search

mooman219 / fontdue

The fastest font renderer in the world, written in pure rust.
Apache License 2.0
1.44k stars 72 forks source link

Microoptimization: Make some functions const #89

Closed deeprobin closed 2 years ago

deeprobin commented 3 years ago

Have marked a few functions as const, if that is possible in the stable version.

So this does not include functions that contain floating point arithmetic or a mut reference.

criterion Benchmarks

Before (Branch master)

layout/fontdue/100      time:   [3.9368 us 3.9979 us 4.0702 us]
Found 14 outliers among 250 measurements (5.60%)
  10 (4.00%) high mild
  4 (1.60%) high severe
layout/fontdue/500      time:   [20.980 us 21.243 us 21.535 us]
Found 20 outliers among 250 measurements (8.00%)
  14 (5.60%) high mild
  6 (2.40%) high severe
layout/fontdue/1000     time:   [42.722 us 43.300 us 43.956 us]
Found 27 outliers among 250 measurements (10.80%)
  12 (4.80%) high mild
  15 (6.00%) high severelayout/glyph_brush_layout/100
                        time:   [14.262 us 14.453 us 14.680 us]
Found 16 outliers among 250 measurements (6.40%)
  9 (3.60%) high mild
  7 (2.80%) high severe
layout/glyph_brush_layout/500
                        time:   [74.289 us 75.880 us 77.730 us]
Found 10 outliers among 250 measurements (4.00%)
  7 (2.80%) high mild
  3 (1.20%) high severe
Benchmarking layout/glyph_brush_layout/1000: Warming up for 3.0000 s
Warning: Unable to complete 250 samples in 4.0s. You may wish to increase target time to 4.6s, enable flat sampling, or reduce sample count to 160.
layout/glyph_brush_layout/1000
                        time:   [144.82 us 147.18 us 149.85 us]
Found 27 outliers among 250 measurements (10.80%)
  16 (6.40%) high mild
  11 (4.40%) high severe     Running unittests (target\release\deps\load-1ebd4c7446ed2dd1.exe)
WARNING: HTML report generation will become a non-default optional feature in Criterion.rs 0.4.0.
This feature is being moved to cargo-criterion (https://github.com/bheisler/cargo-criterion) and will be optional in a future version of Criterion.rs. To silence this warning, either switch to cargo-criterion or enable the 'html_reports' feature in your Cargo.toml.Gnuplot not found, using plotters backend
load/rusttype roboto    time:   [894.43 ns 909.01 ns 924.52 ns]
Found 12 outliers among 100 measurements (12.00%)
  6 (6.00%) high mild
  6 (6.00%) high severe
load/ab_glyph roboto    time:   [1.0128 us 1.0385 us 1.0705 us]
Found 9 outliers among 100 measurements (9.00%)
  2 (2.00%) high mild
  7 (7.00%) high severe
load/fontdue roboto     time:   [4.1027 ms 4.1663 ms 4.2406 ms]
Found 9 outliers among 100 measurements (9.00%)
  6 (6.00%) high mild
  3 (3.00%) high severe     Running unittests (target\release\deps\rasterize-6f6795b8152ccfd3.exe)
WARNING: HTML report generation will become a non-default optional feature in Criterion.rs 0.4.0.
This feature is being moved to cargo-criterion (https://github.com/bheisler/cargo-criterion) and will be optional in a future version of Criterion.rs. To silence this warning, either switch to cargo-criterion or enable the 'html_reports' feature in your Cargo.toml.Gnuplot not found, using plotters backend
rasterize/rusttype truetype 10px
                        time:   [52.563 us 53.613 us 54.892 us]
Found 12 outliers among 100 measurements (12.00%)
  5 (5.00%) high mild
  7 (7.00%) high severe
rasterize/rusttype truetype 20px
                        time:   [70.946 us 71.995 us 73.158 us]
Found 7 outliers among 100 measurements (7.00%)
  5 (5.00%) high mild
  2 (2.00%) high severe
rasterize/rusttype truetype 40px
                        time:   [121.94 us 123.44 us 125.13 us]
Found 12 outliers among 100 measurements (12.00%)
