mirror of
https://github.com/golang/go
synced 2024-11-18 10:14:45 -07:00
ddc7d2a80c
Usually optimization rules have corresponding priorities, some need to
be run first, some run next, and some run last, which produces the best
code. But currently our optimization rules have no priority, this CL
adds a late lower pass that runs those rules that need to be run at last,
such as split unreasonable constant folding. This pass can be seen as
the second round of the lower pass.
For example:
func foo(a, b uint64) uint64 {
d := a+0x1234568
d1 := b+0x1234568
return d&d1
}
The code generated by the master branch:
0x0004 00004 ADD $19088744, R0, R2 // movz+movk+add
0x0010 00016 ADD $19088744, R1, R1 // movz+movk+add
0x001c 00028 AND R1, R2, R0
This is because the current constant folding optimization rules do not
take into account the range of constants, causing the constant to be
loaded repeatedly. This CL splits these unreasonable constants folding
in the late lower pass. With this CL the generated code:
0x0004 00004 MOVD $19088744, R2 // movz+movk
0x000c 00012 ADD R0, R2, R3
0x0010 00016 ADD R1, R2, R1
0x0014 00020 AND R1, R3, R0
This CL also adds constant folding optimization for ADDS instruction.
In addition, in order not to introduce the codegen regression, an
optimization rule is added to change the addition of a negative number
into a subtraction of a positive number.
go1 benchmarks:
name old time/op new time/op delta
BinaryTree17-8 1.22s ± 1% 1.24s ± 0% +1.56% (p=0.008 n=5+5)
Fannkuch11-8 1.54s ± 0% 1.53s ± 0% -0.69% (p=0.016 n=4+5)
FmtFprintfEmpty-8 14.1ns ± 0% 14.1ns ± 0% ~ (p=0.079 n=4+5)
FmtFprintfString-8 26.0ns ± 0% 26.1ns ± 0% +0.23% (p=0.008 n=5+5)
FmtFprintfInt-8 32.3ns ± 0% 32.9ns ± 1% +1.72% (p=0.008 n=5+5)
FmtFprintfIntInt-8 54.5ns ± 0% 55.5ns ± 0% +1.83% (p=0.008 n=5+5)
FmtFprintfPrefixedInt-8 61.5ns ± 0% 62.0ns ± 0% +0.93% (p=0.008 n=5+5)
FmtFprintfFloat-8 72.0ns ± 0% 73.6ns ± 0% +2.24% (p=0.008 n=5+5)
FmtManyArgs-8 221ns ± 0% 224ns ± 0% +1.22% (p=0.008 n=5+5)
GobDecode-8 1.91ms ± 0% 1.93ms ± 0% +0.98% (p=0.008 n=5+5)
GobEncode-8 1.40ms ± 1% 1.39ms ± 0% -0.79% (p=0.032 n=5+5)
Gzip-8 115ms ± 0% 117ms ± 1% +1.17% (p=0.008 n=5+5)
Gunzip-8 19.4ms ± 1% 19.3ms ± 0% -0.71% (p=0.016 n=5+4)
HTTPClientServer-8 27.0µs ± 0% 27.3µs ± 0% +0.80% (p=0.008 n=5+5)
JSONEncode-8 3.36ms ± 1% 3.33ms ± 0% ~ (p=0.056 n=5+5)
JSONDecode-8 17.5ms ± 2% 17.8ms ± 0% +1.71% (p=0.016 n=5+4)
Mandelbrot200-8 2.29ms ± 0% 2.29ms ± 0% ~ (p=0.151 n=5+5)
GoParse-8 1.35ms ± 1% 1.36ms ± 1% ~ (p=0.056 n=5+5)
RegexpMatchEasy0_32-8 24.5ns ± 0% 24.5ns ± 0% ~ (p=0.444 n=4+5)
RegexpMatchEasy0_1K-8 131ns ±11% 118ns ± 6% ~ (p=0.056 n=5+5)
RegexpMatchEasy1_32-8 22.9ns ± 0% 22.9ns ± 0% ~ (p=0.905 n=4+5)
RegexpMatchEasy1_1K-8 126ns ± 0% 127ns ± 0% ~ (p=0.063 n=4+5)
RegexpMatchMedium_32-8 486ns ± 5% 483ns ± 0% ~ (p=0.381 n=5+4)
RegexpMatchMedium_1K-8 15.4µs ± 1% 15.5µs ± 0% ~ (p=0.151 n=5+5)
RegexpMatchHard_32-8 687ns ± 0% 686ns ± 0% ~ (p=0.103 n=5+5)
RegexpMatchHard_1K-8 20.7µs ± 0% 20.7µs ± 1% ~ (p=0.151 n=5+5)
Revcomp-8 175ms ± 2% 176ms ± 3% ~ (p=1.000 n=5+5)
Template-8 20.4ms ± 6% 20.1ms ± 2% ~ (p=0.151 n=5+5)
TimeParse-8 112ns ± 0% 113ns ± 0% +0.97% (p=0.016 n=5+4)
TimeFormat-8 156ns ± 0% 145ns ± 0% -7.14% (p=0.029 n=4+4)
Change-Id: I3ced26e89041f873ac989586514ccc5ee09f13da
Reviewed-on: https://go-review.googlesource.com/c/go/+/425134
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Eric Fang <eric.fang@arm.com>
377 lines
6.2 KiB
Go
377 lines
6.2 KiB
Go
// asmcheck
|
|
|
|
// Copyright 2018 The Go Authors. All rights reserved.
