Merge pull request #816 from harshavardhana/erasure

Simplify erasure package for OSX
This commit is contained in:
Harshavardhana 2015-09-05 20:58:03 -07:00
commit fa11112226
43 changed files with 238 additions and 510 deletions

View File

@ -61,5 +61,5 @@ install: gomake-all
clean:
@echo "Cleaning up all the generated files:"
@rm -fv cover.out
@rm -fv pkg/utils/split/TESTPREFIX.*
@rm -fv minio
@rm -fv pkg/erasure/*.syso

View File

@ -118,14 +118,11 @@ GLOBAL FLAGS:
{{range .Flags}}{{.}}
{{end}}{{end}}
VERSION:
` + getFormattedVersion() +
`
{{range $key, $value := ExtraInfo}}
` + getFormattedVersion() +
`{{range $key, $value := ExtraInfo}}
{{$key}}:
{{$value}}
{{end}}
`
app.CommandNotFound = func(ctx *cli.Context, command string) {
Fatalf("Command not found: %s\n", command)

View File

@ -33,64 +33,6 @@
%define WRT_OPT
%endif
%ifidn __OUTPUT_FORMAT__, macho64
%define EC_ENCODE_DATA_SSE _ec_encode_data_sse
%define EC_ENCODE_DATA_AVX _ec_encode_data_avx
%define EC_ENCODE_DATA_AVX2 _ec_encode_data_avx2
%define EC_ENCODE_DATA_BASE _ec_encode_data_base
%define EC_ENCODE_DATA _ec_encode_data
%define EC_ENCODE_DATA_UPDATE_BASE _ec_encode_data_update_base
%define EC_ENCODE_DATA_UPDATE_SSE _ec_encode_data_update_sse
%define EC_ENCODE_DATA_UPDATE_AVX _ec_encode_data_update_avx
%define EC_ENCODE_DATA_UPDATE_AVX2 _ec_encode_data_update_avx2
%define GF_VECT_MAD_BASE _gf_vect_mad_base
%define GF_VECT_MAD_SSE _gf_vect_mad_sse
%define GF_VECT_MAD_AVX _gf_vect_mad_avx
%define GF_VECT_MAD_AVX2 _gf_vect_mad_avx2
%define GF_VECT_MUL_SSE _gf_vect_mul_sse
%define GF_VECT_MUL_AVX _gf_vect_mul_avx
%define GF_VECT_MUL_BASE _gf_vect_mul_base
%define GF_VECT_MUL _gf_vect_mul
%define GF_VECT_DOT_PROD_SSE _gf_vect_dot_prod_sse
%define GF_VECT_DOT_PROD_AVX _gf_vect_dot_prod_avx
%define GF_VECT_DOT_PROD_AVX2 _gf_vect_dot_prod_avx2
%define GF_VECT_DOT_PROD_BASE _gf_vect_dot_prod_base
%define GF_VECT_DOT_PROD _gf_vect_dot_prod
%else
%define EC_ENCODE_DATA_SSE ec_encode_data_sse
%define EC_ENCODE_DATA_AVX ec_encode_data_avx
%define EC_ENCODE_DATA_AVX2 ec_encode_data_avx2
%define EC_ENCODE_DATA_BASE ec_encode_data_base
%define EC_ENCODE_DATA ec_encode_data
%define EC_ENCODE_DATA_UPDATE_BASE ec_encode_data_update_base
%define EC_ENCODE_DATA_UPDATE_SSE ec_encode_data_update_sse
%define EC_ENCODE_DATA_UPDATE_AVX ec_encode_data_update_avx
%define EC_ENCODE_DATA_UPDATE_AVX2 ec_encode_data_update_avx2
%define GF_VECT_MAD_BASE gf_vect_mad_base
%define GF_VECT_MAD_SSE gf_vect_mad_sse
%define GF_VECT_MAD_AVX gf_vect_mad_avx
%define GF_VECT_MAD_AVX2 gf_vect_mad_avx2
%define GF_VECT_MUL_SSE gf_vect_mul_sse
%define GF_VECT_MUL_AVX gf_vect_mul_avx
%define GF_VECT_MUL_BASE gf_vect_mul_base
%define GF_VECT_MUL gf_vect_mul
%define GF_VECT_DOT_PROD_SSE gf_vect_dot_prod_sse
%define GF_VECT_DOT_PROD_AVX gf_vect_dot_prod_avx
%define GF_VECT_DOT_PROD_AVX2 gf_vect_dot_prod_avx2
%define GF_VECT_DOT_PROD_BASE gf_vect_dot_prod_base
%define GF_VECT_DOT_PROD gf_vect_dot_prod
%endif
%include "ec_reg_sizes.asm"
%ifidn __OUTPUT_FORMAT__, elf32
@ -119,29 +61,29 @@
%define arg5 rdx
extern EC_ENCODE_DATA_UPDATE_SSE
extern EC_ENCODE_DATA_UPDATE_AVX
extern EC_ENCODE_DATA_UPDATE_AVX2
extern GF_VECT_MUL_SSE
extern GF_VECT_MUL_AVX
extern ec_encode_data_update_sse
extern ec_encode_data_update_avx
extern ec_encode_data_update_avx2
extern gf_vect_mul_sse
extern gf_vect_mul_avx
extern GF_VECT_MAD_SSE
extern GF_VECT_MAD_AVX
extern GF_VECT_MAD_AVX2
extern gf_vect_mad_sse
extern gf_vect_mad_avx
extern gf_vect_mad_avx2
%endif
extern GF_VECT_MUL_BASE
extern EC_ENCODE_DATA_BASE
extern EC_ENCODE_DATA_UPDATE_BASE
extern GF_VECT_DOT_PROD_BASE
extern GF_VECT_MAD_BASE
extern gf_vect_mul_base
extern ec_encode_data_base
extern ec_encode_data_update_base
