1
0
mirror of https://github.com/golang/go synced 2024-11-08 19:26:14 -07:00
go/test/codegen/memcombine.go

667 lines
23 KiB
Go
Raw Normal View History

// 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 (
"encoding/binary"
"runtime"
)
var sink64 uint64
var sink32 uint32
var sink16 uint16
// ------------- //
// Loading //
// ------------- //
func load_le64(b []byte) {
// amd64:`MOVQ\s\(.*\),`,-`MOV[BWL]\t[^$]`,-`OR`
// s390x:`MOVDBR\s\(.*\),`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`MOVD\s\(R[0-9]+\),`,-`MOV[BHW]`
// ppc64le:`MOVD\s`,-`MOV[BHW]Z`
sink64 = binary.LittleEndian.Uint64(b)
}
func load_le64_idx(b []byte, idx int) {
// amd64:`MOVQ\s\(.*\)\(.*\*1\),`,-`MOV[BWL]\t[^$]`,-`OR`
// s390x:`MOVDBR\s\(.*\)\(.*\*1\),`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`MOVD\s\(R[0-9]+\)\(R[0-9]+\),`,-`MOV[BHW]`
// ppc64le:`MOVD\s`,-`MOV[BHW]Z\s`
sink64 = binary.LittleEndian.Uint64(b[idx:])
}
func load_le32(b []byte) {
// amd64:`MOVL\s\(.*\),`,-`MOV[BW]`,-`OR`
// 386:`MOVL\s\(.*\),`,-`MOV[BW]`,-`OR`
// s390x:`MOVWBR\s\(.*\),`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`MOVWU\s\(R[0-9]+\),`,-`MOV[BH]`
// ppc64le:`MOVWZ\s`,-`MOV[BH]Z\s`
sink32 = binary.LittleEndian.Uint32(b)
}
func load_le32_idx(b []byte, idx int) {
// amd64:`MOVL\s\(.*\)\(.*\*1\),`,-`MOV[BW]`,-`OR`
// 386:`MOVL\s\(.*\)\(.*\*1\),`,-`MOV[BW]`,-`OR`
// s390x:`MOVWBR\s\(.*\)\(.*\*1\),`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`MOVWU\s\(R[0-9]+\)\(R[0-9]+\),`,-`MOV[BH]`
// ppc64le:`MOVWZ\s`,-`MOV[BH]Z\s`
sink32 = binary.LittleEndian.Uint32(b[idx:])
}
func load_le16(b []byte) {
// amd64:`MOVWLZX\s\(.*\),`,-`MOVB`,-`OR`
// ppc64le:`MOVHZ\s`,-`MOVBZ`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`MOVHU\s\(R[0-9]+\),`,-`MOVB`
// s390x:`MOVHBR\s\(.*\),`
sink16 = binary.LittleEndian.Uint16(b)
}
func load_le16_idx(b []byte, idx int) {
// amd64:`MOVWLZX\s\(.*\),`,-`MOVB`,-`OR`
// ppc64le:`MOVHZ\s`,-`MOVBZ`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`MOVHU\s\(R[0-9]+\)\(R[0-9]+\),`,-`MOVB`
// s390x:`MOVHBR\s\(.*\)\(.*\*1\),`
sink16 = binary.LittleEndian.Uint16(b[idx:])
}
func load_be64(b []byte) {
// amd64:`BSWAPQ`,-`MOV[BWL]\t[^$]`,-`OR`
// s390x:`MOVD\s\(.*\),`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`REV`,`MOVD\s\(R[0-9]+\),`,-`MOV[BHW]`,-`REVW`,-`REV16W`
// ppc64le:`MOVDBR`,-`MOV[BHW]Z`
sink64 = binary.BigEndian.Uint64(b)
}
func load_be64_idx(b []byte, idx int) {
// amd64:`BSWAPQ`,-`MOV[BWL]\t[^$]`,-`OR`
// s390x:`MOVD\s\(.*\)\(.*\*1\),`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`REV`,`MOVD\s\(R[0-9]+\)\(R[0-9]+\),`,-`MOV[WHB]`,-`REVW`,-`REV16W`
// ppc64le:`MOVDBR`,-`MOV[BHW]Z`
sink64 = binary.BigEndian.Uint64(b[idx:])
}
func load_be32(b []byte) {
// amd64:`BSWAPL`,-`MOV[BW]`,-`OR`
// s390x:`MOVWZ\s\(.*\),`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`REVW`,`MOVWU\s\(R[0-9]+\),`,-`MOV[BH]`,-`REV16W`
// ppc64le:`MOVWBR`,-`MOV[BH]Z`
sink32 = binary.BigEndian.Uint32(b)
}
func load_be32_idx(b []byte, idx int) {
// amd64:`BSWAPL`,-`MOV[BW]`,-`OR`
// s390x:`MOVWZ\s\(.*\)\(.*\*1\),`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`REVW`,`MOVWU\s\(R[0-9]+\)\(R[0-9]+\),`,-`MOV[HB]`,-`REV16W`
// ppc64le:`MOVWBR`,-`MOV[BH]Z`
sink32 = binary.BigEndian.Uint32(b[idx:])
}
func load_be16(b []byte) {
// amd64:`ROLW\s\$8`,-`MOVB`,-`OR`
// arm64:`REV16W`,`MOVHU\s\(R[0-9]+\),`,-`MOVB`
cmd/compile,test: combine byte loads and stores on ppc64le CL 74410 added rules to combine consecutive byte loads and stores when the byte order was little endian for ppc64le. This is the corresponding change for bytes that are in big endian order. These rules are all intended for a little endian target arch. This adds new testcases in test/codegen/memcombine.go Fixes #22496 Updates #24242 Benchmark improvement for encoding/binary: name old time/op new time/op delta ReadSlice1000Int32s-16 11.0µs ± 0% 9.0µs ± 0% -17.47% (p=0.029 n=4+4) ReadStruct-16 2.47µs ± 1% 2.48µs ± 0% +0.67% (p=0.114 n=4+4) ReadInts-16 642ns ± 1% 630ns ± 1% -2.02% (p=0.029 n=4+4) WriteInts-16 654ns ± 0% 653ns ± 1% -0.08% (p=0.629 n=4+4) WriteSlice1000Int32s-16 8.75µs ± 0% 8.20µs ± 0% -6.19% (p=0.029 n=4+4) PutUint16-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint32-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint64-16 1.85ns ± 0% 0.93ns ± 0% -49.73% (p=0.029 n=4+4) LittleEndianPutUint16-16 1.03ns ± 0% 0.93ns ± 0% -9.71% (p=0.029 n=4+4) LittleEndianPutUint32-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) LittleEndianPutUint64-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) PutUvarint32-16 43.0ns ± 0% 43.1ns ± 0% +0.12% (p=0.429 n=4+4) PutUvarint64-16 174ns ± 0% 175ns ± 0% +0.29% (p=0.429 n=4+4) Updates made to functions in gcm.go to enable their matching. An existing testcase prevents these functions from being replaced by those in encoding/binary due to import dependencies. Change-Id: Idb3bd1e6e7b12d86cd828fb29cb095848a3e485a Reviewed-on: https://go-review.googlesource.com/98136 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-03-01 09:40:36 -07:00
// ppc64le:`MOVHBR`
// s390x:`MOVHZ\s\(.*\),`,-`OR`,-`ORW`,-`SLD`,-`SLW`
sink16 = binary.BigEndian.Uint16(b)
}
func load_be16_idx(b []byte, idx int) {
// amd64:`ROLW\s\$8`,-`MOVB`,-`OR`
// arm64:`REV16W`,`MOVHU\s\(R[0-9]+\)\(R[0-9]+\),`,-`MOVB`
cmd/compile,test: combine byte loads and stores on ppc64le CL 74410 added rules to combine consecutive byte loads and stores when the byte order was little endian for ppc64le. This is the corresponding change for bytes that are in big endian order. These rules are all intended for a little endian target arch. This adds new testcases in test/codegen/memcombine.go Fixes #22496 Updates #24242 Benchmark improvement for encoding/binary: name old time/op new time/op delta ReadSlice1000Int32s-16 11.0µs ± 0% 9.0µs ± 0% -17.47% (p=0.029 n=4+4) ReadStruct-16 2.47µs ± 1% 2.48µs ± 0% +0.67% (p=0.114 n=4+4) ReadInts-16 642ns ± 1% 630ns ± 1% -2.02% (p=0.029 n=4+4) WriteInts-16 654ns ± 0% 653ns ± 1% -0.08% (p=0.629 n=4+4) WriteSlice1000Int32s-16 8.75µs ± 0% 8.20µs ± 0% -6.19% (p=0.029 n=4+4) PutUint16-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint32-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint64-16 1.85ns ± 0% 0.93ns ± 0% -49.73% (p=0.029 n=4+4) LittleEndianPutUint16-16 1.03ns ± 0% 0.93ns ± 0% -9.71% (p=0.029 n=4+4) LittleEndianPutUint32-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) LittleEndianPutUint64-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) PutUvarint32-16 43.0ns ± 0% 43.1ns ± 0% +0.12% (p=0.429 n=4+4) PutUvarint64-16 174ns ± 0% 175ns ± 0% +0.29% (p=0.429 n=4+4) Updates made to functions in gcm.go to enable their matching. An existing testcase prevents these functions from being replaced by those in encoding/binary due to import dependencies. Change-Id: Idb3bd1e6e7b12d86cd828fb29cb095848a3e485a Reviewed-on: https://go-review.googlesource.com/98136 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-03-01 09:40:36 -07:00
// ppc64le:`MOVHBR`
// s390x:`MOVHZ\s\(.*\)\(.*\*1\),`,-`OR`,-`ORW`,-`SLD`,-`SLW`
sink16 = binary.BigEndian.Uint16(b[idx:])
}
func load_le_byte2_uint16(s []byte) uint16 {
// arm64:`MOVHU\t\(R[0-9]+\)`,-`ORR`,-`MOVB`
// 386:`MOVWLZX\s\([A-Z]+\)`,-`MOVB`,-`OR`
// amd64:`MOVWLZX\s\([A-Z]+\)`,-`MOVB`,-`OR`
// ppc64le:`MOVHZ\t\(R[0-9]+\)`,-`MOVBZ`
return uint16(s[0]) | uint16(s[1])<<8
}
func load_le_byte2_uint16_inv(s []byte) uint16 {
// arm64:`MOVHU\t\(R[0-9]+\)`,-`ORR`,-`MOVB`
// 386:`MOVWLZX\s\([A-Z]+\)`,-`MOVB`,-`OR`
// amd64:`MOVWLZX\s\([A-Z]+\)`,-`MOVB`,-`OR`
// ppc64le:`MOVHZ\t\(R[0-9]+\)`,-`MOVDZ`
return uint16(s[1])<<8 | uint16(s[0])
}
func load_le_byte4_uint32(s []byte) uint32 {
// arm64:`MOVWU\t\(R[0-9]+\)`,-`ORR`,-`MOV[BH]`
// 386:`MOVL\s\([A-Z]+\)`,-`MOV[BW]`,-`OR`
// amd64:`MOVL\s\([A-Z]+\)`,-`MOV[BW]`,-`OR`
// ppc64le:`MOVWZ\t\(R[0-9]+\)`,-`MOV[BH]Z`
return uint32(s[0]) | uint32(s[1])<<8 | uint32(s[2])<<16 | uint32(s[3])<<24
}
func load_le_byte4_uint32_inv(s []byte) uint32 {
// arm64:`MOVWU\t\(R[0-9]+\)`,-`ORR`,-`MOV[BH]`
return uint32(s[3])<<24 | uint32(s[2])<<16 | uint32(s[1])<<8 | uint32(s[0])
}
func load_le_byte8_uint64(s []byte) uint64 {
// arm64:`MOVD\t\(R[0-9]+\)`,-`ORR`,-`MOV[BHW]`
// amd64:`MOVQ\s\([A-Z]+\),\s[A-Z]+`,-`MOV[BWL]\t[^$]`,-`OR`
// ppc64le:`MOVD\t\(R[0-9]+\)`,-`MOV[BHW]Z`
return uint64(s[0]) | uint64(s[1])<<8 | uint64(s[2])<<16 | uint64(s[3])<<24 | uint64(s[4])<<32 | uint64(s[5])<<40 | uint64(s[6])<<48 | uint64(s[7])<<56
}
func load_le_byte8_uint64_inv(s []byte) uint64 {
// arm64:`MOVD\t\(R[0-9]+\)`,-`ORR`,-`MOV[BHW]`
return uint64(s[7])<<56 | uint64(s[6])<<48 | uint64(s[5])<<40 | uint64(s[4])<<32 | uint64(s[3])<<24 | uint64(s[2])<<16 | uint64(s[1])<<8 | uint64(s[0])
}
func load_be_byte2_uint16(s []byte) uint16 {
// arm64:`MOVHU\t\(R[0-9]+\)`,`REV16W`,-`ORR`,-`MOVB`
// amd64:`MOVWLZX\s\([A-Z]+\)`,`ROLW`,-`MOVB`,-`OR`
// ppc64le:`MOVHBR\t\(R[0-9]+\)`,-`MOVBZ`
return uint16(s[0])<<8 | uint16(s[1])
}
func load_be_byte2_uint16_inv(s []byte) uint16 {
// arm64:`MOVHU\t\(R[0-9]+\)`,`REV16W`,-`ORR`,-`MOVB`
// amd64:`MOVWLZX\s\([A-Z]+\)`,`ROLW`,-`MOVB`,-`OR`
// ppc64le:`MOVHBR\t\(R[0-9]+\)`,-`MOVBZ`
return uint16(s[1]) | uint16(s[0])<<8
}
func load_be_byte4_uint32(s []byte) uint32 {
// arm64:`MOVWU\t\(R[0-9]+\)`,`REVW`,-`ORR`,-`REV16W`,-`MOV[BH]`
return uint32(s[0])<<24 | uint32(s[1])<<16 | uint32(s[2])<<8 | uint32(s[3])
