764 lines
19 KiB
ArmAsm
764 lines
19 KiB
ArmAsm
|
|
# qhasm: int64 rp
|
|
|
|
# qhasm: int64 n
|
|
|
|
# qhasm: input rp
|
|
|
|
# qhasm: input n
|
|
|
|
# qhasm: int64 r0
|
|
|
|
# qhasm: int64 r1
|
|
|
|
# qhasm: int64 r2
|
|
|
|
# qhasm: int64 r3
|
|
|
|
# qhasm: int64 r4
|
|
|
|
# qhasm: int64 c1
|
|
|
|
# qhasm: int64 c2
|
|
|
|
# qhasm: int64 c3
|
|
|
|
# qhasm: int64 c4
|
|
|
|
# qhasm: int64 c5
|
|
|
|
# qhasm: int64 c6
|
|
|
|
# qhasm: int64 c7
|
|
|
|
# qhasm: caller c1
|
|
|
|
# qhasm: caller c2
|
|
|
|
# qhasm: caller c3
|
|
|
|
# qhasm: caller c4
|
|
|
|
# qhasm: caller c5
|
|
|
|
# qhasm: caller c6
|
|
|
|
# qhasm: caller c7
|
|
|
|
# qhasm: stack64 c1_stack
|
|
|
|
# qhasm: stack64 c2_stack
|
|
|
|
# qhasm: stack64 c3_stack
|
|
|
|
# qhasm: stack64 c4_stack
|
|
|
|
# qhasm: stack64 c5_stack
|
|
|
|
# qhasm: stack64 c6_stack
|
|
|
|
# qhasm: stack64 c7_stack
|
|
|
|
# qhasm: stack64 x119_stack
|
|
|
|
# qhasm: stack64 x219_stack
|
|
|
|
# qhasm: stack64 x319_stack
|
|
|
|
# qhasm: stack64 x419_stack
|
|
|
|
# qhasm: int64 squarer01
|
|
|
|
# qhasm: int64 squarer11
|
|
|
|
# qhasm: int64 squarer21
|
|
|
|
# qhasm: int64 squarer31
|
|
|
|
# qhasm: int64 squarer41
|
|
|
|
# qhasm: int64 squarerax
|
|
|
|
# qhasm: int64 squarerdx
|
|
|
|
# qhasm: int64 squaret
|
|
|
|
# qhasm: int64 squareredmask
|
|
|
|
# qhasm: stack64 n_stack
|
|
|
|
# qhasm: enter crypto_sign_ed25519_amd64_51_30k_batch_fe25519_nsquare
|
|
.text
|
|
.p2align 5
|
|
.globl _crypto_sign_ed25519_amd64_51_30k_batch_fe25519_nsquare
|
|
.globl crypto_sign_ed25519_amd64_51_30k_batch_fe25519_nsquare
|
|
_crypto_sign_ed25519_amd64_51_30k_batch_fe25519_nsquare:
|
|
crypto_sign_ed25519_amd64_51_30k_batch_fe25519_nsquare:
|
|
mov %rsp,%r11
|
|
and $31,%r11
|
|
add $64,%r11
|
|
sub %r11,%rsp
|
|
|
|
# qhasm: c1_stack = c1
|
|
# asm 1: movq <c1=int64#9,>c1_stack=stack64#1
|
|
# asm 2: movq <c1=%r11,>c1_stack=0(%rsp)
|
|
movq %r11,0(%rsp)
|
|
|
|
# qhasm: c2_stack = c2
|
|
# asm 1: movq <c2=int64#10,>c2_stack=stack64#2
|
|
# asm 2: movq <c2=%r12,>c2_stack=8(%rsp)
|
|
movq %r12,8(%rsp)
|
|
|
|
# qhasm: c3_stack = c3
|
|
# asm 1: movq <c3=int64#11,>c3_stack=stack64#3
|
|
# asm 2: movq <c3=%r13,>c3_stack=16(%rsp)
|
|
movq %r13,16(%rsp)
|
|
|
|
# qhasm: c4_stack = c4
|
|
# asm 1: movq <c4=int64#12,>c4_stack=stack64#4
|
|
# asm 2: movq <c4=%r14,>c4_stack=24(%rsp)
|
|
movq %r14,24(%rsp)
|
|
|
|
# qhasm: c5_stack = c5
|
|
# asm 1: movq <c5=int64#13,>c5_stack=stack64#5
|
|
# asm 2: movq <c5=%r15,>c5_stack=32(%rsp)
|
|
movq %r15,32(%rsp)
|
|
|
|
# qhasm: c6_stack = c6
|
|
# asm 1: movq <c6=int64#14,>c6_stack=stack64#6
|
|
# asm 2: movq <c6=%rbx,>c6_stack=40(%rsp)
|
|
movq %rbx,40(%rsp)
|
|
|
|
# qhasm: c7_stack = c7
|
|
