TrueCloudLab/tzhash
15
0
Fork 6
Code Issues 3 Pull requests 1 Releases 2 Activity

Fix formatting

Browse source
This commit is contained in:
Evgeniy Kulikov 2019-01-03 11:04:43 +03:00
parent 5cf44c62ac
commit 42499b9eb0
No known key found for this signature in database
GPG key ID: BF6AEE0A2A699BF2
2 changed files with 77 additions and 77 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

134
gf127/gf127_amd64.s
View file

@ -2,80 +2,80 @@
// func Add(a, b, c *[2]uint64)
TEXT ·Add(SB),NOSPLIT,$0
MOVQ a+0(FP), AX
MOVUPD (AX), X0
MOVQ b+8(FP), BX
MOVUPD (BX), X1
XORPD X1, X0
MOVQ c+16(FP), CX
MOVUPD X0, (CX)
RET
MOVQ a+0(FP), AX
MOVUPD (AX), X0
MOVQ b+8(FP), BX
MOVUPD (BX), X1
XORPD X1, X0
MOVQ c+16(FP), CX
MOVUPD X0, (CX)
RET
// func Mul10(a, b *[2]uint64)
TEXT ·Mul10(SB),NOSPLIT,$0
MOVQ a+0(FP), AX
MOVUPD (AX), X0
VPSLLQ $1, X0, X1
VPALIGNR $8, X1, X0, X2
PSRLQ $63, X2
MOVUPD ·x127x63(SB), X3
ANDPD X1, X3
VPUNPCKHQDQ X3, X3, X3
XORPD X2, X1
XORPD X3, X1
MOVQ b+8(FP), AX
MOVUPD X1, (AX)
RET
MOVQ a+0(FP), AX
MOVUPD (AX), X0
VPSLLQ $1, X0, X1
VPALIGNR $8, X1, X0, X2
PSRLQ $63, X2
MOVUPD ·x127x63(SB), X3
ANDPD X1, X3
VPUNPCKHQDQ X3, X3, X3
XORPD X2, X1
XORPD X3, X1
MOVQ b+8(FP), AX
MOVUPD X1, (AX)
RET
// func Mul11(a, b *[2]uint64)
TEXT ·Mul11(SB),NOSPLIT,$0
MOVQ a+0(FP), AX
MOVUPD (AX), X0
VPSLLQ $1, X0, X1
VPALIGNR $8, X1, X0, X2
PSRLQ $63, X2
MOVUPD ·x127x63(SB), X3
ANDPD X1, X3
VPUNPCKHQDQ X3, X3, X3
XORPD X2, X1
XORPD X3, X1
XORPD X0, X1
MOVQ b+8(FP), AX
MOVUPD X1, (AX)
RET
MOVQ a+0(FP), AX
MOVUPD (AX), X0
VPSLLQ $1, X0, X1
VPALIGNR $8, X1, X0, X2
PSRLQ $63, X2
MOVUPD ·x127x63(SB), X3
ANDPD X1, X3
VPUNPCKHQDQ X3, X3, X3
XORPD X2, X1
XORPD X3, X1
XORPD X0, X1
MOVQ b+8(FP), AX
MOVUPD X1, (AX)
RET
// func Mul(a, b, c *[2]uint64)
TEXT ·Mul(SB),NOSPLIT,$0
MOVQ a+0(FP), AX // X0 = a0 . a1
MOVUPD (AX), X0 // X0 = a0 . a1
MOVQ b+8(FP), BX // X1 = b0 . b1
MOVUPD (BX), X1 // X1 = b0 . b1
VPUNPCKLQDQ X1, X0, X2 // X2 = a0 . b0
VPUNPCKHQDQ X1, X0, X3 // X3 = a1 . b1
XORPD X2, X3 // X3 = (a0 + a1) . (b0 + b1)
PCLMULQDQ $0x10, X3, X3 // X3 = (a0 + a1) * (b0 + b1)
VPCLMULQDQ $0x00, X0, X1, X4 // X4 = a0 * b0
VPCLMULQDQ $0x11, X0, X1, X5 // X5 = a1 * b1
XORPD X4, X3 //
XORPD X5, X3 // X3 = a0 * b1 + a1 * b0
VPSLLDQ $8, X3, X2 //
XORPD X2, X4 // X4 = a0 * b0 + lo(X3)
VPSRLDQ $8, X3, X6 //
XORPD X6, X5 // X5 = a1 * b1 + hi(X3)
MOVQ a+0(FP), AX // X0 = a0 . a1
MOVUPD (AX), X0 // X0 = a0 . a1
MOVQ b+8(FP), BX // X1 = b0 . b1
MOVUPD (BX), X1 // X1 = b0 . b1
VPUNPCKLQDQ X1, X0, X2 // X2 = a0 . b0
VPUNPCKHQDQ X1, X0, X3 // X3 = a1 . b1
XORPD X2, X3 // X3 = (a0 + a1) . (b0 + b1)
PCLMULQDQ $0x10, X3, X3 // X3 = (a0 + a1) * (b0 + b1)
VPCLMULQDQ $0x00, X0, X1, X4 // X4 = a0 * b0
VPCLMULQDQ $0x11, X0, X1, X5 // X5 = a1 * b1
XORPD X4, X3 //
XORPD X5, X3 // X3 = a0 * b1 + a1 * b0
VPSLLDQ $8, X3, X2 //
XORPD X2, X4 // X4 = a0 * b0 + lo(X3)
VPSRLDQ $8, X3, X6 //
XORPD X6, X5 // X5 = a1 * b1 + hi(X3)
// at this point, a * b = X4 . X5 (as 256-bit number)
// reduction modulo x^127 + x^63 + 1
VPALIGNR $8, X4, X5, X3
XORPD X5, X3
PSLLQ $1, X5
XORPD X5, X4
VPUNPCKHQDQ X3, X5, X5
XORPD X5, X4
PSRLQ $63, X3
XORPD X3, X4
VPUNPCKLQDQ X3, X3, X5
PSLLQ $63, X5
XORPD X5, X4
MOVQ c+16(FP), CX
MOVUPD X4, (CX)
RET
// at this point, a * b = X4 . X5 (as 256-bit number)
// reduction modulo x^127 + x^63 + 1
VPALIGNR $8, X4, X5, X3
XORPD X5, X3
PSLLQ $1, X5
XORPD X5, X4
VPUNPCKHQDQ X3, X5, X5
XORPD X5, X4
PSRLQ $63, X3
XORPD X3, X4
VPUNPCKLQDQ X3, X3, X5
PSLLQ $63, X5
XORPD X5, X4
MOVQ c+16(FP), CX
MOVUPD X4, (CX)
RET