extern gf_vect_dot_prod_base
extern gf_vect_mad_base
extern GF_VECT_DOT_PROD_SSE
extern GF_VECT_DOT_PROD_AVX
extern GF_VECT_DOT_PROD_AVX2
extern EC_ENCODE_DATA_SSE
extern EC_ENCODE_DATA_AVX
extern EC_ENCODE_DATA_AVX2
extern gf_vect_dot_prod_sse
extern gf_vect_dot_prod_avx
extern gf_vect_dot_prod_avx2
extern ec_encode_data_sse
extern ec_encode_data_avx
extern ec_encode_data_avx2
section .data
@ -167,11 +109,11 @@ section .text
;;;;
; ec_encode_data multibinary function
;;;;
global EC_ENCODE_DATA:function
global ec_encode_data:function
ec_encode_data_mbinit:
call ec_encode_data_dispatch_init
EC_ENCODE_DATA:
ec_encode_data:
jmp wrd_sz [ec_encode_data_dispatched]
ec_encode_data_dispatch_init:
@ -180,17 +122,17 @@ ec_encode_data_dispatch_init:
push arg3
push arg4
push arg5
lea arg1, [EC_ENCODE_DATA_BASE WRT_OPT] ; Default
lea arg1, [ec_encode_data_base WRT_OPT] ; Default
mov eax, 1
cpuid
lea arg3, [EC_ENCODE_DATA_SSE WRT_OPT]
lea arg3, [ec_encode_data_sse WRT_OPT]
test ecx, FLAG_CPUID1_ECX_SSE4_1
cmovne arg1, arg3
and ecx, (FLAG_CPUID1_ECX_AVX | FLAG_CPUID1_ECX_OSXSAVE)
cmp ecx, (FLAG_CPUID1_ECX_AVX | FLAG_CPUID1_ECX_OSXSAVE)
lea arg3, [EC_ENCODE_DATA_AVX WRT_OPT]
lea arg3, [ec_encode_data_avx WRT_OPT]
jne _done_ec_encode_data_init
mov arg1, arg3
@ -200,7 +142,7 @@ ec_encode_data_dispatch_init:
mov eax, 7
cpuid
test ebx, FLAG_CPUID1_EBX_AVX2
lea arg3, [EC_ENCODE_DATA_AVX2 WRT_OPT]
lea arg3, [ec_encode_data_avx2 WRT_OPT]
cmovne arg1, arg3
;; Does it have xmm and ymm support
xor ecx, ecx
@ -208,7 +150,7 @@ ec_encode_data_dispatch_init:
and eax, FLAG_XGETBV_EAX_XMM_YMM
cmp eax, FLAG_XGETBV_EAX_XMM_YMM
je _done_ec_encode_data_init
lea arg1, [EC_ENCODE_DATA_SSE WRT_OPT]
lea arg1, [ec_encode_data_sse WRT_OPT]
_done_ec_encode_data_init:
pop arg5
@ -222,28 +164,28 @@ _done_ec_encode_data_init:
;;;;
; gf_vect_mul multibinary function
;;;;
global GF_VECT_MUL:function
global gf_vect_mul:function
gf_vect_mul_mbinit:
call gf_vect_mul_dispatch_init
GF_VECT_MUL:
gf_vect_mul:
jmp wrd_sz [gf_vect_mul_dispatched]
gf_vect_mul_dispatch_init:
push arg1
%ifidn __OUTPUT_FORMAT__, elf32 ;; 32-bit check
lea arg1, [GF_VECT_MUL_BASE]
lea arg1, [gf_vect_mul_base]
%else
push rax
push rbx
push rcx
push rdx
lea arg1, [GF_VECT_MUL_BASE WRT_OPT] ; Default
lea arg1, [gf_vect_mul_base WRT_OPT] ; Default
mov eax, 1
cpuid
test ecx, FLAG_CPUID1_ECX_SSE4_2
lea rbx, [GF_VECT_MUL_SSE WRT_OPT]
lea rbx, [gf_vect_mul_sse WRT_OPT]
je _done_gf_vect_mul_dispatch_init
mov arg1, rbx
@ -258,7 +200,7 @@ gf_vect_mul_dispatch_init:
and eax, FLAG_XGETBV_EAX_XMM_YMM
cmp eax, FLAG_XGETBV_EAX_XMM_YMM
jne _done_gf_vect_mul_dispatch_init
lea arg1, [GF_VECT_MUL_AVX WRT_OPT]
lea arg1, [gf_vect_mul_avx WRT_OPT]
_done_gf_vect_mul_dispatch_init:
pop rdx
@ -273,11 +215,11 @@ _done_gf_vect_mul_dispatch_init:
;;;;
; ec_encode_data_update multibinary function
;;;;
global EC_ENCODE_DATA_UPDATE:function
global ec_encode_data_update:function
ec_encode_data_update_mbinit:
call ec_encode_data_update_dispatch_init
EC_ENCODE_DATA_UPDATE:
ec_encode_data_update:
jmp wrd_sz [ec_encode_data_update_dispatched]
ec_encode_data_update_dispatch_init:
@ -289,17 +231,17 @@ ec_encode_data_update_dispatch_init:
push rbx
push rcx
push rdx
lea arg1, [EC_ENCODE_DATA_UPDATE_BASE WRT_OPT] ; Default
lea arg1, [ec_encode_data_update_base WRT_OPT] ; Default
mov eax, 1
cpuid
lea rbx, [EC_ENCODE_DATA_UPDATE_SSE WRT_OPT]
lea rbx, [ec_encode_data_update_sse WRT_OPT]
test ecx, FLAG_CPUID1_ECX_SSE4_1
cmovne arg1, rbx
and ecx, (FLAG_CPUID1_ECX_AVX | FLAG_CPUID1_ECX_OSXSAVE)
cmp ecx, (FLAG_CPUID1_ECX_AVX | FLAG_CPUID1_ECX_OSXSAVE)
lea rbx, [EC_ENCODE_DATA_UPDATE_AVX WRT_OPT]
lea rbx, [ec_encode_data_update_avx WRT_OPT]
jne _done_ec_encode_data_update_init
