// Code generated by command: go run gen_amd64_compress_asm.go -out ../compress/blocks_amd64.s -stubs ../compress/blocks_amd64.go -pkg compress. DO NOT EDIT. //go:build !purego #include "textflag.h" DATA first_8_blake_consts<>+0(SB)/8, $0x85a308d3243f6a88 DATA first_8_blake_consts<>+8(SB)/8, $0x0370734413198a2e DATA first_8_blake_consts<>+16(SB)/8, $0x299f31d0a4093822 DATA first_8_blake_consts<>+24(SB)/8, $0xec4e6c89082efa98 GLOBL first_8_blake_consts<>(SB), RODATA|NOPTR, $32 DATA permuted_blake_consts<>+0(SB)/8, $0x0370734485a308d3 DATA permuted_blake_consts<>+8(SB)/8, $0xec4e6c89299f31d0 DATA permuted_blake_consts<>+16(SB)/8, $0x13198a2e243f6a88 DATA permuted_blake_consts<>+24(SB)/8, $0x082efa98a4093822 DATA permuted_blake_consts<>+32(SB)/8, $0x34e90c6c38d01377 DATA permuted_blake_consts<>+40(SB)/8, $0xb5470917c97c50dd DATA permuted_blake_consts<>+48(SB)/8, $0xbe5466cf452821e6 DATA permuted_blake_consts<>+56(SB)/8, $0x3f84d5b5c0ac29b7 DATA permuted_blake_consts<>+64(SB)/8, $0x452821e6be5466cf DATA permuted_blake_consts<>+72(SB)/8, $0x082efa98b5470917 DATA permuted_blake_consts<>+80(SB)/8, $0xa40938223f84d5b5 DATA permuted_blake_consts<>+88(SB)/8, $0xc97c50dd38d01377 DATA permuted_blake_consts<>+96(SB)/8, $0x13198a2ec0ac29b7 DATA permuted_blake_consts<>+104(SB)/8, $0x03707344ec4e6c89 DATA permuted_blake_consts<>+112(SB)/8, $0x243f6a8885a308d3 DATA permuted_blake_consts<>+120(SB)/8, $0x299f31d034e90c6c DATA permuted_blake_consts<>+128(SB)/8, $0x243f6a88452821e6 DATA permuted_blake_consts<>+136(SB)/8, $0xc97c50dd13198a2e DATA permuted_blake_consts<>+144(SB)/8, $0xc0ac29b734e90c6c DATA permuted_blake_consts<>+152(SB)/8, $0xb5470917299f31d0 DATA permuted_blake_consts<>+160(SB)/8, $0x082efa983f84d5b5 DATA permuted_blake_consts<>+168(SB)/8, $0xa409382285a308d3 DATA permuted_blake_consts<>+176(SB)/8, $0x03707344be5466cf DATA permuted_blake_consts<>+184(SB)/8, $0x38d01377ec4e6c89 DATA permuted_blake_consts<>+192(SB)/8, $0x85a308d338d01377 DATA permuted_blake_consts<>+200(SB)/8, $0x3f84d5b5c0ac29b7 DATA permuted_blake_consts<>+208(SB)/8, $0x03707344ec4e6c89 DATA permuted_blake_consts<>+216(SB)/8, $0x34e90c6cc97c50dd DATA permuted_blake_consts<>+224(SB)/8, $0xbe5466cf082efa98 DATA permuted_blake_consts<>+232(SB)/8, $0x452821e6243f6a88 DATA permuted_blake_consts<>+240(SB)/8, $0x299f31d013198a2e DATA permuted_blake_consts<>+248(SB)/8, $0xb5470917a4093822 DATA permuted_blake_consts<>+256(SB)/8, $0xec4e6c89243f6a88 DATA permuted_blake_consts<>+264(SB)/8, $0xb5470917a4093822 DATA permuted_blake_consts<>+272(SB)/8, $0x299f31d038d01377 DATA permuted_blake_consts<>+280(SB)/8, $0xbe5466cf13198a2e DATA permuted_blake_consts<>+288(SB)/8, $0xc0ac29b785a308d3 DATA permuted_blake_consts<>+296(SB)/8, $0xc97c50dd452821e6 DATA permuted_blake_consts<>+304(SB)/8, $0x34e90c6c3f84d5b5 DATA permuted_blake_consts<>+312(SB)/8, $0x03707344082efa98 DATA permuted_blake_consts<>+320(SB)/8, $0xbe5466cfc0ac29b7 DATA permuted_blake_consts<>+328(SB)/8, $0x0370734434e90c6c DATA permuted_blake_consts<>+336(SB)/8, $0x082efa9813198a2e DATA permuted_blake_consts<>+344(SB)/8, $0x452821e6243f6a88 DATA permuted_blake_consts<>+352(SB)/8, $0x299f31d0c97c50dd DATA permuted_blake_consts<>+360(SB)/8, $0x38d013773f84d5b5 DATA permuted_blake_consts<>+368(SB)/8, $0xec4e6c89a4093822 DATA permuted_blake_consts<>+376(SB)/8, $0x85a308d3b5470917 DATA permuted_blake_consts<>+384(SB)/8, $0xb5470917299f31d0 DATA permuted_blake_consts<>+392(SB)/8, $0xbe5466cfc97c50dd DATA permuted_blake_consts<>+400(SB)/8, $0x85a308d3c0ac29b7 DATA permuted_blake_consts<>+408(SB)/8, $0xa40938223f84d5b5 DATA permuted_blake_consts<>+416(SB)/8, $0x03707344ec4e6c89 DATA permuted_blake_consts<>+424(SB)/8, $0x34e90c6c13198a2e DATA permuted_blake_consts<>+432(SB)/8, $0x082efa98243f6a88 DATA permuted_blake_consts<>+440(SB)/8, $0x452821e638d01377 DATA permuted_blake_consts<>+448(SB)/8, $0x3f84d5b534e90c6c DATA permuted_blake_consts<>+456(SB)/8, $0x38d0137785a308d3 DATA permuted_blake_consts<>+464(SB)/8, $0xec4e6c89c97c50dd DATA permuted_blake_consts<>+472(SB)/8, $0x03707344c0ac29b7 DATA permuted_blake_consts<>+480(SB)/8, $0xa4093822243f6a88 DATA permuted_blake_consts<>+488(SB)/8, $0xbe5466cf082efa98 DATA permuted_blake_consts<>+496(SB)/8, $0xb5470917299f31d0 DATA permuted_blake_consts<>+504(SB)/8, $0x13198a2e452821e6 DATA permuted_blake_consts<>+512(SB)/8, $0x38d01377b5470917 DATA permuted_blake_consts<>+520(SB)/8, $0x452821e603707344 DATA permuted_blake_consts<>+528(SB)/8, $0x3f84d5b5082efa98 DATA permuted_blake_consts<>+536(SB)/8, $0x243f6a8834e90c6c DATA permuted_blake_consts<>+544(SB)/8, $0xec4e6c8913198a2e DATA permuted_blake_consts<>+552(SB)/8, $0x299f31d0a4093822 DATA permuted_blake_consts<>+560(SB)/8, $0xc97c50ddc0ac29b7 DATA permuted_blake_consts<>+568(SB)/8, $0xbe5466cf85a308d3 DATA permuted_blake_consts<>+576(SB)/8, $0xa409382213198a2e DATA permuted_blake_consts<>+584(SB)/8, $0x299f31d0082efa98 DATA permuted_blake_consts<>+592(SB)/8, $0x452821e6be5466cf DATA permuted_blake_consts<>+600(SB)/8, $0x85a308d3ec4e6c89 DATA permuted_blake_consts<>+608(SB)/8, $0x3f84d5b534e90c6c DATA permuted_blake_consts<>+616(SB)/8, $0x243f6a88c0ac29b7 DATA permuted_blake_consts<>+624(SB)/8, $0x38d01377b5470917 DATA permuted_blake_consts<>+632(SB)/8, $0xc97c50dd03707344 GLOBL permuted_blake_consts<>(SB), RODATA|NOPTR, $640 DATA shuffle_rotr8_4x32<>+0(SB)/8, $0x0407060500030201 DATA shuffle_rotr8_4x32<>+8(SB)/8, $0x0c0f0e0d080b0a09 GLOBL shuffle_rotr8_4x32<>(SB), RODATA|NOPTR, $16 DATA shuffle_rotr16_4x32<>+0(SB)/8, $0x0504070601000302 DATA shuffle_rotr16_4x32<>+8(SB)/8, $0x0d0c0f0e09080b0a GLOBL shuffle_rotr16_4x32<>(SB), RODATA|NOPTR, $16 DATA shuffle_le_to_be_4x32<>+0(SB)/8, $0x0405060700010203 DATA shuffle_le_to_be_4x32<>+8(SB)/8, $0x0c0d0e0f08090a0b GLOBL shuffle_le_to_be_4x32<>(SB), RODATA|NOPTR, $16 // func blocksSSE2(state *State, msg []byte, counter uint64) // Requires: SSE2 TEXT ·blocksSSE2(SB), $64-40 MOVQ state+0(FP), AX MOVQ counter+32(FP), CX MOVQ msg_base+8(FP), DX MOVQ msg_len+16(FP), BX // Convert message len to number of blocks for loop counter. SHRQ $0x06, BX // Initialize state matrix. // row0 = |v0 v1 v2 v3| | h0 h1 h2 h3 | // row1 = |v4 v5 v6 v7| | h4 h5 h6 h7 | MOVOU 32(AX), X0 MOVOU (AX), X1 MOVOU 16(AX), X2 compressLoop: // row2 = |v8 v9 va vb| = |s0^c0 s1^c1 s2^c2 s3^c3| // row3 = |vc vd ve vf| |t0^c4 t0^c5 t1^c6 t1^c7| MOVOU first_8_blake_consts<>+0(SB), X3 PXOR X0, X3 MOVD CX, X4 PSHUFD $0x50, X4, X4 PXOR first_8_blake_consts<>+16(SB), X4 MOVO X1, X5 MOVO X2, X6 // Convert message to big endian. MOVL (DX), SI MOVL 4(DX), DI MOVL 8(DX), R8 MOVL 12(DX), R9 MOVL 16(DX), R10 MOVL 20(DX), R11 MOVL 24(DX), R12 MOVL 28(DX), R13 BSWAPL SI MOVL SI, (SP) BSWAPL DI MOVL DI, 4(SP) BSWAPL R8 MOVL R8, 8(SP) BSWAPL R9 MOVL R9, 12(SP) BSWAPL R10 MOVL R10, 16(SP) BSWAPL R11 MOVL R11, 20(SP) BSWAPL R12 MOVL R12, 24(SP) BSWAPL R13 MOVL R13, 28(SP) MOVL 32(DX), SI MOVL 36(DX), DI MOVL 40(DX), R8 MOVL 44(DX), R9 MOVL 48(DX), R10 MOVL 52(DX), R11 MOVL 56(DX), R12 MOVL 60(DX), R13 BSWAPL SI MOVL SI, 32(SP) BSWAPL DI MOVL DI, 36(SP) BSWAPL R8 MOVL R8, 40(SP) BSWAPL R9 MOVL R9, 44(SP) BSWAPL R10 MOVL R10, 48(SP) BSWAPL R11 MOVL R11, 52(SP) BSWAPL R12 MOVL R12, 56(SP) BSWAPL R13 MOVL R13, 60(SP) // Round 1 column step. MOVD 24(SP), X9 MOVD 16(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 8(SP), X7 MOVD (SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+0(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 28(SP), X9 MOVD 20(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 12(SP), X7 MOVD 4(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+16(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 1 diagonal step part 1: diagonalize. PSHUFD $0x39, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x93, X4, X4 // Round 1 diagonal step part 2: column step. MOVD 56(SP), X9 MOVD 48(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 40(SP), X7 MOVD 32(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+32(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 60(SP), X9 MOVD 52(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 44(SP), X7 MOVD 36(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+48(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 1 diagonal step part 3: undiagonalize. PSHUFD $0x93, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x39, X4, X4 // Round 2 column step. MOVD 52(SP), X9 MOVD 36(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 16(SP), X7 MOVD 56(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+64(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 24(SP), X9 MOVD 60(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 32(SP), X7 MOVD 40(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+80(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 2 diagonal step part 1: diagonalize. PSHUFD $0x39, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x93, X4, X4 // Round 2 diagonal step part 2: column step. MOVD 20(SP), X9 MOVD 44(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD (SP), X7 MOVD 4(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+96(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 12(SP), X9 MOVD 28(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 8(SP), X7 MOVD 48(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+112(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 2 diagonal step part 3: undiagonalize. PSHUFD $0x93, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x39, X4, X4 // Round 3 column step. MOVD 60(SP), X9 MOVD 20(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 48(SP), X7 MOVD 44(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+128(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 52(SP), X9 MOVD 8(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD (SP), X7 MOVD 32(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+144(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 3 diagonal step part 1: diagonalize. PSHUFD $0x39, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x93, X4, X4 // Round 3 diagonal step part 2: column step. MOVD 36(SP), X9 MOVD 28(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 12(SP), X7 MOVD 40(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+160(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 16(SP), X9 MOVD 4(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 24(SP), X7 MOVD 56(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+176(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 3 diagonal step part 3: undiagonalize. PSHUFD $0x93, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x39, X4, X4 // Round 4 column step. MOVD 44(SP), X9 MOVD 52(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 12(SP), X7 MOVD 28(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+192(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 56(SP), X9 MOVD 48(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 4(SP), X7 MOVD 36(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+208(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 4 diagonal step part 1: diagonalize. PSHUFD $0x39, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x93, X4, X4 // Round 4 diagonal step part 2: column step. MOVD 60(SP), X9 MOVD 16(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 20(SP), X7 MOVD 8(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+224(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 32(SP), X9 MOVD (SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 40(SP), X7 MOVD 24(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+240(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 4 diagonal step part 3: undiagonalize. PSHUFD $0x93, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x39, X4, X4 // Round 5 column step. MOVD 40(SP), X9 MOVD 8(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 20(SP), X7 MOVD 36(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+256(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 60(SP), X9 MOVD 16(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 28(SP), X7 MOVD (SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+272(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 5 diagonal step part 1: diagonalize. PSHUFD $0x39, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x93, X4, X4 // Round 5 diagonal step part 2: column step. MOVD 12(SP), X9 MOVD 24(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 44(SP), X7 MOVD 56(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+288(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 52(SP), X9 MOVD 32(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 48(SP), X7 MOVD 4(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+304(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 5 diagonal step part 3: undiagonalize. PSHUFD $0x93, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x39, X4, X4 // Round 6 column step. MOVD 32(SP), X9 MOVD (SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 24(SP), X7 MOVD 8(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+320(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 12(SP), X9 MOVD 44(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 40(SP), X7 MOVD 48(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+336(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 6 diagonal step part 1: diagonalize. PSHUFD $0x39, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x93, X4, X4 // Round 6 diagonal step part 2: column step. MOVD 4(SP), X9 MOVD 60(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 28(SP), X7 MOVD 16(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+352(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 36(SP), X9 MOVD 56(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 20(SP), X7 MOVD 52(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+368(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 6 diagonal step part 3: undiagonalize. PSHUFD $0x93, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x39, X4, X4 // Round 7 column step. MOVD 16(SP), X9 MOVD 56(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 4(SP), X7 MOVD 48(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+384(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 40(SP), X9 MOVD 52(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 60(SP), X7 MOVD 20(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+400(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 7 diagonal step part 1: diagonalize. PSHUFD $0x39, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x93, X4, X4 // Round 7 diagonal step part 2: column step. MOVD 32(SP), X9 MOVD 36(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 24(SP), X7 MOVD (SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+416(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 44(SP), X9 MOVD 