# asm 1: movq <c7=int64#15,>c7_stack=stack64#7
|
|
# asm 2: movq <c7=%rbp,>c7_stack=48(%rsp)
|
|
movq %rbp,48(%rsp)
|
|
|
|
# qhasm: loop:
|
|
._loop:
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 0)
|
|
# asm 1: movq 0(<rp=int64#1),>squarerax=int64#7
|
|
# asm 2: movq 0(<rp=%rdi),>squarerax=%rax
|
|
movq 0(%rdi),%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 0)
|
|
# asm 1: mulq 0(<rp=int64#1)
|
|
# asm 2: mulq 0(<rp=%rdi)
|
|
mulq 0(%rdi)
|
|
|
|
# qhasm: r0 = squarerax
|
|
# asm 1: mov <squarerax=int64#7,>r0=int64#4
|
|
# asm 2: mov <squarerax=%rax,>r0=%rcx
|
|
mov %rax,%rcx
|
|
|
|
# qhasm: squarer01 = squarerdx
|
|
# asm 1: mov <squarerdx=int64#3,>squarer01=int64#5
|
|
# asm 2: mov <squarerdx=%rdx,>squarer01=%r8
|
|
mov %rdx,%r8
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 0)
|
|
# asm 1: movq 0(<rp=int64#1),>squarerax=int64#7
|
|
# asm 2: movq 0(<rp=%rdi),>squarerax=%rax
|
|
movq 0(%rdi),%rax
|
|
|
|
# qhasm: squarerax <<= 1
|
|
# asm 1: shl $1,<squarerax=int64#7
|
|
# asm 2: shl $1,<squarerax=%rax
|
|
shl $1,%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 8)
|
|
# asm 1: mulq 8(<rp=int64#1)
|
|
# asm 2: mulq 8(<rp=%rdi)
|
|
mulq 8(%rdi)
|
|
|
|
# qhasm: r1 = squarerax
|
|
# asm 1: mov <squarerax=int64#7,>r1=int64#6
|
|
# asm 2: mov <squarerax=%rax,>r1=%r9
|
|
mov %rax,%r9
|
|
|
|
# qhasm: squarer11 = squarerdx
|
|
# asm 1: mov <squarerdx=int64#3,>squarer11=int64#8
|
|
# asm 2: mov <squarerdx=%rdx,>squarer11=%r10
|
|
mov %rdx,%r10
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 0)
|
|
# asm 1: movq 0(<rp=int64#1),>squarerax=int64#7
|
|
# asm 2: movq 0(<rp=%rdi),>squarerax=%rax
|
|
movq 0(%rdi),%rax
|
|
|
|
# qhasm: squarerax <<= 1
|
|
# asm 1: shl $1,<squarerax=int64#7
|
|
# asm 2: shl $1,<squarerax=%rax
|
|
shl $1,%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 16)
|
|
# asm 1: mulq 16(<rp=int64#1)
|
|
# asm 2: mulq 16(<rp=%rdi)
|
|
mulq 16(%rdi)
|
|
|
|
# qhasm: r2 = squarerax
|
|
# asm 1: mov <squarerax=int64#7,>r2=int64#9
|
|
# asm 2: mov <squarerax=%rax,>r2=%r11
|
|
mov %rax,%r11
|
|
|
|
# qhasm: squarer21 = squarerdx
|
|
# asm 1: mov <squarerdx=int64#3,>squarer21=int64#10
|
|
# asm 2: mov <squarerdx=%rdx,>squarer21=%r12
|
|
mov %rdx,%r12
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 0)
|
|
# asm 1: movq 0(<rp=int64#1),>squarerax=int64#7
|
|
# asm 2: movq 0(<rp=%rdi),>squarerax=%rax
|
|
movq 0(%rdi),%rax
|
|
|
|
# qhasm: squarerax <<= 1
|
|
# asm 1: shl $1,<squarerax=int64#7
|
|
# asm 2: shl $1,<squarerax=%rax
|
|
shl $1,%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 24)
|
|
# asm 1: mulq 24(<rp=int64#1)
|
|
# asm 2: mulq 24(<rp=%rdi)
|
|
mulq 24(%rdi)
|
|
|
|
# qhasm: r3 = squarerax
|
|
# asm 1: mov <squarerax=int64#7,>r3=int64#11
|
