Nigel Tao
|
385c34930a
|
Check 8-byte-chunk copy distance precondition
|
2021-03-11 22:54:06 +11:00 |
|
Nigel Tao
|
4e3a005d28
|
Use 8-byte-chunk copy for decode_huffman_fast32
On a Raspberry Pi 4 (32-bit armv7l) with -march=native and -mfpu=neon
("native" means "armv8-a+crc+simd"):
wuffs_deflate_decode_1k_full_init/clang9 45.1MB/s ± 5% 47.2MB/s ± 6% ~ (p=0.151 n=5+5)
wuffs_deflate_decode_1k_part_init/clang9 60.6MB/s ± 0% 64.2MB/s ± 0% +5.94% (p=0.008 n=5+5)
wuffs_deflate_decode_10k_full_init/clang9 82.0MB/s ± 1% 104.8MB/s ± 1% +27.75% (p=0.008 n=5+5)
wuffs_deflate_decode_10k_part_init/clang9 87.9MB/s ± 0% 114.4MB/s ± 0% +30.13% (p=0.008 n=5+5)
wuffs_deflate_decode_100k_just_one_read/clang9 111MB/s ± 0% 139MB/s ± 0% +26.03% (p=0.008 n=5+5)
wuffs_deflate_decode_100k_many_big_reads/clang9 60.2MB/s ± 0% 63.3MB/s ± 0% +5.28% (p=0.008 n=5+5)
wuffs_deflate_decode_1k_full_init/gcc8 45.8MB/s ± 3% 47.4MB/s ± 1% +3.38% (p=0.032 n=5+5)
wuffs_deflate_decode_1k_part_init/gcc8 63.0MB/s ± 1% 62.8MB/s ± 1% ~ (p=0.151 n=5+5)
wuffs_deflate_decode_10k_full_init/gcc8 87.9MB/s ± 0% 102.9MB/s ± 1% +17.08% (p=0.008 n=5+5)
wuffs_deflate_decode_10k_part_init/gcc8 95.3MB/s ± 0% 111.5MB/s ± 0% +17.03% (p=0.008 n=5+5)
wuffs_deflate_decode_100k_just_one_read/gcc8 115MB/s ± 0% 130MB/s ± 0% +12.38% (p=0.008 n=5+5)
wuffs_deflate_decode_100k_many_big_reads/gcc8 58.9MB/s ± 0% 60.1MB/s ± 0% +1.94% (p=0.008 n=5+5)
mimic_deflate_decode_1k 77.7MB/s ± 1% 77.2MB/s ± 1% ~ (p=0.222 n=5+5)
mimic_deflate_decode_10k 114MB/s ± 0% 114MB/s ± 1% ~ (p=0.548 n=5+5)
mimic_deflate_decode_100k_just_one_read 158MB/s ± 0% 158MB/s ± 0% ~ (p=0.056 n=5+5)
mimic_deflate_decode_100k_many_big_reads 113MB/s ± 0% 113MB/s ± 0% ~ (p=0.690 n=5+5)
----
wuffs_gzip_decode_10k/clang9 91.4MB/s ± 0% 106.9MB/s ± 0% +16.93% (p=0.008 n=5+5)
wuffs_gzip_decode_100k/clang9 113MB/s ± 0% 128MB/s ± 0% +13.01% (p=0.008 n=5+5)
wuffs_gzip_decode_10k/gcc8 92.2MB/s ± 0% 109.7MB/s ± 0% +19.00% (p=0.008 n=5+5)
wuffs_gzip_decode_100k/gcc8 111MB/s ± 0% 119MB/s ± 0% +7.43% (p=0.008 n=5+5)
mimic_gzip_decode_10k 98.2MB/s ± 0% 98.1MB/s ± 0% ~ (p=0.222 n=5+5)
mimic_gzip_decode_100k 128MB/s ± 0% 128MB/s ± 0% +0.11% (p=0.008 n=5+5)
----
wuffs_zlib_decode_10k/clang9 90.8MB/s ± 0% 111.5MB/s ± 0% +22.86% (p=0.008 n=5+5)
wuffs_zlib_decode_100k/clang9 108MB/s ± 0% 134MB/s ± 0% +23.90% (p=0.008 n=5+5)
wuffs_zlib_decode_10k/gcc8 92.3MB/s ± 0% 108.2MB/s ± 0% +17.15% (p=0.008 n=5+5)
wuffs_zlib_decode_100k/gcc8 112MB/s ± 0% 119MB/s ± 0% +6.90% (p=0.008 n=5+5)
mimic_zlib_decode_10k 108MB/s ± 0% 108MB/s ± 0% +0.39% (p=0.008 n=5+5)
mimic_zlib_decode_100k 143MB/s ± 0% 141MB/s ± 0% -1.28% (p=0.008 n=5+5)
----
wuffs_png_decode_image_19k_8bpp/clang9 72.4MB/s ± 0% 81.9MB/s ± 0% +13.10% (p=0.008 n=5+5)
wuffs_png_decode_image_40k_24bpp/clang9 104MB/s ± 0% 104MB/s ± 0% -0.40% (p=0.008 n=5+5)
wuffs_png_decode_image_77k_8bpp/clang9 274MB/s ± 0% 301MB/s ± 0% +9.80% (p=0.008 n=5+5)
