Add std/adler32 hasher.up_x86_avx2

While the std/adler32 benchmarks got better, the std/zlib and std/png
benchmarks got worse.

This commit will soon be followed by a rollback, but it is committed
anyway so that we can refer to these numbers in the git log.

Binary size, before:
4056 gen/lib/c/clang-9-dynamic/wuffs-std-adler32.lo
4104 gen/lib/c/clang-9-static/wuffs-std-adler32.o
4624 gen/lib/c/gcc-dynamic/wuffs-std-adler32.lo
4680 gen/lib/c/gcc-static/wuffs-std-adler32.o

After:
4952 gen/lib/c/clang-9-dynamic/wuffs-std-adler32.lo
5016 gen/lib/c/clang-9-static/wuffs-std-adler32.o
5512 gen/lib/c/gcc-dynamic/wuffs-std-adler32.lo
5576 gen/lib/c/gcc-static/wuffs-std-adler32.o

name                                                      old speed      new speed      delta

wuffs_adler32_10k/clang9                                  11.5GB/s ± 1%  15.0GB/s ± 5%  +30.73%  (p=0.000 n=9+10)
wuffs_adler32_100k/clang9                                 13.5GB/s ± 1%  15.6GB/s ±14%  +16.06%  (p=0.001 n=9+10)

wuffs_adler32_10k/gcc10                                   11.3GB/s ± 2%  15.6GB/s ± 1%  +38.27%  (p=0.000 n=9+9)
wuffs_adler32_100k/gcc10                                  11.3GB/s ± 1%  19.4GB/s ± 1%  +71.20%  (p=0.000 n=7+9)

wuffs_zlib_decode_10k/clang9                               223MB/s ± 1%   216MB/s ± 2%   -3.45%  (p=0.000 n=10+10)
wuffs_zlib_decode_100k/clang9                              285MB/s ± 1%   284MB/s ± 1%     ~     (p=0.243 n=9+10)

wuffs_zlib_decode_10k/gcc10                                232MB/s ± 1%   216MB/s ± 1%   -6.70%  (p=0.000 n=10+9)
wuffs_zlib_decode_100k/gcc10                               291MB/s ± 1%   280MB/s ± 1%   -3.70%  (p=0.000 n=9+8)

wuffs_png_decode_image_19k_8bpp/clang9                     152MB/s ± 1%   143MB/s ± 3%   -5.89%  (p=0.000 n=8+9)
wuffs_png_decode_image_40k_24bpp/clang9                    171MB/s ± 1%   173MB/s ± 1%     ~     (p=0.094 n=9+9)
wuffs_png_decode_image_77k_8bpp/clang9                     541MB/s ± 1%   509MB/s ± 1%   -5.77%  (p=0.000 n=8+9)
wuffs_png_decode_image_552k_32bpp_ignore_checksum/clang9   479MB/s ± 1%   464MB/s ± 3%   -3.20%  (p=0.000 n=8+10)
wuffs_png_decode_image_552k_32bpp_verify_checksum/clang9   461MB/s ± 1%   452MB/s ± 1%   -2.02%  (p=0.000 n=8+10)
wuffs_png_decode_image_4002k_24bpp/clang9                  177MB/s ± 1%   175MB/s ± 2%   -1.14%  (p=0.004 n=10+10)

wuffs_png_decode_image_19k_8bpp/gcc10                      159MB/s ± 1%   153MB/s ± 1%   -3.54%  (p=0.000 n=8+10)
wuffs_png_decode_image_40k_24bpp/gcc10                     188MB/s ± 1%   187MB/s ± 1%   -0.47%  (p=0.021 n=10+9)
wuffs_png_decode_image_77k_8bpp/gcc10                      557MB/s ± 0%   546MB/s ± 4%   -1.94%  (p=0.000 n=8+10)
wuffs_png_decode_image_552k_32bpp_ignore_checksum/gcc10    510MB/s ± 0%   493MB/s ± 0%   -3.33%  (p=0.000 n=10+9)
wuffs_png_decode_image_552k_32bpp_verify_checksum/gcc10    486MB/s ± 1%   478MB/s ± 0%   -1.77%  (p=0.000 n=8+9)
wuffs_png_decode_image_4002k_24bpp/gcc10                   189MB/s ± 1%   187MB/s ± 1%   -1.26%  (p=0.000 n=10+10)
This commit is contained in:
Nigel Tao committed 2021-04-28 00:18:20 +10:00
1 parent 8ccaf81e96
commit baec831fc8
8 files changed
+367 -1

