mirror
/
go
mirror of https://github.com/golang/go.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

all: align cpu feature variable offset naming 

Add an "offset_" prefix to all cpu feature variable offset constants to
signify that they are not boolean cpu feature variables.

Remove _ from offset constant names.

Change-Id: I6e22a79ebcbe6e2ae54c4ac8764f9260bb3223ff
Reviewed-on: https://go-review.googlesource.com/131215
Run-TryBot: Martin Möhrmann <moehrmann@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
This commit is contained in:
Martin Möhrmann 2018-08-24 11:02:00 +02:00
parent 961eb13b67
commit 05c02444eb
15 changed files with 26 additions and 25 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
src/internal/bytealg/bytealg.go
View File

@ -11,11 +11,12 @@ import (
// Offsets into internal/cpu records for use in assembly.
const (
x86_HasSSE2 = unsafe.Offsetof(cpu.X86.HasSSE2)
x86_HasSSE42 = unsafe.Offsetof(cpu.X86.HasSSE42)
x86_HasAVX2 = unsafe.Offsetof(cpu.X86.HasAVX2)
x86_HasPOPCNT = unsafe.Offsetof(cpu.X86.HasPOPCNT)
s390x_HasVX = unsafe.Offsetof(cpu.S390X.HasVX)
offsetX86HasSSE2 = unsafe.Offsetof(cpu.X86.HasSSE2)
offsetX86HasSSE42 = unsafe.Offsetof(cpu.X86.HasSSE42)
offsetX86HasAVX2 = unsafe.Offsetof(cpu.X86.HasAVX2)
offsetX86HasPOPCNT = unsafe.Offsetof(cpu.X86.HasPOPCNT)
offsetS390xHasVX = unsafe.Offsetof(cpu.S390X.HasVX)
)
// MaxLen is the maximum length of the string to be searched for (argument b) in Index.

2
src/internal/bytealg/compare_amd64.s
View File

@ -47,7 +47,7 @@ TEXT cmpbody<>(SB),NOSPLIT,$0-0
CMPQ R8, $63
JBE loop
CMPB internal∕cpu·X86+const_x86_HasAVX2(SB), $1
CMPB internal∕cpu·X86+const_offsetX86HasAVX2(SB), $1
JEQ big_loop_avx2
JMP big_loop
loop:

6
src/internal/bytealg/count_amd64.s
View File

@ -6,7 +6,7 @@
#include "textflag.h"
TEXT ·Count(SB),NOSPLIT,$0-40
CMPB internal∕cpu·X86+const_x86_HasPOPCNT(SB), $1
CMPB internal∕cpu·X86+const_offsetX86HasPOPCNT(SB), $1
JEQ 2(PC)
JMP ·countGeneric(SB)
MOVQ b_base+0(FP), SI
@ -16,7 +16,7 @@ TEXT ·Count(SB),NOSPLIT,$0-40
JMP countbody<>(SB)
TEXT ·CountString(SB),NOSPLIT,$0-32
CMPB internal∕cpu·X86+const_x86_HasPOPCNT(SB), $1
CMPB internal∕cpu·X86+const_offsetX86HasPOPCNT(SB), $1
JEQ 2(PC)
JMP ·countGenericString(SB)
MOVQ s_base+0(FP), SI
@ -151,7 +151,7 @@ endofpage:
RET
avx2:
CMPB internal∕cpu·X86+const_x86_HasAVX2(SB), $1
CMPB internal∕cpu·X86+const_offsetX86HasAVX2(SB), $1
JNE sse
MOVD AX, X0
LEAQ -32(SI)(BX*1), R11

2
src/internal/bytealg/equal_386.s
View File

@ -80,7 +80,7 @@ TEXT memeqbody<>(SB),NOSPLIT,$0-0
hugeloop:
CMPL BX, $64
JB bigloop
CMPB internal∕cpu·X86+const_x86_HasSSE2(SB), $1
CMPB internal∕cpu·X86+const_offsetX86HasSSE2(SB), $1
JNE bigloop
MOVOU (SI), X0
MOVOU (DI), X1

2
src/internal/bytealg/equal_amd64.s
View File

@ -77,7 +77,7 @@ TEXT memeqbody<>(SB),NOSPLIT,$0-0
JB small
CMPQ BX, $64
JB bigloop
CMPB internal∕cpu·X86+const_x86_HasAVX2(SB), $1
CMPB internal∕cpu·X86+const_offsetX86HasAVX2(SB), $1
JE hugeloop_avx2
// 64 bytes at a time using xmm registers

2
src/internal/bytealg/index_amd64.s
View File

@ -233,7 +233,7 @@ success_avx2:
VZEROUPPER
JMP success
sse42:
CMPB internal∕cpu·X86+const_x86_HasSSE42(SB), $1
CMPB internal∕cpu·X86+const_offsetX86HasSSE42(SB), $1
JNE no_sse42
CMPQ AX, $12
// PCMPESTRI is slower than normal compare,