|
# asm 2: mov <squarerax=%rax,>r3=%r13
|
|
mov %rax,%r13
|
|
|
|
# qhasm: squarer31 = squarerdx
|
|
# asm 1: mov <squarerdx=int64#3,>squarer31=int64#12
|
|
# asm 2: mov <squarerdx=%rdx,>squarer31=%r14
|
|
mov %rdx,%r14
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 0)
|
|
# asm 1: movq 0(<rp=int64#1),>squarerax=int64#7
|
|
# asm 2: movq 0(<rp=%rdi),>squarerax=%rax
|
|
movq 0(%rdi),%rax
|
|
|
|
# qhasm: squarerax <<= 1
|
|
# asm 1: shl $1,<squarerax=int64#7
|
|
# asm 2: shl $1,<squarerax=%rax
|
|
shl $1,%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 32)
|
|
# asm 1: mulq 32(<rp=int64#1)
|
|
# asm 2: mulq 32(<rp=%rdi)
|
|
mulq 32(%rdi)
|
|
|
|
# qhasm: r4 = squarerax
|
|
# asm 1: mov <squarerax=int64#7,>r4=int64#13
|
|
# asm 2: mov <squarerax=%rax,>r4=%r15
|
|
mov %rax,%r15
|
|
|
|
# qhasm: squarer41 = squarerdx
|
|
# asm 1: mov <squarerdx=int64#3,>squarer41=int64#14
|
|
# asm 2: mov <squarerdx=%rdx,>squarer41=%rbx
|
|
mov %rdx,%rbx
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 8)
|
|
# asm 1: movq 8(<rp=int64#1),>squarerax=int64#7
|
|
# asm 2: movq 8(<rp=%rdi),>squarerax=%rax
|
|
movq 8(%rdi),%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 8)
|
|
# asm 1: mulq 8(<rp=int64#1)
|
|
# asm 2: mulq 8(<rp=%rdi)
|
|
mulq 8(%rdi)
|
|
|
|
# qhasm: carry? r2 += squarerax
|
|
# asm 1: add <squarerax=int64#7,<r2=int64#9
|
|
# asm 2: add <squarerax=%rax,<r2=%r11
|
|
add %rax,%r11
|
|
|
|
# qhasm: squarer21 += squarerdx + carry
|
|
# asm 1: adc <squarerdx=int64#3,<squarer21=int64#10
|
|
# asm 2: adc <squarerdx=%rdx,<squarer21=%r12
|
|
adc %rdx,%r12
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 8)
|
|
# asm 1: movq 8(<rp=int64#1),>squarerax=int64#7
|
|
# asm 2: movq 8(<rp=%rdi),>squarerax=%rax
|
|
movq 8(%rdi),%rax
|
|
|
|
# qhasm: squarerax <<= 1
|
|
# asm 1: shl $1,<squarerax=int64#7
|
|
# asm 2: shl $1,<squarerax=%rax
|
|
shl $1,%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 16)
|
|
# asm 1: mulq 16(<rp=int64#1)
|
|
# asm 2: mulq 16(<rp=%rdi)
|
|
mulq 16(%rdi)
|
|
|
|
# qhasm: carry? r3 += squarerax
|
|
# asm 1: add <squarerax=int64#7,<r3=int64#11
|
|
# asm 2: add <squarerax=%rax,<r3=%r13
|
|
add %rax,%r13
|
|
|
|
# qhasm: squarer31 += squarerdx + carry
|
|
# asm 1: adc <squarerdx=int64#3,<squarer31=int64#12
|
|
# asm 2: adc <squarerdx=%rdx,<squarer31=%r14
|
|
adc %rdx,%r14
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 8)
|
|
# asm 1: movq 8(<rp=int64#1),>squarerax=int64#7
|
|
# asm 2: movq 8(<rp=%rdi),>squarerax=%rax
|
|
movq 8(%rdi),%rax
|
|
|
|
# qhasm: squarerax <<= 1
|
|
# asm 1: shl $1,<squarerax=int64#7
|
|
# asm 2: shl $1,<squarerax=%rax
|
|
shl $1,%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 24)
|
|
# asm 1: mulq 24(<rp=int64#1)
|
