Fill count: 1604 -> 1408 (-12.22%) Max live registers: 2086937 -> 2086933 (-0.00%) Red Dead Redemption 2: Totals from 4171 (79.31% of 5259) affected shaders: Instrs: 2619392 -> 2719587 (+3.83%); split: -0.00%, +3.83% Subgroup size: 86416 -> 86432 (+0.02%) Cycle count: 8542836160 -> 8531976886 (-0.13%); split: -0.65%, +0.53% Fill count: 12949 -> 12970 (+0.16%); split: -0.43%, +0.59% Scratch Memory Size: 401408 -> 385024 (-4.08%) Spiderman Remastered: Totals from 6639 (98.94% of 6710) affected shaders: Instrs: 6877980 -> 6800592 (-1.13%); split: -3.11%, +1.98% Cycle count: 282183352210 -> 282100051824 (-0.03%); split: -0.62%, +0.59% Spill count: 63147 -> 64218 (+1.70%); split: -7.12%, +8.82% Fill count: 184931 -> 175591 (-5.05%); split: -10.81%, +5.76% Scratch Memory Size: 5318656 -> 5970944 (+12.26%); split: -5.91%, +18.17% Max live registers: 918240 -> 906604 (-1.27%) Strange Brigade: Totals from 3675 (92.24% of 3984) affected shaders: Instrs: 1462231 -> 1429345 (-2.25%); split: -2.25%, +0.00% Cycle count: 37404050 -> 37345292 (-0.16%); split: -1.25%, +1.09% Max live registers: 361849 -> 351265 (-2.92%) Witcher 3: Totals from 13 (46.43% of 28) affected shaders: Instrs: 593 -> 660 (+11.30%) Cycle count: 28302 -> 28714 (+1.46%) Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io> Reviewed-by: Caio Oliveira <caio.oliveira@intel.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28199>
2024-02-29 20:51:50 +02:00
'brw_opt_address_reg_load.cpp',
'brw_opt_algebraic.cpp',
'brw_opt_bank_conflicts.cpp',
'brw_opt_cmod_propagation.cpp',
'brw_opt_combine_constants.cpp',
'brw_opt_copy_propagation.cpp',
'brw_opt_cse.cpp',
'brw_opt_dead_code_eliminate.cpp',
'brw_opt_register_coalesce.cpp',
'brw_opt_saturate_propagation.cpp',
brw: Combine convergent texture buffer fetches into fewer loads Borderlands 3 (both DX11 and DX12 renderers) have a common pattern across many shaders: con 32x4 %510 = (uint32)txf %2 (handle), %1191 (0x10) (coord), %1 (0x0) (lod), 0 (texture) con 32x4 %512 = (uint32)txf %2 (handle), %1511 (0x11) (coord), %1 (0x0) (lod), 0 (texture) ... con 32x4 %550 = (uint32)txf %2 (handle), %1549 (0x25) (coord), %1 (0x0) (lod), 0 (texture) con 32x4 %552 = (uint32)txf %2 (handle), %1551 (0x26) (coord), %1 (0x0) (lod), 0 (texture) A single basic block contains piles of texelFetches from a 1D buffer texture, with constant coordinates. In most cases, only the .x channel of the result is read. So we have something on the order of 28 sampler messages, each asking for...a single uint32_t scalar value. Because our sampler doesn't have any support for convergent block loads (like the untyped LSC transpose messages for SSBOs)...this means we were emitting SIMD8/16 (or SIMD16/32 on Xe2) sampler messages for every single scalar, replicating what's effectively a SIMD1 value to the entire register. This is hugely wasteful, both in terms of register pressure, and also in back-and-forth sending and receiving memory messages. The good news is we can take advantage of our explicit SIMD model to handle this more efficiently. This patch adds a new optimization pass that detects a series of