8(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 12(SP), X7 MOVD 28(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+432(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 7 diagonal step part 3: undiagonalize. PSHUFD $0x93, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x39, X4, X4 // Round 8 column step. MOVD 12(SP), X9 MOVD 48(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 28(SP), X7 MOVD 52(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+448(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 36(SP), X9 MOVD 4(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 56(SP), X7 MOVD 44(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+464(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 8 diagonal step part 1: diagonalize. PSHUFD $0x39, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x93, X4, X4 // Round 8 diagonal step part 2: column step. MOVD 8(SP), X9 MOVD 32(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 60(SP), X7 MOVD 20(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+480(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 40(SP), X9 MOVD 24(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 16(SP), X7 MOVD (SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+496(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 8 diagonal step part 3: undiagonalize. PSHUFD $0x93, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x39, X4, X4 // Round 9 column step. MOVD (SP), X9 MOVD 44(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 56(SP), X7 MOVD 24(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+512(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 32(SP), X9 MOVD 12(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 36(SP), X7 MOVD 60(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+528(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 9 diagonal step part 1: diagonalize. PSHUFD $0x39, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x93, X4, X4 // Round 9 diagonal step part 2: column step. MOVD 40(SP), X9 MOVD 4(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 52(SP), X7 MOVD 48(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+544(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 20(SP), X9 MOVD 16(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 28(SP), X7 MOVD 8(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+560(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 9 diagonal step part 3: undiagonalize. PSHUFD $0x93, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x39, X4, X4 // Round 10 column step. MOVD 4(SP), X9 MOVD 28(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 32(SP), X7 MOVD 40(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+576(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 20(SP), X9 MOVD 24(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 16(SP), X7 MOVD 8(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+592(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 10 diagonal step part 1: diagonalize. PSHUFD $0x39, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x93, X4, X4 // Round 10 diagonal step part 2: column step. MOVD 52(SP), X9 MOVD 12(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 36(SP), X7 MOVD 60(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+608(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD (SP), X9 MOVD 48(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 56(SP), X7 MOVD 44(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+624(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 10 diagonal step part 3: undiagonalize. PSHUFD $0x93, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x39, X4, X4 // Round 11 column step. MOVD 24(SP), X9 MOVD 16(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 8(SP), X7 MOVD (SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+0(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 28(SP), X9 MOVD 20(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 12(SP), X7 MOVD 4(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+16(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 11 diagonal step part 1: diagonalize. PSHUFD $0x39, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x93, X4, X4 // Round 11 diagonal step part 2: column step. MOVD 56(SP), X9 MOVD 48(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 40(SP), X7 MOVD 32(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+32(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 60(SP), X9 MOVD 52(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 44(SP), X7 MOVD 36(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+48(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 11 diagonal step part 3: undiagonalize. PSHUFD $0x93, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x39, X4, X4 // Round 12 column step. MOVD 52(SP), X9 MOVD 36(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 16(SP), X7 MOVD 56(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+64(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 24(SP), X9 MOVD 60(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 32(SP), X7 MOVD 40(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+80(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 12 diagonal step part 1: diagonalize. PSHUFD $0x39, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x93, X4, X4 // Round 12 diagonal step part 2: column step. MOVD 20(SP), X9 MOVD 44(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD (SP), X7 MOVD 4(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+96(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 12(SP), X9 MOVD 28(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 8(SP), X7 MOVD 48(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+112(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 12 diagonal step part 3: undiagonalize. PSHUFD $0x93, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x39, X4, X4 // Round 13 column step. MOVD 60(SP), X9 MOVD 20(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 48(SP), X7 MOVD 44(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+128(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 52(SP), X9 MOVD 8(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD (SP), X7 MOVD 32(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+144(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 13 diagonal step part 1: diagonalize. PSHUFD $0x39, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x93, X4, X4 // Round 13 diagonal step part 2: column step. MOVD 36(SP), X9 MOVD 28(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 12(SP), X7 MOVD 40(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+160(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 16(SP), X9 MOVD 4(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 24(SP), X7 MOVD 56(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+176(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 13 diagonal step part 3: undiagonalize. PSHUFD $0x93, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x39, X4, X4 // Round 14 column step. MOVD 44(SP), X9 MOVD 52(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 12(SP), X7 MOVD 28(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+192(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 56(SP), X9 MOVD 48(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 4(SP), X7 MOVD 