20
tz/tzbits_amd64.s
View file

@ -24,12 +24,12 @@ TEXT ·mulBitRight(SB),NOSPLIT,$0
XORPD X7, X5
MOVUPD X5, X0
// c00 += c01
XORPD X1, X0
MOVUPD X0, (AX)
// c00 += c01
XORPD X1, X0
MOVUPD X0, (AX)
// c10 *= 2
VPSLLQ $1, X2, X5
// c10 *= 2
VPSLLQ $1, X2, X5
VPALIGNR $8, X5, X2, X6
PSRLQ $63, X6
MOVUPD ·x127x63(SB), X7
@ -39,7 +39,7 @@ TEXT ·mulBitRight(SB),NOSPLIT,$0
XORPD X7, X5
MOVUPD X5, X2
// c10 += c11
// c10 += c11
XORPD X3, X2
MOVUPD X2, (CX)
@ -54,9 +54,9 @@ TEXT ·mulBitRight(SB),NOSPLIT,$0
MOVUPD X1, (BX)
// c11 = c10 + e
VANDPD X2, X5, X3
VANDPD X2, X5, X3
// c11 += X9 (old c10)
XORPD X9, X3
MOVUPD X3, (DX)
RET
XORPD X9, X3
MOVUPD X3, (DX)
RET