mirror of
https://github.com/golang/go
synced 2024-11-16 20:54:48 -07:00
c5f0104daf
The previous change sped up the pure computation form of LeadingZeros8.
This places it somewhat close to the table lookup form.
Depending on something that varies from toolchain to toolchain
(alignment, perhaps?), the slowdown from ditching the table lookup
is either 20% or 5%.
This benchmark is the best case scenario for the table lookup:
It is in the L1 cache already.
I think we're close enough that we can switch to the computational version,
and trust that the memory effects and binary size savings will be worth it.
Code:
func f8(x uint8) { z = bits.LeadingZeros8(x) }
Before:
"".f8 STEXT nosplit size=34 args=0x8 locals=0x0
0x0000 00000 (x.go:7) TEXT "".f8(SB), NOSPLIT, $0-8
0x0000 00000 (x.go:7) FUNCDATA $0, gclocals·2a5305abe05176240e61b8620e19a815(SB)
0x0000 00000 (x.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB)
0x0000 00000 (x.go:7) MOVBLZX "".x+8(SP), AX
0x0005 00005 (x.go:7) MOVBLZX AL, AX
0x0008 00008 (x.go:7) LEAQ math/bits.len8tab(SB), CX
0x000f 00015 (x.go:7) MOVBLZX (CX)(AX*1), AX
0x0013 00019 (x.go:7) ADDQ $-8, AX
0x0017 00023 (x.go:7) NEGQ AX
0x001a 00026 (x.go:7) MOVQ AX, "".z(SB)
0x0021 00033 (x.go:7) RET
After:
"".f8 STEXT nosplit size=30 args=0x8 locals=0x0
0x0000 00000 (x.go:7) TEXT "".f8(SB), NOSPLIT, $0-8
0x0000 00000 (x.go:7) FUNCDATA $0, gclocals·2a5305abe05176240e61b8620e19a815(SB)
0x0000 00000 (x.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB)
0x0000 00000 (x.go:7) MOVBLZX "".x+8(SP), AX
0x0005 00005 (x.go:7) MOVBLZX AL, AX
0x0008 00008 (x.go:7) LEAL 1(AX)(AX*1), AX
0x000c 00012 (x.go:7) BSRL AX, AX
0x000f 00015 (x.go:7) ADDQ $-8, AX
0x0013 00019 (x.go:7) NEGQ AX
0x0016 00022 (x.go:7) MOVQ AX, "".z(SB)
0x001d 00029 (x.go:7) RET
Change-Id: Icc7db50a7820fb9a3da8a816d6b6940d7f8e193e
Reviewed-on: https://go-review.googlesource.com/108942
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
218 lines
4.2 KiB
Go
218 lines
4.2 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/bits"
|
|
|
|
// ----------------------- //
|
|
// bits.LeadingZeros //
|
|
// ----------------------- //
|
|
|
|
func LeadingZeros(n uint) int {
|
|
// amd64:"BSRQ"
|
|
// s390x:"FLOGR"
|
|
// arm:"CLZ" arm64:"CLZ"
|
|
// mips:"CLZ"
|
|
return bits.LeadingZeros(n)
|
|
}
|
|
|
|
func LeadingZeros64(n uint64) int {
|
|
// amd64:"BSRQ"
|
|
// s390x:"FLOGR"
|
|
// arm:"CLZ" arm64:"CLZ"
|
|
// mips:"CLZ"
|
|
return bits.LeadingZeros64(n)
|
|
}
|
|
|
|
func LeadingZeros32(n uint32) int {
|
|
// amd64:"BSRQ","LEAQ",-"CMOVQEQ"
|
|
// s390x:"FLOGR"
|
|
// arm:"CLZ" arm64:"CLZ"
|
|
// mips:"CLZ"
|
|
return bits.LeadingZeros32(n)
|
|
}
|
|
|
|
func LeadingZeros16(n uint16) int {
|
|
// amd64:"BSRL","LEAL",-"CMOVQEQ"
|
|
// s390x:"FLOGR"
|
|
// arm:"CLZ" arm64:"CLZ"
|
|
// mips:"CLZ"
|
|
return bits.LeadingZeros16(n)
|
|
}
|
|
|
|
func LeadingZeros8(n uint8) int {
|
|
// amd64:"BSRL","LEAL",-"CMOVQEQ"
|
|
// s390x:"FLOGR"
|
|
// arm:"CLZ" arm64:"CLZ"
|
|
// mips:"CLZ"
|
|
return bits.LeadingZeros8(n)
|
|
}
|
|
|
|
// --------------- //
|
|
// bits.Len* //
|
|
// --------------- //
|
|
|
|
func Len(n uint) int {
|
|
// amd64:"BSRQ"
|
|
// s390x:"FLOGR"
|
|
// arm:"CLZ" arm64:"CLZ"
|
|
// mips:"CLZ"
|
|
return bits.Len(n)
|
|
}
|
|
|
|
func Len64(n uint64) int {
|
|
// amd64:"BSRQ"
|
|
// s390x:"FLOGR"
|
|
// arm:"CLZ" arm64:"CLZ"
|
|