mov rsi, rbx
@ -309,7 +251,7 @@ ec_encode_data_update_dispatch_init:
mov eax, 7
cpuid
test ebx, FLAG_CPUID1_EBX_AVX2
lea rbx, [EC_ENCODE_DATA_UPDATE_AVX2 WRT_OPT]
lea rbx, [ec_encode_data_update_avx2 WRT_OPT]
cmovne rsi, rbx
;; Does it have xmm and ymm support
@ -318,7 +260,7 @@ ec_encode_data_update_dispatch_init:
and eax, FLAG_XGETBV_EAX_XMM_YMM
cmp eax, FLAG_XGETBV_EAX_XMM_YMM
je _done_ec_encode_data_update_init
lea rsi, [EC_ENCODE_DATA_UPDATE_SSE WRT_OPT]
lea rsi, [ec_encode_data_update_sse WRT_OPT]
_done_ec_encode_data_update_init:
pop rdx
@ -333,11 +275,11 @@ _done_ec_encode_data_update_init:
;;;;
; gf_vect_dot_prod multibinary function
;;;;
global GF_VECT_DOT_PROD:function
global gf_vect_dot_prod:function
gf_vect_dot_prod_mbinit:
call gf_vect_dot_prod_dispatch_init
GF_VECT_DOT_PROD:
gf_vect_dot_prod:
jmp wrd_sz [gf_vect_dot_prod_dispatched]
gf_vect_dot_prod_dispatch_init:
@ -346,17 +288,17 @@ gf_vect_dot_prod_dispatch_init:
push arg3
push arg4
push arg5
lea arg1, [GF_VECT_DOT_PROD_BASE WRT_OPT] ; Default
lea arg1, [gf_vect_dot_prod_base WRT_OPT] ; Default
mov eax, 1
cpuid
lea arg3, [GF_VECT_DOT_PROD_SSE WRT_OPT]
lea arg3, [gf_vect_dot_prod_sse WRT_OPT]
test ecx, FLAG_CPUID1_ECX_SSE4_1
cmovne arg1, arg3
and ecx, (FLAG_CPUID1_ECX_AVX | FLAG_CPUID1_ECX_OSXSAVE)
cmp ecx, (FLAG_CPUID1_ECX_AVX | FLAG_CPUID1_ECX_OSXSAVE)
lea arg3, [GF_VECT_DOT_PROD_AVX WRT_OPT]
lea arg3, [gf_vect_dot_prod_avx WRT_OPT]
jne _done_gf_vect_dot_prod_init
mov arg1, arg3
@ -366,7 +308,7 @@ gf_vect_dot_prod_dispatch_init:
mov eax, 7
cpuid
test ebx, FLAG_CPUID1_EBX_AVX2
lea arg3, [GF_VECT_DOT_PROD_AVX2 WRT_OPT]
lea arg3, [gf_vect_dot_prod_avx2 WRT_OPT]
cmovne arg1, arg3
;; Does it have xmm and ymm support
xor ecx, ecx
@ -374,7 +316,7 @@ gf_vect_dot_prod_dispatch_init:
and eax, FLAG_XGETBV_EAX_XMM_YMM
cmp eax, FLAG_XGETBV_EAX_XMM_YMM
je _done_gf_vect_dot_prod_init
lea arg1, [GF_VECT_DOT_PROD_SSE WRT_OPT]
lea arg1, [gf_vect_dot_prod_sse WRT_OPT]
_done_gf_vect_dot_prod_init:
pop arg5
@ -388,11 +330,11 @@ _done_gf_vect_dot_prod_init:
;;;;
; gf_vect_mad multibinary function
;;;;
global GF_VECT_MAD:function
global gf_vect_mad:function
gf_vect_mad_mbinit:
call gf_vect_mad_dispatch_init
GF_VECT_MAD:
gf_vect_mad:
jmp wrd_sz [gf_vect_mad_dispatched]
gf_vect_mad_dispatch_init:
@ -404,17 +346,17 @@ gf_vect_mad_dispatch_init:
push rbx
push rcx
push rdx
lea arg1, [GF_VECT_MAD_BASE WRT_OPT] ; Default
lea arg1, [gf_vect_mad_base WRT_OPT] ; Default
mov eax, 1
cpuid
lea rbx, [GF_VECT_MAD_SSE WRT_OPT]
lea rbx, [gf_vect_mad_sse WRT_OPT]
test ecx, FLAG_CPUID1_ECX_SSE4_1
cmovne arg1, rbx
and ecx, (FLAG_CPUID1_ECX_AVX | FLAG_CPUID1_ECX_OSXSAVE)
cmp ecx, (FLAG_CPUID1_ECX_AVX | FLAG_CPUID1_ECX_OSXSAVE)
lea rbx, [GF_VECT_MAD_AVX WRT_OPT]
lea rbx, [gf_vect_mad_avx WRT_OPT]
jne _done_gf_vect_mad_init
mov rsi, rbx
@ -424,7 +366,7 @@ gf_vect_mad_dispatch_init:
mov eax, 7
cpuid
test ebx, FLAG_CPUID1_EBX_AVX2
lea rbx, [GF_VECT_MAD_AVX2 WRT_OPT]
lea rbx, [gf_vect_mad_avx2 WRT_OPT]
cmovne rsi, rbx
;; Does it have xmm and ymm support
@ -433,7 +375,7 @@ gf_vect_mad_dispatch_init:
and eax, FLAG_XGETBV_EAX_XMM_YMM
cmp eax, FLAG_XGETBV_EAX_XMM_YMM
je _done_gf_vect_mad_init
lea rsi, [GF_VECT_MAD_SSE WRT_OPT]
lea rsi, [gf_vect_mad_sse WRT_OPT]
_done_gf_vect_mad_init:
pop rdx
@ -455,8 +397,8 @@ global %1_slver
%endmacro
;;; func core, ver, snum
slversion EC_ENCODE_DATA, 00, 03, 0133
slversion GF_VECT_MUL, 00, 02, 0134
slversion EC_ENCODE_DATA_UPDATE, 00, 02, 0212
slversion GF_VECT_DOT_PROD, 00, 02, 0138
slversion GF_VECT_MAD, 00, 01, 0213
slversion ec_encode_data, 00, 03, 0133
slversion gf_vect_mul, 00, 02, 0134
slversion ec_encode_data_update, 00, 02, 0212
slversion gf_vect_dot_prod, 00, 02, 0138
slversion gf_vect_mad, 00, 01, 0213

