tree 8d79857bff334286ea90603b1cc89ccf4c0fe691
parent b61a8273541dd26a4025669cf00f980791c538e9
author Nigel Tao <nigeltao@golang.org> 1611750990 +1100
committer Nigel Tao <nigeltao@golang.org> 1611786877 +1100

Have std/deflate read 64 bits once per inner loop

On a mid-range x86_64 laptop:

wuffs_deflate_decode_1k_full_init/clang9                   101MB/s ± 0%     92MB/s ± 2%   -9.22%  (p=0.008 n=5+5)
wuffs_deflate_decode_1k_part_init/clang9                   117MB/s ± 0%    107MB/s ± 3%   -8.00%  (p=0.008 n=5+5)
wuffs_deflate_decode_10k_full_init/clang9                  129MB/s ± 0%    140MB/s ± 1%   +8.36%  (p=0.008 n=5+5)
wuffs_deflate_decode_10k_part_init/clang9                  131MB/s ± 0%    142MB/s ± 1%   +8.59%  (p=0.008 n=5+5)
wuffs_deflate_decode_100k_just_one_read/clang9             158MB/s ± 0%    184MB/s ± 1%  +16.29%  (p=0.008 n=5+5)
wuffs_deflate_decode_100k_many_big_reads/clang9            133MB/s ± 0%    155MB/s ± 1%  +16.26%  (p=0.008 n=5+5)

wuffs_deflate_decode_1k_full_init/gcc10                   98.0MB/s ± 0%   97.7MB/s ± 0%     ~     (p=0.222 n=5+5)
wuffs_deflate_decode_1k_part_init/gcc10                    112MB/s ± 0%    111MB/s ± 0%   -1.04%  (p=0.008 n=5+5)
wuffs_deflate_decode_10k_full_init/gcc10                   138MB/s ± 0%    179MB/s ± 1%  +29.91%  (p=0.008 n=5+5)
wuffs_deflate_decode_10k_part_init/gcc10                   140MB/s ± 0%    183MB/s ± 1%  +30.52%  (p=0.008 n=5+5)
wuffs_deflate_decode_100k_just_one_read/gcc10              188MB/s ± 0%    211MB/s ± 1%  +12.32%  (p=0.008 n=5+5)
wuffs_deflate_decode_100k_many_big_reads/gcc10             156MB/s ± 0%    172MB/s ± 1%  +10.00%  (p=0.008 n=5+5)

mimic_deflate_decode_1k                                    128MB/s ± 0%    127MB/s ± 1%   -0.99%  (p=0.016 n=5+5)
mimic_deflate_decode_10k                                   143MB/s ± 0%    142MB/s ± 1%     ~     (p=0.056 n=5+5)
mimic_deflate_decode_100k_just_one_read                    176MB/s ± 0%    174MB/s ± 1%     ~     (p=0.151 n=5+5)
mimic_deflate_decode_100k_many_big_reads                   135MB/s ± 0%    134MB/s ± 1%     ~     (p=0.151 n=5+5)

wuffs_gzip_decode_10k/clang9                               134MB/s ± 1%    127MB/s ± 0%   -5.23%  (p=0.008 n=5+5)
wuffs_gzip_decode_100k/clang9                              150MB/s ± 0%    162MB/s ± 0%   +8.23%  (p=0.008 n=5+5)

wuffs_gzip_decode_10k/gcc10                                131MB/s ± 0%    152MB/s ± 0%  +16.36%  (p=0.008 n=5+5)
wuffs_gzip_decode_100k/gcc10                               159MB/s ± 2%    188MB/s ± 0%  +17.78%  (p=0.008 n=5+5)

mimic_gzip_decode_10k                                      121MB/s ± 0%    121MB/s ± 0%     ~     (p=0.841 n=5+5)
mimic_gzip_decode_100k                                     144MB/s ± 0%    143MB/s ± 0%   -0.11%  (p=0.008 n=5+5)