36(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+208(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 14 diagonal step part 1: diagonalize. PSHUFD $0x39, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x93, X4, X4 // Round 14 diagonal step part 2: column step. MOVD 60(SP), X9 MOVD 16(SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 20(SP), X7 MOVD 8(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+224(SB), X8 PXOR X9, X8 PADDD X8, X1 MOVD 32(SP), X9 MOVD (SP), X7 MOVOA X7, X8 PUNPCKLLQ X9, X8 MOVD 40(SP), X7 MOVD 24(SP), X9 PUNPCKLLQ X7, X9 PUNPCKLQDQ X8, X9 MOVOU permuted_blake_consts<>+240(SB), X8 PXOR X9, X8 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x10, X7 PSLLL $0x10, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x0c, X7 PSLLL $0x14, X2 PXOR X7, X2 PADDD X8, X1 PADDD X2, X1 PXOR X1, X4 MOVO X4, X7 PSRLL $0x08, X7 PSLLL $0x18, X4 PXOR X7, X4 PADDD X4, X3 PXOR X3, X2 MOVO X2, X7 PSRLL $0x07, X7 PSLLL $0x19, X2 PXOR X7, X2 // Round 14 diagonal step part 3: undiagonalize. PSHUFD $0x93, X2, X2 PSHUFD $0x4e, X3, X3 PSHUFD $0x39, X4, X4 // Finally the chain value is defined as: // h'0 = h0^s0^v0^v8 // h'1 = h1^s1^v1^v9 // h'2 = h2^s2^v2^va // h'3 = h3^s3^v3^vb // h'4 = h4^s0^v4^vc // h'5 = h5^s1^v5^vd // h'6 = h6^s2^v6^ve // h'7 = h7^s3^v7^vf PXOR X5, X1 PXOR X0, X1 PXOR X3, X1 PXOR X6, X2 PXOR X0, X2 PXOR X4, X2 // Either terminate the loop when there are no more full blocks // to compress or move the message pointer to the next block of // bytes to compress, increment the message bits counter // accordingly, and loop back around to compress it. DECQ BX JZ done LEAQ 64(DX), DX ADDQ $0x00000200, CX JMP compressLoop done: // Output the resulting chain value. MOVOU X1, (AX) MOVOU X2, 16(AX) RET // func blocksSSE41(state *State, msg []byte, counter uint64) // Requires: SSE2, SSE4.1, SSSE3 TEXT ·blocksSSE41(SB), NOSPLIT, $0-40 MOVQ state+0(FP), AX MOVQ counter+32(FP), CX MOVQ msg_base+8(FP), DX MOVQ msg_len+16(FP), BX // Populate registers for faster right rotations. MOVOU shuffle_rotr8_4x32<>+0(SB), X4 MOVOU shuffle_rotr16_4x32<>+0(SB), X5 // Convert message len to number of blocks for loop counter. SHRQ $0x06, BX // Initialize state matrix. // row0 = |v0 v1 v2 v3| | h0 h1 h2 h3 | // row1 = |v4 v5 v6 v7| | h4 h5 h6 h7 | MOVOU 32(AX), X6 MOVOU (AX), X7 MOVOU 16(AX), X8 compressLoop: // row2 = |v8 v9 va vb| = |s0^c0 s1^c1 s2^c2 s3^c3| // row3 = |vc vd ve vf| |t0^c4 t0^c5 t1^c6 t1^c7| MOVOU first_8_blake_consts<>+0(SB), X9 PXOR X6, X9 MOVD CX, X10 PSHUFD $0x50, X10, X10 PXOR first_8_blake_consts<>+16(SB), X10 MOVO X7, X11 MOVO X8, X12 // Convert message to big endian. MOVOU shuffle_le_to_be_4x32<>+0(SB), X13 MOVOU (DX), X0 PSHUFB X13, X0 MOVOU 16(DX), X1 PSHUFB X13, X1 MOVOU 32(DX), X2 PSHUFB X13, X2 MOVOU 48(DX), X3 PSHUFB X13, X3 // Round 1 column step. PSHUFD $0x08, X0, X14 PSHUFD $0x80, X1, X13 PBLENDW $0xf0, X13, X14 MOVOU permuted_blake_consts<>+0(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x0d, X0, X14 PSHUFD $0xd0, X1, X13 PBLENDW $0xf0, X13, X14 MOVOU permuted_blake_consts<>+16(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 1 diagonal step part 1: diagonalize. PSHUFD $0x39, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x93, X10, X10 // Round 1 diagonal step part 2: column step. PSHUFD $0x08, X2, X14 PSHUFD $0x80, X3, X13 PBLENDW $0xf0, X13, X14 MOVOU permuted_blake_consts<>+32(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x0d, X2, X14 PSHUFD $0xd0, X3, X13 PBLENDW $0xf0, X13, X14 MOVOU permuted_blake_consts<>+48(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 1 diagonal step part 3: undiagonalize. PSHUFD $0x93, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x39, X10, X10 // Round 2 column step. PSHUFD $0x00, X1, X14 PSHUFD $0x10, X2, X13 PBLENDW $0x30, X13, X14 PSHUFD $0x42, X3, X13 PBLENDW $0xc3, X13, X14 MOVOU permuted_blake_consts<>+64(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x80, X1, X14 PSHUFD $0x02, X2, X13 PBLENDW $0x0f, X13, X14 PSHUFD $0x30, X3, X13 PBLENDW $0x30, X13, X14 MOVOU permuted_blake_consts<>+80(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 2 diagonal step part 1: diagonalize. PSHUFD $0x39, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x93, X10, X10 // Round 2 diagonal step part 2: column step. PSHUFD $0x01, X0, X14 PSHUFD $0x40, X1, X13 PBLENDW $0xc0, X13, X14 PSHUFD $0x30, X2, X13 PBLENDW $0x30, X13, X14 MOVOU permuted_blake_consts<>+96(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0xc8, X0, X14 PSHUFD $0x30, X1, X13 PBLENDW $0x30, X13, X14 PSHUFD $0x00, X3, X13 PBLENDW $0x03, X13, X14 MOVOU permuted_blake_consts<>+112(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 2 diagonal step part 3: undiagonalize. PSHUFD $0x93, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x39, X10, X10 // Round 3 column step. PSHUFD $0x10, X1, X14 PSHUFD $0x03, X2, X13 PBLENDW $0x03, X13, X14 PSHUFD $0xc0, X3, X13 PBLENDW $0xcc, X13, X14 MOVOU permuted_blake_consts<>+128(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x20, X0, X14 PSHUFD $0x00, X2, X13 PBLENDW $0x03, X13, X14 PSHUFD $0x40, X3, X13 PBLENDW $0xc0, X13, X14 MOVOU permuted_blake_consts<>+144(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 3 diagonal step part 1: diagonalize. PSHUFD $0x39, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x93, X10, X10 // Round 3 diagonal step part 2: column step. PSHUFD $0x0c, X0, X14 PSHUFD $0x30, X1, X13 PBLENDW $0x30, X13, X14 PSHUFD $0x42, X2, X13 PBLENDW $0xc3, X13, X14 MOVOU permuted_blake_consts<>+160(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x10, X0, X14 PSHUFD $0x08, X1, X13 PBLENDW $0xcc, X13, X14 PSHUFD $0x02, X3, X13 PBLENDW $0x03, X13, X14 MOVOU permuted_blake_consts<>+176(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 3 diagonal step part 3: undiagonalize. PSHUFD $0x93, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x39, X10, X10 // Round 4 column step. PSHUFD $0x0c, X0, X14 PSHUFD $0x03, X1, X13 PBLENDW $0x03, X13, X14 PSHUFD $0xc0, X2, X13 PBLENDW $0xc0, X13, X14 PSHUFD $0x10, X3, X13 PBLENDW $0x30, X13, X14 MOVOU permuted_blake_consts<>+192(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x04, X0, X14 PSHUFD $0x01, X2, X13 PBLENDW $0x03, X13, X14 PSHUFD $0x80, X3, X13 PBLENDW $0xf0, X13, X14 MOVOU permuted_blake_consts<>+208(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 4 diagonal step part 1: diagonalize. PSHUFD $0x39, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x93, X10, X10 // Round 4 diagonal step part 2: column step. PSHUFD $0x02, X0, X14 PSHUFD $0x04, X1, X13 PBLENDW $0x3c, X13, X14 PSHUFD $0xc0, X3, X13 PBLENDW $0xc0, X13, X14 MOVOU permuted_blake_consts<>+224(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x00, X0, X14 PSHUFD $0x02, X1, X13 PBLENDW $0x03, X13, X14 PSHUFD $0x08, X2, X13 PBLENDW $0xcc, X13, X14 MOVOU permuted_blake_consts<>+240(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 4 diagonal step part 3: undiagonalize. PSHUFD $0x93, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x39, X10, X10 // Round 5 column step. PSHUFD $0x20, X0, X14 PSHUFD $0x04, X1, X13 PBLENDW $0x0c, X13, X14 PSHUFD $0x81, X2, X13 PBLENDW $0xc3, X13, X14 MOVOU permuted_blake_consts<>+256(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x00, X0, X14 PSHUFD $0x0c, X1, X13 PBLENDW $0x3c, X13, X14 PSHUFD $0xc0, X3, X13 PBLENDW $0xc0, X13, X14 MOVOU permuted_blake_consts<>+272(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 