mirror of
https://github.com/golang/go
synced 2024-11-05 17:36:15 -07:00
cmd/compile: specialize Move up to 79B on amd64
Move currently uses mov instructions directly up to 31 bytes and then
switches to duffcopy. Moving 31 bytes is 4 instructions corresponding to
two loads and two stores, (or 6 if !useSSE) depending on the usage,
duffcopy is five (one or two mov, two or three lea, one call).
This adds direct mov instructions for Move's of size 32, 48, and 64 with
sse and for only size 32 without.
With useSSE:
- 32 is 4 instructions (byte +/- comparison below)
- 33 thru 48 is 6
- 49 thru 64 is 8
Without:
- 32 is 8
Note that the only platform with useSSE set to false is plan 9. I have
built three projects based off tip and tip with this patch and the
project's byte size is equal to or less than they were prior.
The basis of this change is that copying data with instructions directly
is nearly free, whereas calling into duffcopy adds a bit of overhead.
This is most noticeable in range statements where elements are 32+
bytes. For code with the following pattern:
func Benchmark32Range(b *testing.B) {
var f s32
for _, count := range []int{10, 100, 1000, 10000} {
name := strconv.Itoa(count)
b.Run(name, func(b *testing.B) {
base := make([]s32, count)
for i := 0; i < b.N; i++ {
for _, v := range base {
f = v
}
}
})
}
_ = f
}
These are the resulting benchmarks:
Benchmark16Range/10-4 19.1 19.1 +0.00%
Benchmark16Range/100-4 169 170 +0.59%
Benchmark16Range/1000-4 1684 1691 +0.42%
Benchmark16Range/10000-4 18147 18124 -0.13%
Benchmark31Range/10-4 141 142 +0.71%
Benchmark31Range/100-4 1407 1410 +0.21%
Benchmark31Range/1000-4 14070 14074 +0.03%
Benchmark31Range/10000-4 141781 141759 -0.02%
Benchmark32Range/10-4 71.4 32.2 -54.90%
Benchmark32Range/100-4 695 326 -53.09%
Benchmark32Range/1000-4 7166 3313 -53.77%
Benchmark32Range/10000-4 72571 35425 -51.19%
Benchmark64Range/10-4 87.8 64.9 -26.08%
Benchmark64Range/100-4 868 629 -27.53%
Benchmark64Range/1000-4 9355 6907 -26.17%
Benchmark64Range/10000-4 94463 70385 -25.49%
Benchmark79Range/10-4 177 152 -14.12%
Benchmark79Range/100-4 1769 1531 -13.45%
Benchmark79Range/1000-4 17893 15532 -13.20%
Benchmark79Range/10000-4 178947 155551 -13.07%
Benchmark80Range/10-4 99.6 99.7 +0.10%
Benchmark80Range/100-4 987 985 -0.20%
Benchmark80Range/1000-4 10573 10560 -0.12%
Benchmark80Range/10000-4 106792 106639 -0.14%
For runtime's BenchCopyFat* benchmarks:
CopyFat8-4 0.40ns ± 0% 0.40ns ± 0% ~ (all equal)
CopyFat12-4 0.40ns ± 0% 0.80ns ± 0% +100.00% (p=0.000 n=9+9)
CopyFat16-4 0.40ns ± 0% 0.80ns ± 0% +100.00% (p=0.000 n=10+8)
CopyFat24-4 0.80ns ± 0% 0.40ns ± 0% -50.00% (p=0.001 n=8+9)
CopyFat32-4 2.01ns ± 0% 0.40ns ± 0% -80.10% (p=0.000 n=8+8)
CopyFat64-4 2.87ns ± 0% 0.40ns ± 0% -86.07% (p=0.000 n=8+10)
CopyFat128-4 4.82ns ± 0% 4.82ns ± 0% ~ (p=1.000 n=8+8)
CopyFat256-4 8.83ns ± 0% 8.83ns ± 0% ~ (p=1.000 n=8+8)
CopyFat512-4 16.9ns ± 0% 16.9ns ± 0% ~ (all equal)
CopyFat520-4 14.6ns ± 0% 14.6ns ± 1% ~ (p=0.529 n=8+9)
CopyFat1024-4 32.9ns ± 0% 33.0ns ± 0% +0.20% (p=0.041 n=8+9)
Function calls are not benefitted as much due how they are compiled, but
other benchmarks I ran show that calling function with 64 byte elements
is marginally improved.
The main downside with this change is that it may increase binary sizes
depending on the size of the copy, but this change also decreases
binaries for moves of 48 bytes or less.
