blob: e862c8204a47423c0680dfb43a672b957c85e0ab [file] [log] [blame]
#if defined(__i386__)
.text
.globl _GFp_aesni_encrypt
.private_extern _GFp_aesni_encrypt
.align 4
_GFp_aesni_encrypt:
L_GFp_aesni_encrypt_begin:
movl 4(%esp),%eax
movl 12(%esp),%edx
movups (%eax),%xmm2
movl 240(%edx),%ecx
movl 8(%esp),%eax
movups (%edx),%xmm0
movups 16(%edx),%xmm1
leal 32(%edx),%edx
xorps %xmm0,%xmm2
L000enc1_loop_1:
.byte 102,15,56,220,209
decl %ecx
movups (%edx),%xmm1
leal 16(%edx),%edx
jnz L000enc1_loop_1
.byte 102,15,56,221,209
pxor %xmm0,%xmm0
pxor %xmm1,%xmm1
movups %xmm2,(%eax)
pxor %xmm2,%xmm2
ret
.private_extern __aesni_encrypt2
.align 4
__aesni_encrypt2:
movups (%edx),%xmm0
shll $4,%ecx
movups 16(%edx),%xmm1
xorps %xmm0,%xmm2
pxor %xmm0,%xmm3
movups 32(%edx),%xmm0
leal 32(%edx,%ecx,1),%edx
negl %ecx
addl $16,%ecx
L001enc2_loop:
.byte 102,15,56,220,209
.byte 102,15,56,220,217
movups (%edx,%ecx,1),%xmm1
addl $32,%ecx
.byte 102,15,56,220,208
.byte 102,15,56,220,216
movups -16(%edx,%ecx,1),%xmm0
jnz L001enc2_loop
.byte 102,15,56,220,209
.byte 102,15,56,220,217
.byte 102,15,56,221,208
.byte 102,15,56,221,216
ret
.private_extern __aesni_encrypt3
.align 4
__aesni_encrypt3:
movups (%edx),%xmm0
shll $4,%ecx
movups 16(%edx),%xmm1
xorps %xmm0,%xmm2
pxor %xmm0,%xmm3
pxor %xmm0,%xmm4
movups 32(%edx),%xmm0
leal 32(%edx,%ecx,1),%edx
negl %ecx
addl $16,%ecx
L002enc3_loop:
.byte 102,15,56,220,209
.byte 102,15,56,220,217
.byte 102,15,56,220,225
movups (%edx,%ecx,1),%xmm1
addl $32,%ecx
.byte 102,15,56,220,208
.byte 102,15,56,220,216
.byte 102,15,56,220,224
movups -16(%edx,%ecx,1),%xmm0
jnz L002enc3_loop
.byte 102,15,56,220,209
.byte 102,15,56,220,217
.byte 102,15,56,220,225
.byte 102,15,56,221,208
.byte 102,15,56,221,216
.byte 102,15,56,221,224
ret
.private_extern __aesni_encrypt4
.align 4
__aesni_encrypt4:
movups (%edx),%xmm0
movups 16(%edx),%xmm1
shll $4,%ecx
xorps %xmm0,%xmm2
pxor %xmm0,%xmm3
pxor %xmm0,%xmm4
pxor %xmm0,%xmm5
movups 32(%edx),%xmm0
leal 32(%edx,%ecx,1),%edx
negl %ecx
.byte 15,31,64,0
addl $16,%ecx
L003enc4_loop:
.byte 102,15,56,220,209
.byte 102,15,56,220,217
.byte 102,15,56,220,225
.byte 102,15,56,220,233
movups (%edx,%ecx,1),%xmm1
addl $32,%ecx
.byte 102,15,56,220,208
.byte 102,15,56,220,216
.byte 102,15,56,220,224
.byte 102,15,56,220,232
movups -16(%edx,%ecx,1),%xmm0
jnz L003enc4_loop
.byte 102,15,56,220,209
.byte 102,15,56,220,217
.byte 102,15,56,220,225
.byte 102,15,56,220,233
.byte 102,15,56,221,208
.byte 102,15,56,221,216
.byte 102,15,56,221,224
.byte 102,15,56,221,232
ret
.private_extern __aesni_encrypt6
.align 4
__aesni_encrypt6:
movups (%edx),%xmm0
shll $4,%ecx
movups 16(%edx),%xmm1
xorps %xmm0,%xmm2
pxor %xmm0,%xmm3
pxor %xmm0,%xmm4