}
func load_be_byte4_uint32_inv(s []byte) uint32 {
// arm64:`MOVWU\t\(R[0-9]+\)`,`REVW`,-`ORR`,-`REV16W`,-`MOV[BH]`
// amd64:`MOVL\s\([A-Z]+\)`,`BSWAPL`,-`MOV[BW]`,-`OR`
return uint32(s[3]) | uint32(s[2])<<8 | uint32(s[1])<<16 | uint32(s[0])<<24
}
func load_be_byte8_uint64(s []byte) uint64 {
// arm64:`MOVD\t\(R[0-9]+\)`,`REV`,-`ORR`,-`REVW`,-`REV16W`,-`MOV[BHW]`
// ppc64le:`MOVDBR\t\(R[0-9]+\)`,-`MOV[BHW]Z`
return uint64(s[0])<<56 | uint64(s[1])<<48 | uint64(s[2])<<40 | uint64(s[3])<<32 | uint64(s[4])<<24 | uint64(s[5])<<16 | uint64(s[6])<<8 | uint64(s[7])
}
func load_be_byte8_uint64_inv(s []byte) uint64 {
// arm64:`MOVD\t\(R[0-9]+\)`,`REV`,-`ORR`,-`REVW`,-`REV16W`,-`MOV[BHW]`
// amd64:`MOVQ\s\([A-Z]+\),\s[A-Z]+`,`BSWAPQ`,-`MOV[BWL]\t[^$]`,-`OR`
// ppc64le:`MOVDBR\t\(R[0-9]+\)`,-`MOV[BHW]Z`
return uint64(s[7]) | uint64(s[6])<<8 | uint64(s[5])<<16 | uint64(s[4])<<24 | uint64(s[3])<<32 | uint64(s[2])<<40 | uint64(s[1])<<48 | uint64(s[0])<<56
}
func load_le_byte2_uint16_idx(s []byte, idx int) uint16 {
// arm64:`MOVHU\s\(R[0-9]+\)\(R[0-9]+\)`,-`ORR`,-`MOVB`
// 386:`MOVWLZX\s\([A-Z]+\)\([A-Z]+`,-`ORL`,-`MOVB`
// amd64:`MOVWLZX\s\([A-Z]+\)\([A-Z]+`,-`MOVB`,-`OR`
return uint16(s[idx]) | uint16(s[idx+1])<<8
}
func load_le_byte2_uint16_idx_inv(s []byte, idx int) uint16 {
// arm64:`MOVHU\s\(R[0-9]+\)\(R[0-9]+\)`,-`ORR`,-`MOVB`
// 386:`MOVWLZX\s\([A-Z]+\)\([A-Z]+`,-`ORL`,-`MOVB`
// amd64:`MOVWLZX\s\([A-Z]+\)\([A-Z]+`,-`MOVB`,-`OR`
return uint16(s[idx+1])<<8 | uint16(s[idx])
}
func load_le_byte4_uint32_idx(s []byte, idx int) uint32 {
// arm64:`MOVWU\s\(R[0-9]+\)\(R[0-9]+\)`,-`ORR`,-`MOV[BH]`
// amd64:`MOVL\s\([A-Z]+\)\([A-Z]+`,-`MOV[BW]`,-`OR`
return uint32(s[idx]) | uint32(s[idx+1])<<8 | uint32(s[idx+2])<<16 | uint32(s[idx+3])<<24
}
func load_le_byte4_uint32_idx_inv(s []byte, idx int) uint32 {
// arm64:`MOVWU\s\(R[0-9]+\)\(R[0-9]+\)`,-`ORR`,-`MOV[BH]`
return uint32(s[idx+3])<<24 | uint32(s[idx+2])<<16 | uint32(s[idx+1])<<8 | uint32(s[idx])
}
func load_le_byte8_uint64_idx(s []byte, idx int) uint64 {
// arm64:`MOVD\s\(R[0-9]+\)\(R[0-9]+\)`,-`ORR`,-`MOV[BHW]`
// amd64:`MOVQ\s\([A-Z]+\)\([A-Z]+`,-`MOV[BWL]`,-`OR`
return uint64(s[idx]) | uint64(s[idx+1])<<8 | uint64(s[idx+2])<<16 | uint64(s[idx+3])<<24 | uint64(s[idx+4])<<32 | uint64(s[idx+5])<<40 | uint64(s[idx+6])<<48 | uint64(s[idx+7])<<56
}
func load_le_byte8_uint64_idx_inv(s []byte, idx int) uint64 {
// arm64:`MOVD\s\(R[0-9]+\)\(R[0-9]+\)`,-`ORR`,-`MOV[BHW]`
return uint64(s[idx+7])<<56 | uint64(s[idx+6])<<48 | uint64(s[idx+5])<<40 | uint64(s[idx+4])<<32 | uint64(s[idx+3])<<24 | uint64(s[idx+2])<<16 | uint64(s[idx+1])<<8 | uint64(s[idx])
}
func load_be_byte2_uint16_idx(s []byte, idx int) uint16 {
// arm64:`MOVHU\s\(R[0-9]+\)\(R[0-9]+\)`,`REV16W`,-`ORR`,-`MOVB`
// amd64:`MOVWLZX\s\([A-Z]+\)\([A-Z]+`,-`MOVB`,-`OR`
return uint16(s[idx])<<8 | uint16(s[idx+1])
}
func load_be_byte2_uint16_idx_inv(s []byte, idx int) uint16 {
// arm64:`MOVHU\s\(R[0-9]+\)\(R[0-9]+\)`,`REV16W`,-`ORR`,-`MOVB`
// amd64:`MOVWLZX\s\([A-Z]+\)\([A-Z]+`,-`MOVB`,-`OR`
return uint16(s[idx+1]) | uint16(s[idx])<<8
}
func load_be_byte4_uint32_idx(s []byte, idx int) uint32 {
// arm64:`MOVWU\s\(R[0-9]+\)\(R[0-9]+\)`,`REVW`,-`ORR`,-`MOV[BH]`,-`REV16W`
return uint32(s[idx])<<24 | uint32(s[idx+1])<<16 | uint32(s[idx+2])<<8 | uint32(s[idx+3])
}
func load_be_byte8_uint64_idx(s []byte, idx int) uint64 {
// arm64:`MOVD\s\(R[0-9]+\)\(R[0-9]+\)`,`REV`,-`ORR`,-`MOV[BHW]`,-`REVW`,-`REV16W`
return uint64(s[idx])<<56 | uint64(s[idx+1])<<48 | uint64(s[idx+2])<<40 | uint64(s[idx+3])<<32 | uint64(s[idx+4])<<24 | uint64(s[idx+5])<<16 | uint64(s[idx+6])<<8 | uint64(s[idx+7])
}
func load_le_byte2_uint16_idx2(s []byte, idx int) uint16 {
cmd/compile: optimize ARM64 with shifted register indexed load/store ARM64 supports efficient instructions which combine shift, addition, load/store together. Such as "MOVD (R0)(R1<<3), R2" and "MOVWU R6, (R4)(R1<<2)". This CL optimizes the compiler to emit such efficient instuctions. And below is some test data. 1. binary size before/after binary size change pkg/linux_arm64 +80.1KB pkg/tool/linux_arm64 +121.9KB go -4.3KB gofmt -64KB 2. go1 benchmark There is big improvement for the test case Fannkuch11, and slight improvement for sme others, excluding noise. name old time/op new time/op delta BinaryTree17-4 43.9s ± 2% 44.0s ± 2% ~ (p=0.820 n=30+30) Fannkuch11-4 30.6s ± 2% 24.5s ± 3% -19.93% (p=0.000 n=25+30) FmtFprintfEmpty-4 500ns ± 0% 499ns ± 0% -0.11% (p=0.000 n=23+25) FmtFprintfString-4 1.03µs ± 0% 1.04µs ± 3% ~ (p=0.065 n=29+30) FmtFprintfInt-4 1.15µs ± 3% 1.15µs ± 4% -0.56% (p=0.000 n=30+30) FmtFprintfIntInt-4 1.80µs ± 5% 1.82µs ± 0% ~ (p=0.094 n=30+24) FmtFprintfPrefixedInt-4 2.17µs ± 5% 2.20µs ± 0% ~ (p=0.100 n=30+23) FmtFprintfFloat-4 3.08µs ± 3% 3.09µs ± 4% ~ (p=0.123 n=30+30) FmtManyArgs-4 7.41µs ± 4% 7.17µs ± 1% -3.26% (p=0.000 n=30+23) GobDecode-4 93.7ms ± 0% 94.7ms ± 4% ~ (p=0.685 n=24+30) GobEncode-4 78.7ms ± 7% 77.1ms ± 0% ~ (p=0.729 n=30+23) Gzip-4 4.01s ± 0% 3.97s ± 5% -1.11% (p=0.037 n=24+30) Gunzip-4 389ms ± 4% 384ms ± 0% ~ (p=0.155 n=30+23) HTTPClientServer-4 536µs ± 1% 537µs ± 1% ~ (p=0.236 n=30+30) JSONEncode-4 179ms ± 1% 182ms ± 6% ~ (p=0.763 n=24+30) JSONDecode-4 843ms ± 0% 839ms ± 6% -0.42% (p=0.003 n=25+30) Mandelbrot200-4 46.5ms ± 0% 46.5ms ± 0% +0.02% (p=0.000 n=26+26) GoParse-4 44.3ms ± 6% 43.3ms ± 0% ~ (p=0.067 n=30+27) RegexpMatchEasy0_32-4 1.07µs ± 7% 1.07µs ± 4% ~ (p=0.835 n=30+30) RegexpMatchEasy0_1K-4 5.51µs ± 0% 5.49µs ± 0% -0.35% (p=0.000 n=23+26) RegexpMatchEasy1_32-4 1.01µs ± 0% 1.02µs ± 4% +0.96% (p=0.014 n=24+30) RegexpMatchEasy1_1K-4 7.43µs ± 0% 7.18µs ± 0% -3.41% (p=0.000 n=23+24) RegexpMatchMedium_32-4 1.78µs ± 0% 1.81µs ± 4% +1.47% (p=0.012 n=23+30) RegexpMatchMedium_1K-4 547µs ± 1% 542µs ± 3% -0.90% (p=0.003 n=24+30) RegexpMatchHard_32-4 30.4µs ± 0% 29.7µs ± 0% -2.15% (p=0.000 n=19+23) RegexpMatchHard_1K-4 913µs ± 0% 915µs ± 6% +0.25% (p=0.012 n=24+30) Revcomp-4 6.32s ± 1% 6.42s ± 4% ~ (p=0.342 n=25+30) Template-4 868ms ± 6% 878ms ± 6% +1.15% (p=0.000 n=30+30) TimeParse-4 4.57µs ± 4% 4.59µs ± 3% +0.65% (p=0.010 n=29+30) TimeFormat-4 4.51µs ± 0% 4.50µs ± 0% -0.27% (p=0.000 n=27+24) [Geo mean] 695µs 689µs -0.92% name old speed new speed delta GobDecode-4 8.19MB/s ± 0% 8.12MB/s ± 4% ~ (p=0.680 n=24+30) GobEncode-4 9.76MB/s ± 7% 9.96MB/s ± 0% ~ (p=0.616 n=30+23) Gzip-4 4.84MB/s ± 0% 4.89MB/s ± 4% +1.16% (p=0.030 n=24+30) Gunzip-4 49.9MB/s ± 4% 50.6MB/s ± 0% ~ (p=0.162 n=30+23) JSONEncode-4 10.9MB/s ± 1% 10.7MB/s ± 6% ~ (p=0.575 n=24+30) JSONDecode-4 2.30MB/s ± 0% 2.32MB/s ± 5% +0.72% (p=0.003 n=22+30) GoParse-4 1.31MB/s ± 6% 1.34MB/s ± 0% +2.26% (p=0.002 n=30+27) RegexpMatchEasy0_32-4 30.0MB/s ± 6% 30.0MB/s ± 4% ~ (p=1.000 n=30+30) RegexpMatchEasy0_1K-4 186MB/s ± 0% 187MB/s ± 0% +0.35% (p=0.000 n=23+26) RegexpMatchEasy1_32-4 31.8MB/s ± 0% 31.5MB/s ± 4% -0.92% (p=0.012 n=25+30) RegexpMatchEasy1_1K-4 138MB/s ± 0% 143MB/s ± 0% +3.53% (p=0.000 n=23+24) RegexpMatchMedium_32-4 560kB/s ± 0% 553kB/s ± 4% -1.19% (p=0.005 n=23+30) RegexpMatchMedium_1K-4 1.87MB/s ± 0% 1.89MB/s ± 3% +1.04% (p=0.002 n=24+30) RegexpMatchHard_32-4 1.05MB/s ± 0% 1.08MB/s ± 0% +2.40% (p=0.000 n=19+23) RegexpMatchHard_1K-4 1.12MB/s ± 0% 1.12MB/s ± 5% +0.12% (p=0.006 n=25+30) Revcomp-4 40.2MB/s ± 1% 39.6MB/s ± 4% ~ (p=0.242 n=25+30) Template-4 2.24MB/s ± 6% 2.21MB/s ± 6% -1.15% (p=0.000 n=30+30) [Geo mean] 7.87MB/s 7.91MB/s +0.44% Change-Id: If374cb7abf83537aa0a176f73c0f736f7800db03 Reviewed-on: https://go-review.googlesource.com/108735 Reviewed-by: Cherry Zhang <cherryyz@google.com> Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2018-04-21 18:51:00 -06:00
// arm64:`MOVHU\s\(R[0-9]+\)\(R[0-9]+<<1\)`,-`ORR`,-`MOVB`
return uint16(s[idx<<1]) | uint16(s[(idx<<1)+1])<<8
}
func load_le_byte2_uint16_idx2_inv(s []byte, idx int) uint16 {
// arm64:`MOVHU\s\(R[0-9]+\)\(R[0-9]+<<1\)`,-`ORR`,-`MOVB`
return uint16(s[(idx<<1)+1])<<8 | uint16(s[idx<<1])
}
func load_le_byte4_uint32_idx4(s []byte, idx int) uint32 {