For the following code:
package main
type size [32]byte
//go:noinline
func use(t size) {
}
//go:noinline
func get() size {
var z size
return z
}
func main() {
var a size
use(a)
}
Changing size around gives the following assembly leading up to the call
(the initialization and actual call are removed):
tip func call with 32B arg: 27B
48 89 e7 mov %rsp,%rdi
48 8d 74 24 20 lea 0x20(%rsp),%rsi
48 89 6c 24 f0 mov %rbp,-0x10(%rsp)
48 8d 6c 24 f0 lea -0x10(%rsp),%rbp
e8 53 ab ff ff callq 448964 <runtime.duffcopy+0x364>
48 8b 6d 00 mov 0x0(%rbp),%rbp
modified: 19B (-8B)
0f 10 44 24 20 movups 0x20(%rsp),%xmm0
0f 11 04 24 movups %xmm0,(%rsp)
0f 10 44 24 30 movups 0x30(%rsp),%xmm0
0f 11 44 24 10 movups %xmm0,0x10(%rsp)
-
tip with 47B arg: 29B
48 8d 7c 24 0f lea 0xf(%rsp),%rdi
48 8d 74 24 40 lea 0x40(%rsp),%rsi
48 89 6c 24 f0 mov %rbp,-0x10(%rsp)
48 8d 6c 24 f0 lea -0x10(%rsp),%rbp
e8 43 ab ff ff callq 448964 <runtime.duffcopy+0x364>
48 8b 6d 00 mov 0x0(%rbp),%rbp
modified: 20B (-9B)
0f 10 44 24 40 movups 0x40(%rsp),%xmm0
0f 11 44 24 0f movups %xmm0,0xf(%rsp)
0f 10 44 24 50 movups 0x50(%rsp),%xmm0
0f 11 44 24 1f movups %xmm0,0x1f(%rsp)
-
tip with 64B arg: 27B
48 89 e7 mov %rsp,%rdi
48 8d 74 24 40 lea 0x40(%rsp),%rsi
48 89 6c 24 f0 mov %rbp,-0x10(%rsp)
48 8d 6c 24 f0 lea -0x10(%rsp),%rbp
e8 1f ab ff ff callq 448948 <runtime.duffcopy+0x348>
48 8b 6d 00 mov 0x0(%rbp),%rbp
modified: 39B [+12B]
0f 10 44 24 40 movups 0x40(%rsp),%xmm0
0f 11 04 24 movups %xmm0,(%rsp)
0f 10 44 24 50 movups 0x50(%rsp),%xmm0
0f 11 44 24 10 movups %xmm0,0x10(%rsp)
0f 10 44 24 60 movups 0x60(%rsp),%xmm0
0f 11 44 24 20 movups %xmm0,0x20(%rsp)
0f 10 44 24 70 movups 0x70(%rsp),%xmm0
0f 11 44 24 30 movups %xmm0,0x30(%rsp)
-
tip with 79B arg: 29B
48 8d 7c 24 0f lea 0xf(%rsp),%rdi
48 8d 74 24 60 lea 0x60(%rsp),%rsi
48 89 6c 24 f0 mov %rbp,-0x10(%rsp)
48 8d 6c 24 f0 lea -0x10(%rsp),%rbp
e8 09 ab ff ff callq 448948 <runtime.duffcopy+0x348>
48 8b 6d 00 mov 0x0(%rbp),%rbp
modified: 46B [+17B]
0f 10 44 24 60 movups 0x60(%rsp),%xmm0
0f 11 44 24 0f movups %xmm0,0xf(%rsp)
0f 10 44 24 70 movups 0x70(%rsp),%xmm0
0f 11 44 24 1f movups %xmm0,0x1f(%rsp)
0f 10 84 24 80 00 00 movups 0x80(%rsp),%xmm0
00
0f 11 44 24 2f movups %xmm0,0x2f(%rsp)
0f 10 84 24 90 00 00 movups 0x90(%rsp),%xmm0
00
0f 11 44 24 3f movups %xmm0,0x3f(%rsp)
So, at best we save 9B, at worst we gain 17. I do not think that copying
around 65+B sized types is common enough to bloat program sizes. Using
bincmp on the go binary itself shows a zero byte difference; there are
gains and losses all over. One of the largest gains in binary size comes
from cmd/go/internal/cache.(*Cache).Get, which passes around a 64 byte
sized type -- this is one of the cases I would expect to be benefitted
by this change.
I think that this marginal improvement in struct copying for 64 byte
structs is worth it: most data structs / work items I use in my programs
are small, but few are smaller than 32 bytes: with one slice, the budget
is up. The 32 rule alone would allow another 16 bytes, the 48 and 64
rules allow another 32 and 48.