// mips:"CLZ"
|
|
return bits.Len64(n)
|
|
}
|
|
|
|
func Len32(n uint32) int {
|
|
// amd64:"BSRQ","LEAQ",-"CMOVQEQ"
|
|
// s390x:"FLOGR"
|
|
// arm:"CLZ" arm64:"CLZ"
|
|
// mips:"CLZ"
|
|
return bits.Len32(n)
|
|
}
|
|
|
|
func Len16(n uint16) int {
|
|
// amd64:"BSRL","LEAL",-"CMOVQEQ"
|
|
// s390x:"FLOGR"
|
|
// arm:"CLZ" arm64:"CLZ"
|
|
// mips:"CLZ"
|
|
return bits.Len16(n)
|
|
}
|
|
|
|
func Len8(n uint8) int {
|
|
// amd64:"BSRL","LEAL",-"CMOVQEQ"
|
|
// s390x:"FLOGR"
|
|
// arm:"CLZ" arm64:"CLZ"
|
|
// mips:"CLZ"
|
|
return bits.Len8(n)
|
|
}
|
|
|
|
// -------------------- //
|
|
// bits.OnesCount //
|
|
// -------------------- //
|
|
|
|
func OnesCount(n uint) int {
|
|
// amd64:"POPCNTQ",".*support_popcnt"
|
|
// arm64:"VCNT","VUADDLV"
|
|
return bits.OnesCount(n)
|
|
}
|
|
|
|
func OnesCount64(n uint64) int {
|
|
// amd64:"POPCNTQ",".*support_popcnt"
|
|
// arm64:"VCNT","VUADDLV"
|
|
return bits.OnesCount64(n)
|
|
}
|
|
|
|
func OnesCount32(n uint32) int {
|
|
// amd64:"POPCNTL",".*support_popcnt"
|
|
// arm64:"VCNT","VUADDLV"
|
|
return bits.OnesCount32(n)
|
|
}
|
|
|
|
func OnesCount16(n uint16) int {
|
|
// amd64:"POPCNTL",".*support_popcnt"
|
|
// arm64:"VCNT","VUADDLV"
|
|
return bits.OnesCount16(n)
|
|
}
|
|
|
|
// ----------------------- //
|
|
// bits.ReverseBytes //
|
|
// ----------------------- //
|
|
|
|
func ReverseBytes(n uint) uint {
|
|
// amd64:"BSWAPQ"
|
|
// s390x:"MOVDBR"
|
|
// arm64:"REV"
|
|
return bits.ReverseBytes(n)
|
|
}
|
|
|
|
func ReverseBytes64(n uint64) uint64 {
|
|
// amd64:"BSWAPQ"
|
|
// s390x:"MOVDBR"
|
|
// arm64:"REV"
|
|
return bits.ReverseBytes64(n)
|
|
}
|
|
|
|
func ReverseBytes32(n uint32) uint32 {
|
|
// amd64:"BSWAPL"
|
|
// s390x:"MOVWBR"
|
|
// arm64:"REVW"
|
|
return bits.ReverseBytes32(n)
|
|
}
|
|
|
|
func ReverseBytes16(n uint16) uint16 {
|
|
// amd64:"ROLW"
|
|
return bits.ReverseBytes16(n)
|
|
}
|
|
|
|
// --------------------- //
|
|
// bits.RotateLeft //
|
|
// --------------------- //
|
|
|
|
func RotateLeft64(n uint64) uint64 {
|
|
// amd64:"ROLQ"
|
|
// arm64:"ROR"
|
|
// ppc64:"ROTL"
|
|
return bits.RotateLeft64(n, 37)
|
|
}
|
|
|
|
func RotateLeft32(n uint32) uint32 {
|
|
// amd64:"ROLL" 386:"ROLL"
|
|
// arm64:"RORW"
|
|
// ppc64:"ROTLW"
|
|
return bits.RotateLeft32(n, 9)
|
|
}
|
|
|
|
func RotateLeft16(n uint16) uint16 {
|
|
// amd64:"ROLW" 386:"ROLW"
|
|
return bits.RotateLeft16(n, 5)
|
|
}
|
|
|
|
func RotateLeft8(n uint8) uint8 {
|
|
// amd64:"ROLB" 386:"ROLB"
|
|
return bits.RotateLeft8(n, 5)
|
|
}
|
|
|
|
// ------------------------ //
|
|
// bits.TrailingZeros //
|
|
// ------------------------ //
|
|
|
|
func TrailingZeros(n uint) int {
|
|
// amd64:"BSFQ","MOVL\t\\$64","CMOVQEQ"
|
|
// s390x:"FLOGR"
|
|
return bits.TrailingZeros(n)
|
|
}
|
|
|
|
func TrailingZeros64(n uint64) int {
|
|
// amd64:"BSFQ","MOVL\t\\$64","CMOVQEQ"
|
|
// s390x:"FLOGR"
|
|
return bits.TrailingZeros64(n)
|
|
}
|
|
|
|
func TrailingZeros32(n uint32) int {
|
|
// amd64:"BTSQ\\t\\$32","BSFQ"
|
|
// s390x:"FLOGR","MOVWZ"
|
|
return bits.TrailingZeros32(n)
|
|
}
|
|
|
|
func TrailingZeros16(n uint16) int {
|
|
// amd64:"BSFL","BTSL\\t\\$16"
|
|
// s390x:"FLOGR","OR\t\\$65536"
|
|
return bits.TrailingZeros16(n)
|
|
}
|
|
|
|
func TrailingZeros8(n uint8) int {
|
|
// amd64:"BSFL","BTSL\\t\\$8"
|
|
// s390x:"FLOGR","OR\t\\$256"
|
|
return bits.TrailingZeros8(n)
|
|
}
|