5 diagonal step part 1: diagonalize. PSHUFD $0x39, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x93, X10, X10 // Round 5 diagonal step part 2: column step. PSHUFD $0xc0, X0, X14 PSHUFD $0x20, X1, X13 PBLENDW $0x30, X13, X14 PSHUFD $0x0c, X2, X13 PBLENDW $0x0c, X13, X14 PSHUFD $0x02, X3, X13 PBLENDW $0x03, X13, X14 MOVOU permuted_blake_consts<>+288(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x01, X0, X14 PSHUFD $0x00, X2, X13 PBLENDW $0x30, X13, X14 PSHUFD $0x40, X3, X13 PBLENDW $0xcc, X13, X14 MOVOU permuted_blake_consts<>+304(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 5 diagonal step part 3: undiagonalize. PSHUFD $0x93, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x39, X10, X10 // Round 6 column step. PSHUFD $0x02, X0, X14 PSHUFD $0x08, X1, X13 PBLENDW $0x0c, X13, X14 PSHUFD $0x00, X2, X13 PBLENDW $0xc0, X13, X14 MOVOU permuted_blake_consts<>+320(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0xc0, X0, X14 PSHUFD $0x38, X2, X13 PBLENDW $0x3c, X13, X14 PSHUFD $0x00, X3, X13 PBLENDW $0x03, X13, X14 MOVOU permuted_blake_consts<>+336(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 6 diagonal step part 1: diagonalize. PSHUFD $0x39, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x93, X10, X10 // Round 6 diagonal step part 2: column step. PSHUFD $0x40, X0, X14 PSHUFD $0x0c, X1, X13 PBLENDW $0x0f, X13, X14 PSHUFD $0x30, X3, X13 PBLENDW $0x30, X13, X14 MOVOU permuted_blake_consts<>+352(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x04, X1, X14 PSHUFD $0x40, X2, X13 PBLENDW $0xc0, X13, X14 PSHUFD $0x21, X3, X13 PBLENDW $0x33, X13, X14 MOVOU permuted_blake_consts<>+368(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 6 diagonal step part 3: undiagonalize. PSHUFD $0x93, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x39, X10, X10 // Round 7 column step. PSHUFD $0x04, X0, X14 PSHUFD $0x00, X1, X13 PBLENDW $0xc0, X13, X14 PSHUFD $0x20, X3, X13 PBLENDW $0x33, X13, X14 MOVOU permuted_blake_consts<>+384(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x01, X1, X14 PSHUFD $0x80, X2, X13 PBLENDW $0xc0, X13, X14 PSHUFD $0x1c, X3, X13 PBLENDW $0x3c, X13, X14 MOVOU permuted_blake_consts<>+400(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 7 diagonal step part 1: diagonalize. PSHUFD $0x39, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x93, X10, X10 // Round 7 diagonal step part 2: column step. PSHUFD $0x00, X0, X14 PSHUFD $0x08, X1, X13 PBLENDW $0x0c, X13, X14 PSHUFD $0x10, X2, X13 PBLENDW $0xf0, X13, X14 MOVOU permuted_blake_consts<>+416(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x2c, X0, X14 PSHUFD $0x03, X1, X13 PBLENDW $0x03, X13, X14 PSHUFD $0xc0, X2, X13 PBLENDW $0xc0, X13, X14 MOVOU permuted_blake_consts<>+432(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 7 diagonal step part 3: undiagonalize. PSHUFD $0x93, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x39, X10, X10 // Round 8 column step. PSHUFD $0xc0, X0, X14 PSHUFD $0x0c, X1, X13 PBLENDW $0x0c, X13, X14 PSHUFD $0x01, X3, X13 PBLENDW $0x33, X13, X14 MOVOU permuted_blake_consts<>+448(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x10, X0, X14 PSHUFD $0x43, X2, X13 PBLENDW $0xc3, X13, X14 PSHUFD $0x08, X3, X13 PBLENDW $0x0c, X13, X14 MOVOU permuted_blake_consts<>+464(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 8 diagonal step part 1: diagonalize. PSHUFD $0x39, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x93, X10, X10 // Round 8 diagonal step part 2: column step. PSHUFD $0x80, X0, X14 PSHUFD $0x01, X1, X13 PBLENDW $0x03, X13, X14 PSHUFD $0x00, X2, X13 PBLENDW $0x30, X13, X14 PSHUFD $0x0c, X3, X13 PBLENDW $0x0c, X13, X14 MOVOU permuted_blake_consts<>+480(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x00, X0, X14 PSHUFD $0x20, X1, X13 PBLENDW $0x3c, X13, X14 PSHUFD $0x80, X2, X13 PBLENDW $0xc0, X13, X14 MOVOU permuted_blake_consts<>+496(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 8 diagonal step part 3: undiagonalize. PSHUFD $0x93, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x39, X10, X10 // Round 9 column step. PSHUFD $0x00, X0, X14 PSHUFD $0x02, X1, X13 PBLENDW $0x03, X13, X14 PSHUFD $0x30, X2, X13 PBLENDW $0x30, X13, X14 PSHUFD $0x08, X3, X13 PBLENDW $0x0c, X13, X14 MOVOU permuted_blake_consts<>+512(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x30, X0, X14 PSHUFD $0x04, X2, X13 PBLENDW $0xcc, X13, X14 PSHUFD $0x03, X3, X13 PBLENDW $0x03, X13, X14 MOVOU permuted_blake_consts<>+528(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 9 diagonal step part 1: diagonalize. PSHUFD $0x39, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x93, X10, X10 // Round 9 diagonal step part 2: column step. PSHUFD $0x10, X0, X14 PSHUFD $0x80, X2, X13 PBLENDW $0xc0, X13, X14 PSHUFD $0x04, X3, X13 PBLENDW $0x0f, X13, X14 MOVOU permuted_blake_consts<>+544(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x02, X0, X14 PSHUFD $0x4c, X1, X13 PBLENDW $0xfc, X13, X14 MOVOU permuted_blake_consts<>+560(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 9 diagonal step part 3: undiagonalize. PSHUFD $0x93, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x39, X10, X10 // Round 10 column step. PSHUFD $0x40, X0, X14 PSHUFD $0x30, X1, X13 PBLENDW $0x30, X13, X14 PSHUFD $0x02, X2, X13 PBLENDW $0x0f, X13, X14 MOVOU permuted_blake_consts<>+576(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x02, X0, X14 PSHUFD $0x60, X1, X13 PBLENDW $0xfc, X13, X14 MOVOU permuted_blake_consts<>+592(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 10 diagonal step part 1: diagonalize. PSHUFD $0x39, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x93, X10, X10 // Round 10 diagonal step part 2: column step. PSHUFD $0x30, X0, X14 PSHUFD $0x04, X2, X13 PBLENDW $0x0c, X13, X14 PSHUFD $0x43, X3, X13 PBLENDW $0xc3, X13, X14 MOVOU permuted_blake_consts<>+608(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x00, X0, X14 PSHUFD $0x03, X2, X13 PBLENDW $0x03, X13, X14 PSHUFD $0x08, X3, X13 PBLENDW $0x3c, X13, X14 MOVOU permuted_blake_consts<>+624(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 10 diagonal step part 3: undiagonalize. PSHUFD $0x93, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x39, X10, X10 // Round 11 column step. PSHUFD $0x08, X0, X14 PSHUFD $0x80, X1, X13 PBLENDW $0xf0, X13, X14 MOVOU permuted_blake_consts<>+0(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x0d, X0, X14 PSHUFD $0xd0, X1, X13 PBLENDW $0xf0, X13, X14 MOVOU permuted_blake_consts<>+16(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 11 diagonal step part 1: diagonalize. PSHUFD $0x39, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x93, X10, X10 // Round 11 diagonal step part 2: column step. PSHUFD $0x08, X2, X14 PSHUFD $0x80, X3, X13 PBLENDW $0xf0, X13, X14 MOVOU permuted_blake_consts<>+32(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x0d, X2, X14 PSHUFD $0xd0, X3, X13 PBLENDW $0xf0, X13, X14 MOVOU permuted_blake_consts<>+48(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 11 diagonal step part 3: undiagonalize. PSHUFD $0x93, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x39, X10, X10 // Round 12 column step. PSHUFD $0x00, X1, X14 PSHUFD $0x10, X2, X13 PBLENDW $0x30, X13, X14 PSHUFD $0x42, X3, X13 PBLENDW $0xc3, X13, X14 MOVOU permuted_blake_consts<>+64(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x80, X1, X14 PSHUFD $0x02, X2, X13 