cmd/compile: optimize ARM64 with shifted register indexed load/store ARM64 supports efficient instructions which combine shift, addition, load/store together. Such as "MOVD (R0)(R1<<3), R2" and "MOVWU R6, (R4)(R1<<2)". This CL optimizes the compiler to emit such efficient instuctions. And below is some test data. 1. binary size before/after binary size change pkg/linux_arm64 +80.1KB pkg/tool/linux_arm64 +121.9KB go -4.3KB gofmt -64KB 2. go1 benchmark There is big improvement for the test case Fannkuch11, and slight improvement for sme others, excluding noise. name old time/op new time/op delta BinaryTree17-4 43.9s ± 2% 44.0s ± 2% ~ (p=0.820 n=30+30) Fannkuch11-4 30.6s ± 2% 24.5s ± 3% -19.93% (p=0.000 n=25+30) FmtFprintfEmpty-4 500ns ± 0% 499ns ± 0% -0.11% (p=0.000 n=23+25) FmtFprintfString-4 1.03µs ± 0% 1.04µs ± 3% ~ (p=0.065 n=29+30) FmtFprintfInt-4 1.15µs ± 3% 1.15µs ± 4% -0.56% (p=0.000 n=30+30) FmtFprintfIntInt-4 1.80µs ± 5% 1.82µs ± 0% ~ (p=0.094 n=30+24) FmtFprintfPrefixedInt-4 2.17µs ± 5% 2.20µs ± 0% ~ (p=0.100 n=30+23) FmtFprintfFloat-4 3.08µs ± 3% 3.09µs ± 4% ~ (p=0.123 n=30+30) FmtManyArgs-4 7.41µs ± 4% 7.17µs ± 1% -3.26% (p=0.000 n=30+23) GobDecode-4 93.7ms ± 0% 94.7ms ± 4% ~ (p=0.685 n=24+30) GobEncode-4 78.7ms ± 7% 77.1ms ± 0% ~ (p=0.729 n=30+23) Gzip-4 4.01s ± 0% 3.97s ± 5% -1.11% (p=0.037 n=24+30) Gunzip-4 389ms ± 4% 384ms ± 0% ~ (p=0.155 n=30+23) HTTPClientServer-4 536µs ± 1% 537µs ± 1% ~ (p=0.236 n=30+30) JSONEncode-4 179ms ± 1% 182ms ± 6% ~ (p=0.763 n=24+30) JSONDecode-4 843ms ± 0% 839ms ± 6% -0.42% (p=0.003 n=25+30) Mandelbrot200-4 46.5ms ± 0% 46.5ms ± 0% +0.02% (p=0.000 n=26+26) GoParse-4 44.3ms ± 6% 43.3ms ± 0% ~ (p=0.067 n=30+27) RegexpMatchEasy0_32-4 1.07µs ± 7% 1.07µs ± 4% ~ (p=0.835 n=30+30) RegexpMatchEasy0_1K-4 5.51µs ± 0% 5.49µs ± 0% -0.35% (p=0.000 n=23+26) RegexpMatchEasy1_32-4 1.01µs ± 0% 1.02µs ± 4% +0.96% (p=0.014 n=24+30) RegexpMatchEasy1_1K-4 7.43µs ± 0% 7.18µs ± 0% -3.41% (p=0.000 n=23+24) RegexpMatchMedium_32-4 1.78µs ± 0% 1.81µs ± 4% +1.47% (p=0.012 n=23+30) RegexpMatchMedium_1K-4 547µs ± 1% 542µs ± 3% -0.90% (p=0.003 n=24+30) RegexpMatchHard_32-4 30.4µs ± 0% 29.7µs ± 0% -2.15% (p=0.000 n=19+23) RegexpMatchHard_1K-4 913µs ± 0% 915µs ± 6% +0.25% (p=0.012 n=24+30) Revcomp-4 6.32s ± 1% 6.42s ± 4% ~ (p=0.342 n=25+30) Template-4 868ms ± 6% 878ms ± 6% +1.15% (p=0.000 n=30+30) TimeParse-4 4.57µs ± 4% 4.59µs ± 3% +0.65% (p=0.010 n=29+30) TimeFormat-4 4.51µs ± 0% 4.50µs ± 0% -0.27% (p=0.000 n=27+24) [Geo mean] 695µs 689µs -0.92% name old speed new speed delta GobDecode-4 8.19MB/s ± 0% 8.12MB/s ± 4% ~ (p=0.680 n=24+30) GobEncode-4 9.76MB/s ± 7% 9.96MB/s ± 0% ~ (p=0.616 n=30+23) Gzip-4 4.84MB/s ± 0% 4.89MB/s ± 4% +1.16% (p=0.030 n=24+30) Gunzip-4 49.9MB/s ± 4% 50.6MB/s ± 0% ~ (p=0.162 n=30+23) JSONEncode-4 10.9MB/s ± 1% 10.7MB/s ± 6% ~ (p=0.575 n=24+30) JSONDecode-4 2.30MB/s ± 0% 2.32MB/s ± 5% +0.72% (p=0.003 n=22+30) GoParse-4 1.31MB/s ± 6% 1.34MB/s ± 0% +2.26% (p=0.002 n=30+27) RegexpMatchEasy0_32-4 30.0MB/s ± 6% 30.0MB/s ± 4% ~ (p=1.000 n=30+30) RegexpMatchEasy0_1K-4 186MB/s ± 0% 187MB/s ± 0% +0.35% (p=0.000 n=23+26) RegexpMatchEasy1_32-4 31.8MB/s ± 0% 31.5MB/s ± 4% -0.92% (p=0.012 n=25+30) RegexpMatchEasy1_1K-4 138MB/s ± 0% 143MB/s ± 0% +3.53% (p=0.000 n=23+24) RegexpMatchMedium_32-4 560kB/s ± 0% 553kB/s ± 4% -1.19% (p=0.005 n=23+30) RegexpMatchMedium_1K-4 1.87MB/s ± 0% 1.89MB/s ± 3% +1.04% (p=0.002 n=24+30) RegexpMatchHard_32-4 1.05MB/s ± 0% 1.08MB/s ± 0% +2.40% (p=0.000 n=19+23) RegexpMatchHard_1K-4 1.12MB/s ± 0% 1.12MB/s ± 5% +0.12% (p=0.006 n=25+30) Revcomp-4 40.2MB/s ± 1% 39.6MB/s ± 4% ~ (p=0.242 n=25+30) Template-4 2.24MB/s ± 6% 2.21MB/s ± 6% -1.15% (p=0.000 n=30+30) [Geo mean] 7.87MB/s 7.91MB/s +0.44% Change-Id: If374cb7abf83537aa0a176f73c0f736f7800db03 Reviewed-on: https://go-review.googlesource.com/108735 Reviewed-by: Cherry Zhang <cherryyz@google.com> Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2018-04-21 18:51:00 -06:00
// arm64:`MOVWU\s\(R[0-9]+\)\(R[0-9]+<<2\)`,-`ORR`,-`MOV[BH]`
return uint32(s[idx<<2]) | uint32(s[(idx<<2)+1])<<8 | uint32(s[(idx<<2)+2])<<16 | uint32(s[(idx<<2)+3])<<24
}
func load_le_byte4_uint32_idx4_inv(s []byte, idx int) uint32 {
// arm64:`MOVWU\s\(R[0-9]+\)\(R[0-9]+<<2\)`,-`ORR`,-`MOV[BH]`
return uint32(s[(idx<<2)+3])<<24 | uint32(s[(idx<<2)+2])<<16 | uint32(s[(idx<<2)+1])<<8 | uint32(s[idx<<2])
}
func load_le_byte8_uint64_idx8(s []byte, idx int) uint64 {
cmd/compile: optimize ARM64 with shifted register indexed load/store ARM64 supports efficient instructions which combine shift, addition, load/store together. Such as "MOVD (R0)(R1<<3), R2" and "MOVWU R6, (R4)(R1<<2)". This CL optimizes the compiler to emit such efficient instuctions. And below is some test data. 1. binary size before/after binary size change pkg/linux_arm64 +80.1KB pkg/tool/linux_arm64 +121.9KB go -4.3KB gofmt -64KB 2. go1 benchmark There is big improvement for the test case Fannkuch11, and slight improvement for sme others, excluding noise. name old time/op new time/op delta BinaryTree17-4 43.9s ± 2% 44.0s ± 2% ~ (p=0.820 n=30+30) Fannkuch11-4 30.6s ± 2% 24.5s ± 3% -19.93% (p=0.000 n=25+30) FmtFprintfEmpty-4 500ns ± 0% 499ns ± 0% -0.11% (p=0.000 n=23+25) FmtFprintfString-4 1.03µs ± 0% 1.04µs ± 3% ~ (p=0.065 n=29+30) FmtFprintfInt-4 1.15µs ± 3% 1.15µs ± 4% -0.56% (p=0.000 n=30+30) FmtFprintfIntInt-4 1.80µs ± 5% 1.82µs ± 0% ~ (p=0.094 n=30+24) FmtFprintfPrefixedInt-4 2.17µs ± 5% 2.20µs ± 0% ~ (p=0.100 n=30+23) FmtFprintfFloat-4 3.08µs ± 3% 3.09µs ± 4% ~ (p=0.123 n=30+30) FmtManyArgs-4 7.41µs ± 4% 7.17µs ± 1% -3.26% (p=0.000 n=30+23) GobDecode-4 93.7ms ± 0% 94.7ms ± 4% ~ (p=0.685 n=24+30) GobEncode-4 78.7ms ± 7% 77.1ms ± 0% ~ (p=0.729 n=30+23) Gzip-4 4.01s ± 0% 3.97s ± 5% -1.11% (p=0.037 n=24+30) Gunzip-4 389ms ± 4% 384ms ± 0% ~ (p=0.155 n=30+23) HTTPClientServer-4 536µs ± 1% 537µs ± 1% ~ (p=0.236 n=30+30) JSONEncode-4 179ms ± 1% 182ms ± 6% ~ (p=0.763 n=24+30) JSONDecode-4 843ms ± 0% 839ms ± 6% -0.42% (p=0.003 n=25+30) Mandelbrot200-4 46.5ms ± 0% 46.5ms ± 0% +0.02% (p=0.000 n=26+26) GoParse-4 44.3ms ± 6% 43.3ms ± 0% ~ (p=0.067 n=30+27) RegexpMatchEasy0_32-4 1.07µs ± 7% 1.07µs ± 4% ~ (p=0.835 n=30+30) RegexpMatchEasy0_1K-4 5.51µs ± 0% 5.49µs ± 0% -0.35% (p=0.000 n=23+26) RegexpMatchEasy1_32-4 1.01µs ± 0% 1.02µs ± 4% +0.96% (p=0.014 n=24+30) RegexpMatchEasy1_1K-4 7.43µs ± 0% 7.18µs ± 0% -3.41% (p=0.000 n=23+24) RegexpMatchMedium_32-4 1.78µs ± 0% 1.81µs ± 4% +1.47% (p=0.012 n=23+30) RegexpMatchMedium_1K-4 547µs ± 1% 542µs ± 3% -0.90% (p=0.003 n=24+30) RegexpMatchHard_32-4 30.4µs ± 0% 29.7µs ± 0% -2.15% (p=0.000 n=19+23) RegexpMatchHard_1K-4 913µs ± 0% 915µs ± 6% +0.25% (p=0.012 n=24+30) Revcomp-4 6.32s ± 1% 6.42s ± 4% ~ (p=0.342 n=25+30) Template-4 868ms ± 6% 878ms ± 6% +1.15% (p=0.000 n=30+30) TimeParse-4 4.57µs ± 4% 4.59µs ± 3% +0.65% (p=0.010 n=29+30) TimeFormat-4 4.51µs ± 0% 4.50µs ± 0% -0.27% (p=0.000 n=27+24) [Geo mean] 695µs 689µs -0.92% name old speed new speed delta GobDecode-4 8.19MB/s ± 0% 8.12MB/s ± 4% ~ (p=0.680 n=24+30) GobEncode-4 9.76MB/s ± 7% 9.96MB/s ± 0% ~ (p=0.616 n=30+23) Gzip-4 4.84MB/s ± 0% 4.89MB/s ± 4% +1.16% (p=0.030 n=24+30) Gunzip-4 49.9MB/s ± 4% 50.6MB/s ± 0% ~ (p=0.162 n=30+23) JSONEncode-4 10.9MB/s ± 1% 10.7MB/s ± 6% ~ (p=0.575 n=24+30) JSONDecode-4 2.30MB/s ± 0% 2.32MB/s ± 5% +0.72% (p=0.003 n=22+30) GoParse-4 1.31MB/s ± 6% 1.34MB/s ± 0% +2.26% (p=0.002 n=30+27) RegexpMatchEasy0_32-4 30.0MB/s ± 6% 30.0MB/s ± 4% ~ (p=1.000 n=30+30) RegexpMatchEasy0_1K-4 186MB/s ± 0% 187MB/s ± 0% +0.35% (p=0.000 n=23+26) RegexpMatchEasy1_32-4 31.8MB/s ± 0% 31.5MB/s ± 4% -0.92% (p=0.012 n=25+30) RegexpMatchEasy1_1K-4 138MB/s ± 0% 143MB/s ± 0% +3.53% (p=0.000 n=23+24) RegexpMatchMedium_32-4 560kB/s ± 0% 553kB/s ± 4% -1.19% (p=0.005 n=23+30) RegexpMatchMedium_1K-4 1.87MB/s ± 0% 1.89MB/s ± 3% +1.04% (p=0.002 n=24+30) RegexpMatchHard_32-4 1.05MB/s ± 0% 1.08MB/s ± 0% +2.40% (p=0.000 n=19+23) RegexpMatchHard_1K-4 1.12MB/s ± 0% 1.12MB/s ± 5% +0.12% (p=0.006 n=25+30) Revcomp-4 40.2MB/s ± 1% 39.6MB/s ± 4% ~ (p=0.242 n=25+30) Template-4 2.24MB/s ± 6% 2.21MB/s ± 6% -1.15% (p=0.000 n=30+30) [Geo mean] 7.87MB/s 7.91MB/s +0.44% Change-Id: If374cb7abf83537aa0a176f73c0f736f7800db03 Reviewed-on: https://go-review.googlesource.com/108735 Reviewed-by: Cherry Zhang <cherryyz@google.com> Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2018-04-21 18:51:00 -06:00
// arm64:`MOVD\s\(R[0-9]+\)\(R[0-9]+<<3\)`,-`ORR`,-`MOV[BHW]`
return uint64(s[idx<<3]) | uint64(s[(idx<<3)+1])<<8 | uint64(s[(idx<<3)+2])<<16 | uint64(s[(idx<<3)+3])<<24 | uint64(s[(idx<<3)+4])<<32 | uint64(s[(idx<<3)+5])<<40 | uint64(s[(idx<<3)+6])<<48 | uint64(s[(idx<<3)+7])<<56
}
func load_le_byte8_uint64_idx8_inv(s []byte, idx int) uint64 {
// arm64:`MOVD\s\(R[0-9]+\)\(R[0-9]+<<3\)`,-`ORR`,-`MOV[BHW]`
return uint64(s[(idx<<3)+7])<<56 | uint64(s[(idx<<3)+6])<<48 | uint64(s[(idx<<3)+5])<<40 | uint64(s[(idx<<3)+4])<<32 | uint64(s[(idx<<3)+3])<<24 | uint64(s[(idx<<3)+2])<<16 | uint64(s[(idx<<3)+1])<<8 | uint64(s[idx<<3])
}
func load_be_byte2_uint16_idx2(s []byte, idx int) uint16 {
// arm64:`MOVHU\s\(R[0-9]+\)\(R[0-9]+<<1\)`,`REV16W`,-`ORR`,-`MOVB`
return uint16(s[idx<<1])<<8 | uint16(s[(idx<<1)+1])
}
func load_be_byte2_uint16_idx2_inv(s []byte, idx int) uint16 {
// arm64:`MOVHU\s\(R[0-9]+\)\(R[0-9]+<<1\)`,`REV16W`,-`ORR`,-`MOVB`
return uint16(s[(idx<<1)+1]) | uint16(s[idx<<1])<<8
}
func load_be_byte4_uint32_idx4(s []byte, idx int) uint32 {
// arm64:`MOVWU\s\(R[0-9]+\)\(R[0-9]+<<2\)`,`REVW`,-`ORR`,-`MOV[BH]`,-`REV16W`
return uint32(s[idx<<2])<<24 | uint32(s[(idx<<2)+1])<<16 | uint32(s[(idx<<2)+2])<<8 | uint32(s[(idx<<2)+3])
}
func load_be_byte8_uint64_idx8(s []byte, idx int) uint64 {
// arm64:`MOVD\s\(R[0-9]+\)\(R[0-9]+<<3\)`,`REV`,-`ORR`,-`MOV[BHW]`,-`REVW`,-`REV16W`
return uint64(s[idx<<3])<<56 | uint64(s[(idx<<3)+1])<<48 | uint64(s[(idx<<3)+2])<<40 | uint64(s[(idx<<3)+3])<<32 | uint64(s[(idx<<3)+4])<<24 | uint64(s[(idx<<3)+5])<<16 | uint64(s[(idx<<3)+6])<<8 | uint64(s[(idx<<3)+7])
}
// Check load combining across function calls.
func fcall_byte(a, b byte) (byte, byte) {
return fcall_byte(fcall_byte(a, b)) // amd64:`MOVW`
}
func fcall_uint16(a, b uint16) (uint16, uint16) {
return fcall_uint16(fcall_uint16(a, b)) // amd64:`MOVL`
}
func fcall_uint32(a, b uint32) (uint32, uint32) {
return fcall_uint32(fcall_uint32(a, b)) // amd64:`MOVQ`
}
// We want to merge load+op in the first function, but not in the
// second. See Issue 19595.
func load_op_merge(p, q *int) {
x := *p // amd64:`ADDQ\t\(`
*q += x // The combined nilcheck and load would normally have this line number, but we want that combined operation to have the line number of the nil check instead (see #33724).