|
|
// Use of this source code is governed by a BSD-style
|
|
// license that can be found in the LICENSE file.
|
|
|
|
package codegen
|
|
|
|
import "math/bits"
|
|
|
|
/************************************
|
|
* 64-bit instructions
|
|
************************************/
|
|
|
|
func bitcheck64_constleft(a uint64) (n int) {
|
|
// amd64:"BTQ\t[$]63"
|
|
if a&(1<<63) != 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTQ\t[$]60"
|
|
if a&(1<<60) != 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]0"
|
|
if a&(1<<0) != 0 {
|
|
return 1
|
|
}
|
|
return 0
|
|
}
|
|
|
|
func bitcheck64_constright(a [8]uint64) (n int) {
|
|
// amd64:"BTQ\t[$]63"
|
|
if (a[0]>>63)&1 != 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTQ\t[$]63"
|
|
if a[1]>>63 != 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTQ\t[$]63"
|
|
if a[2]>>63 == 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTQ\t[$]60"
|
|
if (a[3]>>60)&1 == 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]1"
|
|
if (a[4]>>1)&1 == 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]0"
|
|
if (a[5]>>0)&1 == 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]7"
|
|
if (a[6]>>5)&4 == 0 {
|
|
return 1
|
|
}
|
|
return 0
|
|
}
|
|
|
|
func bitcheck64_var(a, b uint64) (n int) {
|
|
// amd64:"BTQ"
|
|
if a&(1<<(b&63)) != 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTQ",-"BT.\t[$]0"
|
|
if (b>>(a&63))&1 != 0 {
|
|
return 1
|
|
}
|
|
return 0
|
|
}
|
|
|
|
func bitcheck64_mask(a uint64) (n int) {
|
|
// amd64:"BTQ\t[$]63"
|
|
if a&0x8000000000000000 != 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTQ\t[$]59"
|
|
if a&0x800000000000000 != 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]0"
|
|
if a&0x1 != 0 {
|
|
return 1
|
|
}
|
|
return 0
|
|
}
|
|
|
|
func biton64(a, b uint64) (n uint64) {
|
|
// amd64:"BTSQ"
|
|
n += b | (1 << (a & 63))
|
|
|
|
// amd64:"BTSQ\t[$]63"
|
|
n += a | (1 << 63)
|
|
|
|
// amd64:"BTSQ\t[$]60"
|
|
n += a | (1 << 60)
|
|
|
|
// amd64:"ORQ\t[$]1"
|
|
n += a | (1 << 0)
|
|
|
|
return n
|
|
}
|
|
|
|
func bitoff64(a, b uint64) (n uint64) {
|
|
// amd64:"BTRQ"
|
|
n += b &^ (1 << (a & 63))
|
|
|
|
// amd64:"BTRQ\t[$]63"
|
|
n += a &^ (1 << 63)
|
|
|
|
// amd64:"BTRQ\t[$]60"
|
|
n += a &^ (1 << 60)
|
|
|
|
// amd64:"ANDQ\t[$]-2"
|
|
n += a &^ (1 << 0)
|
|
|
|
return n
|
|
}
|
|
|
|
func bitcompl64(a, b uint64) (n uint64) {
|
|
// amd64:"BTCQ"
|
|
n += b ^ (1 << (a & 63))
|
|
|
|
// amd64:"BTCQ\t[$]63"
|
|
n += a ^ (1 << 63)
|
|
|
|
// amd64:"BTCQ\t[$]60"
|
|
n += a ^ (1 << 60)
|
|
|
|
// amd64:"XORQ\t[$]1"
