feedc0de/wolfssl
1
0
Fork 0
forked from wolfSSL/wolfssl
Code Pull Requests Activity

Get Mac OS X working with the x86_64 assembly files

Browse Source
This commit is contained in:
Sean Parkinson
2019-02-15 15:08:47 +10:00
parent 64cb07557d
commit 16f31cf8c6
8 changed files with 1690 additions and 97 deletions
Download Patch File Download Diff File Expand all files Collapse all files

122
wolfcrypt/src/chacha_asm.S
View File

@@ -26,10 +26,18 @@
#define HAVE_INTEL_AVX2
#endif /* HAVE_INTEL_AVX2 */
#ifndef __APPLE__
.text
.globl chacha_encrypt_x64
.type chacha_encrypt_x64,@function
.align 4
chacha_encrypt_x64:
#else
.section __TEXT,__text
.globl _chacha_encrypt_x64
.p2align 2
_chacha_encrypt_x64:
#endif /* __APPLE__ */
pushq %rbx
pushq %rbp
pushq %r12
@@ -440,24 +448,70 @@ L_chacha_x64_done:
popq %rbp
popq %rbx
repz retq
#ifndef __APPLE__
.size chacha_encrypt_x64,.-chacha_encrypt_x64
#endif /* __APPLE__ */
#ifdef HAVE_INTEL_AVX1
.align 16
#ifndef __APPLE__
.data
#else
.section __DATA,__data
#endif /* __APPLE__ */
#ifndef __APPLE__
.align 16
#else
.p2align 4
#endif /* __APPLE__ */
L_chacha20_avx1_rotl8:
.quad 0x605040702010003, 0xe0d0c0f0a09080b
.align 16
#ifndef __APPLE__
.data
#else
.section __DATA,__data
#endif /* __APPLE__ */
#ifndef __APPLE__
.align 16
#else
.p2align 4
#endif /* __APPLE__ */
L_chacha20_avx1_rotl16:
.quad 0x504070601000302, 0xd0c0f0e09080b0a
.align 16
#ifndef __APPLE__
.data
#else
.section __DATA,__data
#endif /* __APPLE__ */
#ifndef __APPLE__
.align 16
#else
.p2align 4
#endif /* __APPLE__ */
L_chacha20_avx1_add:
.quad 0x100000000, 0x300000002
.align 16
#ifndef __APPLE__
.data
#else
.section __DATA,__data
#endif /* __APPLE__ */
#ifndef __APPLE__
.align 16
#else
.p2align 4
#endif /* __APPLE__ */
L_chacha20_avx1_four:
.quad 0x400000004, 0x400000004
#ifndef __APPLE__
.text
.globl chacha_encrypt_avx1
.type chacha_encrypt_avx1,@function
.align 4
chacha_encrypt_avx1:
#else
.section __TEXT,__text
.globl _chacha_encrypt_avx1
.p2align 2
_chacha_encrypt_avx1:
#endif /* __APPLE__ */
subq $0x190, %rsp
movq %rsp, %r9
leaq 256(%rsp), %r10
@@ -943,29 +997,75 @@ L_chacha20_avx1_partial_end64:
L_chacha20_avx1_partial_done:
addq $0x190, %rsp
repz retq
#ifndef __APPLE__
.size chacha_encrypt_avx1,.-chacha_encrypt_avx1
#endif /* __APPLE__ */
#endif /* HAVE_INTEL_AVX1 */
#ifdef HAVE_INTEL_AVX2
.align 32
#ifndef __APPLE__
.data
#else
.section __DATA,__data
#endif /* __APPLE__ */
#ifndef __APPLE__
.align 32
#else
.p2align 5
#endif /* __APPLE__ */
L_chacha20_avx2_rotl8:
.quad 0x605040702010003, 0xe0d0c0f0a09080b
.quad 0x605040702010003, 0xe0d0c0f0a09080b
.align 32
#ifndef __APPLE__
.data
#else
.section __DATA,__data
#endif /* __APPLE__ */
#ifndef __APPLE__
.align 32
#else
.p2align 5
#endif /* __APPLE__ */
L_chacha20_avx2_rotl16:
.quad 0x504070601000302, 0xd0c0f0e09080b0a
.quad 0x504070601000302, 0xd0c0f0e09080b0a
.align 32
#ifndef __APPLE__
.data
#else
.section __DATA,__data
#endif /* __APPLE__ */
#ifndef __APPLE__
.align 32
#else
.p2align 5
#endif /* __APPLE__ */
L_chacha20_avx2_add:
.quad 0x100000000, 0x300000002
.quad 0x500000004, 0x700000006
.align 32
#ifndef __APPLE__
.data
#else
.section __DATA,__data
#endif /* __APPLE__ */
#ifndef __APPLE__
.align 32
#else
.p2align 5
#endif /* __APPLE__ */
L_chacha20_avx2_eight:
.quad 0x800000008, 0x800000008
.quad 0x800000008, 0x800000008
#ifndef __APPLE__
.text
.globl chacha_encrypt_avx2
.type chacha_encrypt_avx2,@function
.align 4
chacha_encrypt_avx2:
#else
.section __TEXT,__text
.globl _chacha_encrypt_avx2
.p2align 2
_chacha_encrypt_avx2:
#endif /* __APPLE__ */
subq $0x310, %rsp
movq %rsp, %r9
leaq 512(%rsp), %r10
@@ -1307,8 +1407,14 @@ L_chacha20_avx2_done256:
shl $3, %eax
addl %eax, 48(%rdi)
L_chacha20_avx2_end256:
#ifndef __APPLE__
callq chacha_encrypt_avx1@plt
#else
callq _chacha_encrypt_avx1
#endif /* __APPLE__ */
addq $0x310, %rsp
repz retq
#ifndef __APPLE__
.size chacha_encrypt_avx2,.-chacha_encrypt_avx2
#endif /* __APPLE__ */
#endif /* HAVE_INTEL_AVX2 */