}
func load_op_no_merge(p, q *int) {
x := *p
for i := 0; i < 10; i++ {
*q += x // amd64:`ADDQ\t[A-Z]`
}
}
// Make sure offsets are folded into loads and stores.
func offsets_fold(_, a [20]byte) (b [20]byte) {
// arm64:`MOVD\t""\.a\+[0-9]+\(FP\), R[0-9]+`,`MOVD\tR[0-9]+, ""\.b\+[0-9]+\(FP\)`
b = a
return
}
// Make sure we don't put pointers in SSE registers across safe
// points.
func safe_point(p, q *[2]*int) {
a, b := p[0], p[1] // amd64:-`MOVUPS`
runtime.GC()
q[0], q[1] = a, b // amd64:-`MOVUPS`
}
// ------------- //
// Storing //
// ------------- //
func store_le64(b []byte) {
// amd64:`MOVQ\s.*\(.*\)$`,-`SHR.`
// arm64:`MOVD`,-`MOV[WBH]`
// ppc64le:`MOVD\s`,-`MOV[BHW]\s`
// s390x:`MOVDBR\s.*\(.*\)$`
binary.LittleEndian.PutUint64(b, sink64)
}
func store_le64_idx(b []byte, idx int) {
// amd64:`MOVQ\s.*\(.*\)\(.*\*1\)$`,-`SHR.`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`MOVD\sR[0-9]+,\s\(R[0-9]+\)\(R[0-9]+\)`,-`MOV[BHW]`
// ppc64le:`MOVD\s`,-`MOV[BHW]\s`
// s390x:`MOVDBR\s.*\(.*\)\(.*\*1\)$`
binary.LittleEndian.PutUint64(b[idx:], sink64)
}
func store_le32(b []byte) {
// amd64:`MOVL\s`
// arm64:`MOVW`,-`MOV[BH]`
// ppc64le:`MOVW\s`
// s390x:`MOVWBR\s.*\(.*\)$`
binary.LittleEndian.PutUint32(b, sink32)
}
func store_le32_idx(b []byte, idx int) {
// amd64:`MOVL\s`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`MOVW\sR[0-9]+,\s\(R[0-9]+\)\(R[0-9]+\)`,-`MOV[BH]`
// ppc64le:`MOVW\s`
// s390x:`MOVWBR\s.*\(.*\)\(.*\*1\)$`
binary.LittleEndian.PutUint32(b[idx:], sink32)
}
func store_le16(b []byte) {
// amd64:`MOVW\s`
// arm64:`MOVH`,-`MOVB`
cmd/compile,test: combine byte loads and stores on ppc64le CL 74410 added rules to combine consecutive byte loads and stores when the byte order was little endian for ppc64le. This is the corresponding change for bytes that are in big endian order. These rules are all intended for a little endian target arch. This adds new testcases in test/codegen/memcombine.go Fixes #22496 Updates #24242 Benchmark improvement for encoding/binary: name old time/op new time/op delta ReadSlice1000Int32s-16 11.0µs ± 0% 9.0µs ± 0% -17.47% (p=0.029 n=4+4) ReadStruct-16 2.47µs ± 1% 2.48µs ± 0% +0.67% (p=0.114 n=4+4) ReadInts-16 642ns ± 1% 630ns ± 1% -2.02% (p=0.029 n=4+4) WriteInts-16 654ns ± 0% 653ns ± 1% -0.08% (p=0.629 n=4+4) WriteSlice1000Int32s-16 8.75µs ± 0% 8.20µs ± 0% -6.19% (p=0.029 n=4+4) PutUint16-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint32-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint64-16 1.85ns ± 0% 0.93ns ± 0% -49.73% (p=0.029 n=4+4) LittleEndianPutUint16-16 1.03ns ± 0% 0.93ns ± 0% -9.71% (p=0.029 n=4+4) LittleEndianPutUint32-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) LittleEndianPutUint64-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) PutUvarint32-16 43.0ns ± 0% 43.1ns ± 0% +0.12% (p=0.429 n=4+4) PutUvarint64-16 174ns ± 0% 175ns ± 0% +0.29% (p=0.429 n=4+4) Updates made to functions in gcm.go to enable their matching. An existing testcase prevents these functions from being replaced by those in encoding/binary due to import dependencies. Change-Id: Idb3bd1e6e7b12d86cd828fb29cb095848a3e485a Reviewed-on: https://go-review.googlesource.com/98136 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-03-01 09:40:36 -07:00
// ppc64le:`MOVH\s`
// s390x:`MOVHBR\s.*\(.*\)$`
binary.LittleEndian.PutUint16(b, sink16)
}
func store_le16_idx(b []byte, idx int) {
// amd64:`MOVW\s`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`MOVH\sR[0-9]+,\s\(R[0-9]+\)\(R[0-9]+\)`,-`MOVB`
cmd/compile,test: combine byte loads and stores on ppc64le CL 74410 added rules to combine consecutive byte loads and stores when the byte order was little endian for ppc64le. This is the corresponding change for bytes that are in big endian order. These rules are all intended for a little endian target arch. This adds new testcases in test/codegen/memcombine.go Fixes #22496 Updates #24242 Benchmark improvement for encoding/binary: name old time/op new time/op delta ReadSlice1000Int32s-16 11.0µs ± 0% 9.0µs ± 0% -17.47% (p=0.029 n=4+4) ReadStruct-16 2.47µs ± 1% 2.48µs ± 0% +0.67% (p=0.114 n=4+4) ReadInts-16 642ns ± 1% 630ns ± 1% -2.02% (p=0.029 n=4+4) WriteInts-16 654ns ± 0% 653ns ± 1% -0.08% (p=0.629 n=4+4) WriteSlice1000Int32s-16 8.75µs ± 0% 8.20µs ± 0% -6.19% (p=0.029 n=4+4) PutUint16-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint32-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint64-16 1.85ns ± 0% 0.93ns ± 0% -49.73% (p=0.029 n=4+4) LittleEndianPutUint16-16 1.03ns ± 0% 0.93ns ± 0% -9.71% (p=0.029 n=4+4) LittleEndianPutUint32-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) LittleEndianPutUint64-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) PutUvarint32-16 43.0ns ± 0% 43.1ns ± 0% +0.12% (p=0.429 n=4+4) PutUvarint64-16 174ns ± 0% 175ns ± 0% +0.29% (p=0.429 n=4+4) Updates made to functions in gcm.go to enable their matching. An existing testcase prevents these functions from being replaced by those in encoding/binary due to import dependencies. Change-Id: Idb3bd1e6e7b12d86cd828fb29cb095848a3e485a Reviewed-on: https://go-review.googlesource.com/98136 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-03-01 09:40:36 -07:00
// ppc64le:`MOVH\s`
// s390x:`MOVHBR\s.*\(.*\)\(.*\*1\)$`
binary.LittleEndian.PutUint16(b[idx:], sink16)
}
func store_be64(b []byte) {
// amd64:`BSWAPQ`,-`SHR.`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`MOVD`,`REV`,-`MOV[WBH]`,-`REVW`,-`REV16W`
cmd/compile,test: combine byte loads and stores on ppc64le CL 74410 added rules to combine consecutive byte loads and stores when the byte order was little endian for ppc64le. This is the corresponding change for bytes that are in big endian order. These rules are all intended for a little endian target arch. This adds new testcases in test/codegen/memcombine.go Fixes #22496 Updates #24242 Benchmark improvement for encoding/binary: name old time/op new time/op delta ReadSlice1000Int32s-16 11.0µs ± 0% 9.0µs ± 0% -17.47% (p=0.029 n=4+4) ReadStruct-16 2.47µs ± 1% 2.48µs ± 0% +0.67% (p=0.114 n=4+4) ReadInts-16 642ns ± 1% 630ns ± 1% -2.02% (p=0.029 n=4+4) WriteInts-16 654ns ± 0% 653ns ± 1% -0.08% (p=0.629 n=4+4) WriteSlice1000Int32s-16 8.75µs ± 0% 8.20µs ± 0% -6.19% (p=0.029 n=4+4) PutUint16-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint32-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint64-16 1.85ns ± 0% 0.93ns ± 0% -49.73% (p=0.029 n=4+4) LittleEndianPutUint16-16 1.03ns ± 0% 0.93ns ± 0% -9.71% (p=0.029 n=4+4) LittleEndianPutUint32-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) LittleEndianPutUint64-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) PutUvarint32-16 43.0ns ± 0% 43.1ns ± 0% +0.12% (p=0.429 n=4+4) PutUvarint64-16 174ns ± 0% 175ns ± 0% +0.29% (p=0.429 n=4+4) Updates made to functions in gcm.go to enable their matching. An existing testcase prevents these functions from being replaced by those in encoding/binary due to import dependencies. Change-Id: Idb3bd1e6e7b12d86cd828fb29cb095848a3e485a Reviewed-on: https://go-review.googlesource.com/98136 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-03-01 09:40:36 -07:00
// ppc64le:`MOVDBR`
// s390x:`MOVD\s.*\(.*\)$`,-`SRW\s`,-`SRD\s`
binary.BigEndian.PutUint64(b, sink64)
}
func store_be64_idx(b []byte, idx int) {
// amd64:`BSWAPQ`,-`SHR.`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`REV`,`MOVD\sR[0-9]+,\s\(R[0-9]+\)\(R[0-9]+\)`,-`MOV[BHW]`,-`REV16W`,-`REVW`
cmd/compile,test: combine byte loads and stores on ppc64le CL 74410 added rules to combine consecutive byte loads and stores when the byte order was little endian for ppc64le. This is the corresponding change for bytes that are in big endian order. These rules are all intended for a little endian target arch. This adds new testcases in test/codegen/memcombine.go Fixes #22496 Updates #24242 Benchmark improvement for encoding/binary: name old time/op new time/op delta ReadSlice1000Int32s-16 11.0µs ± 0% 9.0µs ± 0% -17.47% (p=0.029 n=4+4) ReadStruct-16 2.47µs ± 1% 2.48µs ± 0% +0.67% (p=0.114 n=4+4) ReadInts-16 642ns ± 1% 630ns ± 1% -2.02% (p=0.029 n=4+4) WriteInts-16 654ns ± 0% 653ns ± 1% -0.08% (p=0.629 n=4+4) WriteSlice1000Int32s-16 8.75µs ± 0% 8.20µs ± 0% -6.19% (p=0.029 n=4+4) PutUint16-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint32-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint64-16 1.85ns ± 0% 0.93ns ± 0% -49.73% (p=0.029 n=4+4) LittleEndianPutUint16-16 1.03ns ± 0% 0.93ns ± 0% -9.71% (p=0.029 n=4+4) LittleEndianPutUint32-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) LittleEndianPutUint64-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) PutUvarint32-16 43.0ns ± 0% 43.1ns ± 0% +0.12% (p=0.429 n=4+4) PutUvarint64-16 174ns ± 0% 175ns ± 0% +0.29% (p=0.429 n=4+4) Updates made to functions in gcm.go to enable their matching. An existing testcase prevents these functions from being replaced by those in encoding/binary due to import dependencies. Change-Id: Idb3bd1e6e7b12d86cd828fb29cb095848a3e485a Reviewed-on: https://go-review.googlesource.com/98136 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-03-01 09:40:36 -07:00
// ppc64le:`MOVDBR`
// s390x:`MOVD\s.*\(.*\)\(.*\*1\)$`,-`SRW\s`,-`SRD\s`
binary.BigEndian.PutUint64(b[idx:], sink64)
}
