diff options
Diffstat (limited to 'arch/x86/crypto')
| -rw-r--r-- | arch/x86/crypto/Makefile | 1 | ||||
| -rw-r--r-- | arch/x86/crypto/chacha20-avx2-x86_64.S | 4 | ||||
| -rw-r--r-- | arch/x86/crypto/chacha20-ssse3-x86_64.S | 4 | ||||
| -rw-r--r-- | arch/x86/crypto/salsa20-x86_64-asm_64.S | 1 | ||||
| -rw-r--r-- | arch/x86/crypto/sha1-mb/Makefile | 1 | ||||
| -rw-r--r-- | arch/x86/crypto/sha1-mb/sha1_mb_mgr_flush_avx2.S | 12 | ||||
| -rw-r--r-- | arch/x86/crypto/sha256-mb/Makefile | 1 | ||||
| -rw-r--r-- | arch/x86/crypto/sha256-mb/sha256_mb_mgr_flush_avx2.S | 12 | ||||
| -rw-r--r-- | arch/x86/crypto/sha512-mb/Makefile | 1 |
9 files changed, 21 insertions, 16 deletions
diff --git a/arch/x86/crypto/Makefile b/arch/x86/crypto/Makefile index 9e32d40..5f07333 100644 --- a/arch/x86/crypto/Makefile +++ b/arch/x86/crypto/Makefile @@ -1,3 +1,4 @@ +# SPDX-License-Identifier: GPL-2.0 # # Arch-specific CryptoAPI modules. # diff --git a/arch/x86/crypto/chacha20-avx2-x86_64.S b/arch/x86/crypto/chacha20-avx2-x86_64.S index 3a2dc3d..f3cd26f 100644 --- a/arch/x86/crypto/chacha20-avx2-x86_64.S +++ b/arch/x86/crypto/chacha20-avx2-x86_64.S @@ -45,7 +45,7 @@ ENTRY(chacha20_8block_xor_avx2) vzeroupper # 4 * 32 byte stack, 32-byte aligned - mov %rsp, %r8 + lea 8(%rsp),%r10 and $~31, %rsp sub $0x80, %rsp @@ -443,6 +443,6 @@ ENTRY(chacha20_8block_xor_avx2) vmovdqu %ymm15,0x01e0(%rsi) vzeroupper - mov %r8,%rsp + lea -8(%r10),%rsp ret ENDPROC(chacha20_8block_xor_avx2) diff --git a/arch/x86/crypto/chacha20-ssse3-x86_64.S b/arch/x86/crypto/chacha20-ssse3-x86_64.S index 3f511a7..512a2b5 100644 --- a/arch/x86/crypto/chacha20-ssse3-x86_64.S +++ b/arch/x86/crypto/chacha20-ssse3-x86_64.S @@ -160,7 +160,7 @@ ENTRY(chacha20_4block_xor_ssse3) # done with the slightly better performing SSSE3 byte shuffling, # 7/12-bit word rotation uses traditional shift+OR. - mov %rsp,%r11 + lea 8(%rsp),%r10 sub $0x80,%rsp and $~63,%rsp @@ -625,6 +625,6 @@ ENTRY(chacha20_4block_xor_ssse3) pxor %xmm1,%xmm15 movdqu %xmm15,0xf0(%rsi) - mov %r11,%rsp + lea -8(%r10),%rsp ret ENDPROC(chacha20_4block_xor_ssse3) diff --git a/arch/x86/crypto/salsa20-x86_64-asm_64.S b/arch/x86/crypto/salsa20-x86_64-asm_64.S index 9279e0b..10db30d 100644 --- a/arch/x86/crypto/salsa20-x86_64-asm_64.S +++ b/arch/x86/crypto/salsa20-x86_64-asm_64.S @@ -1,3 +1,4 @@ +/* SPDX-License-Identifier: GPL-2.0 */ #include <linux/linkage.h> # enter salsa20_encrypt_bytes diff --git a/arch/x86/crypto/sha1-mb/Makefile b/arch/x86/crypto/sha1-mb/Makefile index 2e14acc..815ded3 100644 --- a/arch/x86/crypto/sha1-mb/Makefile +++ b/arch/x86/crypto/sha1-mb/Makefile @@ -1,3 +1,4 @@ +# SPDX-License-Identifier: GPL-2.0 # # Arch-specific CryptoAPI modules. # diff --git a/arch/x86/crypto/sha1-mb/sha1_mb_mgr_flush_avx2.S b/arch/x86/crypto/sha1-mb/sha1_mb_mgr_flush_avx2.S index 