wuffs_png_decode_image_552k_32bpp_ignore_checksum/clang9 165MB/s ± 0% 183MB/s ± 0% +10.96% (p=0.016 n=4+5)
wuffs_png_decode_image_552k_32bpp_verify_checksum/clang9 160MB/s ± 1% 178MB/s ± 0% +11.28% (p=0.008 n=5+5)
wuffs_png_decode_image_4002k_24bpp/clang9 98.1MB/s ± 0% 98.9MB/s ± 0% +0.82% (p=0.016 n=5+4)
wuffs_png_decode_image_19k_8bpp/gcc8 73.5MB/s ± 0% 80.4MB/s ± 0% +9.33% (p=0.008 n=5+5)
wuffs_png_decode_image_40k_24bpp/gcc8 104MB/s ± 0% 106MB/s ± 0% +1.73% (p=0.008 n=5+5)
wuffs_png_decode_image_77k_8bpp/gcc8 248MB/s ± 0% 272MB/s ± 0% +9.67% (p=0.008 n=5+5)
wuffs_png_decode_image_552k_32bpp_ignore_checksum/gcc8 163MB/s ± 0% 177MB/s ± 0% +8.89% (p=0.008 n=5+5)
wuffs_png_decode_image_552k_32bpp_verify_checksum/gcc8 157MB/s ± 0% 170MB/s ± 0% +8.42% (p=0.008 n=5+5)
wuffs_png_decode_image_4002k_24bpp/gcc8 98.6MB/s ± 0% 100.9MB/s ± 0% +2.29% (p=0.008 n=5+5)
mimic_png_decode_image_19k_8bpp 43.7MB/s ± 0% 43.7MB/s ± 0% ~ (p=0.690 n=5+5)
mimic_png_decode_image_40k_24bpp 54.6MB/s ± 0% 54.6MB/s ± 0% ~ (p=0.151 n=5+5)
mimic_png_decode_image_77k_8bpp 125MB/s ± 0% 124MB/s ± 0% -0.28% (p=0.008 n=5+5)
mimic_png_decode_image_552k_32bpp_ignore_checksum skipped
mimic_png_decode_image_552k_32bpp_verify_checksum 102MB/s ± 0% 102MB/s ± 0% ~ (p=0.238 n=5+5)
mimic_png_decode_image_4002k_24bpp 81.7MB/s ± 0% 81.8MB/s ± 0% +0.09% (p=0.008 n=5+5)
|
2021-03-08 15:41:48 +11:00 |
|
Nigel Tao
|
8595a3b696
|
Add std/deflate decode_huffman_fast32
This undoes commit 1a63b53e "Have std/deflate read 64 bits once per
inner loop" for 32-bit CPUs.
On a Raspberry Pi 4 (32-bit armv7l):
wuffs_deflate_decode_1k_full_init/clang9 44.9MB/s ± 1% 50.1MB/s ± 1% +11.54% (p=0.008 n=5+5)
wuffs_deflate_decode_1k_part_init/clang9 52.8MB/s ± 0% 61.1MB/s ± 0% +15.61% (p=0.008 n=5+5)
wuffs_deflate_decode_10k_full_init/clang9 81.4MB/s ± 0% 84.5MB/s ± 0% +3.89% (p=0.008 n=5+5)
wuffs_deflate_decode_10k_part_init/clang9 87.9MB/s ± 0% 91.1MB/s ± 0% +3.67% (p=0.008 n=5+5)
wuffs_deflate_decode_100k_just_one_read/clang9 106MB/s ± 0% 113MB/s ± 0% +5.70% (p=0.008 n=5+5)
wuffs_deflate_decode_100k_many_big_reads/clang9 59.6MB/s ± 0% 59.1MB/s ± 0% -0.81% (p=0.008 n=5+5)
wuffs_deflate_decode_1k_full_init/gcc8 44.7MB/s ± 1% 50.9MB/s ± 1% +13.91% (p=0.008 n=5+5)
wuffs_deflate_decode_1k_part_init/gcc8 52.6MB/s ± 0% 61.5MB/s ± 0% +17.06% (p=0.008 n=5+5)
wuffs_deflate_decode_10k_full_init/gcc8 76.5MB/s ± 0% 85.6MB/s ± 1% +11.89% (p=0.008 n=5+5)
wuffs_deflate_decode_10k_part_init/gcc8 82.3MB/s ± 0% 92.2MB/s ± 0% +12.07% (p=0.008 n=5+5)
wuffs_deflate_decode_100k_just_one_read/gcc8 97.8MB/s ± 0% 112.1MB/s ± 0% +14.62% (p=0.008 n=5+5)
wuffs_deflate_decode_100k_many_big_reads/gcc8 58.0MB/s ± 0% 61.6MB/s ± 0% +6.09% (p=0.008 n=5+5)
|
2021-02-08 23:19:00 +11:00 |
|
Nigel Tao
|
279d69bb38
|
Prepare to add std/deflate decode_huffman_fast32
|
2021-02-08 22:51:57 +11:00 |
|