|
# asm 2: mulq 24(<rp=%rdi)
|
|
mulq 24(%rdi)
|
|
|
|
# qhasm: carry? r4 += squarerax
|
|
# asm 1: add <squarerax=int64#7,<r4=int64#13
|
|
# asm 2: add <squarerax=%rax,<r4=%r15
|
|
add %rax,%r15
|
|
|
|
# qhasm: squarer41 += squarerdx + carry
|
|
# asm 1: adc <squarerdx=int64#3,<squarer41=int64#14
|
|
# asm 2: adc <squarerdx=%rdx,<squarer41=%rbx
|
|
adc %rdx,%rbx
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 8)
|
|
# asm 1: movq 8(<rp=int64#1),>squarerax=int64#3
|
|
# asm 2: movq 8(<rp=%rdi),>squarerax=%rdx
|
|
movq 8(%rdi),%rdx
|
|
|
|
# qhasm: squarerax *= 38
|
|
# asm 1: imulq $38,<squarerax=int64#3,>squarerax=int64#7
|
|
# asm 2: imulq $38,<squarerax=%rdx,>squarerax=%rax
|
|
imulq $38,%rdx,%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 32)
|
|
# asm 1: mulq 32(<rp=int64#1)
|
|
# asm 2: mulq 32(<rp=%rdi)
|
|
mulq 32(%rdi)
|
|
|
|
# qhasm: carry? r0 += squarerax
|
|
# asm 1: add <squarerax=int64#7,<r0=int64#4
|
|
# asm 2: add <squarerax=%rax,<r0=%rcx
|
|
add %rax,%rcx
|
|
|
|
# qhasm: squarer01 += squarerdx + carry
|
|
# asm 1: adc <squarerdx=int64#3,<squarer01=int64#5
|
|
# asm 2: adc <squarerdx=%rdx,<squarer01=%r8
|
|
adc %rdx,%r8
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 16)
|
|
# asm 1: movq 16(<rp=int64#1),>squarerax=int64#7
|
|
# asm 2: movq 16(<rp=%rdi),>squarerax=%rax
|
|
movq 16(%rdi),%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 16)
|
|
# asm 1: mulq 16(<rp=int64#1)
|
|
# asm 2: mulq 16(<rp=%rdi)
|
|
mulq 16(%rdi)
|
|
|
|
# qhasm: carry? r4 += squarerax
|
|
# asm 1: add <squarerax=int64#7,<r4=int64#13
|
|
# asm 2: add <squarerax=%rax,<r4=%r15
|
|
add %rax,%r15
|
|
|
|
# qhasm: squarer41 += squarerdx + carry
|
|
# asm 1: adc <squarerdx=int64#3,<squarer41=int64#14
|
|
# asm 2: adc <squarerdx=%rdx,<squarer41=%rbx
|
|
adc %rdx,%rbx
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 16)
|
|
# asm 1: movq 16(<rp=int64#1),>squarerax=int64#3
|
|
# asm 2: movq 16(<rp=%rdi),>squarerax=%rdx
|
|
movq 16(%rdi),%rdx
|
|
|
|
# qhasm: squarerax *= 38
|
|
# asm 1: imulq $38,<squarerax=int64#3,>squarerax=int64#7
|
|
# asm 2: imulq $38,<squarerax=%rdx,>squarerax=%rax
|
|
imulq $38,%rdx,%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 24)
|
|
# asm 1: mulq 24(<rp=int64#1)
|
|
# asm 2: mulq 24(<rp=%rdi)
|
|
mulq 24(%rdi)
|
|
|
|
# qhasm: carry? r0 += squarerax
|
|
# asm 1: add <squarerax=int64#7,<r0=int64#4
|
|
# asm 2: add <squarerax=%rax,<r0=%rcx
|
|
add %rax,%rcx
|
|
|
|
# qhasm: squarer01 += squarerdx + carry
|
|
# asm 1: adc <squarerdx=int64#3,<squarer01=int64#5
|
|
# asm 2: adc <squarerdx=%rdx,<squarer01=%r8
|
|
adc %rdx,%r8
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 16)
|
|
# asm 1: movq 16(<rp=int64#1),>squarerax=int64#3
|
|
