Add crc32.ieee_hasher.up_x86_avx2

name                                                      old speed      new speed      delta

wuffs_crc32_ieee_10k/clang9                               8.42GB/s ± 1%  8.12GB/s ± 1%   -3.55%  (p=0.000 n=9+8)
wuffs_crc32_ieee_100k/clang9                              11.0GB/s ± 1%  10.9GB/s ± 1%     ~     (p=0.931 n=9+9)

wuffs_crc32_ieee_10k/gcc10                                9.20GB/s ± 0%  9.71GB/s ± 0%   +5.54%  (p=0.000 n=8+10)
wuffs_crc32_ieee_100k/gcc10                               11.9GB/s ± 2%  13.1GB/s ± 1%  +10.22%  (p=0.000 n=10+8)

wuffs_gzip_decode_10k/clang9                               219MB/s ± 1%   218MB/s ± 2%     ~     (p=0.497 n=9+10)
wuffs_gzip_decode_100k/clang9                              283MB/s ± 2%   281MB/s ± 1%     ~     (p=0.063 n=10+10)

wuffs_gzip_decode_10k/gcc10                                224MB/s ± 1%   223MB/s ± 1%     ~     (p=0.489 n=9+9)
wuffs_gzip_decode_100k/gcc10                               285MB/s ± 4%   290MB/s ± 0%   +1.68%  (p=0.004 n=10+8)

wuffs_png_decode_image_19k_8bpp/clang9                     138MB/s ± 2%   138MB/s ± 1%     ~     (p=0.156 n=10+9)
wuffs_png_decode_image_40k_24bpp/clang9                    171MB/s ± 2%   172MB/s ± 1%     ~     (p=0.971 n=10+10)
wuffs_png_decode_image_77k_8bpp/clang9                     495MB/s ± 4%   504MB/s ± 1%   +1.79%  (p=0.000 n=9+9)
wuffs_png_decode_image_552k_32bpp_ignore_checksum/clang9   457MB/s ± 4%   463MB/s ± 0%   +1.35%  (p=0.002 n=9+9)
wuffs_png_decode_image_552k_32bpp_verify_checksum/clang9   435MB/s ± 5%   445MB/s ± 1%   +2.31%  (p=0.000 n=10+9)
wuffs_png_decode_image_4002k_24bpp/clang9                  172MB/s ± 2%   174MB/s ± 1%     ~     (p=0.053 n=10+9)

wuffs_png_decode_image_19k_8bpp/gcc10                      157MB/s ± 1%   157MB/s ± 1%     ~     (p=0.549 n=9+10)
wuffs_png_decode_image_40k_24bpp/gcc10                     187MB/s ± 1%   186MB/s ± 1%     ~     (p=0.182 n=10+9)
wuffs_png_decode_image_77k_8bpp/gcc10                      554MB/s ± 1%   559MB/s ± 0%   +0.84%  (p=0.000 n=9+9)
wuffs_png_decode_image_552k_32bpp_ignore_checksum/gcc10    502MB/s ± 1%   500MB/s ± 0%   -0.38%  (p=0.011 n=10+10)
wuffs_png_decode_image_552k_32bpp_verify_checksum/gcc10    481MB/s ± 1%   478MB/s ± 0%   -0.46%  (p=0.002 n=8+9)
wuffs_png_decode_image_4002k_24bpp/gcc10                   189MB/s ± 2%   190MB/s ± 0%   +0.62%  (p=0.008 n=10+9)
This commit is contained in:
Nigel Tao committed 2021-04-09 23:12:32 +10:00
1 parent 1dd7ad4413
commit 2dddd388bc
7 files changed
+333 -3

