Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

zstd: Shave some instructions off the amd64 asm #720

Merged
merged 1 commit into from
Dec 29, 2022
Merged

zstd: Shave some instructions off the amd64 asm #720

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 8 additions & 9 deletions zstd/_generate/gen.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -502,20 +502,20 @@ func (o options) generateBody(name string, executeSingleTriple func(ctx *execute
o.returnWithCode(errorNotEnoughLiterals)
}

Comment("Return with not enough output space error")
{
if !o.useSeqs {
Comment("Return with not enough output space error")
Label("error_not_enough_space")
if !o.useSeqs {
{
ctx := Dereference(Param("ctx"))
tmp := GP64()
MOVQ(llP, tmp)
Store(tmp, ctx.Field("ll"))
MOVQ(mlP, tmp)
Store(tmp, ctx.Field("ml"))
Store(ec.outPosition, ctx.Field("outPosition"))
}

o.returnWithCode(errorNotEnoughSpace)
o.returnWithCode(errorNotEnoughSpace)
}
}
}

Expand Down Expand Up @@ -1062,10 +1062,9 @@ func (e executeSimple) generateProcedure(name string) {
Store(seqIndex, ctx.Field("seqIndex"))
Store(outPosition, ctx.Field("outPosition"))

// compute litPosition
tmp := GP64()
Load(ctx.Field("literals").Base(), tmp)
SUBQ(tmp, literals) // litPosition := current - initial literals pointer
// litPosition := current - initial literals pointer
litField, _ := ctx.Field("literals").Base().Resolve()
SUBQ(litField.Addr, literals)
Store(literals, ctx.Field("litPosition"))
}
Label("loop_finished")
Expand Down
28 changes: 4 additions & 24 deletions zstd/seqdec_amd64.s
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -320,10 +320,6 @@ error_not_enough_literals:
MOVQ $0x00000004, ret+24(FP)
RET

// Return with not enough output space error
MOVQ $0x00000005, ret+24(FP)
RET

// func sequenceDecs_decode_56_amd64(s *sequenceDecs, br *bitReader, ctx *decodeAsmContext) int
// Requires: CMOV
TEXT ·sequenceDecs_decode_56_amd64(SB), $8-32
Expand Down Expand Up @@ -617,10 +613,6 @@ error_not_enough_literals:
MOVQ $0x00000004, ret+24(FP)
RET

// Return with not enough output space error
MOVQ $0x00000005, ret+24(FP)
RET

// func sequenceDecs_decode_bmi2(s *sequenceDecs, br *bitReader, ctx *decodeAsmContext) int
// Requires: BMI, BMI2, CMOV
TEXT ·sequenceDecs_decode_bmi2(SB), $8-32
Expand Down Expand Up @@ -897,10 +889,6 @@ error_not_enough_literals:
MOVQ $0x00000004, ret+24(FP)
RET

// Return with not enough output space error
MOVQ $0x00000005, ret+24(FP)
RET

// func sequenceDecs_decode_56_bmi2(s *sequenceDecs, br *bitReader, ctx *decodeAsmContext) int
// Requires: BMI, BMI2, CMOV
TEXT ·sequenceDecs_decode_56_bmi2(SB), $8-32
Expand Down Expand Up @@ -1152,10 +1140,6 @@ error_not_enough_literals:
MOVQ $0x00000004, ret+24(FP)
RET

// Return with not enough output space error
MOVQ $0x00000005, ret+24(FP)
RET

// func sequenceDecs_executeSimple_amd64(ctx *executeAsmContext) bool
// Requires: SSE
TEXT ·sequenceDecs_executeSimple_amd64(SB), $8-9
Expand Down Expand Up @@ -1389,8 +1373,7 @@ loop_finished:
MOVQ ctx+0(FP), AX
MOVQ DX, 24(AX)
MOVQ DI, 104(AX)
MOVQ 80(AX), CX
SUBQ CX, SI
SUBQ 80(AX), SI
MOVQ SI, 112(AX)
RET

Expand All @@ -1402,8 +1385,7 @@ error_match_off_too_big:
MOVQ ctx+0(FP), AX
MOVQ DX, 24(AX)
MOVQ DI, 104(AX)
MOVQ 80(AX), CX
SUBQ CX, SI
SUBQ 80(AX), SI
MOVQ SI, 112(AX)
RET

Expand Down Expand Up @@ -1747,8 +1729,7 @@ loop_finished:
MOVQ ctx+0(FP), AX
MOVQ DX, 24(AX)
MOVQ DI, 104(AX)
MOVQ 80(AX), CX
SUBQ CX, SI
SUBQ 80(AX), SI
MOVQ SI, 112(AX)
RET

Expand All @@ -1760,8 +1741,7 @@ error_match_off_too_big:
MOVQ ctx+0(FP), AX
MOVQ DX, 24(AX)
MOVQ DI, 104(AX)
MOVQ 80(AX), CX
SUBQ CX, SI
SUBQ 80(AX), SI
MOVQ SI, 112(AX)
RET

Expand Down