diff options
Diffstat (limited to 'blake2s/blake2s_loong64.s')
| -rw-r--r-- | blake2s/blake2s_loong64.s | 196 |
1 files changed, 0 insertions, 196 deletions
diff --git a/blake2s/blake2s_loong64.s b/blake2s/blake2s_loong64.s deleted file mode 100644 index c222144..0000000 --- a/blake2s/blake2s_loong64.s +++ /dev/null @@ -1,196 +0,0 @@ -// Copyright 2025 The Go Authors. All rights reserved. -// Use of this source code is governed by a BSD-style -// license that can be found in the LICENSE file. - -//go:build loong64 && gc && !purego - -#include "textflag.h" - -DATA ·iv0<>+0(SB)/4, $0x6a09e667 -DATA ·iv0<>+4(SB)/4, $0xbb67ae85 -DATA ·iv0<>+8(SB)/4, $0x3c6ef372 -DATA ·iv0<>+12(SB)/4, $0xa54ff53a -GLOBL ·iv0<>(SB), RODATA|NOPTR, $16 - -DATA ·iv1<>+0(SB)/4, $0x510e527f -DATA ·iv1<>+4(SB)/4, $0x9b05688c -DATA ·iv1<>+8(SB)/4, $0x1f83d9ab -DATA ·iv1<>+12(SB)/4, $0x5be0cd19 -GLOBL ·iv1<>(SB), RODATA|NOPTR, $16 - -#define SHUFFLE_1 \ - VSHUF4IW $57, V1, V1; \ - VSHUF4IW $78, V2, V2; \ - VSHUF4IW $147, V3, V3; \ - -#define SHUFFLE_2 \ - VSHUF4IW $147, V1, V1; \ - VSHUF4IW $78, V2, V2; \ - VSHUF4IW $57, V3, V3; \ - -#define LOAD_M(a, b, c, d, e, f, g, h) \ - VMOVQ a, V8.W[0]; \ - VMOVQ b, V8.W[1]; \ - VMOVQ c, V8.W[2]; \ - VMOVQ d, V8.W[3]; \ - VMOVQ e, V9.W[0]; \ - VMOVQ f, V9.W[1]; \ - VMOVQ g, V9.W[2]; \ - VMOVQ h, V9.W[3]; \ - -#define ROUND_0 \ - VADDW V0, V8, V0; \ - VADDW V0, V1, V0; \ - VXORV V3, V0, V3; \ - VROTRW $16, V3, V3; \ - VADDW V2, V3, V2; \ - VXORV V1, V2, V1; \ - VROTRW $12, V1, V1; \ - VADDW V0, V9, V0; \ - VADDW V0, V1, V0; \ - VXORV V3, V0, V3; \ - VROTRW $8, V3, V3; \ - VADDW V2, V3, V2; \ - VXORV V1, V2, V1; \ - VROTRW $7, V1, V1; \ - -#define ROUND_8 ROUND_0 - -// func hashBlocksVX(h *[8]uint32, c *[2]uint32, flag uint32, blocks []byte) -TEXT ·hashBlocksVX(SB), NOSPLIT, $0-48 - MOVV h+0(FP), R4 - MOVV c+8(FP), R5 - MOVWU flag+16(FP), R6 - MOVV blocks_base+24(FP), R7 - MOVV blocks_len+32(FP), R8 - MOVW (R5), R9 // c0 - MOVW 4(R5), R10 // c1 - -loop: - ADD $0x40, R9 - SGTU $0x40, R9, R11 - ADD R10, R11, R10 - - MOVV $·iv0<>(SB), R11 - MOVV $·iv1<>(SB), R12 - MOVWU 0(R12), R13 // v12 - MOVWU 4(R12), R14 // v13 - MOVWU 8(R12), R15 // v14 - MOVWU 12(R12), R16 // v15 - XOR R13, R9, R13 - XOR R14, R10, R14 - XOR R15, R6, R15 - - VMOVQ (R4), V0 - VMOVQ 16(R4), V1 - VMOVQ (R11), V2 - VMOVQ R16, V3.W[3] - VMOVQ R13, V3.W[0] - VMOVQ R14, V3.W[1] - VMOVQ R15, V3.W[2] - - MOVWU (R7), R11 - MOVWU 4(R7), R12 - MOVWU 8(R7), R13 - MOVWU 12(R7), R14 - MOVWU 16(R7), R15 - MOVWU 20(R7), R16 - MOVWU 24(R7), R17 - MOVWU 28(R7), R18 - MOVWU 32(R7), R19 - MOVWU 36(R7), R24 - MOVWU 40(R7), R25 - MOVWU 44(R7), R26 - MOVWU 48(R7), R27 - MOVWU 52(R7), R28 - MOVWU 56(R7), R29 - MOVWU 60(R7), R30 - - LOAD_M(R11, R13, R15, R17, R12, R14, R16, R18) - ROUND_0 - SHUFFLE_1 - LOAD_M(R19, R25, R27, R29, R24, R26, R28, R30) - ROUND_8 - SHUFFLE_2 - - LOAD_M(R29, R15, R24, R28, R25, R19, R30, R17) - ROUND_0 - SHUFFLE_1 - LOAD_M(R12, R11, R26, R16, R27, R13, R18, R14) - ROUND_8 - SHUFFLE_2 - - LOAD_M(R26, R27, R16, R30, R19, R11, R13, R28) - ROUND_0 - SHUFFLE_1 - LOAD_M(R25, R14, R18, R24, R29, R17, R12, R15) - ROUND_8 - SHUFFLE_2 - - LOAD_M(R18, R14, R28, R26, R24, R12, R27, R29) - ROUND_0 - SHUFFLE_1 - LOAD_M(R13, R16, R15, R30, R17, R25, R11, R19) - ROUND_8 - SHUFFLE_2 - - LOAD_M(R24, R16, R13, R25, R11, R18, R15, R30) - ROUND_0 - SHUFFLE_1 - LOAD_M(R29, R26, R17, R14, R12, R27, R19, R28) - ROUND_8 - SHUFFLE_2 - - LOAD_M(R13, R17, R11, R19, R27, R25, R26, R14) - ROUND_0 - SHUFFLE_1 - LOAD_M(R15, R18, R30, R12, R28, R16, R29, R24) - ROUND_8 - SHUFFLE_2 - - LOAD_M(R27, R12, R29, R15, R16, R30, R28, R25) - ROUND_0 - SHUFFLE_1 - LOAD_M(R11, R17, R24, R19, R18, R14, R13, R26) - ROUND_8 - SHUFFLE_2 - - LOAD_M(R28, R18, R27, R14, R26, R29, R12, R24) - ROUND_0 - SHUFFLE_1 - LOAD_M(R16, R30, R19, R13, R11, R15, R17, R25) - ROUND_8 - SHUFFLE_2 - - LOAD_M(R17, R29, R26, R11, R30, R24, R14, R19) - ROUND_0 - SHUFFLE_1 - LOAD_M(R27, R28, R12, R25, R13, R18, R15, R16) - ROUND_8 - SHUFFLE_2 - - LOAD_M(R25, R19, R18, R12, R13, R15, R17, R16) - ROUND_0 - SHUFFLE_1 - LOAD_M(R30, R24, R14, R28, R26, R29, R27, R11) - ROUND_8 - SHUFFLE_2 - - VMOVQ (R4), V8 - VMOVQ 16(R4), V9 - VXORV V8, V0, V8 - VXORV V9, V1, V9 - VXORV V8, V2, V8 - VXORV V9, V3, V9 - VMOVQ V8, (R4) - VMOVQ V9, 16(R4) - - SUBV $64, R8 - ADDV $64, R7 - BNE R8, R0, loop - - MOVW R9, (R5) - MOVW R10, 4(R5) - - RET - |