func store_be32(b []byte) {
// amd64:`BSWAPL`,-`SHR.`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`MOVW`,`REVW`,-`MOV[BH]`,-`REV16W`
cmd/compile,test: combine byte loads and stores on ppc64le CL 74410 added rules to combine consecutive byte loads and stores when the byte order was little endian for ppc64le. This is the corresponding change for bytes that are in big endian order. These rules are all intended for a little endian target arch. This adds new testcases in test/codegen/memcombine.go Fixes #22496 Updates #24242 Benchmark improvement for encoding/binary: name old time/op new time/op delta ReadSlice1000Int32s-16 11.0µs ± 0% 9.0µs ± 0% -17.47% (p=0.029 n=4+4) ReadStruct-16 2.47µs ± 1% 2.48µs ± 0% +0.67% (p=0.114 n=4+4) ReadInts-16 642ns ± 1% 630ns ± 1% -2.02% (p=0.029 n=4+4) WriteInts-16 654ns ± 0% 653ns ± 1% -0.08% (p=0.629 n=4+4) WriteSlice1000Int32s-16 8.75µs ± 0% 8.20µs ± 0% -6.19% (p=0.029 n=4+4) PutUint16-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint32-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint64-16 1.85ns ± 0% 0.93ns ± 0% -49.73% (p=0.029 n=4+4) LittleEndianPutUint16-16 1.03ns ± 0% 0.93ns ± 0% -9.71% (p=0.029 n=4+4) LittleEndianPutUint32-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) LittleEndianPutUint64-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) PutUvarint32-16 43.0ns ± 0% 43.1ns ± 0% +0.12% (p=0.429 n=4+4) PutUvarint64-16 174ns ± 0% 175ns ± 0% +0.29% (p=0.429 n=4+4) Updates made to functions in gcm.go to enable their matching. An existing testcase prevents these functions from being replaced by those in encoding/binary due to import dependencies. Change-Id: Idb3bd1e6e7b12d86cd828fb29cb095848a3e485a Reviewed-on: https://go-review.googlesource.com/98136 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-03-01 09:40:36 -07:00
// ppc64le:`MOVWBR`
// s390x:`MOVW\s.*\(.*\)$`,-`SRW\s`,-`SRD\s`
binary.BigEndian.PutUint32(b, sink32)
}
func store_be32_idx(b []byte, idx int) {
// amd64:`BSWAPL`,-`SHR.`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`REVW`,`MOVW\sR[0-9]+,\s\(R[0-9]+\)\(R[0-9]+\)`,-`MOV[BH]`,-`REV16W`
cmd/compile,test: combine byte loads and stores on ppc64le CL 74410 added rules to combine consecutive byte loads and stores when the byte order was little endian for ppc64le. This is the corresponding change for bytes that are in big endian order. These rules are all intended for a little endian target arch. This adds new testcases in test/codegen/memcombine.go Fixes #22496 Updates #24242 Benchmark improvement for encoding/binary: name old time/op new time/op delta ReadSlice1000Int32s-16 11.0µs ± 0% 9.0µs ± 0% -17.47% (p=0.029 n=4+4) ReadStruct-16 2.47µs ± 1% 2.48µs ± 0% +0.67% (p=0.114 n=4+4) ReadInts-16 642ns ± 1% 630ns ± 1% -2.02% (p=0.029 n=4+4) WriteInts-16 654ns ± 0% 653ns ± 1% -0.08% (p=0.629 n=4+4) WriteSlice1000Int32s-16 8.75µs ± 0% 8.20µs ± 0% -6.19% (p=0.029 n=4+4) PutUint16-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint32-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint64-16 1.85ns ± 0% 0.93ns ± 0% -49.73% (p=0.029 n=4+4) LittleEndianPutUint16-16 1.03ns ± 0% 0.93ns ± 0% -9.71% (p=0.029 n=4+4) LittleEndianPutUint32-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) LittleEndianPutUint64-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) PutUvarint32-16 43.0ns ± 0% 43.1ns ± 0% +0.12% (p=0.429 n=4+4) PutUvarint64-16 174ns ± 0% 175ns ± 0% +0.29% (p=0.429 n=4+4) Updates made to functions in gcm.go to enable their matching. An existing testcase prevents these functions from being replaced by those in encoding/binary due to import dependencies. Change-Id: Idb3bd1e6e7b12d86cd828fb29cb095848a3e485a Reviewed-on: https://go-review.googlesource.com/98136 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-03-01 09:40:36 -07:00
// ppc64le:`MOVWBR`
// s390x:`MOVW\s.*\(.*\)\(.*\*1\)$`,-`SRW\s`,-`SRD\s`
binary.BigEndian.PutUint32(b[idx:], sink32)
}
func store_be16(b []byte) {
// amd64:`ROLW\s\$8`,-`SHR.`
// arm64:`MOVH`,`REV16W`,-`MOVB`
cmd/compile,test: combine byte loads and stores on ppc64le CL 74410 added rules to combine consecutive byte loads and stores when the byte order was little endian for ppc64le. This is the corresponding change for bytes that are in big endian order. These rules are all intended for a little endian target arch. This adds new testcases in test/codegen/memcombine.go Fixes #22496 Updates #24242 Benchmark improvement for encoding/binary: name old time/op new time/op delta ReadSlice1000Int32s-16 11.0µs ± 0% 9.0µs ± 0% -17.47% (p=0.029 n=4+4) ReadStruct-16 2.47µs ± 1% 2.48µs ± 0% +0.67% (p=0.114 n=4+4) ReadInts-16 642ns ± 1% 630ns ± 1% -2.02% (p=0.029 n=4+4) WriteInts-16 654ns ± 0% 653ns ± 1% -0.08% (p=0.629 n=4+4) WriteSlice1000Int32s-16 8.75µs ± 0% 8.20µs ± 0% -6.19% (p=0.029 n=4+4) PutUint16-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint32-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint64-16 1.85ns ± 0% 0.93ns ± 0% -49.73% (p=0.029 n=4+4) LittleEndianPutUint16-16 1.03ns ± 0% 0.93ns ± 0% -9.71% (p=0.029 n=4+4) LittleEndianPutUint32-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) LittleEndianPutUint64-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) PutUvarint32-16 43.0ns ± 0% 43.1ns ± 0% +0.12% (p=0.429 n=4+4) PutUvarint64-16 174ns ± 0% 175ns ± 0% +0.29% (p=0.429 n=4+4) Updates made to functions in gcm.go to enable their matching. An existing testcase prevents these functions from being replaced by those in encoding/binary due to import dependencies. Change-Id: Idb3bd1e6e7b12d86cd828fb29cb095848a3e485a Reviewed-on: https://go-review.googlesource.com/98136 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-03-01 09:40:36 -07:00
// ppc64le:`MOVHBR`
// s390x:`MOVH\s.*\(.*\)$`,-`SRW\s`,-`SRD\s`
binary.BigEndian.PutUint16(b, sink16)
}
func store_be16_idx(b []byte, idx int) {
// amd64:`ROLW\s\$8`,-`SHR.`
cmd/compile: optimize ARM64 with register indexed load/store ARM64 supports load/store instructions with a memory operand that the address is calculated by base register + index register. In this CL, 1. Some rules are added to the compile's ARM64 backend to emit such efficient instructions. 2. A wrong rule of load combination is fixed. The go1 benchmark does show improvement. name old time/op new time/op delta BinaryTree17-4 44.5s ± 2% 44.1s ± 1% -0.81% (p=0.000 n=28+29) Fannkuch11-4 32.7s ± 3% 30.5s ± 0% -6.79% (p=0.000 n=30+26) FmtFprintfEmpty-4 499ns ± 0% 506ns ± 5% +1.39% (p=0.003 n=25+30) FmtFprintfString-4 1.07µs ± 0% 1.04µs ± 4% -3.17% (p=0.000 n=23+30) FmtFprintfInt-4 1.15µs ± 4% 1.13µs ± 0% -1.55% (p=0.000 n=30+23) FmtFprintfIntInt-4 1.77µs ± 4% 1.74µs ± 0% -1.71% (p=0.000 n=30+24) FmtFprintfPrefixedInt-4 2.37µs ± 5% 2.12µs ± 0% -10.56% (p=0.000 n=30+23) FmtFprintfFloat-4 3.03µs ± 1% 3.03µs ± 4% -0.13% (p=0.003 n=25+30) FmtManyArgs-4 7.38µs ± 1% 7.43µs ± 4% +0.59% (p=0.003 n=25+30) GobDecode-4 101ms ± 6% 95ms ± 5% -5.55% (p=0.000 n=30+30) GobEncode-4 78.0ms ± 4% 78.8ms ± 6% +1.05% (p=0.000 n=30+30) Gzip-4 4.25s ± 0% 4.27s ± 4% +0.45% (p=0.003 n=24+30) Gunzip-4 428ms ± 1% 420ms ± 0% -1.88% (p=0.000 n=23+23) HTTPClientServer-4 549µs ± 1% 541µs ± 1% -1.56% (p=0.000 n=29+29) JSONEncode-4 194ms ± 0% 188ms ± 4% ~ (p=0.417 n=23+30) JSONDecode-4 890ms ± 5% 831ms ± 0% -6.55% (p=0.000 n=30+23) Mandelbrot200-4 47.3ms ± 2% 46.5ms ± 0% ~ (p=0.980 n=30+26) GoParse-4 43.1ms ± 6% 43.8ms ± 6% +1.65% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 1.06µs ± 0% 1.07µs ± 3% ~ (p=0.092 n=23+30) RegexpMatchEasy0_1K-4 5.53µs ± 0% 5.51µs ± 0% -0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 1.02µs ± 3% 1.01µs ± 0% -1.27% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 7.26µs ± 0% 7.33µs ± 0% +0.95% (p=0.000 n=23+26) RegexpMatchMedium_32-4 1.84µs ± 7% 1.79µs ± 1% ~ (p=0.333 n=30+23) RegexpMatchMedium_1K-4 553µs ± 0% 547µs ± 0% -1.14% (p=0.000 n=24+22) RegexpMatchHard_32-4 30.8µs ± 1% 30.3µs ± 0% -1.40% (p=0.000 n=24+24) RegexpMatchHard_1K-4 928µs ± 0% 929µs ± 5% +0.12% (p=0.013 n=23+30) Revcomp-4 8.13s ± 4% 6.32s ± 1% -22.23% (p=0.000 n=30+23) Template-4 899ms ± 6% 854ms ± 1% -5.01% (p=0.000 n=30+24) TimeParse-4 4.66µs ± 4% 4.59µs ± 1% -1.57% (p=0.000 n=30+23) TimeFormat-4 4.58µs ± 0% 4.61µs ± 0% +0.57% (p=0.000 n=26+24) [Geo mean] 717µs 698µs -2.55% name old speed new speed delta GobDecode-4 7.63MB/s ± 6% 8.08MB/s ± 5% +5.88% (p=0.000 n=30+30) GobEncode-4 9.85MB/s ± 4% 9.75MB/s ± 6% -1.04% (p=0.000 n=30+30) Gzip-4 4.56MB/s ± 0% 4.55MB/s ± 4% -0.36% (p=0.003 n=24+30) Gunzip-4 45.3MB/s ± 1% 46.2MB/s ± 0% +1.92% (p=0.000 n=23+23) JSONEncode-4 10.0MB/s ± 0% 10.4MB/s ± 4% ~ (p=0.403 n=23+30) JSONDecode-4 2.18MB/s ± 5% 2.33MB/s ± 0% +6.91% (p=0.000 n=30+23) GoParse-4 1.34MB/s ± 5% 1.32MB/s ± 5% -1.66% (p=0.000 n=30+30) RegexpMatchEasy0_32-4 30.2MB/s ± 0% 29.8MB/s ± 3% ~ (p=0.099 n=23+30) RegexpMatchEasy0_1K-4 185MB/s ± 0% 186MB/s ± 0% +0.24% (p=0.000 n=25+25) RegexpMatchEasy1_32-4 31.4MB/s ± 3% 31.8MB/s ± 0% +1.24% (p=0.000 n=30+24) RegexpMatchEasy1_1K-4 141MB/s ± 0% 140MB/s ± 0% -0.94% (p=0.000 n=23+26) RegexpMatchMedium_32-4 541kB/s ± 6% 560kB/s ± 0% +3.45% (p=0.000 n=30+23) RegexpMatchMedium_1K-4 1.85MB/s ± 0% 1.87MB/s ± 0% +1.08% (p=0.000 n=24+23) RegexpMatchHard_32-4 1.04MB/s ± 1% 1.06MB/s ± 1% +1.48% (p=0.000 n=24+24) RegexpMatchHard_1K-4 1.10MB/s ± 0% 1.10MB/s ± 5% +0.15% (p=0.004 n=23+30) Revcomp-4 31.3MB/s ± 4% 40.2MB/s ± 1% +28.52% (p=0.000 n=30+23) Template-4 2.16MB/s ± 6% 2.27MB/s ± 1% +5.18% (p=0.000 n=30+24) [Geo mean] 7.57MB/s 7.79MB/s +2.98% fixes #24907 Change-Id: I94afd0e3f53d62a1cf5e452f3dd6daf61be21785 Reviewed-on: https://go-review.googlesource.com/107376 Run-TryBot: Cherry Zhang <cherryyz@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cherry Zhang <cherryyz@google.com>