|
|
n += a ^ (1 << 0)
|
|
|
|
return n
|
|
}
|
|
|
|
/************************************
|
|
* 32-bit instructions
|
|
************************************/
|
|
|
|
func bitcheck32_constleft(a uint32) (n int) {
|
|
// amd64:"BTL\t[$]31"
|
|
if a&(1<<31) != 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]28"
|
|
if a&(1<<28) != 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]0"
|
|
if a&(1<<0) != 0 {
|
|
return 1
|
|
}
|
|
return 0
|
|
}
|
|
|
|
func bitcheck32_constright(a [8]uint32) (n int) {
|
|
// amd64:"BTL\t[$]31"
|
|
if (a[0]>>31)&1 != 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]31"
|
|
if a[1]>>31 != 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]31"
|
|
if a[2]>>31 == 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]28"
|
|
if (a[3]>>28)&1 == 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]1"
|
|
if (a[4]>>1)&1 == 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]0"
|
|
if (a[5]>>0)&1 == 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]7"
|
|
if (a[6]>>5)&4 == 0 {
|
|
return 1
|
|
}
|
|
return 0
|
|
}
|
|
|
|
func bitcheck32_var(a, b uint32) (n int) {
|
|
// amd64:"BTL"
|
|
if a&(1<<(b&31)) != 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL",-"BT.\t[$]0"
|
|
if (b>>(a&31))&1 != 0 {
|
|
return 1
|
|
}
|
|
return 0
|
|
}
|
|
|
|
func bitcheck32_mask(a uint32) (n int) {
|
|
// amd64:"BTL\t[$]31"
|
|
if a&0x80000000 != 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]27"
|
|
if a&0x8000000 != 0 {
|
|
return 1
|
|
}
|
|
// amd64:"BTL\t[$]0"
|
|
if a&0x1 != 0 {
|
|
return 1
|
|
}
|
|
return 0
|
|
}
|
|
|
|
func biton32(a, b uint32) (n uint32) {
|
|
// amd64:"BTSL"
|
|
n += b | (1 << (a & 31))
|
|
|
|
// amd64:"BTSL\t[$]31"
|
|
n += a | (1 << 31)
|
|
|
|
// amd64:"BTSL\t[$]28"
|
|
n += a | (1 << 28)
|
|
|
|
// amd64:"ORL\t[$]1"
|
|
n += a | (1 << 0)
|
|
|
|
return n
|
|
}
|
|
|
|
func bitoff32(a, b uint32) (n uint32) {
|
|
// amd64:"BTRL"
|
|
n += b &^ (1 << (a & 31))
|
|
|
|
// amd64:"BTRL\t[$]31"
|
|
n += a &^ (1 << 31)
|
|
|
|
// amd64:"BTRL\t[$]28"
|
|
n += a &^ (1 << 28)
|
|
|
|
// amd64:"ANDL\t[$]-2"
|
|
n += a &^ (1 << 0)
|
|
|
|
return n
|
|
}
|
|
|
|
func bitcompl32(a, b uint32) (n uint32) {
|
|
// amd64:"BTCL"
|
|
n += b ^ (1 << (a & 31))
|
|
|
|
// amd64:"BTCL\t[$]31"
|
|
n += a ^ (1 << 31)
|
|
|
|
// amd64:"BTCL\t[$]28"
|
|
n += a ^ (1 << 28)
|
|
|
|
// amd64:"XORL\t[$]1"
|
|
n += a ^ (1 << 0)
|
|
|
|
return n
|
|
}
|
|
|
|
// check direct operation on memory with constant and shifted constant sources
|
|
