diff options
Diffstat (limited to 'secure/lib/libcrypto/amd64/x86_64-mont.S')
| -rw-r--r-- | secure/lib/libcrypto/amd64/x86_64-mont.S | 358 |
1 files changed, 357 insertions, 1 deletions
diff --git a/secure/lib/libcrypto/amd64/x86_64-mont.S b/secure/lib/libcrypto/amd64/x86_64-mont.S index d3c1c71..9a83800 100644 --- a/secure/lib/libcrypto/amd64/x86_64-mont.S +++ b/secure/lib/libcrypto/amd64/x86_64-mont.S @@ -1,4 +1,5 @@ - # $FreeBSD$ +# $FreeBSD$ +# Do not modify. This file is auto-generated from x86_64-mont.pl. .text @@ -11,6 +12,7 @@ bn_mul_mont: jnz .Lmul_enter cmpl $8,%r9d jb .Lmul_enter + movl OPENSSL_ia32cap_P+8(%rip),%r11d cmpq %rsi,%rdx jne .Lmul4x_enter testl $7,%r9d @@ -230,6 +232,9 @@ bn_mul_mont: .align 16 bn_mul4x_mont: .Lmul4x_enter: + andl $0x80100,%r11d + cmpl $0x80100,%r11d + je .Lmulx4x_enter pushq %rbx pushq %rbp pushq %r12 @@ -634,6 +639,7 @@ bn_mul4x_mont: .size bn_mul4x_mont,.-bn_mul4x_mont + .type bn_sqr8x_mont,@function .align 32 bn_sqr8x_mont: @@ -696,6 +702,25 @@ bn_sqr8x_mont: pxor %xmm0,%xmm0 .byte 102,72,15,110,207 .byte 102,73,15,110,218 + movl OPENSSL_ia32cap_P+8(%rip),%eax + andl $0x80100,%eax + cmpl $0x80100,%eax + jne .Lsqr8x_nox + + call bn_sqrx8x_internal + + + + + leaq (%r8,%rcx,1),%rbx + movq %rcx,%r9 + movq %rcx,%rdx +.byte 102,72,15,126,207 + sarq $3+2,%rcx + jmp .Lsqr8x_sub + +.align 32 +.Lsqr8x_nox: call bn_sqr8x_internal @@ -774,5 +799,336 @@ bn_sqr8x_mont: .Lsqr8x_epilogue: .byte 0xf3,0xc3 .size bn_sqr8x_mont,.-bn_sqr8x_mont +.type bn_mulx4x_mont,@function +.align 32 +bn_mulx4x_mont: +.Lmulx4x_enter: + movq %rsp,%rax + pushq %rbx + pushq %rbp + pushq %r12 + pushq %r13 + pushq %r14 + pushq %r15 + + shll $3,%r9d +.byte 0x67 + xorq %r10,%r10 + subq %r9,%r10 + movq (%r8),%r8 + leaq -72(%rsp,%r10,1),%rsp + andq $-128,%rsp + movq %rax,%r11 + subq %rsp,%r11 + andq $-4096,%r11 +.Lmulx4x_page_walk: + movq (%rsp,%r11,1),%r10 + subq $4096,%r11 +.byte 0x66,0x2e + jnc .Lmulx4x_page_walk + + leaq (%rdx,%r9,1),%r10 + + + + + + + + + + + + + movq %r9,0(%rsp) + shrq $5,%r9 + movq %r10,16(%rsp) + subq $1,%r9 + movq %r8,24(%rsp) + movq %rdi,32(%rsp) + movq %rax,40(%rsp) + movq %r9,48(%rsp) + jmp .Lmulx4x_body + +.align 32 +.Lmulx4x_body: + leaq 8(%rdx),%rdi + movq (%rdx),%rdx + leaq 64+32(%rsp),%rbx + movq %rdx,%r9 + + mulxq 0(%rsi),%r8,%rax + mulxq 8(%rsi),%r11,%r14 + addq %rax,%r11 + movq %rdi,8(%rsp) + mulxq 16(%rsi),%r12,%r13 + adcq %r14,%r12 + adcq $0,%r13 + + movq %r8,%rdi + imulq 24(%rsp),%r8 + xorq %rbp,%rbp + + mulxq 24(%rsi),%rax,%r14 + movq %r8,%rdx + leaq 32(%rsi),%rsi + adcxq %rax,%r13 + adcxq %rbp,%r14 + + mulxq 0(%rcx),%rax,%r10 + adcxq %rax,%rdi + adoxq %r11,%r10 + mulxq 8(%rcx),%rax,%r11 + adcxq %rax,%r10 + adoxq %r12,%r11 +.byte 0xc4,0x62,0xfb,0xf6,0xa1,0x10,0x00,0x00,0x00 + movq 48(%rsp),%rdi + movq %r10,-32(%rbx) + adcxq %rax,%r11 + adoxq %r13,%r12 + mulxq 24(%rcx),%rax,%r15 + movq %r9,%rdx + movq %r11,-24(%rbx) + adcxq %rax,%r12 + adoxq %rbp,%r15 + leaq 32(%rcx),%rcx + movq %r12,-16(%rbx) + + jmp .Lmulx4x_1st + +.align 32 +.Lmulx4x_1st: + adcxq %rbp,%r15 + mulxq 0(%rsi),%r10,%rax + adcxq %r14,%r10 + mulxq 8(%rsi),%r11,%r14 + adcxq %rax,%r11 + mulxq 16(%rsi),%r12,%rax + adcxq %r14,%r12 + mulxq 24(%rsi),%r13,%r14 +.byte 0x67,0x67 + movq %r8,%rdx + adcxq %rax,%r13 + adcxq %rbp,%r14 + leaq 32(%rsi),%rsi + leaq 32(%rbx),%rbx + + adoxq %r15,%r10 + mulxq 0(%rcx),%rax,%r15 + adcxq %rax,%r10 + adoxq %r15,%r11 + mulxq 8(%rcx),%rax,%r15 + adcxq %rax,%r11 + adoxq %r15,%r12 + mulxq 16(%rcx),%rax,%r15 + movq %r10,-40(%rbx) + adcxq %rax,%r12 + movq %r11,-32(%rbx) + adoxq %r15,%r13 + mulxq 24(%rcx),%rax,%r15 + movq %r9,%rdx + movq %r12,-24(%rbx) + adcxq %rax,%r13 + adoxq %rbp,%r15 + leaq 32(%rcx),%rcx + movq %r13,-16(%rbx) + + decq %rdi + jnz .Lmulx4x_1st + + movq 0(%rsp),%rax + movq 8(%rsp),%rdi + adcq %rbp,%r15 + addq %r15,%r14 + sbbq %r15,%r15 + movq %r14,-8(%rbx) + jmp .Lmulx4x_outer + +.align 32 +.Lmulx4x_outer: + movq (%rdi),%rdx + leaq 8(%rdi),%rdi + subq %rax,%rsi + movq %r15,(%rbx) + leaq 64+32(%rsp),%rbx + subq %rax,%rcx + + mulxq 0(%rsi),%r8,%r11 + xorl %ebp,%ebp + movq %rdx,%r9 + mulxq 8(%rsi),%r14,%r12 + adoxq -32(%rbx),%r8 + adcxq %r14,%r11 + mulxq 16(%rsi),%r15,%r13 + adoxq -24(%rbx),%r11 + adcxq %r15,%r12 + adoxq %rbp,%r12 + adcxq %rbp,%r13 + + movq %rdi,8(%rsp) +.byte 0x67 + movq %r8,%r15 + imulq 24(%rsp),%r8 + xorl %ebp,%ebp + + mulxq 24(%rsi),%rax,%r14 + movq %r8,%rdx + adoxq -16(%rbx),%r12 + adcxq %rax,%r13 + adoxq -8(%rbx),%r13 + adcxq %rbp,%r14 + leaq 32(%rsi),%rsi + adoxq %rbp,%r14 + + mulxq 0(%rcx),%rax,%r10 + adcxq %rax,%r15 + adoxq %r11,%r10 + mulxq 8(%rcx),%rax,%r11 + adcxq %rax,%r10 + adoxq %r12,%r11 + mulxq 16(%rcx),%rax,%r12 + movq %r10,-32(%rbx) + adcxq %rax,%r11 + adoxq %r13,%r12 + mulxq 24(%rcx),%rax,%r15 + movq %r9,%rdx + movq %r11,-24(%rbx) + leaq 32(%rcx),%rcx + adcxq %rax,%r12 + adoxq %rbp,%r15 + movq 48(%rsp),%rdi + movq %r12,-16(%rbx) + + jmp .Lmulx4x_inner + +.align 32 +.Lmulx4x_inner: + mulxq 0(%rsi),%r10,%rax + adcxq %rbp,%r15 + adoxq %r14,%r10 + mulxq 8(%rsi),%r11,%r14 + adcxq 0(%rbx),%r10 + adoxq %rax,%r11 + mulxq 16(%rsi),%r12,%rax + adcxq 8(%rbx),%r11 + adoxq %r14,%r12 + mulxq 24(%rsi),%r13,%r14 + movq %r8,%rdx + adcxq 16(%rbx),%r12 + adoxq %rax,%r13 + adcxq 24(%rbx),%r13 + adoxq %rbp,%r14 + leaq 32(%rsi),%rsi + leaq 32(%rbx),%rbx + adcxq %rbp,%r14 + + adoxq %r15,%r10 + mulxq 0(%rcx),%rax,%r15 + adcxq %rax,%r10 + adoxq %r15,%r11 + mulxq 8(%rcx),%rax,%r15 + adcxq %rax,%r11 + adoxq %r15,%r12 + mulxq 16(%rcx),%rax,%r15 + movq %r10,-40(%rbx) + adcxq %rax,%r12 + adoxq %r15,%r13 + mulxq 24(%rcx),%rax,%r15 + movq %r9,%rdx + movq %r11,-32(%rbx) + movq %r12,-24(%rbx) + adcxq %rax,%r13 + adoxq %rbp,%r15 + leaq 32(%rcx),%rcx + movq %r13,-16(%rbx) + + decq %rdi + jnz .Lmulx4x_inner + + movq 0(%rsp),%rax + movq 8(%rsp),%rdi + adcq %rbp,%r15 + subq 0(%rbx),%rbp + adcq %r15,%r14 + sbbq %r15,%r15 + movq %r14,-8(%rbx) + + cmpq 16(%rsp),%rdi + jne .Lmulx4x_outer + + leaq 64(%rsp),%rbx + subq %rax,%rcx + negq %r15 + movq %rax,%rdx + shrq $3+2,%rax + movq 32(%rsp),%rdi + jmp .Lmulx4x_sub + +.align 32 +.Lmulx4x_sub: + movq 0(%rbx),%r11 + movq 8(%rbx),%r12 + movq 16(%rbx),%r13 + movq 24(%rbx),%r14 + leaq 32(%rbx),%rbx + sbbq 0(%rcx),%r11 + sbbq 8(%rcx),%r12 + sbbq 16(%rcx),%r13 + sbbq 24(%rcx),%r14 + leaq 32(%rcx),%rcx + movq %r11,0(%rdi) + movq %r12,8(%rdi) + movq %r13,16(%rdi) + movq %r14,24(%rdi) + leaq 32(%rdi),%rdi + decq %rax + jnz .Lmulx4x_sub + + sbbq $0,%r15 + leaq 64(%rsp),%rbx + subq %rdx,%rdi + +.byte 102,73,15,110,207 + pxor %xmm0,%xmm0 + pshufd $0,%xmm1,%xmm1 + movq 40(%rsp),%rsi + jmp .Lmulx4x_cond_copy + +.align 32 +.Lmulx4x_cond_copy: + movdqa 0(%rbx),%xmm2 + movdqa 16(%rbx),%xmm3 + leaq 32(%rbx),%rbx + movdqu 0(%rdi),%xmm4 + movdqu 16(%rdi),%xmm5 + leaq 32(%rdi),%rdi + movdqa %xmm0,-32(%rbx) + movdqa %xmm0,-16(%rbx) + pcmpeqd %xmm1,%xmm0 + pand %xmm1,%xmm2 + pand %xmm1,%xmm3 + pand %xmm0,%xmm4 + pand %xmm0,%xmm5 + pxor %xmm0,%xmm0 + por %xmm2,%xmm4 + por %xmm3,%xmm5 + movdqu %xmm4,-32(%rdi) + movdqu %xmm5,-16(%rdi) + subq $32,%rdx + jnz .Lmulx4x_cond_copy + + movq %rdx,(%rbx) + + movq $1,%rax + movq -48(%rsi),%r15 + movq -40(%rsi),%r14 + movq -32(%rsi),%r13 + movq -24(%rsi),%r12 + movq -16(%rsi),%rbp + movq -8(%rsi),%rbx + leaq (%rsi),%rsp +.Lmulx4x_epilogue: + .byte 0xf3,0xc3 +.size bn_mulx4x_mont,.-bn_mulx4x_mont .byte 77,111,110,116,103,111,109,101,114,121,32,77,117,108,116,105,112,108,105,99,97,116,105,111,110,32,102,111,114,32,120,56,54,95,54,52,44,32,67,82,89,80,84,79,71,65,77,83,32,98,121,32,60,97,112,112,114,111,64,111,112,101,110,115,115,108,46,111,114,103,62,0 .align 16 |
