mirror of
https://github.com/golang/go
synced 2024-09-06 00:19:45 +00:00
![Richard Musiol](/assets/img/avatar_default.png)
This commit adds compiler intrinsics for the packages math and math/bits on the wasm architecture for better performance. benchmark old ns/op new ns/op delta BenchmarkCeil 8.31 3.21 -61.37% BenchmarkCopysign 5.24 3.88 -25.95% BenchmarkAbs 5.42 3.34 -38.38% BenchmarkFloor 8.29 3.18 -61.64% BenchmarkRoundToEven 9.76 3.26 -66.60% BenchmarkSqrtLatency 8.13 4.88 -39.98% BenchmarkSqrtPrime 5246 3535 -32.62% BenchmarkTrunc 8.29 3.15 -62.00% BenchmarkLeadingZeros 13.0 4.23 -67.46% BenchmarkLeadingZeros8 4.65 4.42 -4.95% BenchmarkLeadingZeros16 7.60 4.38 -42.37% BenchmarkLeadingZeros32 10.7 4.48 -58.13% BenchmarkLeadingZeros64 12.9 4.31 -66.59% BenchmarkTrailingZeros 6.52 4.04 -38.04% BenchmarkTrailingZeros8 4.57 4.14 -9.41% BenchmarkTrailingZeros16 6.69 4.16 -37.82% BenchmarkTrailingZeros32 6.97 4.23 -39.31% BenchmarkTrailingZeros64 6.59 4.00 -39.30% BenchmarkOnesCount 7.93 3.30 -58.39% BenchmarkOnesCount8 3.56 3.19 -10.39% BenchmarkOnesCount16 4.85 3.19 -34.23% BenchmarkOnesCount32 7.27 3.19 -56.12% BenchmarkOnesCount64 8.08 3.28 -59.41% BenchmarkRotateLeft 4.88 3.80 -22.13% BenchmarkRotateLeft64 5.03 3.63 -27.83% Change-Id: Ic1e0c2984878be8defb6eb7eb6ee63765c793222 Reviewed-on: https://go-review.googlesource.com/c/go/+/165177 Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
178 lines
4.4 KiB
Go
178 lines
4.4 KiB
Go
// asmcheck
|
|
|
|
// Copyright 2018 The Go Authors. All rights reserved.
|
|
// Use of this source code is governed by a BSD-style
|
|
// license that can be found in the LICENSE file.
|
|
|
|
package codegen
|
|
|
|
import "math"
|
|
|
|
var sink64 [8]float64
|
|
|
|
func approx(x float64) {
|
|
// s390x:"FIDBR\t[$]6"
|
|
// arm64:"FRINTPD"
|
|
// ppc64:"FRIP"
|
|
// ppc64le:"FRIP"
|
|
// wasm:"F64Ceil"
|
|
sink64[0] = math.Ceil(x)
|
|
|
|
// s390x:"FIDBR\t[$]7"
|
|
// arm64:"FRINTMD"
|
|
// ppc64:"FRIM"
|
|
// ppc64le:"FRIM"
|
|
// wasm:"F64Floor"
|
|
sink64[1] = math.Floor(x)
|
|
|
|
// s390x:"FIDBR\t[$]1"
|
|
// arm64:"FRINTAD"
|
|
// ppc64:"FRIN"
|
|
// ppc64le:"FRIN"
|
|
sink64[2] = math.Round(x)
|
|
|
|
// s390x:"FIDBR\t[$]5"
|
|
// arm64:"FRINTZD"
|
|
// ppc64:"FRIZ"
|
|
// ppc64le:"FRIZ"
|
|
// wasm:"F64Trunc"
|
|
sink64[3] = math.Trunc(x)
|
|
|
|
// s390x:"FIDBR\t[$]4"
|
|
// arm64:"FRINTND"
|
|
// wasm:"F64Nearest"
|
|
sink64[4] = math.RoundToEven(x)
|
|
}
|
|
|
|
func sqrt(x float64) float64 {
|
|
// amd64:"SQRTSD"
|
|
// 386/387:"FSQRT" 386/sse2:"SQRTSD"
|
|
// arm64:"FSQRTD"
|
|
// arm/7:"SQRTD"
|
|
// mips/hardfloat:"SQRTD" mips/softfloat:-"SQRTD"
|
|
// mips64/hardfloat:"SQRTD" mips64/softfloat:-"SQRTD"
|
|
// wasm:"F64Sqrt"
|
|
return math.Sqrt(x)
|
|
}
|
|
|
|
// Check that it's using integer registers
|
|
func abs(x, y float64) {
|
|
// amd64:"BTRQ\t[$]63"
|
|
// arm64:"FABSD\t"
|
|
// s390x:"LPDFR\t",-"MOVD\t" (no integer load/store)
|
|
// ppc64:"FABS\t"
|
|
// ppc64le:"FABS\t"
|
|
// wasm:"F64Abs"
|
|
sink64[0] = math.Abs(x)
|
|
|
|
// amd64:"BTRQ\t[$]63","PXOR" (TODO: this should be BTSQ)
|
|
// s390x:"LNDFR\t",-"MOVD\t" (no integer load/store)
|
|
// ppc64:"FNABS\t"
|
|
// ppc64le:"FNABS\t"
|
|
sink64[1] = -math.Abs(y)
|
|
}
|
|
|
|
// Check that it's using integer registers
|
|
func abs32(x float32) float32 {
|
|
// s390x:"LPDFR",-"LDEBR",-"LEDBR" (no float64 conversion)
|
|
return float32(math.Abs(float64(x)))
|
|
}
|
|
|
|
// Check that it's using integer registers
|
|
func copysign(a, b, c float64) {
|
|
// amd64:"BTRQ\t[$]63","SHRQ\t[$]63","SHLQ\t[$]63","ORQ"
|
|
// s390x:"CPSDR",-"MOVD" (no integer load/store)
|
|
// ppc64:"FCPSGN"
|
|
// ppc64le:"FCPSGN"
|
|
// wasm:"F64Copysign"
|
|
sink64[0] = math.Copysign(a, b)
|
|
|
|
// amd64:"BTSQ\t[$]63"
|
|
// s390x:"LNDFR\t",-"MOVD\t" (no integer load/store)
|
|
// ppc64:"FCPSGN"
|
|
// ppc64le:"FCPSGN"
|
|
// arm64:"ORR", -"AND"
|
|
sink64[1] = math.Copysign(c, -1)
|
|
|
|
// Like math.Copysign(c, -1), but with integer operations. Useful
|
|
// for platforms that have a copysign opcode to see if it's detected.