No files matched your search

+100
View File
@@ -20559,6 +20559,13 @@ wuffs_adler32__hasher__up_arm_neon(
wuffs_base__slice_u8 a_x);
#endif // defined(WUFFS_BASE__CPU_ARCH__ARM_NEON)
#if defined(WUFFS_BASE__CPU_ARCH__X86_64)
static wuffs_base__empty_struct
wuffs_adler32__hasher__up_x86_avx2(
wuffs_adler32__hasher* self,
wuffs_base__slice_u8 a_x);
#endif // defined(WUFFS_BASE__CPU_ARCH__X86_64)
#if defined(WUFFS_BASE__CPU_ARCH__X86_64)
static wuffs_base__empty_struct
wuffs_adler32__hasher__up_x86_sse42(
@@ -20680,6 +20687,9 @@ wuffs_adler32__hasher__update_u32(
#if defined(WUFFS_BASE__CPU_ARCH__ARM_NEON)
wuffs_base__cpu_arch__have_arm_neon() ? &wuffs_adler32__hasher__up_arm_neon :
#endif
#if defined(WUFFS_BASE__CPU_ARCH__X86_64)
wuffs_base__cpu_arch__have_x86_avx2() ? &wuffs_adler32__hasher__up_x86_avx2 :
#endif
#if defined(WUFFS_BASE__CPU_ARCH__X86_64)
wuffs_base__cpu_arch__have_x86_sse42() ? &wuffs_adler32__hasher__up_x86_sse42 :
#endif
@@ -20869,6 +20879,96 @@ wuffs_adler32__hasher__up_arm_neon(
#endif // defined(WUFFS_BASE__CPU_ARCH__ARM_NEON)
// ‼ WUFFS MULTI-FILE SECTION -arm_neon
// ‼ WUFFS MULTI-FILE SECTION +x86_avx2
// -------- func adler32.hasher.up_x86_avx2
#if defined(WUFFS_BASE__CPU_ARCH__X86_64)
WUFFS_BASE__MAYBE_ATTRIBUTE_TARGET("pclmul,popcnt,sse4.2,avx2")
static wuffs_base__empty_struct
wuffs_adler32__hasher__up_x86_avx2(
wuffs_adler32__hasher* self,
wuffs_base__slice_u8 a_x) {
uint32_t v_s1 = 0;
uint32_t v_s2 = 0;
wuffs_base__slice_u8 v_remaining = {0};
wuffs_base__slice_u8 v_p = {0};
__m256i v_zeroes = {0};
__m256i v_ones = {0};
__m256i v_weights = {0};
__m256i v_q = {0};
__m256i v_v1 = {0};
__m256i v_v2 = {0};
__m256i v_v2j = {0};
__m256i v_v2k = {0};
__m128i v_h1 = {0};
__m128i v_h2 = {0};
uint32_t v_num_iterate_bytes = 0;
uint64_t v_tail_index = 0;
v_zeroes = _mm256_set1_epi16((int16_t)(0));
v_ones = _mm256_set1_epi16((int16_t)(1));
v_weights = _mm256_set_epi8((int8_t)(1), (int8_t)(2), (int8_t)(3), (int8_t)(4), (int8_t)(5), (int8_t)(6), (int8_t)(7), (int8_t)(8), (int8_t)(9), (int8_t)(10), (int8_t)(11), (int8_t)(12), (int8_t)(13), (int8_t)(14), (int8_t)(15), (int8_t)(16), (int8_t)(17), (int8_t)(18), (int8_t)(19), (int8_t)(20), (int8_t)(21), (int8_t)(22), (int8_t)(23), (int8_t)(24), (int8_t)(25), (int8_t)(26), (int8_t)(27), (int8_t)(28), (int8_t)(29), (int8_t)(30), (int8_t)(31), (int8_t)(32));
v_s1 = ((self->private_impl.f_state) & 0xFFFF);