SHADER_OPCODE_TXF_LOGICAL, in the same basic block, with constant offsets, from the same texture. It constructs a new divergent coordinate where each channel is one of the constants (i.e <10, 11, 12, ..., 26> in the above example). It issues a new NoMask divergent texel fetch which loads N useful channels in one go, and replaces the rest with expansion MOVs that splat the SIMD1 result back to the full SIMD width. (These get copy propagated away.) We can pick the SIMD size of the load independently of the native shader width as well. On Xe2, those 28 convergent loads become a single SIMD32 ld message. On earlier hardware, we use 2 SIMD16 messages. Or we can use a smaller size when there aren't many to combine. In fossil-db, this cuts 27% of send messages in affected shaders, 3-6% of cycles, 2-3% of instructions, and 8-12% of live registers. On A770, this improves performance of Borderlands 3 by roughly 2.5-3.5%. Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32573>
2024-12-09 13:25:18 -08:00
'brw_opt_txf_combiner.cpp',
'brw_opt_virtual_grfs.cpp',
'brw_packed_float.c',
'brw_print.cpp',
'brw_prim.h',
'brw_private.h',
'brw_reg.cpp',
'brw_reg.h',
'brw_reg_allocate.cpp',
'brw_reg_type.c',
'brw_reg_type.h',
'brw_rt.h',
'brw_schedule_instructions.cpp',
'brw_simd_selection.cpp',
'brw_spirv.c',
'brw_thread_payload.cpp',
'brw_thread_payload.h',
'brw_validate.cpp',
'brw_vue_map.c',
'brw_workaround.cpp',
)
brw_device_sha1_gen_src = custom_target('brw_device_sha1_gen.c',
input : ['brw_device_sha1_gen_c.py', '../dev/intel_device_info.py'],
output : ['brw_device_sha1_gen.c'],
command : [prog_python, '@INPUT0@', '--outdir', meson.current_build_dir()])
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>
2024-05-01 21:02:13 -07:00
brw_nir_lower_fsign = custom_target(
'brw_nir_lower_fsign.c',
input : 'brw_nir_lower_fsign.py',
output : 'brw_nir_lower_fsign.c',
command : [
prog_python, '@INPUT@', '-p', dir_compiler_nir,
],
depend_files : nir_algebraic_depends,
capture : true,
)
brw_nir_trig = custom_target(
'brw_nir_trig_workarounds.c',
input : 'brw_nir_trig_workarounds.py',
output : 'brw_nir_trig_workarounds.c',
command : [
prog_python, '@INPUT@', '-p', dir_compiler_nir,
],
depend_files : nir_algebraic_depends,
capture : true,
)
libintel_compiler_brw = static_library(
'intel_compiler',
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>
2024-05-01 21:02:13 -07:00
[libintel_compiler_brw_files, intel_nir_files, brw_nir_lower_fsign, brw_nir_trig, ir_expression_operation_h, [brw_device_sha1_gen_src]],
include_directories : [inc_include, inc_src, inc_intel],
c_args : [no_override_init_args],
gnu_symbol_visibility : 'hidden',
dependencies : [idep_nir_headers, idep_mesautil, idep_intel_dev],
build_by_default : false,
)
idep_intel_compiler_brw = declare_dependency(
link_with : [libintel_compiler_brw],
dependencies : [
idep_nir,
idep_mesautil,
idep_vtn,
],
)
# For now this tool is only going to be used by Anv
if with_intel_bvh_grl
if get_option('intel-clc') == 'system' or get_option('precomp-compiler') == 'system'
prog_intel_clc = find_program('intel_clc', native : true)
dep_prog_intel_clc = []