# asm 2: movq 16(<rp=%rdi),>squarerax=%rdx
|
|
movq 16(%rdi),%rdx
|
|
|
|
# qhasm: squarerax *= 38
|
|
# asm 1: imulq $38,<squarerax=int64#3,>squarerax=int64#7
|
|
# asm 2: imulq $38,<squarerax=%rdx,>squarerax=%rax
|
|
imulq $38,%rdx,%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 32)
|
|
# asm 1: mulq 32(<rp=int64#1)
|
|
# asm 2: mulq 32(<rp=%rdi)
|
|
mulq 32(%rdi)
|
|
|
|
# qhasm: carry? r1 += squarerax
|
|
# asm 1: add <squarerax=int64#7,<r1=int64#6
|
|
# asm 2: add <squarerax=%rax,<r1=%r9
|
|
add %rax,%r9
|
|
|
|
# qhasm: squarer11 += squarerdx + carry
|
|
# asm 1: adc <squarerdx=int64#3,<squarer11=int64#8
|
|
# asm 2: adc <squarerdx=%rdx,<squarer11=%r10
|
|
adc %rdx,%r10
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 24)
|
|
# asm 1: movq 24(<rp=int64#1),>squarerax=int64#3
|
|
# asm 2: movq 24(<rp=%rdi),>squarerax=%rdx
|
|
movq 24(%rdi),%rdx
|
|
|
|
# qhasm: squarerax *= 19
|
|
# asm 1: imulq $19,<squarerax=int64#3,>squarerax=int64#7
|
|
# asm 2: imulq $19,<squarerax=%rdx,>squarerax=%rax
|
|
imulq $19,%rdx,%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 24)
|
|
# asm 1: mulq 24(<rp=int64#1)
|
|
# asm 2: mulq 24(<rp=%rdi)
|
|
mulq 24(%rdi)
|
|
|
|
# qhasm: carry? r1 += squarerax
|
|
# asm 1: add <squarerax=int64#7,<r1=int64#6
|
|
# asm 2: add <squarerax=%rax,<r1=%r9
|
|
add %rax,%r9
|
|
|
|
# qhasm: squarer11 += squarerdx + carry
|
|
# asm 1: adc <squarerdx=int64#3,<squarer11=int64#8
|
|
# asm 2: adc <squarerdx=%rdx,<squarer11=%r10
|
|
adc %rdx,%r10
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 24)
|
|
# asm 1: movq 24(<rp=int64#1),>squarerax=int64#3
|
|
# asm 2: movq 24(<rp=%rdi),>squarerax=%rdx
|
|
movq 24(%rdi),%rdx
|
|
|
|
# qhasm: squarerax *= 38
|
|
# asm 1: imulq $38,<squarerax=int64#3,>squarerax=int64#7
|
|
# asm 2: imulq $38,<squarerax=%rdx,>squarerax=%rax
|
|
imulq $38,%rdx,%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 32)
|
|
# asm 1: mulq 32(<rp=int64#1)
|
|
# asm 2: mulq 32(<rp=%rdi)
|
|
mulq 32(%rdi)
|
|
|
|
# qhasm: carry? r2 += squarerax
|
|
# asm 1: add <squarerax=int64#7,<r2=int64#9
|
|
# asm 2: add <squarerax=%rax,<r2=%r11
|
|
add %rax,%r11
|
|
|
|
# qhasm: squarer21 += squarerdx + carry
|
|
# asm 1: adc <squarerdx=int64#3,<squarer21=int64#10
|
|
# asm 2: adc <squarerdx=%rdx,<squarer21=%r12
|
|
adc %rdx,%r12
|
|
|
|
# qhasm: squarerax = *(uint64 *)(rp + 32)
|
|
# asm 1: movq 32(<rp=int64#1),>squarerax=int64#3
|
|
# asm 2: movq 32(<rp=%rdi),>squarerax=%rdx
|
|
movq 32(%rdi),%rdx
|
|
|
|
# qhasm: squarerax *= 19
|
|
# asm 1: imulq $19,<squarerax=int64#3,>squarerax=int64#7
|
|
# asm 2: imulq $19,<squarerax=%rdx,>squarerax=%rax
|
|
imulq $19,%rdx,%rax
|
|
|
|
# qhasm: (uint128) squarerdx squarerax = squarerax * *(uint64 *)(rp + 32)
|
|