2018-04-16 08:04:26 -06:00
// arm64:`MOVH\sR[0-9]+,\s\(R[0-9]+\)\(R[0-9]+\)`,`REV16W`,-`MOVB`
cmd/compile,test: combine byte loads and stores on ppc64le CL 74410 added rules to combine consecutive byte loads and stores when the byte order was little endian for ppc64le. This is the corresponding change for bytes that are in big endian order. These rules are all intended for a little endian target arch. This adds new testcases in test/codegen/memcombine.go Fixes #22496 Updates #24242 Benchmark improvement for encoding/binary: name old time/op new time/op delta ReadSlice1000Int32s-16 11.0µs ± 0% 9.0µs ± 0% -17.47% (p=0.029 n=4+4) ReadStruct-16 2.47µs ± 1% 2.48µs ± 0% +0.67% (p=0.114 n=4+4) ReadInts-16 642ns ± 1% 630ns ± 1% -2.02% (p=0.029 n=4+4) WriteInts-16 654ns ± 0% 653ns ± 1% -0.08% (p=0.629 n=4+4) WriteSlice1000Int32s-16 8.75µs ± 0% 8.20µs ± 0% -6.19% (p=0.029 n=4+4) PutUint16-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint32-16 1.16ns ± 0% 0.93ns ± 0% -19.83% (p=0.029 n=4+4) PutUint64-16 1.85ns ± 0% 0.93ns ± 0% -49.73% (p=0.029 n=4+4) LittleEndianPutUint16-16 1.03ns ± 0% 0.93ns ± 0% -9.71% (p=0.029 n=4+4) LittleEndianPutUint32-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) LittleEndianPutUint64-16 0.93ns ± 0% 0.93ns ± 0% ~ (all equal) PutUvarint32-16 43.0ns ± 0% 43.1ns ± 0% +0.12% (p=0.429 n=4+4) PutUvarint64-16 174ns ± 0% 175ns ± 0% +0.29% (p=0.429 n=4+4) Updates made to functions in gcm.go to enable their matching. An existing testcase prevents these functions from being replaced by those in encoding/binary due to import dependencies. Change-Id: Idb3bd1e6e7b12d86cd828fb29cb095848a3e485a Reviewed-on: https://go-review.googlesource.com/98136 Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-03-01 09:40:36 -07:00
// ppc64le:`MOVHBR`
// s390x:`MOVH\s.*\(.*\)\(.*\*1\)$`,-`SRW\s`,-`SRD\s`
binary.BigEndian.PutUint16(b[idx:], sink16)
}
func store_le_byte_2(b []byte, val uint16) {
_ = b[2]
// arm64:`MOVH\sR[0-9]+,\s1\(R[0-9]+\)`,-`MOVB`
cmd/compile: optimize 386's load/store combination This CL adds more combinations of two consequtive MOVBload/MOVBstore to a unique MOVWload/MOVWstore. 1. The size of the go executable decreases about 4KB, and the total size of pkg/linux_386 (excluding cmd/compile) decreases about 1.5KB. 2. There is no regression in the go1 benchmark result, excluding noise. name old time/op new time/op delta BinaryTree17-4 3.28s ± 2% 3.29s ± 2% ~ (p=0.151 n=40+40) Fannkuch11-4 3.52s ± 1% 3.51s ± 1% -0.28% (p=0.002 n=40+40) FmtFprintfEmpty-4 45.4ns ± 4% 45.0ns ± 4% -0.89% (p=0.019 n=40+40) FmtFprintfString-4 81.9ns ± 7% 81.3ns ± 1% ~ (p=0.660 n=40+25) FmtFprintfInt-4 91.9ns ± 9% 91.4ns ± 9% ~ (p=0.249 n=40+40) FmtFprintfIntInt-4 143ns ± 4% 143ns ± 4% ~ (p=0.760 n=40+40) FmtFprintfPrefixedInt-4 184ns ± 3% 183ns ± 4% ~ (p=0.485 n=40+40) FmtFprintfFloat-4 408ns ± 3% 409ns ± 3% ~ (p=0.961 n=40+40) FmtManyArgs-4 597ns ± 4% 602ns ± 3% ~ (p=0.413 n=40+40) GobDecode-4 7.13ms ± 6% 7.14ms ± 6% ~ (p=0.859 n=40+40) GobEncode-4 6.86ms ± 9% 6.94ms ± 7% ~ (p=0.162 n=40+40) Gzip-4 395ms ± 4% 396ms ± 3% ~ (p=0.099 n=40+40) Gunzip-4 40.9ms ± 4% 41.1ms ± 3% ~ (p=0.064 n=40+40) HTTPClientServer-4 63.6µs ± 2% 63.6µs ± 3% ~ (p=0.832 n=36+39) JSONEncode-4 16.1ms ± 3% 15.8ms ± 3% -1.60% (p=0.001 n=40+40) JSONDecode-4 61.0ms ± 3% 61.5ms ± 4% ~ (p=0.065 n=40+40) Mandelbrot200-4 5.16ms ± 3% 5.18ms ± 3% ~ (p=0.056 n=40+40) GoParse-4 3.25ms ± 2% 3.23ms ± 3% ~ (p=0.727 n=40+40) RegexpMatchEasy0_32-4 90.2ns ± 3% 89.3ns ± 6% -0.98% (p=0.002 n=40+40) RegexpMatchEasy0_1K-4 812ns ± 3% 815ns ± 3% ~ (p=0.309 n=40+40) RegexpMatchEasy1_32-4 103ns ± 6% 103ns ± 5% ~ (p=0.680 n=40+40) RegexpMatchEasy1_1K-4 1.01µs ± 4% 1.02µs ± 3% ~ (p=0.326 n=40+33) RegexpMatchMedium_32-4 120ns ± 4% 120ns ± 5% ~ (p=0.834 n=40+40) RegexpMatchMedium_1K-4 40.1µs ± 3% 39.5µs ± 4% -1.35% (p=0.000 n=40+40) RegexpMatchHard_32-4 2.27µs ± 6% 2.23µs ± 4% -1.67% (p=0.011 n=40+40) RegexpMatchHard_1K-4 67.2µs ± 3% 67.2µs ± 3% ~ (p=0.149 n=40+40) Revcomp-4 1.84s ± 2% 1.86s ± 3% +0.70% (p=0.020 n=40+40) Template-4 69.0ms ± 4% 69.8ms ± 3% +1.20% (p=0.003 n=40+40) TimeParse-4 438ns ± 3% 439ns ± 4% ~ (p=0.650 n=40+40) TimeFormat-4 412ns ± 3% 412ns ± 3% ~ (p=0.888 n=40+40) [Geo mean] 65.2µs 65.2µs -0.04% name old speed new speed delta GobDecode-4 108MB/s ± 6% 108MB/s ± 6% ~ (p=0.855 n=40+40) GobEncode-4 112MB/s ± 9% 111MB/s ± 8% ~ (p=0.159 n=40+40) Gzip-4 49.2MB/s ± 4% 49.1MB/s ± 3% ~ (p=0.102 n=40+40) Gunzip-4 474MB/s ± 3% 472MB/s ± 3% ~ (p=0.063 n=40+40) JSONEncode-4 121MB/s ± 3% 123MB/s ± 3% +1.62% (p=0.001 n=40+40) JSONDecode-4 31.9MB/s ± 3% 31.6MB/s ± 4% ~ (p=0.070 n=40+40) GoParse-4 17.9MB/s ± 2% 17.9MB/s ± 3% ~ (p=0.696 n=40+40) RegexpMatchEasy0_32-4 355MB/s ± 3% 358MB/s ± 5% +0.99% (p=0.002 n=40+40) RegexpMatchEasy0_1K-4 1.26GB/s ± 3% 1.26GB/s ± 3% ~ (p=0.381 n=40+40) RegexpMatchEasy1_32-4 310MB/s ± 5% 310MB/s ± 4% ~ (p=0.655 n=40+40) RegexpMatchEasy1_1K-4 1.01GB/s ± 4% 1.01GB/s ± 3% ~ (p=0.351 n=40+33) RegexpMatchMedium_32-4 8.32MB/s ± 4% 8.34MB/s ± 5% ~ (p=0.696 n=40+40) RegexpMatchMedium_1K-4 25.6MB/s ± 3% 25.9MB/s ± 4% +1.36% (p=0.000 n=40+40) RegexpMatchHard_32-4 14.1MB/s ± 6% 14.3MB/s ± 4% +1.64% (p=0.011 n=40+40) RegexpMatchHard_1K-4 15.2MB/s ± 3% 15.2MB/s ± 3% ~ (p=0.147 n=40+40) Revcomp-4 138MB/s ± 2% 137MB/s ± 3% -0.70% (p=0.021 n=40+40) Template-4 28.1MB/s ± 4% 27.8MB/s ± 3% -1.19% (p=0.003 n=40+40) [Geo mean] 83.7MB/s 83.7MB/s +0.03% Change-Id: I2a2b3a942b5c45467491515d201179fd192e65c9 Reviewed-on: https://go-review.googlesource.com/c/141650 Run-TryBot: Ben Shi <powerman1st@163.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-10-14 19:19:48 -06:00
// 386:`MOVW\s[A-Z]+,\s1\([A-Z]+\)`,-`MOVB`
// amd64:`MOVW\s[A-Z]+,\s1\([A-Z]+\)`,-`MOVB`
b[1], b[2] = byte(val), byte(val>>8)
}
cmd/compile: optimize 386's load/store combination This CL adds more combinations of two consequtive MOVBload/MOVBstore to a unique MOVWload/MOVWstore. 1. The size of the go executable decreases about 4KB, and the total size of pkg/linux_386 (excluding cmd/compile) decreases about 1.5KB. 2. There is no regression in the go1 benchmark result, excluding noise. name old time/op new time/op delta BinaryTree17-4 3.28s ± 2% 3.29s ± 2% ~ (p=0.151 n=40+40) Fannkuch11-4 3.52s ± 1% 3.51s ± 1% -0.28% (p=0.002 n=40+40) FmtFprintfEmpty-4 45.4ns ± 4% 45.0ns ± 4% -0.89% (p=0.019 n=40+40) FmtFprintfString-4 81.9ns ± 7% 81.3ns ± 1% ~ (p=0.660 n=40+25) FmtFprintfInt-4 91.9ns ± 9% 91.4ns ± 9% ~ (p=0.249 n=40+40) FmtFprintfIntInt-4 143ns ± 4% 143ns ± 4% ~ (p=0.760 n=40+40) FmtFprintfPrefixedInt-4 184ns ± 3% 183ns ± 4% ~ (p=0.485 n=40+40) FmtFprintfFloat-4 408ns ± 3% 409ns ± 3% ~ (p=0.961 n=40+40) FmtManyArgs-4 597ns ± 4% 602ns ± 3% ~ (p=0.413 n=40+40) GobDecode-4 7.13ms ± 6% 7.14ms ± 6% ~ (p=0.859 n=40+40) GobEncode-4 6.86ms ± 9% 6.94ms ± 7% ~ (p=0.162 n=40+40) Gzip-4 395ms ± 4% 396ms ± 3% ~ (p=0.099 n=40+40) Gunzip-4 40.9ms ± 4% 41.1ms ± 3% ~ (p=0.064 n=40+40) HTTPClientServer-4 63.6µs ± 2% 63.6µs ± 3% ~ (p=0.832 n=36+39) JSONEncode-4 16.1ms ± 3% 15.8ms ± 3% -1.60% (p=0.001 n=40+40) JSONDecode-4 61.0ms ± 3% 61.5ms ± 4% ~ (p=0.065 n=40+40) Mandelbrot200-4 5.16ms ± 3% 5.18ms ± 3% ~ (p=0.056 n=40+40) GoParse-4 3.25ms ± 2% 3.23ms ± 3% ~ (p=0.727 n=40+40) RegexpMatchEasy0_32-4 90.2ns ± 3% 89.3ns ± 6% -0.98% (p=0.002 n=40+40) RegexpMatchEasy0_1K-4 812ns ± 3% 815ns ± 3% ~ (p=0.309 n=40+40) RegexpMatchEasy1_32-4 103ns ± 6% 103ns ± 5% ~ (p=0.680 n=40+40) RegexpMatchEasy1_1K-4 1.01µs ± 4% 1.02µs ± 3% ~ (p=0.326 n=40+33) RegexpMatchMedium_32-4 120ns ± 4% 120ns ± 5% ~ (p=0.834 n=40+40) RegexpMatchMedium_1K-4 40.1µs ± 3% 39.5µs ± 4% -1.35% (p=0.000 n=40+40) RegexpMatchHard_32-4 2.27µs ± 6% 2.23µs ± 4% -1.67% (p=0.011 n=40+40) RegexpMatchHard_1K-4 67.2µs ± 3% 67.2µs ± 3% ~ (p=0.149 n=40+40) Revcomp-4 1.84s ± 2% 1.86s ± 3% +0.70% (p=0.020 n=40+40) Template-4 69.0ms ± 4% 69.8ms ± 3% +1.20% (p=0.003 n=40+40) TimeParse-4 438ns ± 3% 439ns ± 4% ~ (p=0.650 n=40+40) TimeFormat-4 412ns ± 3% 412ns ± 3% ~ (p=0.888 n=40+40) [Geo mean] 65.2µs 65.2µs -0.04% name old speed new speed delta GobDecode-4 108MB/s ± 6% 108MB/s ± 6% ~ (p=0.855 n=40+40) GobEncode-4 112MB/s ± 9% 111MB/s ± 8% ~ (p=0.159 n=40+40) Gzip-4 49.2MB/s ± 4% 49.1MB/s ± 3% ~ (p=0.102 n=40+40) Gunzip-4 474MB/s ± 3% 472MB/s ± 3% ~ (p=0.063 n=40+40) JSONEncode-4 121MB/s ± 3% 123MB/s ± 3% +1.62% (p=0.001 n=40+40) JSONDecode-4 31.9MB/s ± 3% 31.6MB/s ± 4% ~ (p=0.070 n=40+40) GoParse-4 17.9MB/s ± 2% 17.9MB/s ± 3% ~ (p=0.696 n=40+40) RegexpMatchEasy0_32-4 355MB/s ± 3% 358MB/s ± 5% +0.99% (p=0.002 n=40+40) RegexpMatchEasy0_1K-4 1.26GB/s ± 3% 1.26GB/s ± 3% ~ (p=0.381 n=40+40) RegexpMatchEasy1_32-4 310MB/s ± 5% 310MB/s ± 4% ~ (p=0.655 n=40+40) RegexpMatchEasy1_1K-4 1.01GB/s ± 4% 1.01GB/s ± 3% ~ (p=0.351 n=40+33) RegexpMatchMedium_32-4 8.32MB/s ± 4% 8.34MB/s ± 5% ~ (p=0.696 n=40+40) RegexpMatchMedium_1K-4 25.6MB/s ± 3% 25.9MB/s ± 4% +1.36% (p=0.000 n=40+40) RegexpMatchHard_32-4 14.1MB/s ± 6% 14.3MB/s ± 4% +1.64% (p=0.011 n=40+40) RegexpMatchHard_1K-4 15.2MB/s ± 3% 15.2MB/s ± 3% ~ (p=0.147 n=40+40) Revcomp-4 138MB/s ± 2% 137MB/s ± 3% -0.70% (p=0.021 n=40+40) Template-4 28.1MB/s ± 4% 27.8MB/s ± 3% -1.19% (p=0.003 n=40+40) [Geo mean] 83.7MB/s 83.7MB/s +0.03% Change-Id: I2a2b3a942b5c45467491515d201179fd192e65c9 Reviewed-on: https://go-review.googlesource.com/c/141650 Run-TryBot: Ben Shi <powerman1st@163.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-10-14 19:19:48 -06:00