elif with_intel_clc
prog_intel_clc = executable(
'intel_clc',
[
'intel_clc.c',
'brw_kernel.c',
# Use just the nir_options part of ELK instead of fully linking.
'elk/elk_nir_options.h',
'elk/elk_nir_options.c',
'elk/elk_spirv.c',
],
link_with : [libisl],
include_directories : [inc_include, inc_src, inc_intel],
c_args : [pre_args, no_override_init_args],
link_args : [ld_args_build_id],
dependencies : [idep_nir, idep_mesaclc, idep_mesautil, idep_intel_dev,
idep_intel_compiler_brw],
# If we can run host binaries directly, just build intel_clc for the host.
# Most commonly this happens when doing a cross compile from an x86_64 build
# machine to an x86 host
native : not meson.can_run_host_binaries(),
install : get_option('install-intel-clc') or get_option('install-precomp-compiler'),
)
dep_prog_intel_clc = [prog_intel_clc]
endif
else
dep_prog_intel_clc = []
endif
if with_tests
test(
'intel_compiler_brw_tests',
executable(
'intel_compiler_brw_tests',
files(
'test_eu_compact.cpp',
'test_eu_validate.cpp',
'test_lower_scoreboard.cpp',
'test_opt_cmod_propagation.cpp',
'test_opt_combine_constants.cpp',
'test_opt_copy_propagation.cpp',
'test_opt_cse.cpp',
'test_opt_saturate_propagation.cpp',
'test_simd_selection.cpp',
'test_vf_float_conversions.cpp',
),
ir_expression_operation_h,
include_directories : [inc_include, inc_src, inc_intel],
link_with : libisl,
dependencies : [idep_gtest, idep_nir, idep_mesautil, idep_intel_dev,
idep_intel_compiler_brw],
),
suite : ['intel'],
protocol : 'gtest',
)
endif
if with_intel_tools
bison_command = []
if yacc_is_bison
bison_command = [
prog_bison, '@INPUT@', '--defines=@OUTPUT1@',
'--output=@OUTPUT0@'
]
else
bison_command = [
prog_bison, '-H', '@OUTPUT1@',
'-o', '@OUTPUT0@', '@INPUT@'
]
endif
brw_gram_tab = custom_target(
'brw_gram.tab.[ch]',
input : 'brw_gram.y',
output : ['brw_gram.tab.c', 'brw_gram.tab.h'],
command : bison_command
)
brw_lex_yy_c = custom_target(
'brw_lex.yy.c',
input : 'brw_lex.l',
output : 'brw_lex.yy.c',
command : [prog_flex, '-o', '@OUTPUT@', '@INPUT@']
)
brw_asm_deps = [
dep_thread,
idep_intel_compiler_brw,
idep_intel_dev,
idep_mesautil,
]
brw_asm = static_library(
'brw_asm',
['brw_asm.c', brw_gram_tab[0], brw_gram_tab[1], brw_lex_yy_c],
dependencies : brw_asm_deps,
include_directories : [inc_include, inc_src, inc_intel],
c_args : [no_override_init_args],
gnu_symbol_visibility : 'hidden',
build_by_default : false,
)
idep_brw_asm = declare_dependency(
link_with : brw_asm,
dependencies : brw_asm_deps,
)
brw_asm_tool = executable(
'brw_asm',
['brw_asm_tool.c'],
dependencies : idep_brw_asm,
include_directories : [inc_include, inc_src, inc_intel],
c_args : [no_override_init_args],
gnu_symbol_visibility : 'hidden',
install : true
)
asm_testcases = [
['skl', 'gfx9'],
['icl', 'gfx11'],
['tgl', 'gfx12'],
['dg2', 'gfx12.5'],
]
test_runner = find_program('tests/run-test.py')
foreach testcase : asm_testcases
_gen_name = testcase[0]
_gen_num = testcase[1]
_gen_folder = join_paths(meson.current_source_dir(), 'tests',
_gen_num.replace('gfx', 'gen'))
test(
'brw_asm_' + _gen_num, test_runner,
args : [
'--brw_asm', brw_asm_tool,
'--gen_name', _gen_name,
'--gen_folder', _gen_folder,
],
suite : 'intel',
)
endforeach
brw_disasm_tool = executable(
'brw_disasm',
files('brw_disasm_tool.c'),
dependencies : [idep_mesautil, dep_thread, idep_intel_dev,
idep_intel_compiler_brw],
include_directories : [inc_include, inc_src, inc_intel],
c_args : [no_override_init_args],
gnu_symbol_visibility : 'hidden',
install : true
)
endif
if with_intel_elk
subdir('elk')
else
idep_intel_compiler_elk = null_dep
endif