# asm 1: mulq 32(<rp=int64#1)
|
|
# asm 2: mulq 32(<rp=%rdi)
|
|
mulq 32(%rdi)
|
|
|
|
# qhasm: carry? r3 += squarerax
|
|
# asm 1: add <squarerax=int64#7,<r3=int64#11
|
|
# asm 2: add <squarerax=%rax,<r3=%r13
|
|
add %rax,%r13
|
|
|
|
# qhasm: squarer31 += squarerdx + carry
|
|
# asm 1: adc <squarerdx=int64#3,<squarer31=int64#12
|
|
# asm 2: adc <squarerdx=%rdx,<squarer31=%r14
|
|
adc %rdx,%r14
|
|
|
|
# qhasm: squareredmask = *(uint64 *) &crypto_sign_ed25519_amd64_51_30k_batch_REDMASK51
|
|
# asm 1: movq crypto_sign_ed25519_amd64_51_30k_batch_REDMASK51,>squareredmask=int64#3
|
|
# asm 2: movq crypto_sign_ed25519_amd64_51_30k_batch_REDMASK51,>squareredmask=%rdx
|
|
movq crypto_sign_ed25519_amd64_51_30k_batch_REDMASK51,%rdx
|
|
|
|
# qhasm: squarer01 = (squarer01.r0) << 13
|
|
# asm 1: shld $13,<r0=int64#4,<squarer01=int64#5
|
|
# asm 2: shld $13,<r0=%rcx,<squarer01=%r8
|
|
shld $13,%rcx,%r8
|
|
|
|
# qhasm: r0 &= squareredmask
|
|
# asm 1: and <squareredmask=int64#3,<r0=int64#4
|
|
# asm 2: and <squareredmask=%rdx,<r0=%rcx
|
|
and %rdx,%rcx
|
|
|
|
# qhasm: squarer11 = (squarer11.r1) << 13
|
|
# asm 1: shld $13,<r1=int64#6,<squarer11=int64#8
|
|
# asm 2: shld $13,<r1=%r9,<squarer11=%r10
|
|
shld $13,%r9,%r10
|
|
|
|
# qhasm: r1 &= squareredmask
|
|
# asm 1: and <squareredmask=int64#3,<r1=int64#6
|
|
# asm 2: and <squareredmask=%rdx,<r1=%r9
|
|
and %rdx,%r9
|
|
|
|
# qhasm: r1 += squarer01
|
|
# asm 1: add <squarer01=int64#5,<r1=int64#6
|
|
# asm 2: add <squarer01=%r8,<r1=%r9
|
|
add %r8,%r9
|
|
|
|
# qhasm: squarer21 = (squarer21.r2) << 13
|
|
# asm 1: shld $13,<r2=int64#9,<squarer21=int64#10
|
|
# asm 2: shld $13,<r2=%r11,<squarer21=%r12
|
|
shld $13,%r11,%r12
|
|
|
|
# qhasm: r2 &= squareredmask
|
|
# asm 1: and <squareredmask=int64#3,<r2=int64#9
|
|
# asm 2: and <squareredmask=%rdx,<r2=%r11
|
|
and %rdx,%r11
|
|
|
|
# qhasm: r2 += squarer11
|
|
# asm 1: add <squarer11=int64#8,<r2=int64#9
|
|
# asm 2: add <squarer11=%r10,<r2=%r11
|
|
add %r10,%r11
|
|
|
|
# qhasm: squarer31 = (squarer31.r3) << 13
|
|
# asm 1: shld $13,<r3=int64#11,<squarer31=int64#12
|
|
# asm 2: shld $13,<r3=%r13,<squarer31=%r14
|
|
shld $13,%r13,%r14
|
|
|
|
# qhasm: r3 &= squareredmask
|
|
# asm 1: and <squareredmask=int64#3,<r3=int64#11
|
|
# asm 2: and <squareredmask=%rdx,<r3=%r13
|
|
and %rdx,%r13
|
|
|
|
# qhasm: r3 += squarer21
|
|
# asm 1: add <squarer21=int64#10,<r3=int64#11
|
|
# asm 2: add <squarer21=%r12,<r3=%r13
|
|
add %r12,%r13
|
|
|
|
# qhasm: squarer41 = (squarer41.r4) << 13
|
|
# asm 1: shld $13,<r4=int64#13,<squarer41=int64#14
|
|
# asm 2: shld $13,<r4=%r15,<squarer41=%rbx
|
|
shld $13,%r15,%rbx
|
|
|
|
# qhasm: r4 &= squareredmask
|
|
# asm 1: and <squareredmask=int64#3,<r4=int64#13
|
|
