diff options
Diffstat (limited to 'arch/x86/crypto/sha256-avx2-asm.S')
| -rw-r--r-- | arch/x86/crypto/sha256-avx2-asm.S | 54 |
1 files changed, 30 insertions, 24 deletions
diff --git a/arch/x86/crypto/sha256-avx2-asm.S b/arch/x86/crypto/sha256-avx2-asm.S index 3eada9416852..9918212faf91 100644 --- a/arch/x86/crypto/sha256-avx2-asm.S +++ b/arch/x86/crypto/sha256-avx2-asm.S @@ -538,12 +538,12 @@ SYM_TYPED_FUNC_START(sha256_transform_rorx) and $-32, %rsp # align rsp to 32 byte boundary shl $6, NUM_BLKS # convert to bytes - jz done_hash + jz .Ldone_hash lea -64(INP, NUM_BLKS), NUM_BLKS # pointer to last block mov NUM_BLKS, _INP_END(%rsp) cmp NUM_BLKS, INP - je only_one_block + je .Lonly_one_block ## load initial digest mov (CTX), a @@ -561,7 +561,7 @@ SYM_TYPED_FUNC_START(sha256_transform_rorx) mov CTX, _CTX(%rsp) -loop0: +.Lloop0: ## Load first 16 dwords from two blocks VMOVDQ 0*32(INP),XTMP0 VMOVDQ 1*32(INP),XTMP1 @@ -580,7 +580,7 @@ loop0: vperm2i128 $0x20, XTMP3, XTMP1, X2 vperm2i128 $0x31, XTMP3, XTMP1, X3 -last_block_enter: +.Llast_block_enter: add $64, INP mov INP, _INP(%rsp) @@ -588,34 +588,40 @@ last_block_enter: xor SRND, SRND .align 16 -loop1: - vpaddd K256+0*32(SRND), X0, XFER +.Lloop1: + leaq K256+0*32(%rip), INP ## reuse INP as scratch reg + vpaddd (INP, SRND), X0, XFER vmovdqa XFER, 0*32+_XFER(%rsp, SRND) FOUR_ROUNDS_AND_SCHED _XFER + 0*32 - vpaddd K256+1*32(SRND), X0, XFER + leaq K256+1*32(%rip), INP + vpaddd (INP, SRND), X0, XFER vmovdqa XFER, 1*32+_XFER(%rsp, SRND) FOUR_ROUNDS_AND_SCHED _XFER + 1*32 - vpaddd K256+2*32(SRND), X0, XFER + leaq K256+2*32(%rip), INP + vpaddd (INP, SRND), X0, XFER vmovdqa XFER, 2*32+_XFER(%rsp, SRND) FOUR_ROUNDS_AND_SCHED _XFER + 2*32 - vpaddd K256+3*32(SRND), X0, XFER + leaq K256+3*32(%rip), INP + vpaddd (INP, SRND), X0, XFER vmovdqa XFER, 3*32+_XFER(%rsp, SRND) FOUR_ROUNDS_AND_SCHED _XFER + 3*32 add $4*32, SRND cmp $3*4*32, SRND - jb loop1 + jb .Lloop1 -loop2: +.Lloop2: ## Do last 16 rounds with no scheduling - vpaddd K256+0*32(SRND), X0, XFER + leaq K256+0*32(%rip), INP + vpaddd (INP, SRND), X0, XFER vmovdqa XFER, 0*32+_XFER(%rsp, SRND) DO_4ROUNDS _XFER + 0*32 - vpaddd K256+1*32(SRND), X1, XFER + leaq K256+1*32(%rip), INP + vpaddd (INP, SRND), X1, XFER vmovdqa XFER, 1*32+_XFER(%rsp, SRND) DO_4ROUNDS _XFER + 1*32 add $2*32, SRND @@ -624,7 +630,7 @@ loop2: vmovdqa X3, X1 cmp $4*4*32, SRND - jb loop2 + jb .Lloop2 mov _CTX(%rsp), CTX mov _INP(%rsp), INP @@ -639,17 +645,17 @@ loop2: addm (4*7)(CTX),h cmp _INP_END(%rsp), INP - ja done_hash + ja .Ldone_hash #### Do second block using previously scheduled results xor SRND, SRND .align 16 -loop3: +.Lloop3: DO_4ROUNDS _XFER + 0*32 + 16 DO_4ROUNDS _XFER + 1*32 + 16 add $2*32, SRND cmp $4*4*32, SRND - jb loop3 + jb .Lloop3 mov _CTX(%rsp), CTX mov _INP(%rsp), INP @@ -665,10 +671,10 @@ loop3: addm (4*7)(CTX),h cmp _INP_END(%rsp), INP - jb loop0 - ja done_hash + jb .Lloop0 + ja .Ldone_hash -do_last_block: +.Ldo_last_block: VMOVDQ 0*16(INP),XWORD0 VMOVDQ 1*16(INP),XWORD1 VMOVDQ 2*16(INP),XWORD2 @@ -679,9 +685,9 @@ do_last_block: vpshufb X_BYTE_FLIP_MASK, XWORD2, XWORD2 vpshufb X_BYTE_FLIP_MASK, XWORD3, XWORD3 - jmp last_block_enter + jmp .Llast_block_enter -only_one_block: +.Lonly_one_block: ## load initial digest mov (4*0)(CTX),a @@ -698,9 +704,9 @@ only_one_block: vmovdqa _SHUF_DC00(%rip), SHUF_DC00 mov CTX, _CTX(%rsp) - jmp do_last_block + jmp .Ldo_last_block -done_hash: +.Ldone_hash: mov %rbp, %rsp pop %rbp |