No files matched your search

+1
View File
@@ -27,6 +27,7 @@ pub func ieee_hasher.update_u32!(x: slice base.u8) base.u32 {
if this.state == 0 {
choose up = [
up_arm_crc32,
up_x86_avx2,
up_x86_sse42]
}
this.up!(x: args.x)
+138
View File
@@ -0,0 +1,138 @@
// Copyright 2021 The Wuffs Authors.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// https://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
// --------
// See "SIMD Implementations" in README.md for a link to Gopal et al. "Fast CRC
// Computation for Generic Polynomials Using PCLMULQDQ Instruction".
// up_x86_avx2 is exactly the same as up_x86_sse42 except for the "choose
// cpu_arch >= x86_avx2". With AVX, PCLMULQDQ has a three-operand form, not
// just a two-operand form: https://www.felixcloutier.com/x86/pclmulqdq
pri func ieee_hasher.up_x86_avx2!(x: slice base.u8),
choose cpu_arch >= x86_avx2,
{
var s : base.u32
var p : slice base.u8
var util : base.x86_sse42_utility
var k : base.x86_m128i
var x0 : base.x86_m128i
var x1 : base.x86_m128i
var x2 : base.x86_m128i
var x3 : base.x86_m128i
var y0 : base.x86_m128i
var y1 : base.x86_m128i
var y2 : base.x86_m128i
var y3 : base.x86_m128i
var tail_index : base.u64
s = 0xFFFF_FFFF ^ this.state
// Align to a 16-byte boundary.
while (args.x.length() > 0) and ((15 & args.x.uintptr_low_12_bits()) <> 0) {
s = IEEE_TABLE[0][((s & 0xFF) as base.u8) ^ args.x[0]] ^ (s >> 8)
args.x = args.x[1 ..]
} endwhile
// For short inputs, just do a simple loop.
if args.x.length() < 64 {
iterate (p = args.x)(length: 1, advance: 1, unroll: 1) {
s = IEEE_TABLE[0][((s & 0xFF) as base.u8) ^ p[0]] ^ (s >> 8)
}
this.state = 0xFFFF_FFFF ^ s
return nothing
}
// Load 128×4 = 512 bits from the first 64-byte chunk.
x0 = util.make_m128i_slice128(a: args.x[0x00 .. 0x10])
x1 = util.make_m128i_slice128(a: args.x[0x10 .. 0x20])
x2 = util.make_m128i_slice128(a: args.x[0x20 .. 0x30])
x3 = util.make_m128i_slice128(a: args.x[0x30 .. 0x40])
// Combine with the initial state.
x0 = x0._mm_xor_si128(b: util.make_m128i_single_u32(a: s))
// Process the remaining 64-byte chunks.
k = util.make_m128i_slice128(a: IEEE_X86_SSE42_K1K2[.. 16])
iterate (p = args.x[64 ..])(length: 64, advance: 64, unroll: 1) {
y0 = x0._mm_clmulepi64_si128(b: k, imm8: 0x00)
y1 = x1._mm_clmulepi64_si128(b: k, imm8: 0x00)
y2 = x2._mm_clmulepi64_si128(b: k, imm8: 0x00)
y3 = x3._mm_clmulepi64_si128(b: k, imm8: 0x00)
x0 = x0._mm_clmulepi64_si128(b: k, imm8: 0x11)
x1 = x1._mm_clmulepi64_si128(b: k, imm8: 0x11)
x2 = x2._mm_clmulepi64_si128(b: k, imm8: 0x11)
x3 = x3._mm_clmulepi64_si128(b: k, imm8: 0x11)
x0 = x0._mm_xor_si128(b: y0)._mm_xor_si128(b: util.make_m128i_slice128(a: p[0x00 .. 0x10]))
x1 = x1._mm_xor_si128(b: y1)._mm_xor_si128(b: util.make_m128i_slice128(a: p[0x10 .. 0x20]))
x2 = x2._mm_xor_si128(b: y2)._mm_xor_si128(b: util.make_m128i_slice128(a: p[0x20 .. 0x30]))
x3 = x3._mm_xor_si128(b: y3)._mm_xor_si128(b: util.make_m128i_slice128(a: p[0x30 .. 0x40]))
}
// Reduce 128×4 = 512 bits to 128 bits.
k = util.make_m128i_slice128(a: IEEE_X86_SSE42_K3K4[.. 16])
y0 = x0._mm_clmulepi64_si128(b: k, imm8: 0x00)
x0 = x0._mm_clmulepi64_si128(b: k, imm8: 0x11)
x0 = x0._mm_xor_si128(b: x1)
x0 = x0._mm_xor_si128(b: y0)
y0 = x0._mm_clmulepi64_si128(b: k, imm8: 0x00)
x0 = x0._mm_clmulepi64_si128(b: k, imm8: 0x11)
x0 = x0._mm_xor_si128(b: x2)
x0 = x0._mm_xor_si128(b: y0)
y0 = x0._mm_clmulepi64_si128(b: k, imm8: 0x00)
x0 = x0._mm_clmulepi64_si128(b: k, imm8: 0x11)
x0 = x0._mm_xor_si128(b: x3)
x0 = x0._mm_xor_si128(b: y0)
// Reduce 128 bits to 64 bits.
x1 = x0._mm_clmulepi64_si128(b: k, imm8: 0x10)
x2 = util.make_m128i_multiple_u32(
a00: 0xFFFF_FFFF,
a01: 0x0000_0000,
a02: 0xFFFF_FFFF,
a03: 0x0000_0000)
x0 = x0._mm_srli_si128(imm8: 8)
x0 = x0._mm_xor_si128(b: x1)
k = util.make_m128i_slice128(a: IEEE_X86_SSE42_K5ZZ[.. 16])
x1 = x0._mm_srli_si128(imm8: 4)
x0 = x0._mm_and_si128(b: x2)
x0 = x0._mm_clmulepi64_si128(b: k, imm8: 0x00)
x0 = x0._mm_xor_si128(b: x1)
// Reduce 64 bits to 32 bits (Barrett Reduction) and extract.
//
// Barrett Reduction is Algorithm 1 (page 14) of Gopal et al., after
// adjusting for bit-reflection as per Figure 12 (page 21).
k = util.make_m128i_slice128(a: IEEE_X86_SSE42_PXMU[.. 16])
x1 = x0._mm_and_si128(b: x2)
x1 = x1._mm_clmulepi64_si128(b: k, imm8: 0x10)
x1 = x1._mm_and_si128(b: x2)
x1 = x1._mm_clmulepi64_si128(b: k, imm8: 0x00)
x0 = x0._mm_xor_si128(b: x1)
s = x0._mm_extract_epi32(imm8: 1)
// Handle the tail of args.x that wasn't a complete 64-byte chunk.
tail_index = args.x.length() & 0xFFFF_FFFF_FFFF_FFC0 // And-not 64.
if tail_index < args.x.length() {
iterate (p = args.x[tail_index ..])(length: 1, advance: 1, unroll: 1) {
s = IEEE_TABLE[0][((s & 0xFF) as base.u8) ^ p[0]] ^ (s >> 8)
}
}
this.state = 0xFFFF_FFFF ^ s
}