  7 (7.00%) high mild
  5 (5.00%) high severe
rasterize/rusttype truetype 80px
                        time:   [254.45 us 258.14 us 262.78 us]
Found 9 outliers among 100 measurements (9.00%)
  5 (5.00%) high mild
  4 (4.00%) high severe
rasterize/rusttype truetype 160px
                        time:   [736.76 us 743.05 us 749.93 us]
Found 10 outliers among 100 measurements (10.00%)
  9 (9.00%) high mild
  1 (1.00%) high severe
Benchmarking rasterize/rusttype truetype 200px: Warming up for 3.0000 s
Warning: Unable to complete 100 samples in 4.0s. You may wish to increase target time to 5.5s, enable flat sampling, or reduce sample count to 60.
rasterize/rusttype truetype 200px
                        time:   [1.0651 ms 1.0735 ms 1.0841 ms]
Found 6 outliers among 100 measurements (6.00%)
  4 (4.00%) high mild
  2 (2.00%) high severe
rasterize/rusttype opentype 10px
                        time:   [73.679 us 75.298 us 77.094 us]
Found 6 outliers among 100 measurements (6.00%)
  4 (4.00%) high mild
  2 (2.00%) high severe
rasterize/rusttype opentype 20px
                        time:   [92.924 us 94.812 us 96.898 us]
Found 8 outliers among 100 measurements (8.00%)
  7 (7.00%) high mild
  1 (1.00%) high severe
rasterize/rusttype opentype 40px
                        time:   [144.18 us 147.56 us 151.42 us]
Found 6 outliers among 100 measurements (6.00%)
  5 (5.00%) high mild
  1 (1.00%) high severe
rasterize/rusttype opentype 80px
                        time:   [287.16 us 292.25 us 298.09 us]
Found 12 outliers among 100 measurements (12.00%)
  4 (4.00%) high mild
  8 (8.00%) high severe
Benchmarking rasterize/rusttype opentype 160px: Warming up for 3.0000 s
Warning: Unable to complete 100 samples in 4.0s. You may wish to increase target time to 4.0s, enable flat sampling, or reduce sample count to 70.
rasterize/rusttype opentype 160px
                        time:   [781.97 us 791.54 us 802.82 us]
Found 10 outliers among 100 measurements (10.00%)
  6 (6.00%) high mild
  4 (4.00%) high severe
Benchmarking rasterize/rusttype opentype 200px: Warming up for 3.0000 s
Warning: Unable to complete 100 samples in 4.0s. You may wish to increase target time to 5.8s, enable flat sampling, or reduce sample count to 50.
rasterize/rusttype opentype 200px
                        time:   [1.1246 ms 1.1375 ms 1.1534 ms]
Found 11 outliers among 100 measurements (11.00%)
  7 (7.00%) high mild
  4 (4.00%) high severe
rasterize/ab_glyph truetype 10px
                        time:   [70.340 us 71.803 us 73.616 us]
Found 5 outliers among 100 measurements (5.00%)
  4 (4.00%) high mild
  1 (1.00%) high severe
rasterize/ab_glyph truetype 20px
                        time:   [88.384 us 89.413 us 90.523 us]
Found 7 outliers among 100 measurements (7.00%)
  4 (4.00%) high mild
  3 (3.00%) high severe
rasterize/ab_glyph truetype 40px
                        time:   [142.69 us 145.48 us 148.57 us]
Found 7 outliers among 100 measurements (7.00%)
  3 (3.00%) high mild
  4 (4.00%) high severe
rasterize/ab_glyph truetype 80px
                        time:   [282.32 us 287.58 us 294.43 us]
Found 5 outliers among 100 measurements (5.00%)
  4 (4.00%) high mild
  1 (1.00%) high severe
rasterize/ab_glyph truetype 160px
                        time:   [769.58 us 778.66 us 788.26 us]
Found 4 outliers among 100 measurements (4.00%)
  3 (3.00%) high mild
  1 (1.00%) high severe
Benchmarking rasterize/ab_glyph truetype 200px: Warming up for 3.0000 s
Warning: Unable to complete 100 samples in 4.0s. You may wish to increase target time to 5.7s, enable flat sampling, or reduce sample count to 50.
rasterize/ab_glyph truetype 200px
                        time:   [1.0970 ms 1.1072 ms 1.1189 ms]