# asm 2: and <squareredmask=%rdx,<r4=%r15
|
|
and %rdx,%r15
|
|
|
|
# qhasm: r4 += squarer31
|
|
# asm 1: add <squarer31=int64#12,<r4=int64#13
|
|
# asm 2: add <squarer31=%r14,<r4=%r15
|
|
add %r14,%r15
|
|
|
|
# qhasm: squarer41 = squarer41 * 19
|
|
# asm 1: imulq $19,<squarer41=int64#14,>squarer41=int64#5
|
|
# asm 2: imulq $19,<squarer41=%rbx,>squarer41=%r8
|
|
imulq $19,%rbx,%r8
|
|
|
|
# qhasm: r0 += squarer41
|
|
# asm 1: add <squarer41=int64#5,<r0=int64#4
|
|
# asm 2: add <squarer41=%r8,<r0=%rcx
|
|
add %r8,%rcx
|
|
|
|
# qhasm: squaret = r0
|
|
# asm 1: mov <r0=int64#4,>squaret=int64#5
|
|
# asm 2: mov <r0=%rcx,>squaret=%r8
|
|
mov %rcx,%r8
|
|
|
|
# qhasm: (uint64) squaret >>= 51
|
|
# asm 1: shr $51,<squaret=int64#5
|
|
# asm 2: shr $51,<squaret=%r8
|
|
shr $51,%r8
|
|
|
|
# qhasm: squaret += r1
|
|
# asm 1: add <r1=int64#6,<squaret=int64#5
|
|
# asm 2: add <r1=%r9,<squaret=%r8
|
|
add %r9,%r8
|
|
|
|
# qhasm: r0 &= squareredmask
|
|
# asm 1: and <squareredmask=int64#3,<r0=int64#4
|
|
# asm 2: and <squareredmask=%rdx,<r0=%rcx
|
|
and %rdx,%rcx
|
|
|
|
# qhasm: r1 = squaret
|
|
# asm 1: mov <squaret=int64#5,>r1=int64#6
|
|
# asm 2: mov <squaret=%r8,>r1=%r9
|
|
mov %r8,%r9
|
|
|
|
# qhasm: (uint64) squaret >>= 51
|
|
# asm 1: shr $51,<squaret=int64#5
|
|
# asm 2: shr $51,<squaret=%r8
|
|
shr $51,%r8
|
|
|
|
# qhasm: squaret += r2
|
|
# asm 1: add <r2=int64#9,<squaret=int64#5
|
|
# asm 2: add <r2=%r11,<squaret=%r8
|
|
add %r11,%r8
|
|
|
|
# qhasm: r1 &= squareredmask
|
|
# asm 1: and <squareredmask=int64#3,<r1=int64#6
|
|
# asm 2: and <squareredmask=%rdx,<r1=%r9
|
|
and %rdx,%r9
|
|
|
|
# qhasm: r2 = squaret
|
|
# asm 1: mov <squaret=int64#5,>r2=int64#7
|
|
# asm 2: mov <squaret=%r8,>r2=%rax
|
|
mov %r8,%rax
|
|
|
|
# qhasm: (uint64) squaret >>= 51
|
|
# asm 1: shr $51,<squaret=int64#5
|
|
# asm 2: shr $51,<squaret=%r8
|
|
shr $51,%r8
|
|
|
|
# qhasm: squaret += r3
|
|
# asm 1: add <r3=int64#11,<squaret=int64#5
|
|
# asm 2: add <r3=%r13,<squaret=%r8
|
|
add %r13,%r8
|
|
|
|
# qhasm: r2 &= squareredmask
|
|
# asm 1: and <squareredmask=int64#3,<r2=int64#7
|
|
# asm 2: and <squareredmask=%rdx,<r2=%rax
|
|
and %rdx,%rax
|
|
|
|
# qhasm: r3 = squaret
|
|
# asm 1: mov <squaret=int64#5,>r3=int64#8
|
|
# asm 2: mov <squaret=%r8,>r3=%r10
|
|
mov %r8,%r10
|
|
|
|
# qhasm: (uint64) squaret >>= 51
|
|
# asm 1: shr $51,<squaret=int64#5
|
|
# asm 2: shr $51,<squaret=%r8
|
|
shr $51,%r8
|
|
|
|
# qhasm: squaret += r4
|
|
# asm 1: add <r4=int64#13,<squaret=int64#5
|
|
# asm 2: add <r4=%r15,<squaret=%r8
|
|
add %r15,%r8
|
|
|
|
# qhasm: r3 &= squareredmask
|
|
# asm 1: and <squareredmask=int64#3,<r3=int64#8
|
|
# asm 2: and <squareredmask=%rdx,<r3=%r10
|
|
and %rdx,%r10
|
|
|
|
# qhasm: r4 = squaret
|
|