View File

@ -94,3 +94,14 @@
%define BYTE(reg) reg %+ b
%define XWORD(reg) reg %+ x
%ifidn __OUTPUT_FORMAT__,elf32
section .note.GNU-stack noalloc noexec nowrite progbits
section .text
%endif
%ifidn __OUTPUT_FORMAT__,elf64
section .note.GNU-stack noalloc noexec nowrite progbits
section .text
%endif

View File

@ -2,42 +2,42 @@
package erasure
//go:generate yasm -f macho64 ec_multibinary.asm -o ec_multibinary.syso
//go:generate yasm -f macho64 gf_2vect_mad_avx2.asm -o gf_2vect_mad_avx2.syso
//go:generate yasm -f macho64 gf_2vect_mad_avx.asm -o gf_2vect_mad_avx.syso
//go:generate yasm -f macho64 gf_2vect_mad_sse.asm -o gf_2vect_mad_sse.syso
//go:generate yasm -f macho64 gf_3vect_mad_avx2.asm -o gf_3vect_mad_avx2.syso
//go:generate yasm -f macho64 gf_3vect_mad_avx.asm -o gf_3vect_mad_avx.syso
//go:generate yasm -f macho64 gf_3vect_mad_sse.asm -o gf_3vect_mad_sse.syso
//go:generate yasm -f macho64 gf_4vect_mad_avx2.asm -o gf_4vect_mad_avx2.syso
//go:generate yasm -f macho64 gf_4vect_mad_avx.asm -o gf_4vect_mad_avx.syso
//go:generate yasm -f macho64 gf_4vect_mad_sse.asm -o gf_4vect_mad_sse.syso
//go:generate yasm -f macho64 gf_5vect_mad_avx2.asm -o gf_5vect_mad_avx2.syso
//go:generate yasm -f macho64 gf_5vect_mad_avx.asm -o gf_5vect_mad_avx.syso
//go:generate yasm -f macho64 gf_5vect_mad_sse.asm -o gf_5vect_mad_sse.syso
//go:generate yasm -f macho64 gf_6vect_mad_avx2.asm -o gf_6vect_mad_avx2.syso
//go:generate yasm -f macho64 gf_6vect_mad_avx.asm -o gf_6vect_mad_avx.syso
//go:generate yasm -f macho64 gf_6vect_mad_sse.asm -o gf_6vect_mad_sse.syso
//go:generate yasm -f macho64 gf_vect_mad_avx2.asm -o gf_vect_mad_avx2.syso
//go:generate yasm -f macho64 gf_vect_mad_avx.asm -o gf_vect_mad_avx.syso
//go:generate yasm -f macho64 gf_vect_mad_sse.asm -o gf_vect_mad_sse.syso
//go:generate yasm -f macho64 gf_2vect_dot_prod_avx2.asm -o gf_2vect_dot_prod_avx2.syso
//go:generate yasm -f macho64 gf_2vect_dot_prod_avx.asm -o gf_2vect_dot_prod_avx.syso
//go:generate yasm -f macho64 gf_2vect_dot_prod_sse.asm -o gf_2vect_dot_prod_sse.syso
//go:generate yasm -f macho64 gf_3vect_dot_prod_avx2.asm -o gf_3vect_dot_prod_avx2.syso
//go:generate yasm -f macho64 gf_3vect_dot_prod_avx.asm -o gf_3vect_dot_prod_avx.syso
//go:generate yasm -f macho64 gf_3vect_dot_prod_sse.asm -o gf_3vect_dot_prod_sse.syso
//go:generate yasm -f macho64 gf_4vect_dot_prod_avx2.asm -o gf_4vect_dot_prod_avx2.syso
//go:generate yasm -f macho64 gf_4vect_dot_prod_avx.asm -o gf_4vect_dot_prod_avx.syso
//go:generate yasm -f macho64 gf_4vect_dot_prod_sse.asm -o gf_4vect_dot_prod_sse.syso
//go:generate yasm -f macho64 gf_5vect_dot_prod_avx2.asm -o gf_5vect_dot_prod_avx2.syso
//go:generate yasm -f macho64 gf_5vect_dot_prod_avx.asm -o gf_5vect_dot_prod_avx.syso
//go:generate yasm -f macho64 gf_5vect_dot_prod_sse.asm -o gf_5vect_dot_prod_sse.syso
//go:generate yasm -f macho64 gf_6vect_dot_prod_avx2.asm -o gf_6vect_dot_prod_avx2.syso
//go:generate yasm -f macho64 gf_6vect_dot_prod_avx.asm -o gf_6vect_dot_prod_avx.syso
//go:generate yasm -f macho64 gf_6vect_dot_prod_sse.asm -o gf_6vect_dot_prod_sse.syso
//go:generate yasm -f macho64 gf_vect_dot_prod_avx2.asm -o gf_vect_dot_prod_avx2.syso
//go:generate yasm -f macho64 gf_vect_dot_prod_avx.asm -o gf_vect_dot_prod_avx.syso
//go:generate yasm -f macho64 gf_vect_dot_prod_sse.asm -o gf_vect_dot_prod_sse.syso
//go:generate yasm -f macho64 gf_vect_mul_avx.asm -o gf_vect_mul_avx.syso
//go:generate yasm -f macho64 gf_vect_mul_sse.asm -o gf_vect_mul_sse.syso
//go:generate yasm -f macho64 --prefix=_ ec_multibinary.asm -o ec_multibinary.syso
//go:generate yasm -f macho64 --prefix=_ gf_2vect_mad_avx2.asm -o gf_2vect_mad_avx2.syso
//go:generate yasm -f macho64 --prefix=_ gf_2vect_mad_avx.asm -o gf_2vect_mad_avx.syso
//go:generate yasm -f macho64 --prefix=_ gf_2vect_mad_sse.asm -o gf_2vect_mad_sse.syso
//go:generate yasm -f macho64 --prefix=_ gf_3vect_mad_avx2.asm -o gf_3vect_mad_avx2.syso
//go:generate yasm -f macho64 --prefix=_ gf_3vect_mad_avx.asm -o gf_3vect_mad_avx.syso
//go:generate yasm -f macho64 --prefix=_ gf_3vect_mad_sse.asm -o gf_3vect_mad_sse.syso
//go:generate yasm -f macho64 --prefix=_ gf_4vect_mad_avx2.asm -o gf_4vect_mad_avx2.syso
//go:generate yasm -f macho64 --prefix=_ gf_4vect_mad_avx.asm -o gf_4vect_mad_avx.syso
//go:generate yasm -f macho64 --prefix=_ gf_4vect_mad_sse.asm -o gf_4vect_mad_sse.syso
//go:generate yasm -f macho64 --prefix=_ gf_5vect_mad_avx2.asm -o gf_5vect_mad_avx2.syso
//go:generate yasm -f macho64 --prefix=_ gf_5vect_mad_avx.asm -o gf_5vect_mad_avx.syso
//go:generate yasm -f macho64 --prefix=_ gf_5vect_mad_sse.asm -o gf_5vect_mad_sse.syso
//go:generate yasm -f macho64 --prefix=_ gf_6vect_mad_avx2.asm -o gf_6vect_mad_avx2.syso
//go:generate yasm -f macho64 --prefix=_ gf_6vect_mad_avx.asm -o gf_6vect_mad_avx.syso
//go:generate yasm -f macho64 --prefix=_ gf_6vect_mad_sse.asm -o gf_6vect_mad_sse.syso
//go:generate yasm -f macho64 --prefix=_ gf_vect_mad_avx2.asm -o gf_vect_mad_avx2.syso
//go:generate yasm -f macho64 --prefix=_ gf_vect_mad_avx.asm -o gf_vect_mad_avx.syso
//go:generate yasm -f macho64 --prefix=_ gf_vect_mad_sse.asm -o gf_vect_mad_sse.syso
//go:generate yasm -f macho64 --prefix=_ gf_2vect_dot_prod_avx2.asm -o gf_2vect_dot_prod_avx2.syso
//go:generate yasm -f macho64 --prefix=_ gf_2vect_dot_prod_avx.asm -o gf_2vect_dot_prod_avx.syso
//go:generate yasm -f macho64 --prefix=_ gf_2vect_dot_prod_sse.asm -o gf_2vect_dot_prod_sse.syso
//go:generate yasm -f macho64 --prefix=_ gf_3vect_dot_prod_avx2.asm -o gf_3vect_dot_prod_avx2.syso
//go:generate yasm -f macho64 --prefix=_ gf_3vect_dot_prod_avx.asm -o gf_3vect_dot_prod_avx.syso
//go:generate yasm -f macho64 --prefix=_ gf_3vect_dot_prod_sse.asm -o gf_3vect_dot_prod_sse.syso
//go:generate yasm -f macho64 --prefix=_ gf_4vect_dot_prod_avx2.asm -o gf_4vect_dot_prod_avx2.syso
//go:generate yasm -f macho64 --prefix=_ gf_4vect_dot_prod_avx.asm -o gf_4vect_dot_prod_avx.syso
//go:generate yasm -f macho64 --prefix=_ gf_4vect_dot_prod_sse.asm -o gf_4vect_dot_prod_sse.syso
//go:generate yasm -f macho64 --prefix=_ gf_5vect_dot_prod_avx2.asm -o gf_5vect_dot_prod_avx2.syso
//go:generate yasm -f macho64 --prefix=_ gf_5vect_dot_prod_avx.asm -o gf_5vect_dot_prod_avx.syso
//go:generate yasm -f macho64 --prefix=_ gf_5vect_dot_prod_sse.asm -o gf_5vect_dot_prod_sse.syso
//go:generate yasm -f macho64 --prefix=_ gf_6vect_dot_prod_avx2.asm -o gf_6vect_dot_prod_avx2.syso
//go:generate yasm -f macho64 --prefix=_ gf_6vect_dot_prod_avx.asm -o gf_6vect_dot_prod_avx.syso
//go:generate yasm -f macho64 --prefix=_ gf_6vect_dot_prod_sse.asm -o gf_6vect_dot_prod_sse.syso
//go:generate yasm -f macho64 --prefix=_ gf_vect_dot_prod_avx2.asm -o gf_vect_dot_prod_avx2.syso
//go:generate yasm -f macho64 --prefix=_ gf_vect_dot_prod_avx.asm -o gf_vect_dot_prod_avx.syso
//go:generate yasm -f macho64 --prefix=_ gf_vect_dot_prod_sse.asm -o gf_vect_dot_prod_sse.syso
//go:generate yasm -f macho64 --prefix=_ gf_vect_mul_avx.asm -o gf_vect_mul_avx.syso
//go:generate yasm -f macho64 --prefix=_ gf_vect_mul_sse.asm -o gf_vect_mul_sse.syso