v_s2 = ((self->private_impl.f_state) >> (32 - (16)));
while (((uint64_t)(a_x.len)) > 0) {
v_remaining = wuffs_base__slice_u8__subslice_j(a_x, 0);
if (((uint64_t)(a_x.len)) > 5536) {
v_remaining = wuffs_base__slice_u8__subslice_i(a_x, 5536);
a_x = wuffs_base__slice_u8__subslice_j(a_x, 5536);
}
v_num_iterate_bytes = ((uint32_t)((((uint64_t)(a_x.len)) & 4294967264)));
v_s2 += ((uint32_t)(v_s1 * v_num_iterate_bytes));
v_v1 = _mm256_setzero_si256();
v_v2j = _mm256_setzero_si256();
v_v2k = _mm256_setzero_si256();
{
wuffs_base__slice_u8 i_slice_p = a_x;
v_p.ptr = i_slice_p.ptr;
v_p.len = 32;
uint8_t* i_end0_p = v_p.ptr + (((i_slice_p.len - (size_t)(v_p.ptr - i_slice_p.ptr)) / 32) * 32);
while (v_p.ptr < i_end0_p) {
v_q = _mm256_lddqu_si256((const __m256i*)(const void*)(v_p.ptr));
v_v2j = _mm256_add_epi32(v_v2j, v_v1);
v_v1 = _mm256_add_epi32(v_v1, _mm256_sad_epu8(v_q, v_zeroes));
v_v2k = _mm256_add_epi32(v_v2k, _mm256_madd_epi16(v_ones, _mm256_maddubs_epi16(v_q, v_weights)));
v_p.ptr += 32;
}
v_p.len = 0;
}
v_h1 = _mm_add_epi32(_mm256_extracti128_si256(v_v1, (int32_t)(0)), _mm256_extracti128_si256(v_v1, (int32_t)(1)));
v_h1 = _mm_add_epi32(v_h1, _mm_shuffle_epi32(v_h1, (int32_t)(177)));
v_h1 = _mm_add_epi32(v_h1, _mm_shuffle_epi32(v_h1, (int32_t)(78)));
v_s1 += ((uint32_t)(_mm_cvtsi128_si32(v_h1)));
v_v2 = _mm256_add_epi32(v_v2k, _mm256_slli_epi32(v_v2j, (int32_t)(5)));
v_h2 = _mm_add_epi32(_mm256_extracti128_si256(v_v2, (int32_t)(0)), _mm256_extracti128_si256(v_v2, (int32_t)(1)));
v_h2 = _mm_add_epi32(v_h2, _mm_shuffle_epi32(v_h2, (int32_t)(177)));
v_h2 = _mm_add_epi32(v_h2, _mm_shuffle_epi32(v_h2, (int32_t)(78)));
v_s2 += ((uint32_t)(_mm_cvtsi128_si32(v_h2)));
v_tail_index = (((uint64_t)(a_x.len)) & 18446744073709551584u);
if (v_tail_index < ((uint64_t)(a_x.len))) {
{
wuffs_base__slice_u8 i_slice_p = wuffs_base__slice_u8__subslice_i(a_x, v_tail_index);
v_p.ptr = i_slice_p.ptr;
v_p.len = 1;
uint8_t* i_end0_p = i_slice_p.ptr + i_slice_p.len;
while (v_p.ptr < i_end0_p) {
v_s1 += ((uint32_t)(v_p.ptr[0]));
v_s2 += v_s1;
v_p.ptr += 1;
}
v_p.len = 0;
}
}
v_s1 %= 65521;
v_s2 %= 65521;
a_x = v_remaining;
}
self->private_impl.f_state = (((v_s2 & 65535) << 16) | (v_s1 & 65535));
return wuffs_base__make_empty_struct();
}
#endif // defined(WUFFS_BASE__CPU_ARCH__X86_64)
// ‼ WUFFS MULTI-FILE SECTION -x86_avx2
// ‼ WUFFS MULTI-FILE SECTION +x86_sse42
// -------- func adler32.hasher.up_x86_sse42