func bitOpOnMem(a []uint32, b, c, d uint32) {
|
|
// amd64:`ANDL\s[$]200,\s\([A-Z][A-Z0-9]+\)`
|
|
a[0] &= 200
|
|
// amd64:`ORL\s[$]220,\s4\([A-Z][A-Z0-9]+\)`
|
|
a[1] |= 220
|
|
// amd64:`XORL\s[$]240,\s8\([A-Z][A-Z0-9]+\)`
|
|
a[2] ^= 240
|
|
}
|
|
|
|
func bitcheckMostNegative(b uint8) bool {
|
|
// amd64:"TESTB"
|
|
return b&0x80 == 0x80
|
|
}
|
|
|
|
// Check AND masking on arm64 (Issue #19857)
|
|
|
|
func and_mask_1(a uint64) uint64 {
|
|
// arm64:`AND\t`
|
|
return a & ((1 << 63) - 1)
|
|
}
|
|
|
|
func and_mask_2(a uint64) uint64 {
|
|
// arm64:`AND\t`
|
|
return a & (1 << 63)
|
|
}
|
|
|
|
func and_mask_3(a, b uint32) (uint32, uint32) {
|
|
// arm/7:`BIC`,-`AND`
|
|
a &= 0xffffaaaa
|
|
// arm/7:`BFC`,-`AND`,-`BIC`
|
|
b &= 0xffc003ff
|
|
return a, b
|
|
}
|
|
|
|
// Check generation of arm64 BIC/EON/ORN instructions
|
|
|
|
func op_bic(x, y uint32) uint32 {
|
|
// arm64:`BIC\t`,-`AND`
|
|
return x &^ y
|
|
}
|
|
|
|
func op_eon(x, y, z uint32, a []uint32, n, m uint64) uint64 {
|
|
// arm64:`EON\t`,-`EOR`,-`MVN`
|
|
a[0] = x ^ (y ^ 0xffffffff)
|
|
|
|
// arm64:`EON\t`,-`EOR`,-`MVN`
|
|
a[1] = ^(y ^ z)
|
|
|
|
// arm64:`EON\t`,-`XOR`
|
|
a[2] = x ^ ^z
|
|
|
|
// arm64:`EON\t`,-`EOR`,-`MVN`
|
|
return n ^ (m ^ 0xffffffffffffffff)
|
|
}
|
|
|
|
func op_orn(x, y uint32) uint32 {
|
|
// arm64:`ORN\t`,-`ORR`
|
|
return x | ^y
|
|
}
|
|
|
|
// check bitsets
|
|
func bitSetPowerOf2Test(x int) bool {
|
|
// amd64:"BTL\t[$]3"
|
|
return x&8 == 8
|
|
}
|
|
|
|
func bitSetTest(x int) bool {
|
|
// amd64:"ANDL\t[$]9, AX"
|
|
// amd64:"CMPQ\tAX, [$]9"
|
|
return x&9 == 9
|
|
}
|
|
|
|
// mask contiguous one bits
|
|
func cont1Mask64U(x uint64) uint64 {
|
|
// s390x:"RISBGZ\t[$]16, [$]47, [$]0,"
|
|
return x & 0x0000ffffffff0000
|
|
}
|
|
|
|
// mask contiguous zero bits
|
|
func cont0Mask64U(x uint64) uint64 {
|
|
// s390x:"RISBGZ\t[$]48, [$]15, [$]0,"
|
|
return x & 0xffff00000000ffff
|
|
}
|
|
|
|
func issue44228a(a []int64, i int) bool {
|
|
// amd64: "BTQ", -"SHL"
|
|
return a[i>>6]&(1<<(i&63)) != 0
|
|
}
|
|
func issue44228b(a []int32, i int) bool {
|
|
// amd64: "BTL", -"SHL"
|
|
return a[i>>5]&(1<<(i&31)) != 0
|
|
}
|
|
|
|
func issue48467(x, y uint64) uint64 {
|
|
// arm64: -"NEG"
|
|
d, borrow := bits.Sub64(x, y, 0)
|
|
return x - d&(-borrow)
|
|
}
|
|
|
|
func foldConst(x, y uint64) uint64 {
|
|
// arm64: "ADDS\t[$]7",-"MOVD\t[$]7"
|
|
d, b := bits.Add64(x, 7, 0)
|
|
return b & d
|
|
}
|
|
|
|
func foldConstOutOfRange(a uint64) uint64 {
|
|
// arm64: "MOVD\t[$]19088744",-"ADD\t[$]19088744"
|
|
return a + 0x1234568
|
|
}
|