View File

@ -31,12 +31,6 @@
;;; gf_2vect_dot_prod_avx(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_2VECT_DOT_PROD_AVX _gf_2vect_dot_prod_avx
%else
%define GF_2VECT_DOT_PROD_AVX gf_2vect_dot_prod_avx
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -274,9 +268,8 @@ section .text
%endif
align 16
global GF_2VECT_DOT_PROD_AVX:function
func(GF_2VECT_DOT_PROD_AVX)
global gf_2vect_dot_prod_avx:function
func(gf_2vect_dot_prod_avx)
FUNC_SAVE
SLDR len, len_m
sub len, 16
@ -378,4 +371,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_2VECT_DOT_PROD_AVX, 02, 04, 0191
slversion gf_2vect_dot_prod_avx, 02, 04, 0191

View File

@ -31,12 +31,6 @@
;;; gf_2vect_dot_prod_avx2(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_2VECT_DOT_PROD_AVX2 _gf_2vect_dot_prod_avx2
%else
%define GF_2VECT_DOT_PROD_AVX2 gf_2vect_dot_prod_avx2
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -282,9 +276,8 @@ section .text
%endif
align 16
global GF_2VECT_DOT_PROD_AVX2:function
func(GF_2VECT_DOT_PROD_AVX2)
global gf_2vect_dot_prod_avx2:function
func(gf_2vect_dot_prod_avx2)
FUNC_SAVE
SLDR len, len_m
sub len, 32
@ -395,4 +388,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_2VECT_DOT_PROD_AVX2, 04, 04, 0196
slversion gf_2vect_dot_prod_avx2, 04, 04, 0196

View File

@ -31,12 +31,6 @@
;;; gf_2vect_dot_prod_sse(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_2VECT_DOT_PROD_SSE _gf_2vect_dot_prod_sse
%else
%define GF_2VECT_DOT_PROD_SSE gf_2vect_dot_prod_sse
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -274,9 +268,8 @@ section .text
%endif
align 16
global GF_2VECT_DOT_PROD_SSE:function
func(GF_2VECT_DOT_PROD_SSE)
global gf_2vect_dot_prod_sse:function
func(gf_2vect_dot_prod_sse)
FUNC_SAVE
SLDR len, len_m
sub len, 16
@ -380,4 +373,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_2VECT_DOT_PROD_SSE, 00, 03, 0062
slversion gf_2vect_dot_prod_sse, 00, 03, 0062

View File

@ -31,12 +31,6 @@
;;; gf_2vect_mad_avx(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_2VECT_MAD_AVX _gf_2vect_mad_avx
%else
%define GF_2VECT_MAD_AVX gf_2vect_mad_avx
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -104,6 +98,7 @@
%define func(x) x:
%define FUNC_SAVE
%define FUNC_RESTORE
%elifidn __OUTPUT_FORMAT__, macho64
%define arg0 rdi
%define arg0.w edi
@ -175,9 +170,8 @@ section .text
align 16
global GF_2VECT_MAD_AVX:function
func(GF_2VECT_MAD_AVX)
global gf_2vect_mad_avx:function
func(gf_2vect_mad_avx)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -261,4 +255,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_2VECT_MAD_AVX, 02, 00, 0204
slversion gf_2vect_mad_avx, 02, 00, 0204

View File

@ -31,12 +31,6 @@
;;; gf_2vect_mad_avx2(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_2VECT_MAD_AVX2 _gf_2vect_mad_avx2
%else
%define GF_2VECT_MAD_AVX2 gf_2vect_mad_avx2
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -188,9 +182,8 @@ section .text
%define xtmpd2 ymm9
align 16
global GF_2VECT_MAD_AVX2:function
func(GF_2VECT_MAD_AVX2)
global gf_2vect_mad_avx2:function
func(gf_2vect_mad_avx2)
FUNC_SAVE
sub len, 32
jl .return_fail
@ -277,4 +270,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_2VECT_MAD_AVX2, 04, 00, 0205
slversion gf_2vect_mad_avx2, 04, 00, 0205