2
src/internal/bytealg/indexbyte_amd64.s
View File

@ -139,7 +139,7 @@ endofpage:
RET
avx2:
CMPB internal∕cpu·X86+const_x86_HasAVX2(SB), $1
CMPB internal∕cpu·X86+const_offsetX86HasAVX2(SB), $1
JNE sse
MOVD AX, X0
LEAQ -32(SI)(BX*1), R11

2
src/internal/bytealg/indexbyte_s390x.s
View File

@ -64,7 +64,7 @@ notfound:
RET
large:
MOVBZ internal∕cpu·S390X+const_s390x_HasVX(SB), R1
MOVBZ internal∕cpu·S390X+const_offsetS390xHasVX(SB), R1
CMPBNE R1, $0, vectorimpl
srstimpl: // no vector facility

2
src/runtime/asm_386.s
View File

@ -881,7 +881,7 @@ TEXT runtime·stackcheck(SB), NOSPLIT, $0-0
// func cputicks() int64
TEXT runtime·cputicks(SB),NOSPLIT,$0-8
CMPB internal∕cpu·X86+const_offset_x86_HasSSE2(SB), $1
CMPB internal∕cpu·X86+const_offsetX86HasSSE2(SB), $1
JNE done
CMPB runtime·lfenceBeforeRdtsc(SB), $1
JNE mfence

8
src/runtime/cpuflags.go
View File

@ -11,9 +11,9 @@ import (
// Offsets into internal/cpu records for use in assembly.
const (
offset_x86_HasAVX2 = unsafe.Offsetof(cpu.X86.HasAVX2)
offset_x86_HasERMS = unsafe.Offsetof(cpu.X86.HasERMS)
offset_x86_HasSSE2 = unsafe.Offsetof(cpu.X86.HasSSE2)
offsetX86HasAVX2 = unsafe.Offsetof(cpu.X86.HasAVX2)
offsetX86HasERMS = unsafe.Offsetof(cpu.X86.HasERMS)
offsetX86HasSSE2 = unsafe.Offsetof(cpu.X86.HasSSE2)
offset_arm_HasIDIVA = unsafe.Offsetof(cpu.ARM.HasIDIVA)
offsetARMHasIDIVA = unsafe.Offsetof(cpu.ARM.HasIDIVA)
)

2
src/runtime/memclr_386.s
View File

@ -29,7 +29,7 @@ tail:
JBE _5through8
CMPL BX, $16
JBE _9through16
CMPB internal∕cpu·X86+const_offset_x86_HasSSE2(SB), $1
CMPB internal∕cpu·X86+const_offsetX86HasSSE2(SB), $1
JNE nosse2
PXOR X0, X0
CMPL BX, $32

2
src/runtime/memclr_amd64.s
View File

@ -38,7 +38,7 @@ tail:
JBE _65through128
CMPQ BX, $256
JBE _129through256
CMPB internal∕cpu·X86+const_offset_x86_HasAVX2(SB), $1
CMPB internal∕cpu·X86+const_offsetX86HasAVX2(SB), $1
JE loop_preheader_avx2
// TODO: for really big clears, use MOVNTDQ, even without AVX2.

4
src/runtime/memmove_386.s
View File

@ -52,7 +52,7 @@ tail:
JBE move_5through8
CMPL BX, $16
JBE move_9through16
CMPB internal∕cpu·X86+const_offset_x86_HasSSE2(SB), $1
CMPB internal∕cpu·X86+const_offsetX86HasSSE2(SB), $1
JNE nosse2
CMPL BX, $32
JBE move_17through32
@ -73,7 +73,7 @@ nosse2:
*/
forward:
// If REP MOVSB isn't fast, don't use it
CMPB internal∕cpu·X86+const_offset_x86_HasERMS(SB), $1 // enhanced REP MOVSB/STOSB
CMPB internal∕cpu·X86+const_offsetX86HasERMS(SB), $1 // enhanced REP MOVSB/STOSB
JNE fwdBy4
// Check alignment

2
src/runtime/memmove_amd64.s
View File

@ -84,7 +84,7 @@ forward:
JLS move_256through2048
// If REP MOVSB isn't fast, don't use it
CMPB internal∕cpu·X86+const_offset_x86_HasERMS(SB), $1 // enhanced REP MOVSB/STOSB
CMPB internal∕cpu·X86+const_offsetX86HasERMS(SB), $1 // enhanced REP MOVSB/STOSB
JNE fwdBy8
// Check alignment

2
src/runtime/vlop_arm.s
View File

@ -44,7 +44,7 @@
// the RET instruction will clobber R12 on nacl, and the compiler's register
// allocator needs to know.
TEXT runtime·udiv(SB),NOSPLIT|NOFRAME,$0
MOVBU internal∕cpu·ARM+const_offset_arm_HasIDIVA(SB), Ra
MOVBU internal∕cpu·ARM+const_offsetARMHasIDIVA(SB), Ra
CMP $0, Ra
BNE udiv_hardware