# asm 1: mov <squaret=int64#5,>r4=int64#9
|
|
# asm 2: mov <squaret=%r8,>r4=%r11
|
|
mov %r8,%r11
|
|
|
|
# qhasm: (uint64) squaret >>= 51
|
|
# asm 1: shr $51,<squaret=int64#5
|
|
# asm 2: shr $51,<squaret=%r8
|
|
shr $51,%r8
|
|
|
|
# qhasm: squaret *= 19
|
|
# asm 1: imulq $19,<squaret=int64#5,>squaret=int64#5
|
|
# asm 2: imulq $19,<squaret=%r8,>squaret=%r8
|
|
imulq $19,%r8,%r8
|
|
|
|
# qhasm: r0 += squaret
|
|
# asm 1: add <squaret=int64#5,<r0=int64#4
|
|
# asm 2: add <squaret=%r8,<r0=%rcx
|
|
add %r8,%rcx
|
|
|
|
# qhasm: r4 &= squareredmask
|
|
# asm 1: and <squareredmask=int64#3,<r4=int64#9
|
|
# asm 2: and <squareredmask=%rdx,<r4=%r11
|
|
and %rdx,%r11
|
|
|
|
# qhasm: *(uint64 *)(rp + 0) = r0
|
|
# asm 1: movq <r0=int64#4,0(<rp=int64#1)
|
|
# asm 2: movq <r0=%rcx,0(<rp=%rdi)
|
|
movq %rcx,0(%rdi)
|
|
|
|
# qhasm: *(uint64 *)(rp + 8) = r1
|
|
# asm 1: movq <r1=int64#6,8(<rp=int64#1)
|
|
# asm 2: movq <r1=%r9,8(<rp=%rdi)
|
|
movq %r9,8(%rdi)
|
|
|
|
# qhasm: *(uint64 *)(rp + 16) = r2
|
|
# asm 1: movq <r2=int64#7,16(<rp=int64#1)
|
|
# asm 2: movq <r2=%rax,16(<rp=%rdi)
|
|
movq %rax,16(%rdi)
|
|
|
|
# qhasm: *(uint64 *)(rp + 24) = r3
|
|
# asm 1: movq <r3=int64#8,24(<rp=int64#1)
|
|
# asm 2: movq <r3=%r10,24(<rp=%rdi)
|
|
movq %r10,24(%rdi)
|
|
|
|
# qhasm: *(uint64 *)(rp + 32) = r4
|
|
# asm 1: movq <r4=int64#9,32(<rp=int64#1)
|
|
# asm 2: movq <r4=%r11,32(<rp=%rdi)
|
|
movq %r11,32(%rdi)
|
|
|
|
# qhasm: signed>? n -= 1
|
|
# asm 1: sub $1,<n=int64#2
|
|
# asm 2: sub $1,<n=%rsi
|
|
sub $1,%rsi
|
|
# comment:fp stack unchanged by jump
|
|
|
|
# qhasm: goto loop if signed>
|
|
jg ._loop
|
|
|
|
# qhasm: c1 =c1_stack
|
|
# asm 1: movq <c1_stack=stack64#1,>c1=int64#9
|
|
# asm 2: movq <c1_stack=0(%rsp),>c1=%r11
|
|
movq 0(%rsp),%r11
|
|
|
|
# qhasm: c2 =c2_stack
|
|
# asm 1: movq <c2_stack=stack64#2,>c2=int64#10
|
|
# asm 2: movq <c2_stack=8(%rsp),>c2=%r12
|
|
movq 8(%rsp),%r12
|
|
|
|
# qhasm: c3 =c3_stack
|
|
# asm 1: movq <c3_stack=stack64#3,>c3=int64#11
|
|
# asm 2: movq <c3_stack=16(%rsp),>c3=%r13
|
|
movq 16(%rsp),%r13
|
|
|
|
# qhasm: c4 =c4_stack
|
|
# asm 1: movq <c4_stack=stack64#4,>c4=int64#12
|
|
# asm 2: movq <c4_stack=24(%rsp),>c4=%r14
|
|
movq 24(%rsp),%r14
|
|
|
|
# qhasm: c5 =c5_stack
|
|
# asm 1: movq <c5_stack=stack64#5,>c5=int64#13
|
|
# asm 2: movq <c5_stack=32(%rsp),>c5=%r15
|
|
movq 32(%rsp),%r15
|
|
|
|
# qhasm: c6 =c6_stack
|
|
# asm 1: movq <c6_stack=stack64#6,>c6=int64#14
|
|
# asm 2: movq <c6_stack=40(%rsp),>c6=%rbx
|
|
movq 40(%rsp),%rbx
|
|
|
|
# qhasm: c7 =c7_stack
|
|
# asm 1: movq <c7_stack=stack64#7,>c7=int64#15
|
|
# asm 2: movq <c7_stack=48(%rsp),>c7=%rbp
|
|
movq 48(%rsp),%rbp
|
|
|
|
# qhasm: leave
|
|
add %r11,%rsp
|
|
mov %rdi,%rax
|
|
mov %rsi,%rdx
|
|
ret
|