View File

@ -31,12 +31,6 @@
;;; gf_2vect_mad_sse(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_2VECT_MAD_SSE _gf_2vect_mad_sse
%else
%define GF_2VECT_MAD_SSE gf_2vect_mad_sse
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -104,6 +98,7 @@
%define func(x) x:
%define FUNC_SAVE
%define FUNC_RESTORE
%elifidn __OUTPUT_FORMAT__, macho64
%define arg0 rdi
%define arg0.w edi
@ -174,8 +169,8 @@ section .text
align 16
global GF_2VECT_MAD_SSE:function
func(GF_2VECT_MAD_SSE)
global gf_2vect_mad_sse:function
func(gf_2vect_mad_sse)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -264,4 +259,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_2VECT_MAD_SSE, 00, 00, 0203
slversion gf_2vect_mad_sse, 00, 00, 0203

View File

@ -31,12 +31,6 @@
;;; gf_3vect_dot_prod_avx(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_3VECT_DOT_PROD_AVX _gf_3vect_dot_prod_avx
%else
%define GF_3VECT_DOT_PROD_AVX gf_3vect_dot_prod_avx
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -297,9 +291,9 @@ section .text
%endif
align 16
global GF_3VECT_DOT_PROD_AVX:function
func(GF_3VECT_DOT_PROD_AVX)
FUNC_SAVE
global gf_3vect_dot_prod_avx:function
func(gf_3vect_dot_prod_avx)
FUNC_SAVE
SLDR len, len_m
sub len, 16
SSTR len_m, len
@ -418,4 +412,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_3VECT_DOT_PROD_AVX, 02, 04, 0192
slversion gf_3vect_dot_prod_avx, 02, 04, 0192

View File

@ -31,12 +31,6 @@
;;; gf_3vect_dot_prod_avx2(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_3VECT_DOT_PROD_AVX2 _gf_3vect_dot_prod_avx2
%else
%define GF_3VECT_DOT_PROD_AVX2 gf_3vect_dot_prod_avx2
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -305,8 +299,8 @@ section .text
%endif
align 16
global GF_3VECT_DOT_PROD_AVX2:function
func(GF_3VECT_DOT_PROD_AVX2)
global gf_3vect_dot_prod_avx2:function
func(gf_3vect_dot_prod_avx2)
FUNC_SAVE
SLDR len, len_m
sub len, 32
@ -438,4 +432,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_3VECT_DOT_PROD_AVX2, 04, 04, 0197
slversion gf_3vect_dot_prod_avx2, 04, 04, 0197

View File

@ -31,12 +31,6 @@
;;; gf_3vect_dot_prod_sse(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_3VECT_DOT_PROD_SSE _gf_3vect_dot_prod_sse
%else
%define GF_3VECT_DOT_PROD_SSE gf_3vect_dot_prod_sse
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -297,8 +291,8 @@ section .text
%endif
align 16
global GF_3VECT_DOT_PROD_SSE:function
func(GF_3VECT_DOT_PROD_SSE)
global gf_3vect_dot_prod_sse:function
func(gf_3vect_dot_prod_sse)
FUNC_SAVE
SLDR len, len_m
sub len, 16
@ -419,4 +413,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_3VECT_DOT_PROD_SSE, 00, 05, 0063
slversion gf_3vect_dot_prod_sse, 00, 05, 0063

View File

@ -31,12 +31,6 @@
;;; gf_3vect_mad_avx(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_3VECT_MAD_AVX _gf_3vect_mad_avx
%else
%define GF_3VECT_MAD_AVX gf_3vect_mad_avx
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -104,6 +98,7 @@
%define func(x) x:
%define FUNC_SAVE
%define FUNC_RESTORE
%elifidn __OUTPUT_FORMAT__, macho64
%define arg0 rdi
%define arg0.w edi
@ -177,8 +172,8 @@ section .text
%define xd3 xtmph1
align 16
global GF_3VECT_MAD_AVX:function
func(GF_3VECT_MAD_AVX)
global gf_3vect_mad_avx:function
func(gf_3vect_mad_avx)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -312,4 +307,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_3VECT_MAD_AVX, 02, 00, 0207
slversion gf_3vect_mad_avx, 02, 00, 0207

View File

@ -31,12 +31,6 @@
;;; gf_3vect_mad_avx2(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_3VECT_MAD_AVX2 _gf_3vect_mad_avx2
%else
%define GF_3VECT_MAD_AVX2 gf_3vect_mad_avx2
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -110,6 +104,7 @@
%define func(x) x:
%define FUNC_SAVE
%define FUNC_RESTORE
%elifidn __OUTPUT_FORMAT__, macho64
%define arg0 rdi
%define arg0.w edi
@ -187,8 +182,8 @@ section .text
%define xd3 ymm10
align 16
global GF_3VECT_MAD_AVX2:function
func(GF_3VECT_MAD_AVX2)
global gf_3vect_mad_avx2:function
func(gf_3vect_mad_avx2)
FUNC_SAVE
sub len, 32
jl .return_fail
@ -344,4 +339,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_3VECT_MAD_AVX2, 04, 00, 0208
slversion gf_3vect_mad_avx2, 04, 00, 0208

View File

@ -31,12 +31,6 @@
;;; gf_3vect_mad_sse(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_3VECT_MAD_SSE _gf_3vect_mad_sse
%else
%define GF_3VECT_MAD_SSE gf_3vect_mad_sse
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -104,6 +98,7 @@
%define func(x) x:
%define FUNC_SAVE
%define FUNC_RESTORE
%elifidn __OUTPUT_FORMAT__, macho64
%define arg0 rdi
%define arg0.w edi
@ -176,8 +171,8 @@ section .text
%define xd3 xtmph1
align 16
global GF_3VECT_MAD_SSE:function
func(GF_3VECT_MAD_SSE)
global gf_3vect_mad_sse:function
func(gf_3vect_mad_sse)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -323,4 +318,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_3VECT_MAD_SSE, 00, 00, 0206
slversion gf_3vect_mad_sse, 00, 00, 0206

View File

@ -31,12 +31,6 @@
;;; gf_4vect_dot_prod_avx(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_4VECT_DOT_PROD_AVX _gf_4vect_dot_prod_avx
%else
%define GF_4VECT_DOT_PROD_AVX gf_4vect_dot_prod_avx
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -334,8 +328,8 @@ section .text
%define xp4 xmm5
%endif
align 16
global GF_4VECT_DOT_PROD_AVX:function
func(GF_4VECT_DOT_PROD_AVX)
global gf_4vect_dot_prod_avx:function
func(gf_4vect_dot_prod_avx)
FUNC_SAVE
SLDR len, len_m
sub len, 16
@ -486,4 +480,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_4VECT_DOT_PROD_AVX, 02, 04, 0193
slversion gf_4vect_dot_prod_avx, 02, 04, 0193