Found 3 outliers among 100 measurements (3.00%)
  2 (2.00%) high mild
  1 (1.00%) high severe
rasterize/ab_glyph opentype 10px
                        time:   [90.433 us 93.300 us 96.781 us]
Found 4 outliers among 100 measurements (4.00%)
  1 (1.00%) high mild
  3 (3.00%) high severe
rasterize/ab_glyph opentype 20px
                        time:   [83.210 us 84.901 us 87.019 us]
Found 11 outliers among 100 measurements (11.00%)
  6 (6.00%) high mild
  5 (5.00%) high severe
rasterize/ab_glyph opentype 40px
                        time:   [135.41 us 137.98 us 140.95 us]
Found 7 outliers among 100 measurements (7.00%)
  3 (3.00%) high mild
  4 (4.00%) high severe
rasterize/ab_glyph opentype 80px
                        time:   [279.60 us 284.92 us 291.71 us]
Found 7 outliers among 100 measurements (7.00%)
  3 (3.00%) high mild
  4 (4.00%) high severe
rasterize/ab_glyph opentype 160px
                        time:   [774.97 us 782.57 us 791.16 us]
Found 8 outliers among 100 measurements (8.00%)
  4 (4.00%) high mild
  4 (4.00%) high severe
Benchmarking rasterize/ab_glyph opentype 200px: Warming up for 3.0000 s
Warning: Unable to complete 100 samples in 4.0s. You may wish to increase target time to 5.7s, enable flat sampling, or reduce sample count to 50.
rasterize/ab_glyph opentype 200px
                        time:   [1.1276 ms 1.1442 ms 1.1628 ms]
Found 6 outliers among 100 measurements (6.00%)
  4 (4.00%) high mild
  2 (2.00%) high severe
rasterize/fontdue truetype 10px
                        time:   [13.000 us 13.315 us 13.692 us]
Found 7 outliers among 100 measurements (7.00%)
  2 (2.00%) high mild
  5 (5.00%) high severe
rasterize/fontdue truetype 20px
                        time:   [24.334 us 24.611 us 24.923 us]
Found 8 outliers among 100 measurements (8.00%)
  6 (6.00%) high mild
  2 (2.00%) high severe
rasterize/fontdue truetype 40px
                        time:   [26.627 us 27.122 us 27.716 us]
Found 11 outliers among 100 measurements (11.00%)
  3 (3.00%) high mild
  8 (8.00%) high severe
rasterize/fontdue truetype 80px
                        time:   [57.133 us 58.759 us 60.594 us]
Found 8 outliers among 100 measurements (8.00%)
  6 (6.00%) high mild
  2 (2.00%) high severe
rasterize/fontdue truetype 160px
                        time:   [165.12 us 171.35 us 178.09 us]
Found 3 outliers among 100 measurements (3.00%)
  2 (2.00%) high mild
  1 (1.00%) high severe
rasterize/fontdue truetype 200px
                        time:   [220.03 us 224.26 us 228.85 us]
Found 2 outliers among 100 measurements (2.00%)
  2 (2.00%) high mild
rasterize/fontdue opentype 10px
                        time:   [14.551 us 14.901 us 15.325 us]
Found 11 outliers among 100 measurements (11.00%)
  3 (3.00%) high mild
  8 (8.00%) high severe
rasterize/fontdue opentype 20px
                        time:   [17.753 us 18.219 us 18.799 us]
Found 11 outliers among 100 measurements (11.00%)
  5 (5.00%) high mild
  6 (6.00%) high severe
rasterize/fontdue opentype 40px
                        time:   [30.263 us 31.043 us 31.994 us]
Found 5 outliers among 100 measurements (5.00%)
  4 (4.00%) high mild
  1 (1.00%) high severe
rasterize/fontdue opentype 80px
                        time:   [53.996 us 55.609 us 57.353 us]
Found 9 outliers among 100 measurements (9.00%)
  4 (4.00%) high mild
  5 (5.00%) high severe
rasterize/fontdue opentype 160px
                        time:   [165.43 us 171.82 us 178.74 us]
Found 5 outliers among 100 measurements (5.00%)
  4 (4.00%) high mild
  1 (1.00%) high severe
rasterize/fontdue opentype 200px
                        time:   [206.67 us 210.05 us 213.75 us]
Found 5 outliers among 100 measurements (5.00%)
  2 (2.00%) high mild
  3 (3.00%) high severe