93b9455..7cfba73 100644 --- a/arch/x86/crypto/sha1-mb/sha1_mb_mgr_flush_avx2.S +++ b/arch/x86/crypto/sha1-mb/sha1_mb_mgr_flush_avx2.S @@ -157,8 +157,8 @@ LABEL skip_ %I .endr # Find min length - vmovdqa _lens+0*16(state), %xmm0 - vmovdqa _lens+1*16(state), %xmm1 + vmovdqu _lens+0*16(state), %xmm0 + vmovdqu _lens+1*16(state), %xmm1 vpminud %xmm1, %xmm0, %xmm2 # xmm2 has {D,C,B,A} vpalignr $8, %xmm2, %xmm3, %xmm3 # xmm3 has {x,x,D,C} @@ -178,8 +178,8 @@ LABEL skip_ %I vpsubd %xmm2, %xmm0, %xmm0 vpsubd %xmm2, %xmm1, %xmm1 - vmovdqa %xmm0, _lens+0*16(state) - vmovdqa %xmm1, _lens+1*16(state) + vmovdqu %xmm0, _lens+0*16(state) + vmovdqu %xmm1, _lens+1*16(state) # "state" and "args" are the same address, arg1 # len is arg2 @@ -235,8 +235,8 @@ ENTRY(sha1_mb_mgr_get_comp_job_avx2) jc .return_null # Find min length - vmovdqa _lens(state), %xmm0 - vmovdqa _lens+1*16(state), %xmm1 + vmovdqu _lens(state), %xmm0 + vmovdqu _lens+1*16(state), %xmm1 vpminud %xmm1, %xmm0, %xmm2 # xmm2 has {D,C,B,A} vpalignr $8, %xmm2, %xmm3, %xmm3 # xmm3 has {x,x,D,C} diff --git a/arch/x86/crypto/sha256-mb/Makefile b/arch/x86/crypto/sha256-mb/Makefile index 45b4fca..53ad6e7 100644 --- a/arch/x86/crypto/sha256-mb/Makefile +++ b/arch/x86/crypto/sha256-mb/Makefile @@ -1,3 +1,4 @@ +# SPDX-License-Identifier: GPL-2.0 # # Arch-specific CryptoAPI modules. # diff --git a/arch/x86/crypto/sha256-mb/sha256_mb_mgr_flush_avx2.S b/arch/x86/crypto/sha256-mb/sha256_mb_mgr_flush_avx2.S index 8fe6338..16c4ccb 100644 --- a/arch/x86/crypto/sha256-mb/sha256_mb_mgr_flush_avx2.S +++ b/arch/x86/crypto/sha256-mb/sha256_mb_mgr_flush_avx2.S @@ -155,8 +155,8 @@ LABEL skip_ %I .endr # Find min length - vmovdqa _lens+0*16(state), %xmm0 - vmovdqa _lens+1*16(state), %xmm1 + vmovdqu _lens+0*16(state), %xmm0 + vmovdqu _lens+1*16(state), %xmm1 vpminud %xmm1, %xmm0, %xmm2 # xmm2 has {D,C,B,A} vpalignr $8, %xmm2, %xmm3, %xmm3 # xmm3 has {x,x,D,C} @@ -176,8 +176,8 @@ LABEL skip_ %I vpsubd %xmm2, %xmm0, %xmm0 vpsubd %xmm2, %xmm1, %xmm1 - vmovdqa %xmm0, _lens+0*16(state) - vmovdqa %xmm1, _lens+1*16(state) + vmovdqu %xmm0, _lens+0*16(state) + vmovdqu %xmm1, _lens+1*16(state) # "state" and "args" are the same address, arg1 # len is arg2 @@ -234,8 +234,8 @@ ENTRY(sha256_mb_mgr_get_comp_job_avx2) jc .return_null # Find min length - vmovdqa _lens(state), %xmm0 - vmovdqa _lens+1*16(state), %xmm1 + vmovdqu _lens(state), %xmm0 + vmovdqu _lens+1*16(state), %xmm1 vpminud %xmm1, %xmm0, %xmm2 # xmm2 has {D,C,B,A} vpalignr $8, %xmm2, %xmm3, %xmm3 # xmm3 has {x,x,D,C} diff --git a/arch/x86/crypto/sha512-mb/Makefile b/arch/x86/crypto/sha512-mb/Makefile index 0a57e21..90f1ef6 100644 --- a/arch/x86/crypto/sha512-mb/Makefile +++ b/arch/x86/crypto/sha512-mb/Makefile @@ -1,3 +1,4 @@ +# SPDX-License-Identifier: GPL-2.0 # # Arch-specific CryptoAPI modules. # |