func store_le_byte_2_inv(b []byte, val uint16) {
_ = b[2]
// 386:`MOVW\s[A-Z]+,\s1\([A-Z]+\)`,-`MOVB`
// amd64:`MOVW\s[A-Z]+,\s1\([A-Z]+\)`,-`MOVB`
cmd/compile: optimize 386's load/store combination This CL adds more combinations of two consequtive MOVBload/MOVBstore to a unique MOVWload/MOVWstore. 1. The size of the go executable decreases about 4KB, and the total size of pkg/linux_386 (excluding cmd/compile) decreases about 1.5KB. 2. There is no regression in the go1 benchmark result, excluding noise. name old time/op new time/op delta BinaryTree17-4 3.28s ± 2% 3.29s ± 2% ~ (p=0.151 n=40+40) Fannkuch11-4 3.52s ± 1% 3.51s ± 1% -0.28% (p=0.002 n=40+40) FmtFprintfEmpty-4 45.4ns ± 4% 45.0ns ± 4% -0.89% (p=0.019 n=40+40) FmtFprintfString-4 81.9ns ± 7% 81.3ns ± 1% ~ (p=0.660 n=40+25) FmtFprintfInt-4 91.9ns ± 9% 91.4ns ± 9% ~ (p=0.249 n=40+40) FmtFprintfIntInt-4 143ns ± 4% 143ns ± 4% ~ (p=0.760 n=40+40) FmtFprintfPrefixedInt-4 184ns ± 3% 183ns ± 4% ~ (p=0.485 n=40+40) FmtFprintfFloat-4 408ns ± 3% 409ns ± 3% ~ (p=0.961 n=40+40) FmtManyArgs-4 597ns ± 4% 602ns ± 3% ~ (p=0.413 n=40+40) GobDecode-4 7.13ms ± 6% 7.14ms ± 6% ~ (p=0.859 n=40+40) GobEncode-4 6.86ms ± 9% 6.94ms ± 7% ~ (p=0.162 n=40+40) Gzip-4 395ms ± 4% 396ms ± 3% ~ (p=0.099 n=40+40) Gunzip-4 40.9ms ± 4% 41.1ms ± 3% ~ (p=0.064 n=40+40) HTTPClientServer-4 63.6µs ± 2% 63.6µs ± 3% ~ (p=0.832 n=36+39) JSONEncode-4 16.1ms ± 3% 15.8ms ± 3% -1.60% (p=0.001 n=40+40) JSONDecode-4 61.0ms ± 3% 61.5ms ± 4% ~ (p=0.065 n=40+40) Mandelbrot200-4 5.16ms ± 3% 5.18ms ± 3% ~ (p=0.056 n=40+40) GoParse-4 3.25ms ± 2% 3.23ms ± 3% ~ (p=0.727 n=40+40) RegexpMatchEasy0_32-4 90.2ns ± 3% 89.3ns ± 6% -0.98% (p=0.002 n=40+40) RegexpMatchEasy0_1K-4 812ns ± 3% 815ns ± 3% ~ (p=0.309 n=40+40) RegexpMatchEasy1_32-4 103ns ± 6% 103ns ± 5% ~ (p=0.680 n=40+40) RegexpMatchEasy1_1K-4 1.01µs ± 4% 1.02µs ± 3% ~ (p=0.326 n=40+33) RegexpMatchMedium_32-4 120ns ± 4% 120ns ± 5% ~ (p=0.834 n=40+40) RegexpMatchMedium_1K-4 40.1µs ± 3% 39.5µs ± 4% -1.35% (p=0.000 n=40+40) RegexpMatchHard_32-4 2.27µs ± 6% 2.23µs ± 4% -1.67% (p=0.011 n=40+40) RegexpMatchHard_1K-4 67.2µs ± 3% 67.2µs ± 3% ~ (p=0.149 n=40+40) Revcomp-4 1.84s ± 2% 1.86s ± 3% +0.70% (p=0.020 n=40+40) Template-4 69.0ms ± 4% 69.8ms ± 3% +1.20% (p=0.003 n=40+40) TimeParse-4 438ns ± 3% 439ns ± 4% ~ (p=0.650 n=40+40) TimeFormat-4 412ns ± 3% 412ns ± 3% ~ (p=0.888 n=40+40) [Geo mean] 65.2µs 65.2µs -0.04% name old speed new speed delta GobDecode-4 108MB/s ± 6% 108MB/s ± 6% ~ (p=0.855 n=40+40) GobEncode-4 112MB/s ± 9% 111MB/s ± 8% ~ (p=0.159 n=40+40) Gzip-4 49.2MB/s ± 4% 49.1MB/s ± 3% ~ (p=0.102 n=40+40) Gunzip-4 474MB/s ± 3% 472MB/s ± 3% ~ (p=0.063 n=40+40) JSONEncode-4 121MB/s ± 3% 123MB/s ± 3% +1.62% (p=0.001 n=40+40) JSONDecode-4 31.9MB/s ± 3% 31.6MB/s ± 4% ~ (p=0.070 n=40+40) GoParse-4 17.9MB/s ± 2% 17.9MB/s ± 3% ~ (p=0.696 n=40+40) RegexpMatchEasy0_32-4 355MB/s ± 3% 358MB/s ± 5% +0.99% (p=0.002 n=40+40) RegexpMatchEasy0_1K-4 1.26GB/s ± 3% 1.26GB/s ± 3% ~ (p=0.381 n=40+40) RegexpMatchEasy1_32-4 310MB/s ± 5% 310MB/s ± 4% ~ (p=0.655 n=40+40) RegexpMatchEasy1_1K-4 1.01GB/s ± 4% 1.01GB/s ± 3% ~ (p=0.351 n=40+33) RegexpMatchMedium_32-4 8.32MB/s ± 4% 8.34MB/s ± 5% ~ (p=0.696 n=40+40) RegexpMatchMedium_1K-4 25.6MB/s ± 3% 25.9MB/s ± 4% +1.36% (p=0.000 n=40+40) RegexpMatchHard_32-4 14.1MB/s ± 6% 14.3MB/s ± 4% +1.64% (p=0.011 n=40+40) RegexpMatchHard_1K-4 15.2MB/s ± 3% 15.2MB/s ± 3% ~ (p=0.147 n=40+40) Revcomp-4 138MB/s ± 2% 137MB/s ± 3% -0.70% (p=0.021 n=40+40) Template-4 28.1MB/s ± 4% 27.8MB/s ± 3% -1.19% (p=0.003 n=40+40) [Geo mean] 83.7MB/s 83.7MB/s +0.03% Change-Id: I2a2b3a942b5c45467491515d201179fd192e65c9 Reviewed-on: https://go-review.googlesource.com/c/141650 Run-TryBot: Ben Shi <powerman1st@163.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-10-14 19:19:48 -06:00
b[2], b[1] = byte(val>>8), byte(val)
}
func store_le_byte_4(b []byte, val uint32) {
_ = b[4]
// arm64:`MOVW\sR[0-9]+,\s1\(R[0-9]+\)`,-`MOVB`,-`MOVH`
// 386:`MOVL\s[A-Z]+,\s1\([A-Z]+\)`,-`MOVB`,-`MOVW`
// amd64:`MOVL\s[A-Z]+,\s1\([A-Z]+\)`,-`MOVB`,-`MOVW`
b[1], b[2], b[3], b[4] = byte(val), byte(val>>8), byte(val>>16), byte(val>>24)
}
func store_le_byte_8(b []byte, val uint64) {
_ = b[8]
// arm64:`MOVD\sR[0-9]+,\s1\(R[0-9]+\)`,-`MOVB`,-`MOVH`,-`MOVW`
// amd64:`MOVQ\s[A-Z]+,\s1\([A-Z]+\)`,-`MOVB`,-`MOVW`,-`MOVL`
b[1], b[2], b[3], b[4], b[5], b[6], b[7], b[8] = byte(val), byte(val>>8), byte(val>>16), byte(val>>24), byte(val>>32), byte(val>>40), byte(val>>48), byte(val>>56)
}
func store_be_byte_2(b []byte, val uint16) {
_ = b[2]
// arm64:`REV16W`,`MOVH\sR[0-9]+,\s1\(R[0-9]+\)`,-`MOVB`
// amd64:`MOVW\s[A-Z]+,\s1\([A-Z]+\)`,-`MOVB`
b[1], b[2] = byte(val>>8), byte(val)
}
func store_be_byte_4(b []byte, val uint32) {
_ = b[4]
// arm64:`REVW`,`MOVW\sR[0-9]+,\s1\(R[0-9]+\)`,-`MOVB`,-`MOVH`,-`REV16W`
// amd64:`MOVL\s[A-Z]+,\s1\([A-Z]+\)`,-`MOVB`,-`MOVW`
b[1], b[2], b[3], b[4] = byte(val>>24), byte(val>>16), byte(val>>8), byte(val)
}
func store_be_byte_8(b []byte, val uint64) {
_ = b[8]
// arm64:`REV`,`MOVD\sR[0-9]+,\s1\(R[0-9]+\)`,-`MOVB`,-`MOVH`,-`MOVW`,-`REV16W`,-`REVW`
// amd64:`MOVQ\s[A-Z]+,\s1\([A-Z]+\)`,-`MOVB`,-`MOVW`,-`MOVL`
b[1], b[2], b[3], b[4], b[5], b[6], b[7], b[8] = byte(val>>56), byte(val>>48), byte(val>>40), byte(val>>32), byte(val>>24), byte(val>>16), byte(val>>8), byte(val)
}
func store_le_byte_2_idx(b []byte, idx int, val uint16) {
_, _ = b[idx+0], b[idx+1]
// arm64:`MOVH\sR[0-9]+,\s\(R[0-9]+\)\(R[0-9]+\)`,-`MOVB`
// 386:`MOVW\s[A-Z]+,\s\([A-Z]+\)\([A-Z]+`,-`MOVB`
b[idx+1], b[idx+0] = byte(val>>8), byte(val)
}
cmd/compile: optimize 386's load/store combination This CL adds more combinations of two consequtive MOVBload/MOVBstore to a unique MOVWload/MOVWstore. 1. The size of the go executable decreases about 4KB, and the total size of pkg/linux_386 (excluding cmd/compile) decreases about 1.5KB. 2. There is no regression in the go1 benchmark result, excluding noise. name old time/op new time/op delta BinaryTree17-4 3.28s ± 2% 3.29s ± 2% ~ (p=0.151 n=40+40) Fannkuch11-4 3.52s ± 1% 3.51s ± 1% -0.28% (p=0.002 n=40+40) FmtFprintfEmpty-4 45.4ns ± 4% 45.0ns ± 4% -0.89% (p=0.019 n=40+40) FmtFprintfString-4 81.9ns ± 7% 81.3ns ± 1% ~ (p=0.660 n=40+25) FmtFprintfInt-4 91.9ns ± 9% 91.4ns ± 9% ~ (p=0.249 n=40+40) FmtFprintfIntInt-4 143ns ± 4% 143ns ± 4% ~ (p=0.760 n=40+40) FmtFprintfPrefixedInt-4 184ns ± 3% 183ns ± 4% ~ (p=0.485 n=40+40) FmtFprintfFloat-4 408ns ± 3% 409ns ± 3% ~ (p=0.961 n=40+40) FmtManyArgs-4 597ns ± 4% 602ns ± 3% ~ (p=0.413 n=40+40) GobDecode-4 7.13ms ± 6% 7.14ms ± 6% ~ (p=0.859 n=40+40) GobEncode-4 6.86ms ± 9% 6.94ms ± 7% ~ (p=0.162 n=40+40) Gzip-4 395ms ± 4% 396ms ± 3% ~ (p=0.099 n=40+40) Gunzip-4 40.9ms ± 4% 41.1ms ± 3% ~ (p=0.064 n=40+40) HTTPClientServer-4 63.6µs ± 2% 63.6µs ± 3% ~ (p=0.832 n=36+39) JSONEncode-4 16.1ms ± 3% 15.8ms ± 3% -1.60% (p=0.001 n=40+40) JSONDecode-4 61.0ms ± 3% 61.5ms ± 4% ~ (p=0.065 n=40+40) Mandelbrot200-4 5.16ms ± 3% 5.18ms ± 3% ~ (p=0.056 n=40+40) GoParse-4 3.25ms ± 2% 3.23ms ± 3% ~ (p=0.727 n=40+40) RegexpMatchEasy0_32-4 90.2ns ± 3% 89.3ns ± 6% -0.98% (p=0.002 n=40+40) RegexpMatchEasy0_1K-4 812ns ± 3% 815ns ± 3% ~ (p=0.309 n=40+40) RegexpMatchEasy1_32-4 103ns ± 6% 103ns ± 5% ~ (p=0.680 n=40+40) RegexpMatchEasy1_1K-4 1.01µs ± 4% 1.02µs ± 3% ~ (p=0.326 n=40+33) RegexpMatchMedium_32-4 120ns ± 4% 120ns ± 5% ~ (p=0.834 n=40+40) RegexpMatchMedium_1K-4 40.1µs ± 3% 39.5µs ± 4% -1.35% (p=0.000 n=40+40) RegexpMatchHard_32-4 2.27µs ± 6% 2.23µs ± 4% -1.67% (p=0.011 n=40+40) RegexpMatchHard_1K-4 67.2µs ± 3% 67.2µs ± 3% ~ (p=0.149 n=40+40) Revcomp-4 1.84s ± 2% 1.86s ± 3% +0.70% (p=0.020 n=40+40) Template-4 69.0ms ± 4% 69.8ms ± 3% +1.20% (p=0.003 n=40+40) TimeParse-4 438ns ± 3% 439ns ± 4% ~ (p=0.650 n=40+40) TimeFormat-4 412ns ± 3% 412ns ± 3% ~ (p=0.888 n=40+40) [Geo mean] 65.2µs 65.2µs -0.04% name old speed new speed delta GobDecode-4 108MB/s ± 6% 108MB/s ± 6% ~ (p=0.855 n=40+40) GobEncode-4 112MB/s ± 9% 111MB/s ± 8% ~ (p=0.159 n=40+40) Gzip-4 49.2MB/s ± 4% 49.1MB/s ± 3% ~ (p=0.102 n=40+40) Gunzip-4 474MB/s ± 3% 472MB/s ± 3% ~ (p=0.063 n=40+40) JSONEncode-4 121MB/s ± 3% 123MB/s ± 3% +1.62% (p=0.001 n=40+40) JSONDecode-4 31.9MB/s ± 3% 31.6MB/s ± 4% ~ (p=0.070 n=40+40) GoParse-4 17.9MB/s ± 2% 17.9MB/s ± 3% ~ (p=0.696 n=40+40) RegexpMatchEasy0_32-4 355MB/s ± 3% 358MB/s ± 5% +0.99% (p=0.002 n=40+40) RegexpMatchEasy0_1K-4 1.26GB/s ± 3% 1.26GB/s ± 3% ~ (p=0.381 n=40+40) RegexpMatchEasy1_32-4 310MB/s ± 5% 310MB/s ± 4% ~ (p=0.655 n=40+40) RegexpMatchEasy1_1K-4 1.01GB/s ± 4% 1.01GB/s ± 3% ~ (p=0.351 n=40+33) RegexpMatchMedium_32-4 8.32MB/s ± 4% 8.34MB/s ± 5% ~ (p=0.696 n=40+40) RegexpMatchMedium_1K-4 25.6MB/s ± 3% 25.9MB/s ± 4% +1.36% (p=0.000 n=40+40) RegexpMatchHard_32-4 14.1MB/s ± 6% 14.3MB/s ± 4% +1.64% (p=0.011 n=40+40) RegexpMatchHard_1K-4 15.2MB/s ± 3% 15.2MB/s ± 3% ~ (p=0.147 n=40+40) Revcomp-4 138MB/s ± 2% 137MB/s ± 3% -0.70% (p=0.021 n=40+40) Template-4 28.1MB/s ± 4% 27.8MB/s ± 3% -1.19% (p=0.003 n=40+40) [Geo mean] 83.7MB/s 83.7MB/s +0.03% Change-Id: I2a2b3a942b5c45467491515d201179fd192e65c9 Reviewed-on: https://go-review.googlesource.com/c/141650 Run-TryBot: Ben Shi <powerman1st@163.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-10-14 19:19:48 -06:00