After (Branch code-optimization)

layout/fontdue/100      time:   [4.3643 us 4.4218 us 4.4931 us]
Found 29 outliers among 250 measurements (11.60%)
  16 (6.40%) high mild
  13 (5.20%) high severe
layout/fontdue/500      time:   [23.136 us 23.396 us 23.681 us]
Found 18 outliers among 250 measurements (7.20%)
  12 (4.80%) high mild
  6 (2.40%) high severe
layout/fontdue/1000     time:   [46.804 us 47.228 us 47.713 us]
Found 19 outliers among 250 measurements (7.60%)
  15 (6.00%) high mild
  4 (1.60%) high severelayout/glyph_brush_layout/100
                        time:   [14.549 us 14.788 us 15.061 us]
Found 31 outliers among 250 measurements (12.40%)
  15 (6.00%) high mild
  16 (6.40%) high severe
layout/glyph_brush_layout/500
                        time:   [73.153 us 73.955 us 74.840 us]
Found 21 outliers among 250 measurements (8.40%)
  13 (5.20%) high mild
  8 (3.20%) high severe
Benchmarking layout/glyph_brush_layout/1000: Warming up for 3.0000 s
Warning: Unable to complete 250 samples in 4.0s. You may wish to increase target time to 4.7s, enable flat sampling, or reduce sample count to 160.
layout/glyph_brush_layout/1000
                        time:   [145.95 us 148.79 us 152.17 us]
Found 31 outliers among 250 measurements (12.40%)
  9 (3.60%) high mild
  22 (8.80%) high severe     Running unittests (target\release\deps\load-1ebd4c7446ed2dd1.exe)
WARNING: HTML report generation will become a non-default optional feature in Criterion.rs 0.4.0.
This feature is being moved to cargo-criterion (https://github.com/bheisler/cargo-criterion) and will be optional in a future version of Criterion.rs. To silence this warning, either switch to cargo-criterion or enable the 'html_reports' feature in your Cargo.toml.Gnuplot not found, using plotters backend
load/rusttype roboto    time:   [881.63 ns 900.26 ns 922.12 ns]
Found 9 outliers among 100 measurements (9.00%)
  4 (4.00%) high mild
  5 (5.00%) high severe
load/ab_glyph roboto    time:   [990.41 ns 1.0091 us 1.0316 us]
Found 10 outliers among 100 measurements (10.00%)
  5 (5.00%) high mild
  5 (5.00%) high severe
load/fontdue roboto     time:   [4.4020 ms 4.4644 ms 4.5320 ms]
Found 8 outliers among 100 measurements (8.00%)
  6 (6.00%) high mild
  2 (2.00%) high severe     Running unittests (target\release\deps\rasterize-6f6795b8152ccfd3.exe)
WARNING: HTML report generation will become a non-default optional feature in Criterion.rs 0.4.0.
This feature is being moved to cargo-criterion (https://github.com/bheisler/cargo-criterion) and will be optional in a future version of Criterion.rs. To silence this warning, either switch to cargo-criterion or enable the 'html_reports' feature in your Cargo.toml.Gnuplot not found, using plotters backend
rasterize/rusttype truetype 10px
                        time:   [51.405 us 52.240 us 53.267 us]
Found 6 outliers among 100 measurements (6.00%)
  1 (1.00%) high mild
  5 (5.00%) high severe
rasterize/rusttype truetype 20px
                        time:   [71.044 us 71.931 us 72.982 us]
Found 8 outliers among 100 measurements (8.00%)
  4 (4.00%) high mild
  4 (4.00%) high severe
rasterize/rusttype truetype 40px
                        time:   [125.07 us 127.42 us 130.69 us]
Found 4 outliers among 100 measurements (4.00%)
  1 (1.00%) high mild
  3 (3.00%) high severe
rasterize/rusttype truetype 80px
                        time:   [256.21 us 259.25 us 263.01 us]
Found 5 outliers among 100 measurements (5.00%)
  4 (4.00%) high mild
  1 (1.00%) high severe
rasterize/rusttype truetype 160px
                        time:   [734.03 us 738.59 us 743.83 us]
Found 10 outliers among 100 measurements (10.00%)
  4 (4.00%) high mild
  6 (6.00%) high severe
Benchmarking rasterize/rusttype truetype 200px: Warming up for 3.0000 s
Warning: Unable to complete 100 samples in 4.0s. You may wish to increase target time to 5.4s, enable flat sampling, or reduce sample count to 60.