|
|
// s390x:"LNDFR\t",-"MOVD\t" (no integer load/store)
|
|
sink64[2] = math.Float64frombits(math.Float64bits(a) | 1<<63)
|
|
|
|
// amd64:-"SHLQ\t[$]1",-"SHRQ\t[$]1","SHRQ\t[$]63","SHLQ\t[$]63","ORQ"
|
|
// s390x:"CPSDR\t",-"MOVD\t" (no integer load/store)
|
|
// ppc64:"FCPSGN"
|
|
// ppc64le:"FCPSGN"
|
|
sink64[3] = math.Copysign(-1, c)
|
|
}
|
|
|
|
func fromFloat64(f64 float64) uint64 {
|
|
// amd64:"MOVQ\tX.*, [^X].*"
|
|
// arm64:"FMOVD\tF.*, R.*"
|
|
// ppc64:"MFVSRD"
|
|
// ppc64le:"MFVSRD"
|
|
return math.Float64bits(f64+1) + 1
|
|
}
|
|
|
|
func fromFloat32(f32 float32) uint32 {
|
|
// amd64:"MOVL\tX.*, [^X].*"
|
|
// arm64:"FMOVS\tF.*, R.*"
|
|
return math.Float32bits(f32+1) + 1
|
|
}
|
|
|
|
func toFloat64(u64 uint64) float64 {
|
|
// amd64:"MOVQ\t[^X].*, X.*"
|
|
// arm64:"FMOVD\tR.*, F.*"
|
|
// ppc64:"MTVSRD"
|
|
// ppc64le:"MTVSRD"
|
|
return math.Float64frombits(u64+1) + 1
|
|
}
|
|
|
|
func toFloat32(u32 uint32) float32 {
|
|
// amd64:"MOVL\t[^X].*, X.*"
|
|
// arm64:"FMOVS\tR.*, F.*"
|
|
return math.Float32frombits(u32+1) + 1
|
|
}
|
|
|
|
// Test that comparisons with constants converted to float
|
|
// are evaluated at compile-time
|
|
|
|
func constantCheck64() bool {
|
|
// amd64:"MOVB\t[$]0",-"FCMP",-"MOVB\t[$]1"
|
|
// s390x:"MOV(B|BZ|D)\t[$]0,",-"FCMPU",-"MOV(B|BZ|D)\t[$]1,"
|
|
return 0.5 == float64(uint32(1)) || 1.5 > float64(uint64(1<<63)) || math.NaN() == math.NaN()
|
|
}
|
|
|
|
func constantCheck32() bool {
|
|
// amd64:"MOVB\t[$]1",-"FCMP",-"MOVB\t[$]0"
|
|
// s390x:"MOV(B|BZ|D)\t[$]1,",-"FCMPU",-"MOV(B|BZ|D)\t[$]0,"
|
|
return float32(0.5) <= float32(int64(1)) && float32(1.5) >= float32(int32(-1<<31)) && float32(math.NaN()) != float32(math.NaN())
|
|
}
|
|
|
|
// Test that integer constants are converted to floating point constants
|
|
// at compile-time
|
|
|
|
func constantConvert32(x float32) float32 {
|
|
// amd64:"MOVSS\t[$]f32.3f800000\\(SB\\)"
|
|
// s390x:"FMOVS\t[$]f32.3f800000\\(SB\\)"
|
|
// ppc64:"FMOVS\t[$]f32.3f800000\\(SB\\)"
|
|
// ppc64le:"FMOVS\t[$]f32.3f800000\\(SB\\)"
|
|
// arm64:"FMOVS\t[$]\\(1.0\\)"
|
|
if x > math.Float32frombits(0x3f800000) {
|
|
return -x
|
|
}
|
|
return x
|
|
}
|
|
|
|
func constantConvertInt32(x uint32) uint32 {
|
|
// amd64:-"MOVSS"
|
|
// s390x:-"FMOVS"
|
|
// ppc64:-"FMOVS"
|
|
// ppc64le:-"FMOVS"
|
|
// arm64:-"FMOVS"
|
|
if x > math.Float32bits(1) {
|
|
return -x
|
|
}
|
|
return x
|
|
}
|