PBLENDW $0x0f, X13, X14 PSHUFD $0x30, X3, X13 PBLENDW $0x30, X13, X14 MOVOU permuted_blake_consts<>+80(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 12 diagonal step part 1: diagonalize. PSHUFD $0x39, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x93, X10, X10 // Round 12 diagonal step part 2: column step. PSHUFD $0x01, X0, X14 PSHUFD $0x40, X1, X13 PBLENDW $0xc0, X13, X14 PSHUFD $0x30, X2, X13 PBLENDW $0x30, X13, X14 MOVOU permuted_blake_consts<>+96(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0xc8, X0, X14 PSHUFD $0x30, X1, X13 PBLENDW $0x30, X13, X14 PSHUFD $0x00, X3, X13 PBLENDW $0x03, X13, X14 MOVOU permuted_blake_consts<>+112(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 12 diagonal step part 3: undiagonalize. PSHUFD $0x93, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x39, X10, X10 // Round 13 column step. PSHUFD $0x10, X1, X14 PSHUFD $0x03, X2, X13 PBLENDW $0x03, X13, X14 PSHUFD $0xc0, X3, X13 PBLENDW $0xcc, X13, X14 MOVOU permuted_blake_consts<>+128(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x20, X0, X14 PSHUFD $0x00, X2, X13 PBLENDW $0x03, X13, X14 PSHUFD $0x40, X3, X13 PBLENDW $0xc0, X13, X14 MOVOU permuted_blake_consts<>+144(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 13 diagonal step part 1: diagonalize. PSHUFD $0x39, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x93, X10, X10 // Round 13 diagonal step part 2: column step. PSHUFD $0x0c, X0, X14 PSHUFD $0x30, X1, X13 PBLENDW $0x30, X13, X14 PSHUFD $0x42, X2, X13 PBLENDW $0xc3, X13, X14 MOVOU permuted_blake_consts<>+160(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x10, X0, X14 PSHUFD $0x08, X1, X13 PBLENDW $0xcc, X13, X14 PSHUFD $0x02, X3, X13 PBLENDW $0x03, X13, X14 MOVOU permuted_blake_consts<>+176(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 13 diagonal step part 3: undiagonalize. PSHUFD $0x93, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x39, X10, X10 // Round 14 column step. PSHUFD $0x0c, X0, X14 PSHUFD $0x03, X1, X13 PBLENDW $0x03, X13, X14 PSHUFD $0xc0, X2, X13 PBLENDW $0xc0, X13, X14 PSHUFD $0x10, X3, X13 PBLENDW $0x30, X13, X14 MOVOU permuted_blake_consts<>+192(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x04, X0, X14 PSHUFD $0x01, X2, X13 PBLENDW $0x03, X13, X14 PSHUFD $0x80, X3, X13 PBLENDW $0xf0, X13, X14 MOVOU permuted_blake_consts<>+208(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 14 diagonal step part 1: diagonalize. PSHUFD $0x39, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x93, X10, X10 // Round 14 diagonal step part 2: column step. PSHUFD $0x02, X0, X14 PSHUFD $0x04, X1, X13 PBLENDW $0x3c, X13, X14 PSHUFD $0xc0, X3, X13 PBLENDW $0xc0, X13, X14 MOVOU permuted_blake_consts<>+224(SB), X15 PXOR X14, X15 PADDD X15, X7 PSHUFD $0x00, X0, X14 PSHUFD $0x02, X1, X13 PBLENDW $0x03, X13, X14 PSHUFD $0x08, X2, X13 PBLENDW $0xcc, X13, X14 MOVOU permuted_blake_consts<>+240(SB), X15 PXOR X14, X15 PADDD X8, X7 PXOR X7, X10 PSHUFB X5, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x0c, X13 PSLLL $0x14, X8 PXOR X13, X8 PADDD X15, X7 PADDD X8, X7 PXOR X7, X10 PSHUFB X4, X10 PADDD X10, X9 PXOR X9, X8 MOVO X8, X13 PSRLL $0x07, X13 PSLLL $0x19, X8 PXOR X13, X8 // Round 14 diagonal step part 3: undiagonalize. PSHUFD $0x93, X8, X8 PSHUFD $0x4e, X9, X9 PSHUFD $0x39, X10, X10 // Finally the chain value is defined as: // h'0 = h0^s0^v0^v8 // h'1 = h1^s1^v1^v9 // h'2 = h2^s2^v2^va // h'3 = h3^s3^v3^vb // h'4 = h4^s0^v4^vc // h'5 = h5^s1^v5^vd // h'6 = h6^s2^v6^ve // h'7 = h7^s3^v7^vf PXOR X11, X7 PXOR X6, X7 PXOR X9, X7 PXOR X12, X8 PXOR X6, X8 PXOR X10, X8 // Either terminate the loop when there are no more full blocks // to compress or move the message pointer to the next block of // bytes to compress, increment the message bits counter // accordingly, and loop back around to compress it. DECQ BX JZ done LEAQ 64(DX), DX ADDQ $0x00000200, CX JMP compressLoop done: // Output the resulting chain value. MOVOU X7, (AX) MOVOU X8, 16(AX) RET // func blocksAVX(state *State, msg []byte, counter uint64) // Requires: AVX TEXT ·blocksAVX(SB), NOSPLIT, $0-40 MOVQ state+0(FP), AX MOVQ counter+32(FP), CX MOVQ msg_base+8(FP), DX MOVQ msg_len+16(FP), BX // Populate registers for fast right rotations. VMOVDQU shuffle_rotr8_4x32<>+0(SB), X4 VMOVDQU shuffle_rotr16_4x32<>+0(SB), X5 // Convert message len to number of blocks for loop counter. SHRQ $0x06, BX // Initialize state matrix. // row0 = |v0 v1 v2 v3| | h0 h1 h2 h3 | // row1 = |v4 v5 v6 v7| | h4 h5 h6 h7 | VMOVDQU 32(AX), X6 VMOVDQU (AX), X7 VMOVDQU 16(AX), X8 compressLoop: // row2 = |v8 v9 va vb| = |s0^c0 s1^c1 s2^c2 s3^c3| // row3 = |vc vd ve vf| |t0^c4 t0^c5 t1^c6 t1^c7| VMOVDQU first_8_blake_consts<>+0(SB), X9 VPXOR X6, X9, X9 VMOVQ CX, X10 VPSHUFD $0x50, X10, X10 VPXOR first_8_blake_consts<>+16(SB), X10, X10 VMOVDQA X7, X11 VMOVDQA X8, X12 // Convert message to big endian. VMOVDQU shuffle_le_to_be_4x32<>+0(SB), X13 VMOVDQU (DX), X0 VPSHUFB X13, X0, X0 VMOVDQU 16(DX), X1 VPSHUFB X13, X1, X1 VMOVDQU 32(DX), X2 VPSHUFB X13, X2, X2 VMOVDQU 48(DX), X3 VPSHUFB X13, X3, X3 // Round 1 column step. VPSHUFD $0x08, X0, X14 VPSHUFD $0x80, X1, X13 VPBLENDW $0xf0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+0(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x0d, X0, X14 VPSHUFD $0xd0, X1, X13 VPBLENDW $0xf0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+16(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 1 diagonal step part 1: diagonalize. VPSHUFD $0x39, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x93, X10, X10 // Round 1 diagonal step part 2: column step. VPSHUFD $0x08, X2, X14 VPSHUFD $0x80, X3, X13 VPBLENDW $0xf0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+32(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x0d, X2, X14 VPSHUFD $0xd0, X3, X13 VPBLENDW $0xf0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+48(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 1 diagonal step part 3: undiagonalize. VPSHUFD $0x93, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x39, X10, X10 // Round 2 column step. VPSHUFD $0x00, X1, X14 VPSHUFD $0x10, X2, X13 VPBLENDW $0x30, X13, X14, X14 VPSHUFD $0x42, X3, X13 VPBLENDW $0xc3, X13, X14, X14 VMOVDQU permuted_blake_consts<>+64(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x80, X1, X14 VPSHUFD $0x02, X2, X13 VPBLENDW $0x0f, X13, X14, X14 VPSHUFD $0x30, X3, X13 VPBLENDW $0x30, X13, X14, X14 VMOVDQU permuted_blake_consts<>+80(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 2 diagonal step part 1: diagonalize. VPSHUFD $0x39, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x93, X10, X10 // Round 2 diagonal step part 2: column step. VPSHUFD $0x01, X0, X14 VPSHUFD $0x40, X1, X13 VPBLENDW $0xc0, X13, X14, X14 VPSHUFD $0x30, X2, X13 VPBLENDW $0x30, X13, X14, X14 VMOVDQU permuted_blake_consts<>+96(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0xc8, X0, X14 VPSHUFD $0x30, X1, X13 VPBLENDW $0x30, X13, X14, X14 VPSHUFD $0x00, X3, X13 VPBLENDW $0x03, X13, X14, X14 VMOVDQU permuted_blake_consts<>+112(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 2 diagonal step part 3: undiagonalize. VPSHUFD $0x93, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x39, X10, X10 // Round 3 column step. VPSHUFD $0x10, X1, X14 VPSHUFD $0x03, X2, X13 VPBLENDW $0x03, X13, X14, X14 VPSHUFD $0xc0, X3, X13 VPBLENDW $0xcc, X13, X14, X14 VMOVDQU