Change-Id: I19a8f9190d5d41825091f17f268f4763bfc12a62
Reviewed-on: https://go-review.googlesource.com/100718
Reviewed-by: Ilya Tocar <ilya.tocar@intel.com>
Reviewed-by: Keith Randall <khr@golang.org>
This commit is contained in:
Travis Bischel
Ilya Tocar
parent
fc6280d4b0
commit
4f7b774822
@ -328,6 +328,25 @@
|
||||
(Move [16] dst src mem) && !config.useSSE ->
|
||||
(MOVQstore [8] dst (MOVQload [8] src mem)
|
||||
(MOVQstore dst (MOVQload src mem) mem))
|
||||
|
||||
(Move [32] dst src mem) ->
|
||||
(Move [16]
|
||||
(OffPtr <dst.Type> dst [16])
|
||||
(OffPtr <src.Type> src [16])
|
||||
(Move [16] dst src mem))
|
||||
|
||||
(Move [48] dst src mem) && config.useSSE ->
|
||||
(Move [32]
|
||||
(OffPtr <dst.Type> dst [16])
|
||||
(OffPtr <src.Type> src [16])
|
||||
(Move [16] dst src mem))
|
||||
|
||||
(Move [64] dst src mem) && config.useSSE ->
|
||||
(Move [32]
|
||||
(OffPtr <dst.Type> dst [32])
|
||||
(OffPtr <src.Type> src [32])
|
||||
(Move [32] dst src mem))
|
||||
|
||||
(Move [3] dst src mem) ->
|
||||
(MOVBstore [2] dst (MOVBload [2] src mem)
|
||||
(MOVWstore dst (MOVWload src mem) mem))
|
||||
@ -367,7 +386,7 @@
|
||||
|
||||
// Medium copying uses a duff device.
|
||||
(Move [s] dst src mem)
|
||||
&& s >= 32 && s <= 16*64 && s%16 == 0
|
||||
&& s > 64 && s <= 16*64 && s%16 == 0
|
||||
&& !config.noDuffDevice ->
|
||||
(DUFFCOPY [14*(64-s/16)] dst src mem)
|
||||
// 14 and 64 are magic constants. 14 is the number of bytes to encode:
|
||||
|
||||
@ -51484,6 +51484,108 @@ func rewriteValueAMD64_OpMove_0(v *Value) bool {
|
||||
v.AddArg(v1)
|
||||
return true
|
||||
}
|
||||
// match: (Move [32] dst src mem)
|
||||
// cond:
|
||||
// result: (Move [16] (OffPtr <dst.Type> dst [16]) (OffPtr <src.Type> src [16]) (Move [16] dst src mem))
|
||||
for {
|
||||
if v.AuxInt != 32 {
|
||||
break
|
||||
}
|
||||
_ = v.Args[2]
|
||||
dst := v.Args[0]
|
||||
src := v.Args[1]
|
||||
mem := v.Args[2]
|
||||
v.reset(OpMove)
|
||||
v.AuxInt = 16
|
||||
v0 := b.NewValue0(v.Pos, OpOffPtr, dst.Type)
|
||||
v0.AuxInt = 16
|
||||
v0.AddArg(dst)
|
||||
v.AddArg(v0)
|
||||
v1 := b.NewValue0(v.Pos, OpOffPtr, src.Type)
|
||||
v1.AuxInt = 16
|
||||
v1.AddArg(src)
|
||||
v.AddArg(v1)
|
||||
v2 := b.NewValue0(v.Pos, OpMove, types.TypeMem)
|
||||
v2.AuxInt = 16
|
||||
v2.AddArg(dst)
|
||||
v2.AddArg(src)
|
||||
v2.AddArg(mem)
|
||||
v.AddArg(v2)
|
||||
return true
|
||||
}
|
||||
// match: (Move [48] dst src mem)
|
||||
// cond: config.useSSE
|
||||
// result: (Move [32] (OffPtr <dst.Type> dst [16]) (OffPtr <src.Type> src [16]) (Move [16] dst src mem))
|
||||
for {
|
||||
if v.AuxInt != 48 {
|
||||
break
|
||||
}
|
||||
_ = v.Args[2]
|
||||
dst := v.Args[0]
|
||||
src := v.Args[1]
|
||||
mem := v.Args[2]
|
||||
if !(config.useSSE) {
|
||||
break
|
||||
}
|
||||
v.reset(OpMove)
|
||||
v.AuxInt = 32
|
||||
v0 := b.NewValue0(v.Pos, OpOffPtr, dst.Type)
|
||||