.byte 102,15,56,220,209
pxor %xmm0,%xmm5
pxor %xmm0,%xmm6
.byte 102,15,56,220,217
leal 32(%edx,%ecx,1),%edx
negl %ecx
.byte 102,15,56,220,225
pxor %xmm0,%xmm7
movups (%edx,%ecx,1),%xmm0
addl $16,%ecx
jmp L004_aesni_encrypt6_inner
.align 4,0x90
L005enc6_loop:
.byte 102,15,56,220,209
.byte 102,15,56,220,217
.byte 102,15,56,220,225
L004_aesni_encrypt6_inner:
.byte 102,15,56,220,233
.byte 102,15,56,220,241
.byte 102,15,56,220,249
L_aesni_encrypt6_enter:
movups (%edx,%ecx,1),%xmm1
addl $32,%ecx
.byte 102,15,56,220,208
.byte 102,15,56,220,216
.byte 102,15,56,220,224
.byte 102,15,56,220,232
.byte 102,15,56,220,240
.byte 102,15,56,220,248
movups -16(%edx,%ecx,1),%xmm0
jnz L005enc6_loop
.byte 102,15,56,220,209
.byte 102,15,56,220,217
.byte 102,15,56,220,225
.byte 102,15,56,220,233
.byte 102,15,56,220,241
.byte 102,15,56,220,249
.byte 102,15,56,221,208
.byte 102,15,56,221,216
.byte 102,15,56,221,224
.byte 102,15,56,221,232
.byte 102,15,56,221,240
.byte 102,15,56,221,248
ret
.globl _GFp_aesni_ctr32_encrypt_blocks
.private_extern _GFp_aesni_ctr32_encrypt_blocks
.align 4
_GFp_aesni_ctr32_encrypt_blocks:
L_GFp_aesni_ctr32_encrypt_blocks_begin:
pushl %ebp
pushl %ebx
pushl %esi
pushl %edi
movl 20(%esp),%esi
movl 24(%esp),%edi
movl 28(%esp),%eax
movl 32(%esp),%edx
movl 36(%esp),%ebx
movl %esp,%ebp
subl $88,%esp
andl $-16,%esp
movl %ebp,80(%esp)
cmpl $1,%eax
je L006ctr32_one_shortcut
movdqu (%ebx),%xmm7
movl $202182159,(%esp)
movl $134810123,4(%esp)
movl $67438087,8(%esp)
movl $66051,12(%esp)
movl $6,%ecx
xorl %ebp,%ebp
movl %ecx,16(%esp)
movl %ecx,20(%esp)
movl %ecx,24(%esp)
movl %ebp,28(%esp)
.byte 102,15,58,22,251,3
.byte 102,15,58,34,253,3
movl 240(%edx),%ecx
bswap %ebx
pxor %xmm0,%xmm0
pxor %xmm1,%xmm1
movdqa (%esp),%xmm2
.byte 102,15,58,34,195,0
leal 3(%ebx),%ebp
.byte 102,15,58,34,205,0
incl %ebx
.byte 102,15,58,34,195,1
incl %ebp
.byte 102,15,58,34,205,1
incl %ebx
.byte 102,15,58,34,195,2
incl %ebp
.byte 102,15,58,34,205,2
movdqa %xmm0,48(%esp)
.byte 102,15,56,0,194
movdqu (%edx),%xmm6
movdqa %xmm1,64(%esp)
.byte 102,15,56,0,202
pshufd $192,%xmm0,%xmm2
pshufd $128,%xmm0,%xmm3
cmpl $6,%eax
jb L007ctr32_tail
pxor %xmm6,%xmm7
shll $4,%ecx
movl $16,%ebx
movdqa %xmm7,32(%esp)
movl %edx,%ebp
subl %ecx,%ebx
leal 32(%edx,%ecx,1),%edx
subl $6,%eax
jmp L008ctr32_loop6
.align 4,0x90
L008ctr32_loop6:
pshufd $64,%xmm0,%xmm4
movdqa 32(%esp),%xmm0
pshufd $192,%xmm1,%xmm5
pxor %xmm0,%xmm2
pshufd $128,%xmm1,%xmm6
pxor %xmm0,%xmm3
pshufd $64,%xmm1,%xmm7
movups 16(%ebp),%xmm1
pxor %xmm0,%xmm4
pxor %xmm0,%xmm5
.byte 102,15,56,220,209
pxor %xmm0,%xmm6
pxor %xmm0,%xmm7
.byte 102,15,56,220,217
movups 32(%ebp),%xmm0
movl %ebx,%ecx
.byte 102,15,56,220,225
.byte 102,15,56,220,233