View File

@ -31,12 +31,6 @@
;;; gf_4vect_dot_prod_avx2(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_4VECT_DOT_PROD_AVX2 _gf_4vect_dot_prod_avx2
%else
%define GF_4VECT_DOT_PROD_AVX2 gf_4vect_dot_prod_avx2
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -344,8 +338,8 @@ section .text
%define xp4 ymm5
%endif
align 16
global GF_4VECT_DOT_PROD_AVX2:function
func(GF_4VECT_DOT_PROD_AVX2)
global gf_4vect_dot_prod_avx2:function
func(gf_4vect_dot_prod_avx2)
FUNC_SAVE
SLDR len, len_m
sub len, 32
@ -507,4 +501,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_4VECT_DOT_PROD_AVX2, 04, 04, 0198
slversion gf_4vect_dot_prod_avx2, 04, 04, 0198

View File

@ -31,12 +31,6 @@
;;; gf_4vect_dot_prod_sse(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_4VECT_DOT_PROD_SSE _gf_4vect_dot_prod_sse
%else
%define GF_4VECT_DOT_PROD_SSE gf_4vect_dot_prod_sse
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -334,8 +328,8 @@ section .text
%define xp4 xmm5
%endif
align 16
global GF_4VECT_DOT_PROD_SSE:function
func(GF_4VECT_DOT_PROD_SSE)
global gf_4vect_dot_prod_sse:function
func(gf_4vect_dot_prod_sse)
FUNC_SAVE
SLDR len, len_m
sub len, 16
@ -488,4 +482,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_4VECT_DOT_PROD_SSE, 00, 05, 0064
slversion gf_4vect_dot_prod_sse, 00, 05, 0064

View File

@ -31,12 +31,6 @@
;;; gf_4vect_mad_avx(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_4VECT_MAD_AVX _gf_4vect_mad_avx
%else
%define GF_4VECT_MAD_AVX gf_4vect_mad_avx
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -114,6 +108,7 @@
%macro FUNC_RESTORE 0
pop r12
%endmacro
%elifidn __OUTPUT_FORMAT__, macho64
%define arg0 rdi
%define arg0.w edi
@ -194,8 +189,8 @@ section .text
%define xd4 xtmpl1
align 16
global GF_4VECT_MAD_AVX:function
func(GF_4VECT_MAD_AVX)
global gf_4vect_mad_avx:function
func(gf_4vect_mad_avx)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -367,4 +362,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_4VECT_MAD_AVX, 02, 00, 020a
slversion gf_4vect_mad_avx, 02, 00, 020a

View File

@ -30,13 +30,7 @@
;;;
;;; gf_4vect_mad_avx2(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_4VECT_MAD_AVX2 _gf_4vect_mad_avx2
%else
%define GF_4VECT_MAD_AVX2 gf_4vect_mad_avx2
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -108,6 +102,7 @@
%define func(x) x:
%define FUNC_SAVE
%define FUNC_RESTORE
%elifidn __OUTPUT_FORMAT__, macho64
%define arg0 rdi
%define arg0.w edi
@ -186,8 +181,8 @@ section .text
%define xd4 ymm10
align 16
global GF_4VECT_MAD_AVX2:function
func(GF_4VECT_MAD_AVX2)
global gf_4vect_mad_avx2:function
func(gf_4vect_mad_avx2)
FUNC_SAVE
sub len, 32
jl .return_fail
@ -368,4 +363,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_4VECT_MAD_AVX2, 04, 00, 020b
slversion gf_4vect_mad_avx2, 04, 00, 020b

View File

@ -31,12 +31,6 @@
;;; gf_4vect_mad_sse(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_4VECT_MAD_SSE _gf_4vect_mad_sse
%else
%define GF_4VECT_MAD_SSE gf_4vect_mad_sse
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -114,6 +108,7 @@
%macro FUNC_RESTORE 0
pop r12
%endmacro
%elifidn __OUTPUT_FORMAT__, macho64
%define arg0 rdi
%define arg0.w edi
@ -193,8 +188,8 @@ section .text
%define xd4 xtmpl1
align 16
global GF_4VECT_MAD_SSE:function
func(GF_4VECT_MAD_SSE)
global gf_4vect_mad_sse:function
func(gf_4vect_mad_sse)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -372,4 +367,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_4VECT_MAD_SSE, 00, 00, 0209
slversion gf_4vect_mad_sse, 00, 00, 0209

View File

@ -31,12 +31,6 @@
;;; gf_5vect_dot_prod_avx(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_5VECT_DOT_PROD_AVX _gf_5vect_dot_prod_avx
%else
%define GF_5VECT_DOT_PROD_AVX gf_5vect_dot_prod_avx
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -221,8 +215,8 @@ section .text
%define xp5 xmm6
align 16
global GF_5VECT_DOT_PROD_AVX:function
func(GF_5VECT_DOT_PROD_AVX)
global gf_5vect_dot_prod_avx:function
func(gf_5vect_dot_prod_avx)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -345,4 +339,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_5VECT_DOT_PROD_AVX, 02, 03, 0194
slversion gf_5vect_dot_prod_avx, 02, 03, 0194

View File

@ -31,12 +31,6 @@
;;; gf_5vect_dot_prod_avx2(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_5VECT_DOT_PROD_AVX2 _gf_5vect_dot_prod_avx2
%else
%define GF_5VECT_DOT_PROD_AVX2 gf_5vect_dot_prod_avx2
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -228,8 +222,8 @@ section .text
%define xp5 ymm6
align 16
global GF_5VECT_DOT_PROD_AVX2:function
func(GF_5VECT_DOT_PROD_AVX2)
global gf_5vect_dot_prod_avx2:function
func(gf_5vect_dot_prod_avx2)
FUNC_SAVE
sub len, 32
jl .return_fail
@ -359,4 +353,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_5VECT_DOT_PROD_AVX2, 04, 03, 0199
slversion gf_5vect_dot_prod_avx2, 04, 03, 0199

View File

@ -31,12 +31,6 @@
;;; gf_5vect_dot_prod_sse(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_5VECT_DOT_PROD_SSE _gf_5vect_dot_prod_sse
%else
%define GF_5VECT_DOT_PROD_SSE gf_5vect_dot_prod_sse
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -221,8 +215,8 @@ section .text
%define xp5 xmm14
align 16
global GF_5VECT_DOT_PROD_SSE:function
func(GF_5VECT_DOT_PROD_SSE)
global gf_5vect_dot_prod_sse:function
func(gf_5vect_dot_prod_sse)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -346,4 +340,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_5VECT_DOT_PROD_SSE, 00, 04, 0065
slversion gf_5vect_dot_prod_sse, 00, 04, 0065