wuffs_zlib_decode_10k/clang9                               139MB/s ± 0%    135MB/s ± 0%   -2.36%  (p=0.008 n=5+5)
wuffs_zlib_decode_100k/clang9                              155MB/s ± 0%    176MB/s ± 0%  +13.83%  (p=0.008 n=5+5)

wuffs_zlib_decode_10k/gcc10                                141MB/s ± 0%    156MB/s ± 1%  +10.94%  (p=0.008 n=5+5)
wuffs_zlib_decode_100k/gcc10                               172MB/s ± 0%    193MB/s ± 1%  +12.27%  (p=0.016 n=4+5)

mimic_zlib_decode_10k                                      133MB/s ± 0%    132MB/s ± 0%   -1.29%  (p=0.008 n=5+5)
mimic_zlib_decode_100k                                     160MB/s ± 0%    158MB/s ± 0%   -1.02%  (p=0.008 n=5+5)

wuffs_png_decode_image_19k_8bpp/clang9                    94.7MB/s ± 0%  103.4MB/s ± 0%   +9.17%  (p=0.016 n=4+5)
wuffs_png_decode_image_40k_24bpp/clang9                    135MB/s ± 0%    146MB/s ± 0%   +8.03%  (p=0.008 n=5+5)
wuffs_png_decode_image_77k_8bpp/clang9                     351MB/s ± 0%    380MB/s ± 0%   +8.48%  (p=0.008 n=5+5)
wuffs_png_decode_image_552k_32bpp_ignore_checksum/clang9   320MB/s ± 1%    343MB/s ± 0%   +7.20%  (p=0.008 n=5+5)
wuffs_png_decode_image_552k_32bpp_verify_checksum/clang9   308MB/s ± 0%    328MB/s ± 0%   +6.59%  (p=0.008 n=5+5)
wuffs_png_decode_image_4002k_24bpp/clang9                  138MB/s ± 0%    149MB/s ± 0%   +8.02%  (p=0.016 n=5+4)

wuffs_png_decode_image_19k_8bpp/gcc10                      101MB/s ± 1%    116MB/s ± 0%  +14.61%  (p=0.008 n=5+5)
wuffs_png_decode_image_40k_24bpp/gcc10                     136MB/s ± 0%    154MB/s ± 0%  +13.10%  (p=0.008 n=5+5)
wuffs_png_decode_image_77k_8bpp/gcc10                      371MB/s ± 0%    426MB/s ± 0%  +14.80%  (p=0.008 n=5+5)
wuffs_png_decode_image_552k_32bpp_ignore_checksum/gcc10    333MB/s ± 0%    357MB/s ± 0%   +7.31%  (p=0.008 n=5+5)
wuffs_png_decode_image_552k_32bpp_verify_checksum/gcc10    318MB/s ± 0%    341MB/s ± 0%   +7.27%  (p=0.016 n=5+4)
wuffs_png_decode_image_4002k_24bpp/gcc10                   138MB/s ± 0%    157MB/s ± 0%  +13.59%  (p=0.008 n=5+5)

mimic_png_decode_image_19k_8bpp                           58.2MB/s ± 0%   58.1MB/s ± 0%     ~     (p=0.841 n=5+5)
mimic_png_decode_image_40k_24bpp                          73.3MB/s ± 0%   73.2MB/s ± 0%   -0.16%  (p=0.032 n=5+5)
mimic_png_decode_image_77k_8bpp                            177MB/s ± 0%    177MB/s ± 0%     ~     (p=0.690 n=5+5)
mimic_png_decode_image_552k_32bpp_ignore_checksum               skipped
mimic_png_decode_image_552k_32bpp_verify_checksum          146MB/s ± 0%    146MB/s ± 0%     ~     (p=0.690 n=5+5)
mimic_png_decode_image_4002k_24bpp                         104MB/s ± 0%    104MB/s ± 0%     ~     (p=0.841 n=5+5)