permuted_blake_consts<>+128(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x20, X0, X14 VPSHUFD $0x00, X2, X13 VPBLENDW $0x03, X13, X14, X14 VPSHUFD $0x40, X3, X13 VPBLENDW $0xc0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+144(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 3 diagonal step part 1: diagonalize. VPSHUFD $0x39, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x93, X10, X10 // Round 3 diagonal step part 2: column step. VPSHUFD $0x0c, X0, X14 VPSHUFD $0x30, X1, X13 VPBLENDW $0x30, X13, X14, X14 VPSHUFD $0x42, X2, X13 VPBLENDW $0xc3, X13, X14, X14 VMOVDQU permuted_blake_consts<>+160(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x10, X0, X14 VPSHUFD $0x08, X1, X13 VPBLENDW $0xcc, X13, X14, X14 VPSHUFD $0x02, X3, X13 VPBLENDW $0x03, X13, X14, X14 VMOVDQU permuted_blake_consts<>+176(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 3 diagonal step part 3: undiagonalize. VPSHUFD $0x93, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x39, X10, X10 // Round 4 column step. VPSHUFD $0x0c, X0, X14 VPSHUFD $0x03, X1, X13 VPBLENDW $0x03, X13, X14, X14 VPSHUFD $0xc0, X2, X13 VPBLENDW $0xc0, X13, X14, X14 VPSHUFD $0x10, X3, X13 VPBLENDW $0x30, X13, X14, X14 VMOVDQU permuted_blake_consts<>+192(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x04, X0, X14 VPSHUFD $0x01, X2, X13 VPBLENDW $0x03, X13, X14, X14 VPSHUFD $0x80, X3, X13 VPBLENDW $0xf0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+208(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 4 diagonal step part 1: diagonalize. VPSHUFD $0x39, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x93, X10, X10 // Round 4 diagonal step part 2: column step. VPSHUFD $0x02, X0, X14 VPSHUFD $0x04, X1, X13 VPBLENDW $0x3c, X13, X14, X14 VPSHUFD $0xc0, X3, X13 VPBLENDW $0xc0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+224(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x00, X0, X14 VPSHUFD $0x02, X1, X13 VPBLENDW $0x03, X13, X14, X14 VPSHUFD $0x08, X2, X13 VPBLENDW $0xcc, X13, X14, X14 VMOVDQU permuted_blake_consts<>+240(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 4 diagonal step part 3: undiagonalize. VPSHUFD $0x93, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x39, X10, X10 // Round 5 column step. VPSHUFD $0x20, X0, X14 VPSHUFD $0x04, X1, X13 VPBLENDW $0x0c, X13, X14, X14 VPSHUFD $0x81, X2, X13 VPBLENDW $0xc3, X13, X14, X14 VMOVDQU permuted_blake_consts<>+256(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x00, X0, X14 VPSHUFD $0x0c, X1, X13 VPBLENDW $0x3c, X13, X14, X14 VPSHUFD $0xc0, X3, X13 VPBLENDW $0xc0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+272(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 5 diagonal step part 1: diagonalize. VPSHUFD $0x39, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x93, X10, X10 // Round 5 diagonal step part 2: column step. VPSHUFD $0xc0, X0, X14 VPSHUFD $0x20, X1, X13 VPBLENDW $0x30, X13, X14, X14 VPSHUFD $0x0c, X2, X13 VPBLENDW $0x0c, X13, X14, X14 VPSHUFD $0x02, X3, X13 VPBLENDW $0x03, X13, X14, X14 VMOVDQU permuted_blake_consts<>+288(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x01, X0, X14 VPSHUFD $0x00, X2, X13 VPBLENDW $0x30, X13, X14, X14 VPSHUFD $0x40, X3, X13 VPBLENDW $0xcc, X13, X14, X14 VMOVDQU permuted_blake_consts<>+304(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 5 diagonal step part 3: undiagonalize. VPSHUFD $0x93, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x39, X10, X10 // Round 6 column step. VPSHUFD $0x02, X0, X14 VPSHUFD $0x08, X1, X13 VPBLENDW $0x0c, X13, X14, X14 VPSHUFD $0x00, X2, X13 VPBLENDW $0xc0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+320(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0xc0, X0, X14 VPSHUFD $0x38, X2, X13 VPBLENDW $0x3c, X13, X14, X14 VPSHUFD $0x00, X3, X13 VPBLENDW $0x03, X13, X14, X14 VMOVDQU permuted_blake_consts<>+336(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 6 diagonal step part 1: diagonalize. VPSHUFD $0x39, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x93, X10, X10 // Round 6 diagonal step part 2: column step. VPSHUFD $0x40, X0, X14 VPSHUFD $0x0c, X1, X13 VPBLENDW $0x0f, X13, X14, X14 VPSHUFD $0x30, X3, X13 VPBLENDW $0x30, X13, X14, X14 VMOVDQU permuted_blake_consts<>+352(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x04, X1, X14 VPSHUFD $0x40, X2, X13 VPBLENDW $0xc0, X13, X14, X14 VPSHUFD $0x21, X3, X13 VPBLENDW $0x33, X13, X14, X14 VMOVDQU permuted_blake_consts<>+368(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 6 diagonal step part 3: undiagonalize. VPSHUFD $0x93, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x39, X10, X10 // Round 7 column step. VPSHUFD $0x04, X0, X14 VPSHUFD $0x00, X1, X13 VPBLENDW $0xc0, X13, X14, X14 VPSHUFD $0x20, X3, X13 VPBLENDW $0x33, X13, X14, X14 VMOVDQU permuted_blake_consts<>+384(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x01, X1, X14 VPSHUFD $0x80, X2, X13 VPBLENDW $0xc0, X13, X14, X14 VPSHUFD $0x1c, X3, X13 VPBLENDW $0x3c, X13, X14, X14 VMOVDQU permuted_blake_consts<>+400(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 7 diagonal step part 1: diagonalize. VPSHUFD $0x39, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x93, X10, X10 // Round 7 diagonal step part 2: column step. VPSHUFD $0x00, X0, X14 VPSHUFD $0x08, X1, X13 VPBLENDW $0x0c, X13, X14, X14 VPSHUFD $0x10, X2, X13 VPBLENDW $0xf0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+416(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x2c, X0, X14 VPSHUFD $0x03, X1, X13 VPBLENDW $0x03, X13, X14, X14 VPSHUFD $0xc0, X2, X13 VPBLENDW $0xc0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+432(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 7 diagonal step part 3: undiagonalize. VPSHUFD $0x93, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x39, X10, X10 // Round 8 column step. VPSHUFD $0xc0, X0, X14 VPSHUFD $0x0c, X1, X13 VPBLENDW $0x0c, X13, X14, X14 VPSHUFD $0x01, X3, X13 VPBLENDW $0x33, X13, X14, X14 VMOVDQU permuted_blake_consts<>+448(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x10, X0, X14 VPSHUFD $0x43, X2, X13 VPBLENDW $0xc3, X13, X14, X14 VPSHUFD $0x08, X3, X13 VPBLENDW $0x0c, X13, X14, X14 VMOVDQU permuted_blake_consts<>+464(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 8 diagonal step part 1: diagonalize. VPSHUFD $0x39, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x93, X10, X10 // Round 8 diagonal step part 2: column step. VPSHUFD $0x80, X0, X14 VPSHUFD $0x01, X1, X13 VPBLENDW $0x03, X13, X14, X14 VPSHUFD $0x00, X2, X13 VPBLENDW $0x30, X13, X14, X14 VPSHUFD $0x0c, X3, X13 VPBLENDW $0x0c, X13, X14, X14 VMOVDQU permuted_blake_consts<>+480(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x00, X0, X14 VPSHUFD $0x20, X1, X13 VPBLENDW $0x3c, X13, X14, X14 VPSHUFD $0x80, X2, X13 VPBLENDW $0xc0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+496(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 8 diagonal step part 3: undiagonalize. VPSHUFD $0x93, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x39, X10, X10 // Round 9 column step. VPSHUFD $0x00, X0, X14 VPSHUFD $0x02, X1, X13 VPBLENDW $0x03, X13, X14, X14 VPSHUFD $0x30, X2, X13 VPBLENDW $0x30, X13, X14, X14 VPSHUFD $0x08, X3, X13 VPBLENDW $0x0c, X13, X14, X14 VMOVDQU permuted_blake_consts<>+512(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x30, X0, X14 VPSHUFD $0x04, X2, X13 VPBLENDW $0xcc, X13, X14, X14 VPSHUFD $0x03, X3, X13 VPBLENDW $0x03, X13, X14, X14 VMOVDQU