.byte 102,15,56,220,241
.byte 102,15,56,220,249
call L_aesni_encrypt6_enter
movups (%esi),%xmm1
movups 16(%esi),%xmm0
xorps %xmm1,%xmm2
movups 32(%esi),%xmm1
xorps %xmm0,%xmm3
movups %xmm2,(%edi)
movdqa 16(%esp),%xmm0
xorps %xmm1,%xmm4
movdqa 64(%esp),%xmm1
movups %xmm3,16(%edi)
movups %xmm4,32(%edi)
paddd %xmm0,%xmm1
paddd 48(%esp),%xmm0
movdqa (%esp),%xmm2
movups 48(%esi),%xmm3
movups 64(%esi),%xmm4
xorps %xmm3,%xmm5
movups 80(%esi),%xmm3
leal 96(%esi),%esi
movdqa %xmm0,48(%esp)
.byte 102,15,56,0,194
xorps %xmm4,%xmm6
movups %xmm5,48(%edi)
xorps %xmm3,%xmm7
movdqa %xmm1,64(%esp)
.byte 102,15,56,0,202
movups %xmm6,64(%edi)
pshufd $192,%xmm0,%xmm2
movups %xmm7,80(%edi)
leal 96(%edi),%edi
pshufd $128,%xmm0,%xmm3
subl $6,%eax
jnc L008ctr32_loop6
addl $6,%eax
jz L009ctr32_ret
movdqu (%ebp),%xmm7
movl %ebp,%edx
pxor 32(%esp),%xmm7
movl 240(%ebp),%ecx
L007ctr32_tail:
por %xmm7,%xmm2
cmpl $2,%eax
jb L010ctr32_one
pshufd $64,%xmm0,%xmm4
por %xmm7,%xmm3
je L011ctr32_two
pshufd $192,%xmm1,%xmm5
por %xmm7,%xmm4
cmpl $4,%eax
jb L012ctr32_three
pshufd $128,%xmm1,%xmm6
por %xmm7,%xmm5
je L013ctr32_four
por %xmm7,%xmm6
call __aesni_encrypt6
movups (%esi),%xmm1
movups 16(%esi),%xmm0
xorps %xmm1,%xmm2
movups 32(%esi),%xmm1
xorps %xmm0,%xmm3
movups 48(%esi),%xmm0
xorps %xmm1,%xmm4
movups 64(%esi),%xmm1
xorps %xmm0,%xmm5
movups %xmm2,(%edi)
xorps %xmm1,%xmm6
movups %xmm3,16(%edi)
movups %xmm4,32(%edi)
movups %xmm5,48(%edi)
movups %xmm6,64(%edi)
jmp L009ctr32_ret
.align 4,0x90
L006ctr32_one_shortcut:
movups (%ebx),%xmm2
movl 240(%edx),%ecx
L010ctr32_one:
movups (%edx),%xmm0
movups 16(%edx),%xmm1
leal 32(%edx),%edx
xorps %xmm0,%xmm2
L014enc1_loop_2:
.byte 102,15,56,220,209
decl %ecx
movups (%edx),%xmm1
leal 16(%edx),%edx
jnz L014enc1_loop_2
.byte 102,15,56,221,209
movups (%esi),%xmm6
xorps %xmm2,%xmm6
movups %xmm6,(%edi)
jmp L009ctr32_ret
.align 4,0x90
L011ctr32_two:
call __aesni_encrypt2
movups (%esi),%xmm5
movups 16(%esi),%xmm6
xorps %xmm5,%xmm2
xorps %xmm6,%xmm3
movups %xmm2,(%edi)
movups %xmm3,16(%edi)
jmp L009ctr32_ret
.align 4,0x90
L012ctr32_three:
call __aesni_encrypt3
movups (%esi),%xmm5
movups 16(%esi),%xmm6
xorps %xmm5,%xmm2
movups 32(%esi),%xmm7
xorps %xmm6,%xmm3
movups %xmm2,(%edi)
xorps %xmm7,%xmm4
movups %xmm3,16(%edi)
movups %xmm4,32(%edi)
jmp L009ctr32_ret
.align 4,0x90
L013ctr32_four:
call __aesni_encrypt4
movups (%esi),%xmm6
movups 16(%esi),%xmm7
movups 32(%esi),%xmm1
xorps %xmm6,%xmm2
movups 48(%esi),%xmm0
xorps %xmm7,%xmm3
movups %xmm2,(%edi)
xorps %xmm1,%xmm4
movups %xmm3,16(%edi)
xorps %xmm0,%xmm5
movups %xmm4,32(%edi)
movups %xmm5,48(%edi)
L009ctr32_ret:
pxor %xmm0,%xmm0
pxor %xmm1,%xmm1
pxor %xmm2,%xmm2
pxor %xmm3,%xmm3
pxor %xmm4,%xmm4
movdqa %xmm0,32(%esp)