v0.AuxInt = 16
|
||||
v0.AddArg(dst)
|
||||
v.AddArg(v0)
|
||||
v1 := b.NewValue0(v.Pos, OpOffPtr, src.Type)
|
||||
v1.AuxInt = 16
|
||||
v1.AddArg(src)
|
||||
v.AddArg(v1)
|
||||
v2 := b.NewValue0(v.Pos, OpMove, types.TypeMem)
|
||||
v2.AuxInt = 16
|
||||
v2.AddArg(dst)
|
||||
v2.AddArg(src)
|
||||
v2.AddArg(mem)
|
||||
v.AddArg(v2)
|
||||
return true
|
||||
}
|
||||
// match: (Move [64] dst src mem)
|
||||
// cond: config.useSSE
|
||||
// result: (Move [32] (OffPtr <dst.Type> dst [32]) (OffPtr <src.Type> src [32]) (Move [32] dst src mem))
|
||||
for {
|
||||
if v.AuxInt != 64 {
|
||||
break
|
||||
}
|
||||
_ = v.Args[2]
|
||||
dst := v.Args[0]
|
||||
src := v.Args[1]
|
||||
mem := v.Args[2]
|
||||
if !(config.useSSE) {
|
||||
break
|
||||
}
|
||||
v.reset(OpMove)
|
||||
v.AuxInt = 32
|
||||
v0 := b.NewValue0(v.Pos, OpOffPtr, dst.Type)
|
||||
v0.AuxInt = 32
|
||||
v0.AddArg(dst)
|
||||
v.AddArg(v0)
|
||||
v1 := b.NewValue0(v.Pos, OpOffPtr, src.Type)
|
||||
v1.AuxInt = 32
|
||||
v1.AddArg(src)
|
||||
v.AddArg(v1)
|
||||
v2 := b.NewValue0(v.Pos, OpMove, types.TypeMem)
|
||||
v2.AuxInt = 32
|
||||
v2.AddArg(dst)
|
||||
v2.AddArg(src)
|
||||
v2.AddArg(mem)
|
||||
v.AddArg(v2)
|
||||
return true
|
||||
}
|
||||
return false
|
||||
}
|
||||
func rewriteValueAMD64_OpMove_10(v *Value) bool {
|
||||
b := v.Block
|
||||
_ = b
|
||||
config := b.Func.Config
|
||||
_ = config
|
||||
typ := &b.Func.Config.Types
|
||||
_ = typ
|
||||
// match: (Move [3] dst src mem)
|
||||
// cond:
|
||||
// result: (MOVBstore [2] dst (MOVBload [2] src mem) (MOVWstore dst (MOVWload src mem) mem))
|
||||
@ -51571,15 +51673,6 @@ func rewriteValueAMD64_OpMove_0(v *Value) bool {
|
||||
v.AddArg(v1)
|
||||
return true
|
||||
}
|
||||
return false
|
||||
}
|
||||
func rewriteValueAMD64_OpMove_10(v *Value) bool {
|
||||
b := v.Block
|
||||
_ = b
|
||||
config := b.Func.Config
|
||||
_ = config
|
||||
typ := &b.Func.Config.Types
|
||||
_ = typ
|
||||
// match: (Move [7] dst src mem)
|
||||
// cond:
|
||||
// result: (MOVLstore [3] dst (MOVLload [3] src mem) (MOVLstore dst (MOVLload src mem) mem))
|
||||
@ -51745,7 +51838,7 @@ func rewriteValueAMD64_OpMove_10(v *Value) bool {
|
||||
return true
|
||||
}
|
||||
// match: (Move [s] dst src mem)
|
||||
// cond: s >= 32 && s <= 16*64 && s%16 == 0 && !config.noDuffDevice
|
||||
// cond: s > 64 && s <= 16*64 && s%16 == 0 && !config.noDuffDevice
|
||||
// result: (DUFFCOPY [14*(64-s/16)] dst src mem)
|
||||
for {
|
||||
s := v.AuxInt
|
||||
@ -51753,7 +51846,7 @@ func rewriteValueAMD64_OpMove_10(v *Value) bool {
|
||||
dst := v.Args[0]
|
||||
src := v.Args[1]
|
||||
mem := v.Args[2]
|
||||
if !(s >= 32 && s <= 16*64 && s%16 == 0 && !config.noDuffDevice) {
|
||||
if !(s > 64 && s <= 16*64 && s%16 == 0 && !config.noDuffDevice) {
|
||||
break
|
||||
}
|
||||
v.reset(OpAMD64DUFFCOPY)
|
||||
|
||||
Loading…
Reference in New Issue
Block a user