permuted_blake_consts<>+528(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 9 diagonal step part 1: diagonalize. VPSHUFD $0x39, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x93, X10, X10 // Round 9 diagonal step part 2: column step. VPSHUFD $0x10, X0, X14 VPSHUFD $0x80, X2, X13 VPBLENDW $0xc0, X13, X14, X14 VPSHUFD $0x04, X3, X13 VPBLENDW $0x0f, X13, X14, X14 VMOVDQU permuted_blake_consts<>+544(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x02, X0, X14 VPSHUFD $0x4c, X1, X13 VPBLENDW $0xfc, X13, X14, X14 VMOVDQU permuted_blake_consts<>+560(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 9 diagonal step part 3: undiagonalize. VPSHUFD $0x93, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x39, X10, X10 // Round 10 column step. VPSHUFD $0x40, X0, X14 VPSHUFD $0x30, X1, X13 VPBLENDW $0x30, X13, X14, X14 VPSHUFD $0x02, X2, X13 VPBLENDW $0x0f, X13, X14, X14 VMOVDQU permuted_blake_consts<>+576(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x02, X0, X14 VPSHUFD $0x60, X1, X13 VPBLENDW $0xfc, X13, X14, X14 VMOVDQU permuted_blake_consts<>+592(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 10 diagonal step part 1: diagonalize. VPSHUFD $0x39, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x93, X10, X10 // Round 10 diagonal step part 2: column step. VPSHUFD $0x30, X0, X14 VPSHUFD $0x04, X2, X13 VPBLENDW $0x0c, X13, X14, X14 VPSHUFD $0x43, X3, X13 VPBLENDW $0xc3, X13, X14, X14 VMOVDQU permuted_blake_consts<>+608(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x00, X0, X14 VPSHUFD $0x03, X2, X13 VPBLENDW $0x03, X13, X14, X14 VPSHUFD $0x08, X3, X13 VPBLENDW $0x3c, X13, X14, X14 VMOVDQU permuted_blake_consts<>+624(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 10 diagonal step part 3: undiagonalize. VPSHUFD $0x93, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x39, X10, X10 // Round 11 column step. VPSHUFD $0x08, X0, X14 VPSHUFD $0x80, X1, X13 VPBLENDW $0xf0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+0(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x0d, X0, X14 VPSHUFD $0xd0, X1, X13 VPBLENDW $0xf0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+16(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 11 diagonal step part 1: diagonalize. VPSHUFD $0x39, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x93, X10, X10 // Round 11 diagonal step part 2: column step. VPSHUFD $0x08, X2, X14 VPSHUFD $0x80, X3, X13 VPBLENDW $0xf0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+32(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x0d, X2, X14 VPSHUFD $0xd0, X3, X13 VPBLENDW $0xf0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+48(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 11 diagonal step part 3: undiagonalize. VPSHUFD $0x93, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x39, X10, X10 // Round 12 column step. VPSHUFD $0x00, X1, X14 VPSHUFD $0x10, X2, X13 VPBLENDW $0x30, X13, X14, X14 VPSHUFD $0x42, X3, X13 VPBLENDW $0xc3, X13, X14, X14 VMOVDQU permuted_blake_consts<>+64(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x80, X1, X14 VPSHUFD $0x02, X2, X13 VPBLENDW $0x0f, X13, X14, X14 VPSHUFD $0x30, X3, X13 VPBLENDW $0x30, X13, X14, X14 VMOVDQU permuted_blake_consts<>+80(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 12 diagonal step part 1: diagonalize. VPSHUFD $0x39, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x93, X10, X10 // Round 12 diagonal step part 2: column step. VPSHUFD $0x01, X0, X14 VPSHUFD $0x40, X1, X13 VPBLENDW $0xc0, X13, X14, X14 VPSHUFD $0x30, X2, X13 VPBLENDW $0x30, X13, X14, X14 VMOVDQU permuted_blake_consts<>+96(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0xc8, X0, X14 VPSHUFD $0x30, X1, X13 VPBLENDW $0x30, X13, X14, X14 VPSHUFD $0x00, X3, X13 VPBLENDW $0x03, X13, X14, X14 VMOVDQU permuted_blake_consts<>+112(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 12 diagonal step part 3: undiagonalize. VPSHUFD $0x93, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x39, X10, X10 // Round 13 column step. VPSHUFD $0x10, X1, X14 VPSHUFD $0x03, X2, X13 VPBLENDW $0x03, X13, X14, X14 VPSHUFD $0xc0, X3, X13 VPBLENDW $0xcc, X13, X14, X14 VMOVDQU permuted_blake_consts<>+128(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x20, X0, X14 VPSHUFD $0x00, X2, X13 VPBLENDW $0x03, X13, X14, X14 VPSHUFD $0x40, X3, X13 VPBLENDW $0xc0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+144(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 13 diagonal step part 1: diagonalize. VPSHUFD $0x39, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x93, X10, X10 // Round 13 diagonal step part 2: column step. VPSHUFD $0x0c, X0, X14 VPSHUFD $0x30, X1, X13 VPBLENDW $0x30, X13, X14, X14 VPSHUFD $0x42, X2, X13 VPBLENDW $0xc3, X13, X14, X14 VMOVDQU permuted_blake_consts<>+160(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x10, X0, X14 VPSHUFD $0x08, X1, X13 VPBLENDW $0xcc, X13, X14, X14 VPSHUFD $0x02, X3, X13 VPBLENDW $0x03, X13, X14, X14 VMOVDQU permuted_blake_consts<>+176(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 13 diagonal step part 3: undiagonalize. VPSHUFD $0x93, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x39, X10, X10 // Round 14 column step. VPSHUFD $0x0c, X0, X14 VPSHUFD $0x03, X1, X13 VPBLENDW $0x03, X13, X14, X14 VPSHUFD $0xc0, X2, X13 VPBLENDW $0xc0, X13, X14, X14 VPSHUFD $0x10, X3, X13 VPBLENDW $0x30, X13, X14, X14 VMOVDQU permuted_blake_consts<>+192(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x04, X0, X14 VPSHUFD $0x01, X2, X13 VPBLENDW $0x03, X13, X14, X14 VPSHUFD $0x80, X3, X13 VPBLENDW $0xf0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+208(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 14 diagonal step part 1: diagonalize. VPSHUFD $0x39, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x93, X10, X10 // Round 14 diagonal step part 2: column step. VPSHUFD $0x02, X0, X14 VPSHUFD $0x04, X1, X13 VPBLENDW $0x3c, X13, X14, X14 VPSHUFD $0xc0, X3, X13 VPBLENDW $0xc0, X13, X14, X14 VMOVDQU permuted_blake_consts<>+224(SB), X15 VPXOR X14, X15, X15 VPADDD X15, X7, X7 VPSHUFD $0x00, X0, X14 VPSHUFD $0x02, X1, X13 VPBLENDW $0x03, X13, X14, X14 VPSHUFD $0x08, X2, X13 VPBLENDW $0xcc, X13, X14, X14 VMOVDQU permuted_blake_consts<>+240(SB), X15 VPXOR X14, X15, X15 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X5, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x0c, X8, X13 VPSLLD $0x14, X8, X8 VPXOR X13, X8, X8 VPADDD X15, X7, X7 VPADDD X8, X7, X7 VPXOR X7, X10, X10 VPSHUFB X4, X10, X10 VPADDD X10, X9, X9 VPXOR X9, X8, X8 VPSRLD $0x07, X8, X13 VPSLLD $0x19, X8, X8 VPXOR X13, X8, X8 // Round 14 diagonal step part 3: undiagonalize. VPSHUFD $0x93, X8, X8 VPSHUFD $0x4e, X9, X9 VPSHUFD $0x39, X10, X10 // Finally the chain value is defined as: // h'0 = h0^s0^v0^v8 // h'1 = h1^s1^v1^v9 // h'2 = h2^s2^v2^va // h'3 = h3^s3^v3^vb // h'4 = h4^s0^v4^vc // h'5 = h5^s1^v5^vd // h'6 = h6^s2^v6^ve // h'7 = h7^s3^v7^vf VPXOR X11, X7, X7 VPXOR X6, X7, X7 VPXOR X9, X7, X7 VPXOR X12, X8, X8 VPXOR X6, X8, X8 VPXOR X10, X8, X8 // Either terminate the loop when there are no more full blocks // to compress or move the message pointer to the next block of // bytes to compress, increment the message bits counter // accordingly, and loop back around to compress it. DECQ BX JZ done LEAQ 64(DX), DX ADDQ $0x00000200, CX JMP compressLoop done: // Output the resulting chain value. VMOVDQU X7, (AX) VMOVDQU X8, 16(AX) RET