pxor %xmm5,%xmm5
movdqa %xmm0,48(%esp)
pxor %xmm6,%xmm6
movdqa %xmm0,64(%esp)
pxor %xmm7,%xmm7
movl 80(%esp),%esp
popl %edi
popl %esi
popl %ebx
popl %ebp
ret
.private_extern __aesni_set_encrypt_key
.align 4
__aesni_set_encrypt_key:
pushl %ebp
pushl %ebx
testl %eax,%eax
jz L015bad_pointer
testl %edx,%edx
jz L015bad_pointer
call L016pic
L016pic:
popl %ebx
leal Lkey_const-L016pic(%ebx),%ebx
movl L_GFp_ia32cap_P$non_lazy_ptr-Lkey_const(%ebx),%ebp
movups (%eax),%xmm0
xorps %xmm4,%xmm4
movl 4(%ebp),%ebp
leal 16(%edx),%edx
andl $268437504,%ebp
cmpl $256,%ecx
je L01714rounds
cmpl $128,%ecx
jne L018bad_keybits
.align 4,0x90
L01910rounds:
cmpl $268435456,%ebp
je L02010rounds_alt
movl $9,%ecx
movups %xmm0,-16(%edx)
.byte 102,15,58,223,200,1
call L021key_128_cold
.byte 102,15,58,223,200,2
call L022key_128
.byte 102,15,58,223,200,4
call L022key_128
.byte 102,15,58,223,200,8
call L022key_128
.byte 102,15,58,223,200,16
call L022key_128
.byte 102,15,58,223,200,32
call L022key_128
.byte 102,15,58,223,200,64
call L022key_128
.byte 102,15,58,223,200,128
call L022key_128
.byte 102,15,58,223,200,27
call L022key_128
.byte 102,15,58,223,200,54
call L022key_128
movups %xmm0,(%edx)
movl %ecx,80(%edx)
jmp L023good_key
.align 4,0x90
L022key_128:
movups %xmm0,(%edx)
leal 16(%edx),%edx
L021key_128_cold:
shufps $16,%xmm0,%xmm4
xorps %xmm4,%xmm0
shufps $140,%xmm0,%xmm4
xorps %xmm4,%xmm0
shufps $255,%xmm1,%xmm1
xorps %xmm1,%xmm0
ret
.align 4,0x90
L02010rounds_alt:
movdqa (%ebx),%xmm5
movl $8,%ecx
movdqa 32(%ebx),%xmm4
movdqa %xmm0,%xmm2
movdqu %xmm0,-16(%edx)
L024loop_key128:
.byte 102,15,56,0,197
.byte 102,15,56,221,196
pslld $1,%xmm4
leal 16(%edx),%edx
movdqa %xmm2,%xmm3
pslldq $4,%xmm2
pxor %xmm2,%xmm3
pslldq $4,%xmm2
pxor %xmm2,%xmm3
pslldq $4,%xmm2
pxor %xmm3,%xmm2
pxor %xmm2,%xmm0
movdqu %xmm0,-16(%edx)
movdqa %xmm0,%xmm2
decl %ecx
jnz L024loop_key128
movdqa 48(%ebx),%xmm4
.byte 102,15,56,0,197
.byte 102,15,56,221,196
pslld $1,%xmm4
movdqa %xmm2,%xmm3
pslldq $4,%xmm2
pxor %xmm2,%xmm3
pslldq $4,%xmm2
pxor %xmm2,%xmm3
pslldq $4,%xmm2
pxor %xmm3,%xmm2
pxor %xmm2,%xmm0
movdqu %xmm0,(%edx)
movdqa %xmm0,%xmm2
.byte 102,15,56,0,197
.byte 102,15,56,221,196
movdqa %xmm2,%xmm3
pslldq $4,%xmm2
pxor %xmm2,%xmm3
pslldq $4,%xmm2
pxor %xmm2,%xmm3
pslldq $4,%xmm2
pxor %xmm3,%xmm2
pxor %xmm2,%xmm0
movdqu %xmm0,16(%edx)
movl $9,%ecx
movl %ecx,96(%edx)
jmp L023good_key
.align 4,0x90
L01714rounds:
movups 16(%eax),%xmm2
leal 16(%edx),%edx
cmpl $268435456,%ebp
je L02514rounds_alt
movl $13,%ecx
movups %xmm0,-32(%edx)
movups %xmm2,-16(%edx)
.byte 102,15,58,223,202,1
call L026key_256a_cold
.byte 102,15,58,223,200,1
call L027key_256b
.byte 102,15,58,223,202,2
call L028key_256a