rasterize/rusttype truetype 200px
                        time:   [1.0666 ms 1.0737 ms 1.0818 ms]
Found 6 outliers among 100 measurements (6.00%)
  4 (4.00%) high mild
  2 (2.00%) high severe
rasterize/rusttype opentype 10px
                        time:   [72.672 us 74.957 us 77.553 us]
Found 12 outliers among 100 measurements (12.00%)
  6 (6.00%) high mild
  6 (6.00%) high severe
rasterize/rusttype opentype 20px
                        time:   [92.580 us 93.683 us 94.929 us]
Found 9 outliers among 100 measurements (9.00%)
  3 (3.00%) high mild
  6 (6.00%) high severe
rasterize/rusttype opentype 40px
                        time:   [143.39 us 147.03 us 151.34 us]
Found 13 outliers among 100 measurements (13.00%)
  9 (9.00%) high mild
  4 (4.00%) high severe
rasterize/rusttype opentype 80px
                        time:   [287.46 us 291.36 us 295.52 us]
Found 12 outliers among 100 measurements (12.00%)
  6 (6.00%) high mild
  6 (6.00%) high severe
rasterize/rusttype opentype 160px
                        time:   [778.09 us 785.46 us 794.68 us]
Found 10 outliers among 100 measurements (10.00%)
  5 (5.00%) high mild
  5 (5.00%) high severe
Benchmarking rasterize/rusttype opentype 200px: Warming up for 3.0000 s
Warning: Unable to complete 100 samples in 4.0s. You may wish to increase target time to 5.7s, enable flat sampling, or reduce sample count to 50.
rasterize/rusttype opentype 200px
                        time:   [1.1217 ms 1.1321 ms 1.1455 ms]
Found 5 outliers among 100 measurements (5.00%)
  3 (3.00%) high mild
  2 (2.00%) high severe
rasterize/ab_glyph truetype 10px
                        time:   [70.629 us 72.246 us 74.203 us]
Found 6 outliers among 100 measurements (6.00%)
  4 (4.00%) high mild
  2 (2.00%) high severe
rasterize/ab_glyph truetype 20px
                        time:   [90.244 us 92.038 us 94.215 us]
Found 6 outliers among 100 measurements (6.00%)
  3 (3.00%) high mild
  3 (3.00%) high severe
rasterize/ab_glyph truetype 40px
                        time:   [140.04 us 141.98 us 144.43 us]
Found 15 outliers among 100 measurements (15.00%)
  6 (6.00%) high mild
  9 (9.00%) high severe
rasterize/ab_glyph truetype 80px
                        time:   [282.09 us 285.94 us 290.44 us]
Found 10 outliers among 100 measurements (10.00%)
  7 (7.00%) high mild
  3 (3.00%) high severe
rasterize/ab_glyph truetype 160px
                        time:   [747.87 us 753.43 us 759.84 us]
Found 7 outliers among 100 measurements (7.00%)
  6 (6.00%) high mild
  1 (1.00%) high severe
Benchmarking rasterize/ab_glyph truetype 200px: Warming up for 3.0000 s
Warning: Unable to complete 100 samples in 4.0s. You may wish to increase target time to 5.5s, enable flat sampling, or reduce sample count to 50.
rasterize/ab_glyph truetype 200px
                        time:   [1.0902 ms 1.0992 ms 1.1087 ms]
Found 3 outliers among 100 measurements (3.00%)
  3 (3.00%) high mild
rasterize/ab_glyph opentype 10px
                        time:   [63.984 us 65.677 us 67.730 us]
Found 8 outliers among 100 measurements (8.00%)
  4 (4.00%) high mild
  4 (4.00%) high severe
rasterize/ab_glyph opentype 20px
                        time:   [83.610 us 85.011 us 86.719 us]
Found 13 outliers among 100 measurements (13.00%)
  9 (9.00%) high mild
  4 (4.00%) high severe
rasterize/ab_glyph opentype 40px
                        time:   [132.84 us 134.58 us 136.66 us]
Found 7 outliers among 100 measurements (7.00%)
  6 (6.00%) high mild
  1 (1.00%) high severe
rasterize/ab_glyph opentype 80px
                        time:   [284.74 us 292.98 us 302.85 us]
Found 9 outliers among 100 measurements (9.00%)
  3 (3.00%) high mild
  6 (6.00%) high severe
rasterize/ab_glyph opentype 160px
                        time:   [765.39 us 770.66 us 776.43 us]
Found 8 outliers among 100 measurements (8.00%)
  6 (6.00%) high mild
  2 (2.00%) high severe
Benchmarking rasterize/ab_glyph opentype 200px: Warming up for 3.0000 s