func store_le_byte_2_idx_inv(b []byte, idx int, val uint16) {
_, _ = b[idx+0], b[idx+1]
// 386:`MOVW\s[A-Z]+,\s\([A-Z]+\)\([A-Z]+`,-`MOVB`
cmd/compile: optimize 386's load/store combination This CL adds more combinations of two consequtive MOVBload/MOVBstore to a unique MOVWload/MOVWstore. 1. The size of the go executable decreases about 4KB, and the total size of pkg/linux_386 (excluding cmd/compile) decreases about 1.5KB. 2. There is no regression in the go1 benchmark result, excluding noise. name old time/op new time/op delta BinaryTree17-4 3.28s ± 2% 3.29s ± 2% ~ (p=0.151 n=40+40) Fannkuch11-4 3.52s ± 1% 3.51s ± 1% -0.28% (p=0.002 n=40+40) FmtFprintfEmpty-4 45.4ns ± 4% 45.0ns ± 4% -0.89% (p=0.019 n=40+40) FmtFprintfString-4 81.9ns ± 7% 81.3ns ± 1% ~ (p=0.660 n=40+25) FmtFprintfInt-4 91.9ns ± 9% 91.4ns ± 9% ~ (p=0.249 n=40+40) FmtFprintfIntInt-4 143ns ± 4% 143ns ± 4% ~ (p=0.760 n=40+40) FmtFprintfPrefixedInt-4 184ns ± 3% 183ns ± 4% ~ (p=0.485 n=40+40) FmtFprintfFloat-4 408ns ± 3% 409ns ± 3% ~ (p=0.961 n=40+40) FmtManyArgs-4 597ns ± 4% 602ns ± 3% ~ (p=0.413 n=40+40) GobDecode-4 7.13ms ± 6% 7.14ms ± 6% ~ (p=0.859 n=40+40) GobEncode-4 6.86ms ± 9% 6.94ms ± 7% ~ (p=0.162 n=40+40) Gzip-4 395ms ± 4% 396ms ± 3% ~ (p=0.099 n=40+40) Gunzip-4 40.9ms ± 4% 41.1ms ± 3% ~ (p=0.064 n=40+40) HTTPClientServer-4 63.6µs ± 2% 63.6µs ± 3% ~ (p=0.832 n=36+39) JSONEncode-4 16.1ms ± 3% 15.8ms ± 3% -1.60% (p=0.001 n=40+40) JSONDecode-4 61.0ms ± 3% 61.5ms ± 4% ~ (p=0.065 n=40+40) Mandelbrot200-4 5.16ms ± 3% 5.18ms ± 3% ~ (p=0.056 n=40+40) GoParse-4 3.25ms ± 2% 3.23ms ± 3% ~ (p=0.727 n=40+40) RegexpMatchEasy0_32-4 90.2ns ± 3% 89.3ns ± 6% -0.98% (p=0.002 n=40+40) RegexpMatchEasy0_1K-4 812ns ± 3% 815ns ± 3% ~ (p=0.309 n=40+40) RegexpMatchEasy1_32-4 103ns ± 6% 103ns ± 5% ~ (p=0.680 n=40+40) RegexpMatchEasy1_1K-4 1.01µs ± 4% 1.02µs ± 3% ~ (p=0.326 n=40+33) RegexpMatchMedium_32-4 120ns ± 4% 120ns ± 5% ~ (p=0.834 n=40+40) RegexpMatchMedium_1K-4 40.1µs ± 3% 39.5µs ± 4% -1.35% (p=0.000 n=40+40) RegexpMatchHard_32-4 2.27µs ± 6% 2.23µs ± 4% -1.67% (p=0.011 n=40+40) RegexpMatchHard_1K-4 67.2µs ± 3% 67.2µs ± 3% ~ (p=0.149 n=40+40) Revcomp-4 1.84s ± 2% 1.86s ± 3% +0.70% (p=0.020 n=40+40) Template-4 69.0ms ± 4% 69.8ms ± 3% +1.20% (p=0.003 n=40+40) TimeParse-4 438ns ± 3% 439ns ± 4% ~ (p=0.650 n=40+40) TimeFormat-4 412ns ± 3% 412ns ± 3% ~ (p=0.888 n=40+40) [Geo mean] 65.2µs 65.2µs -0.04% name old speed new speed delta GobDecode-4 108MB/s ± 6% 108MB/s ± 6% ~ (p=0.855 n=40+40) GobEncode-4 112MB/s ± 9% 111MB/s ± 8% ~ (p=0.159 n=40+40) Gzip-4 49.2MB/s ± 4% 49.1MB/s ± 3% ~ (p=0.102 n=40+40) Gunzip-4 474MB/s ± 3% 472MB/s ± 3% ~ (p=0.063 n=40+40) JSONEncode-4 121MB/s ± 3% 123MB/s ± 3% +1.62% (p=0.001 n=40+40) JSONDecode-4 31.9MB/s ± 3% 31.6MB/s ± 4% ~ (p=0.070 n=40+40) GoParse-4 17.9MB/s ± 2% 17.9MB/s ± 3% ~ (p=0.696 n=40+40) RegexpMatchEasy0_32-4 355MB/s ± 3% 358MB/s ± 5% +0.99% (p=0.002 n=40+40) RegexpMatchEasy0_1K-4 1.26GB/s ± 3% 1.26GB/s ± 3% ~ (p=0.381 n=40+40) RegexpMatchEasy1_32-4 310MB/s ± 5% 310MB/s ± 4% ~ (p=0.655 n=40+40) RegexpMatchEasy1_1K-4 1.01GB/s ± 4% 1.01GB/s ± 3% ~ (p=0.351 n=40+33) RegexpMatchMedium_32-4 8.32MB/s ± 4% 8.34MB/s ± 5% ~ (p=0.696 n=40+40) RegexpMatchMedium_1K-4 25.6MB/s ± 3% 25.9MB/s ± 4% +1.36% (p=0.000 n=40+40) RegexpMatchHard_32-4 14.1MB/s ± 6% 14.3MB/s ± 4% +1.64% (p=0.011 n=40+40) RegexpMatchHard_1K-4 15.2MB/s ± 3% 15.2MB/s ± 3% ~ (p=0.147 n=40+40) Revcomp-4 138MB/s ± 2% 137MB/s ± 3% -0.70% (p=0.021 n=40+40) Template-4 28.1MB/s ± 4% 27.8MB/s ± 3% -1.19% (p=0.003 n=40+40) [Geo mean] 83.7MB/s 83.7MB/s +0.03% Change-Id: I2a2b3a942b5c45467491515d201179fd192e65c9 Reviewed-on: https://go-review.googlesource.com/c/141650 Run-TryBot: Ben Shi <powerman1st@163.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2018-10-14 19:19:48 -06:00
b[idx+0], b[idx+1] = byte(val), byte(val>>8)
}
func store_le_byte_4_idx(b []byte, idx int, val uint32) {
_, _, _, _ = b[idx+0], b[idx+1], b[idx+2], b[idx+3]
// arm64:`MOVW\sR[0-9]+,\s\(R[0-9]+\)\(R[0-9]+\)`,-`MOVB`,-`MOVH`
b[idx+3], b[idx+2], b[idx+1], b[idx+0] = byte(val>>24), byte(val>>16), byte(val>>8), byte(val)
}
func store_be_byte_2_idx(b []byte, idx int, val uint16) {
_, _ = b[idx+0], b[idx+1]
// arm64:`REV16W`,`MOVH\sR[0-9]+,\s\(R[0-9]+\)\(R[0-9]+\)`,-`MOVB`
b[idx+0], b[idx+1] = byte(val>>8), byte(val)
}
func store_be_byte_4_idx(b []byte, idx int, val uint32) {
_, _, _, _ = b[idx+0], b[idx+1], b[idx+2], b[idx+3]
// arm64:`REVW`,`MOVW\sR[0-9]+,\s\(R[0-9]+\)\(R[0-9]+\)`,-`MOVB`,-`MOVH`,-`REV16W`
b[idx+0], b[idx+1], b[idx+2], b[idx+3] = byte(val>>24), byte(val>>16), byte(val>>8), byte(val)
}
func store_be_byte_2_idx2(b []byte, idx int, val uint16) {
_, _ = b[(idx<<1)+0], b[(idx<<1)+1]
// arm64:`REV16W`,`MOVH\sR[0-9]+,\s\(R[0-9]+\)\(R[0-9]+<<1\)`,-`MOVB`
b[(idx<<1)+0], b[(idx<<1)+1] = byte(val>>8), byte(val)
}
func store_le_byte_2_idx2(b []byte, idx int, val uint16) {
_, _ = b[(idx<<1)+0], b[(idx<<1)+1]
// arm64:`MOVH\sR[0-9]+,\s\(R[0-9]+\)\(R[0-9]+<<1\)`,-`MOVB`
b[(idx<<1)+1], b[(idx<<1)+0] = byte(val>>8), byte(val)
}
func store_be_byte_4_idx4(b []byte, idx int, val uint32) {
_, _, _, _ = b[(idx<<2)+0], b[(idx<<2)+1], b[(idx<<2)+2], b[(idx<<2)+3]
// arm64:`REVW`,`MOVW\sR[0-9]+,\s\(R[0-9]+\)\(R[0-9]+<<2\)`,-`MOVB`,-`MOVH`,-`REV16W`
b[(idx<<2)+0], b[(idx<<2)+1], b[(idx<<2)+2], b[(idx<<2)+3] = byte(val>>24), byte(val>>16), byte(val>>8), byte(val)
}
func store_le_byte_4_idx4_inv(b []byte, idx int, val uint32) {
_, _, _, _ = b[(idx<<2)+0], b[(idx<<2)+1], b[(idx<<2)+2], b[(idx<<2)+3]
// arm64:`MOVW\sR[0-9]+,\s\(R[0-9]+\)\(R[0-9]+<<2\)`,-`MOVB`,-`MOVH`
b[(idx<<2)+3], b[(idx<<2)+2], b[(idx<<2)+1], b[(idx<<2)+0] = byte(val>>24), byte(val>>16), byte(val>>8), byte(val)
}
// ------------- //
// Zeroing //
// ------------- //
// Check that zero stores are combined into larger stores
func zero_byte_2(b1, b2 []byte) {
// bounds checks to guarantee safety of writes below
_, _ = b1[1], b2[1]
// arm64:"MOVH\tZR",-"MOVB"
// amd64:`MOVW\s[$]0,\s\([A-Z]+\)`
// 386:`MOVW\s[$]0,\s\([A-Z]+\)`
b1[0], b1[1] = 0, 0
// arm64:"MOVH\tZR",-"MOVB"
// 386:`MOVW\s[$]0,\s\([A-Z]+\)`
// amd64:`MOVW\s[$]0,\s\([A-Z]+\)`
b2[1], b2[0] = 0, 0
}
func zero_byte_4(b1, b2 []byte) {
_, _ = b1[3], b2[3]
// arm64:"MOVW\tZR",-"MOVB",-"MOVH"
// amd64:`MOVL\s[$]0,\s\([A-Z]+\)`
// 386:`MOVL\s[$]0,\s\([A-Z]+\)`
b1[0], b1[1], b1[2], b1[3] = 0, 0, 0, 0
// arm64:"MOVW\tZR",-"MOVB",-"MOVH"
b2[2], b2[3], b2[1], b2[0] = 0, 0, 0, 0
}
func zero_byte_8(b []byte) {
_ = b[7]
b[0], b[1], b[2], b[3] = 0, 0, 0, 0
b[4], b[5], b[6], b[7] = 0, 0, 0, 0 // arm64:"MOVD\tZR",-"MOVB",-"MOVH",-"MOVW"
}
func zero_byte_16(b []byte) {
_ = b[15]
b[0], b[1], b[2], b[3] = 0, 0, 0, 0
b[4], b[5], b[6], b[7] = 0, 0, 0, 0
b[8], b[9], b[10], b[11] = 0, 0, 0, 0
b[12], b[13], b[14], b[15] = 0, 0, 0, 0 // arm64:"STP",-"MOVB",-"MOVH",-"MOVW"
}
func zero_byte_30(a *[30]byte) {
*a = [30]byte{} // arm64:"STP",-"MOVB",-"MOVH",-"MOVW"
}
func zero_byte_39(a *[39]byte) {
*a = [39]byte{} // arm64:"MOVD",-"MOVB",-"MOVH",-"MOVW"
}
func zero_byte_2_idx(b []byte, idx int) {
_, _ = b[idx+0], b[idx+1]
// arm64:`MOVH\sZR,\s\(R[0-9]+\)\(R[0-9]+\)`,-`MOVB`
b[idx+0], b[idx+1] = 0, 0
}
func zero_byte_2_idx2(b []byte, idx int) {
_, _ = b[(idx<<1)+0], b[(idx<<1)+1]
// arm64:`MOVH\sZR,\s\(R[0-9]+\)\(R[0-9]+<<1\)`,-`MOVB`
b[(idx<<1)+0], b[(idx<<1)+1] = 0, 0
}
func zero_uint16_2(h1, h2 []uint16) {
_, _ = h1[1], h2[1]
// arm64:"MOVW\tZR",-"MOVB",-"MOVH"
// amd64:`MOVL\s[$]0,\s\([A-Z]+\)`
// 386:`MOVL\s[$]0,\s\([A-Z]+\)`
h1[0], h1[1] = 0, 0
// arm64:"MOVW\tZR",-"MOVB",-"MOVH"
// amd64:`MOVL\s[$]0,\s\([A-Z]+\)`
// 386:`MOVL\s[$]0,\s\([A-Z]+\)`
h2[1], h2[0] = 0, 0
}
func zero_uint16_4(h1, h2 []uint16) {
_, _ = h1[3], h2[3]
// arm64:"MOVD\tZR",-"MOVB",-"MOVH",-"MOVW"
// amd64:`MOVQ\s[$]0,\s\([A-Z]+\)`
h1[0], h1[1], h1[2], h1[3] = 0, 0, 0, 0
// arm64:"MOVD\tZR",-"MOVB",-"MOVH",-"MOVW"
h2[2], h2[3], h2[1], h2[0] = 0, 0, 0, 0
}
func zero_uint16_8(h []uint16) {
_ = h[7]
h[0], h[1], h[2], h[3] = 0, 0, 0, 0
h[4], h[5], h[6], h[7] = 0, 0, 0, 0 // arm64:"STP",-"MOVB",-"MOVH"
}
func zero_uint32_2(w1, w2 []uint32) {
_, _ = w1[1], w2[1]
// arm64:"MOVD\tZR",-"MOVB",-"MOVH",-"MOVW"
// amd64:`MOVQ\s[$]0,\s\([A-Z]+\)`
w1[0], w1[1] = 0, 0
// arm64:"MOVD\tZR",-"MOVB",-"MOVH",-"MOVW"
// amd64:`MOVQ\s[$]0,\s\([A-Z]+\)`
w2[1], w2[0] = 0, 0
}
func zero_uint32_4(w1, w2 []uint32) {
_, _ = w1[3], w2[3]
w1[0], w1[1], w1[2], w1[3] = 0, 0, 0, 0 // arm64:"STP",-"MOVB",-"MOVH"
w2[2], w2[3], w2[1], w2[0] = 0, 0, 0, 0 // arm64:"STP",-"MOVB",-"MOVH"
}
func zero_uint64_2(d1, d2 []uint64) {
_, _ = d1[1], d2[1]
d1[0], d1[1] = 0, 0 // arm64:"STP",-"MOVB",-"MOVH"
d2[1], d2[0] = 0, 0 // arm64:"STP",-"MOVB",-"MOVH"
}