.byte 102,15,58,223,200,2
call L027key_256b
.byte 102,15,58,223,202,4
call L028key_256a
.byte 102,15,58,223,200,4
call L027key_256b
.byte 102,15,58,223,202,8
call L028key_256a
.byte 102,15,58,223,200,8
call L027key_256b
.byte 102,15,58,223,202,16
call L028key_256a
.byte 102,15,58,223,200,16
call L027key_256b
.byte 102,15,58,223,202,32
call L028key_256a
.byte 102,15,58,223,200,32
call L027key_256b
.byte 102,15,58,223,202,64
call L028key_256a
movups %xmm0,(%edx)
movl %ecx,16(%edx)
xorl %eax,%eax
jmp L023good_key
.align 4,0x90
L028key_256a:
movups %xmm2,(%edx)
leal 16(%edx),%edx
L026key_256a_cold:
shufps $16,%xmm0,%xmm4
xorps %xmm4,%xmm0
shufps $140,%xmm0,%xmm4
xorps %xmm4,%xmm0
shufps $255,%xmm1,%xmm1
xorps %xmm1,%xmm0
ret
.align 4,0x90
L027key_256b:
movups %xmm0,(%edx)
leal 16(%edx),%edx
shufps $16,%xmm2,%xmm4
xorps %xmm4,%xmm2
shufps $140,%xmm2,%xmm4
xorps %xmm4,%xmm2
shufps $170,%xmm1,%xmm1
xorps %xmm1,%xmm2
ret
.align 4,0x90
L02514rounds_alt:
movdqa (%ebx),%xmm5
movdqa 32(%ebx),%xmm4
movl $7,%ecx
movdqu %xmm0,-32(%edx)
movdqa %xmm2,%xmm1
movdqu %xmm2,-16(%edx)
L029loop_key256:
.byte 102,15,56,0,213
.byte 102,15,56,221,212
movdqa %xmm0,%xmm3
pslldq $4,%xmm0
pxor %xmm0,%xmm3
pslldq $4,%xmm0
pxor %xmm0,%xmm3
pslldq $4,%xmm0
pxor %xmm3,%xmm0
pslld $1,%xmm4
pxor %xmm2,%xmm0
movdqu %xmm0,(%edx)
decl %ecx
jz L030done_key256
pshufd $255,%xmm0,%xmm2
pxor %xmm3,%xmm3
.byte 102,15,56,221,211
movdqa %xmm1,%xmm3
pslldq $4,%xmm1
pxor %xmm1,%xmm3
pslldq $4,%xmm1
pxor %xmm1,%xmm3
pslldq $4,%xmm1
pxor %xmm3,%xmm1
pxor %xmm1,%xmm2
movdqu %xmm2,16(%edx)
leal 32(%edx),%edx
movdqa %xmm2,%xmm1
jmp L029loop_key256
L030done_key256:
movl $13,%ecx
movl %ecx,16(%edx)
L023good_key:
pxor %xmm0,%xmm0
pxor %xmm1,%xmm1
pxor %xmm2,%xmm2
pxor %xmm3,%xmm3
pxor %xmm4,%xmm4
pxor %xmm5,%xmm5
xorl %eax,%eax
popl %ebx
popl %ebp
ret
.align 2,0x90
L015bad_pointer:
movl $-1,%eax
popl %ebx
popl %ebp
ret
.align 2,0x90
L018bad_keybits:
pxor %xmm0,%xmm0
movl $-2,%eax
popl %ebx
popl %ebp
ret
.globl _GFp_aesni_set_encrypt_key
.private_extern _GFp_aesni_set_encrypt_key
.align 4
_GFp_aesni_set_encrypt_key:
L_GFp_aesni_set_encrypt_key_begin:
movl 4(%esp),%eax
movl 8(%esp),%ecx
movl 12(%esp),%edx
call __aesni_set_encrypt_key
ret
.align 6,0x90
Lkey_const:
.long 202313229,202313229,202313229,202313229
.long 67569157,67569157,67569157,67569157
.long 1,1,1,1
.long 27,27,27,27
.byte 65,69,83,32,102,111,114,32,73,110,116,101,108,32,65,69
.byte 83,45,78,73,44,32,67,82,89,80,84,79,71,65,77,83
.byte 32,98,121,32,60,97,112,112,114,111,64,111,112,101,110,115
.byte 115,108,46,111,114,103,62,0
.section __IMPORT,__pointers,non_lazy_symbol_pointers
L_GFp_ia32cap_P$non_lazy_ptr:
.indirect_symbol _GFp_ia32cap_P
.long 0
#endif