Warning: Unable to complete 100 samples in 4.0s. You may wish to increase target time to 5.7s, enable flat sampling, or reduce sample count to 50.
rasterize/ab_glyph opentype 200px
                        time:   [1.1085 ms 1.1182 ms 1.1300 ms]
Found 6 outliers among 100 measurements (6.00%)
  2 (2.00%) high mild
  4 (4.00%) high severe
rasterize/fontdue truetype 10px
                        time:   [35.574 us 36.151 us 36.750 us]
Found 10 outliers among 100 measurements (10.00%)
  5 (5.00%) high mild
  5 (5.00%) high severe
rasterize/fontdue truetype 20px
                        time:   [20.003 us 20.317 us 20.706 us]
Found 8 outliers among 100 measurements (8.00%)
  6 (6.00%) high mild
  2 (2.00%) high severe
rasterize/fontdue truetype 40px
                        time:   [27.129 us 27.646 us 28.238 us]
Found 9 outliers among 100 measurements (9.00%)
  4 (4.00%) high mild
  5 (5.00%) high severe
rasterize/fontdue truetype 80px
                        time:   [60.512 us 62.464 us 64.809 us]
Found 7 outliers among 100 measurements (7.00%)
  1 (1.00%) high mild
  6 (6.00%) high severe
rasterize/fontdue truetype 160px
                        time:   [152.80 us 155.76 us 158.95 us]
Found 2 outliers among 100 measurements (2.00%)
  1 (1.00%) high mild
  1 (1.00%) high severe
rasterize/fontdue truetype 200px
                        time:   [213.14 us 216.81 us 220.91 us]
Found 3 outliers among 100 measurements (3.00%)
  3 (3.00%) high severe
rasterize/fontdue opentype 10px
                        time:   [11.939 us 12.163 us 12.434 us]
Found 10 outliers among 100 measurements (10.00%)
  4 (4.00%) high mild
  6 (6.00%) high severe
rasterize/fontdue opentype 20px
                        time:   [17.800 us 18.161 us 18.581 us]
Found 5 outliers among 100 measurements (5.00%)
  2 (2.00%) high mild
  3 (3.00%) high severe
rasterize/fontdue opentype 40px
                        time:   [29.286 us 29.723 us 30.208 us]
Found 8 outliers among 100 measurements (8.00%)
  4 (4.00%) high mild
  4 (4.00%) high severe
rasterize/fontdue opentype 80px
                        time:   [57.324 us 58.598 us 60.133 us]
Found 12 outliers among 100 measurements (12.00%)
  2 (2.00%) high mild
  10 (10.00%) high severe
rasterize/fontdue opentype 160px
                        time:   [149.25 us 152.62 us 156.81 us]
Found 10 outliers among 100 measurements (10.00%)
  5 (5.00%) high mild
  5 (5.00%) high severe
rasterize/fontdue opentype 200px
                        time:   [207.22 us 212.67 us 219.20 us]
Found 9 outliers among 100 measurements (9.00%)
  7 (7.00%) high mild
  2 (2.00%) high severe

Comparison

mooman219 commented 3 years ago

See CI failures

deeprobin commented 3 years ago

@mooman219 Please re-run the CI :)

deeprobin commented 3 years ago

@mooman219 Checks passed 👍🏼

mooman219 commented 3 years ago

The functions marked as const are used during font loading. The benchmarks you provided show a performance regression after the const changes load/fontdue.

deeprobin commented 3 years ago

The functions marked as const are used during font loading. The benchmarks you provided show a performance regression after the const changes load/fontdue.

Weird. const only ensures that the compiler tries to evaluate the method in compile time, otherwise nothing else should happen.

In my opinion, there should be no regression. Possibly something went wrong with my benchmark. Maybe it would be useful if you or someone else provides a benchmark.

References

mooman219 commented 3 years ago

Being totally honest, +/- 10% between runs with unchanged code is like really common and is up to the alignment of the planets, your os scheduling, and whatever mood lto is in. It may just be that the functions are already being evaluated as const.

deeprobin commented 3 years ago

What is the current status of this :smile:?

mooman219 commented 2 years ago

It can't hurt, I'll merge it.