View File

@ -31,12 +31,6 @@
;;; gf_5vect_mad_avx(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_5VECT_MAD_AVX _gf_5vect_mad_avx
%else
%define GF_5VECT_MAD_AVX gf_5vect_mad_avx
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -206,8 +200,8 @@ section .text
align 16
global GF_5VECT_MAD_AVX:function
func(GF_5VECT_MAD_AVX)
global gf_5vect_mad_avx:function
func(gf_5vect_mad_avx)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -398,4 +392,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_5VECT_MAD_AVX, 02, 00, 020d
slversion gf_5vect_mad_avx, 02, 00, 020d

View File

@ -31,12 +31,6 @@
;;; gf_5vect_mad_avx2(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_5VECT_MAD_AVX2 _gf_5vect_mad_avx2
%else
%define GF_5VECT_MAD_AVX2 gf_5vect_mad_avx2
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -188,8 +182,8 @@ section .text
%define xd5 ymm9
align 16
global GF_5VECT_MAD_AVX2:function
func(GF_5VECT_MAD_AVX2)
global gf_5vect_mad_avx2:function
func(gf_5vect_mad_avx2)
FUNC_SAVE
sub len, 32
jl .return_fail
@ -390,4 +384,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_5VECT_MAD_AVX2, 04, 00, 020e
slversion gf_5vect_mad_avx2, 04, 00, 020e

View File

@ -31,12 +31,6 @@
;;; gf_5vect_mad_sse(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_5VECT_MAD_SSE _gf_5vect_mad_sse
%else
%define GF_5VECT_MAD_SSE gf_5vect_mad_sse
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -205,8 +199,8 @@ section .text
align 16
global GF_5VECT_MAD_SSE:function
func(GF_5VECT_MAD_SSE)
global gf_5vect_mad_sse:function
func(gf_5vect_mad_sse)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -406,4 +400,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_5VECT_MAD_SSE, 00, 00, 020c
slversion gf_5vect_mad_sse, 00, 00, 020c

View File

@ -31,12 +31,6 @@
;;; gf_6vect_dot_prod_avx(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_6VECT_DOT_PROD_AVX _gf_6vect_dot_prod_avx
%else
%define GF_6VECT_DOT_PROD_AVX gf_6vect_dot_prod_avx
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -219,8 +213,8 @@ section .text
%define xp6 xmm7
align 16
global GF_6VECT_DOT_PROD_AVX:function
func(GF_6VECT_DOT_PROD_AVX)
global gf_6vect_dot_prod_avx:function
func(gf_6vect_dot_prod_avx)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -357,4 +351,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_6VECT_DOT_PROD_AVX, 02, 03, 0195
slversion gf_6vect_dot_prod_avx, 02, 03, 0195

View File

@ -31,12 +31,6 @@
;;; gf_6vect_dot_prod_avx2(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_6VECT_DOT_PROD_AVX2 _gf_6vect_dot_prod_avx2
%else
%define GF_6VECT_DOT_PROD_AVX2 gf_6vect_dot_prod_avx2
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -226,8 +220,8 @@ section .text
%define xp6 ymm7
align 16
global GF_6VECT_DOT_PROD_AVX2:function
func(GF_6VECT_DOT_PROD_AVX2)
global gf_6vect_dot_prod_avx2:function
func(gf_6vect_dot_prod_avx2)
FUNC_SAVE
sub len, 32
jl .return_fail
@ -370,4 +364,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_6VECT_DOT_PROD_AVX2, 04, 03, 019a
slversion gf_6vect_dot_prod_avx2, 04, 03, 019a

View File

@ -31,12 +31,6 @@
;;; gf_6vect_dot_prod_sse(len, vec, *g_tbls, **buffs, **dests);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_6VECT_DOT_PROD_SSE _gf_6vect_dot_prod_sse
%else
%define GF_6VECT_DOT_PROD_SSE gf_6vect_dot_prod_sse
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -219,8 +213,8 @@ section .text
%define xp6 xmm13
align 16
global GF_6VECT_DOT_PROD_SSE:function
func(GF_6VECT_DOT_PROD_SSE)
global gf_6vect_dot_prod_sse:function
func(gf_6vect_dot_prod_sse)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -357,4 +351,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_6VECT_DOT_PROD_SSE, 00, 04, 0066
slversion gf_6vect_dot_prod_sse, 00, 04, 0066

View File

@ -31,12 +31,6 @@
;;; gf_6vect_mad_avx(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_6VECT_MAD_AVX _gf_6vect_mad_avx
%else
%define GF_6VECT_MAD_AVX gf_6vect_mad_avx
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -215,8 +209,8 @@ section .text
align 16
global GF_6VECT_MAD_AVX:function
func(GF_6VECT_MAD_AVX)
global gf_6vect_mad_avx:function
func(gf_6vect_mad_avx)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -430,4 +424,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_6VECT_MAD_AVX, 02, 00, 0210
slversion gf_6vect_mad_avx, 02, 00, 0210

View File

@ -31,12 +31,6 @@
;;; gf_6vect_mad_avx2(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_6VECT_MAD_AVX2 _gf_6vect_mad_avx2
%else
%define GF_6VECT_MAD_AVX2 gf_6vect_mad_avx2
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -204,8 +198,8 @@ section .text
%define xd6 xd1
align 16
global GF_6VECT_MAD_AVX2:function
func(GF_6VECT_MAD_AVX2)
global gf_6vect_mad_avx2:function
func(gf_6vect_mad_avx2)
FUNC_SAVE
sub len, 32
jl .return_fail
@ -432,4 +426,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_6VECT_MAD_AVX2, 04, 00, 0211
slversion gf_6vect_mad_avx2, 04, 00, 0211

View File

@ -31,12 +31,6 @@
;;; gf_6vect_mad_sse(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_6VECT_MAD_SSE _gf_6vect_mad_sse
%else
%define GF_6VECT_MAD_SSE gf_6vect_mad_sse
%endif
%define PS 8
%ifidn __OUTPUT_FORMAT__, win64
@ -217,8 +211,8 @@ section .text
align 16
global GF_6VECT_MAD_SSE:function
func(GF_6VECT_MAD_SSE)
global gf_6vect_mad_sse:function
func(gf_6vect_mad_sse)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -443,4 +437,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_6VECT_MAD_SSE, 00, 00, 020f
slversion gf_6vect_mad_sse, 00, 00, 020f

View File

@ -31,12 +31,6 @@
;;; gf_vect_dot_prod_avx(len, vec, *g_tbls, **buffs, *dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_VECT_DOT_PROD_AVX _gf_vect_dot_prod_avx
%else
%define GF_VECT_DOT_PROD_AVX gf_vect_dot_prod_avx
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -218,8 +212,8 @@ section .text
%define xp xmm2
align 16
global GF_VECT_DOT_PROD_AVX:function
func(GF_VECT_DOT_PROD_AVX)
global gf_vect_dot_prod_avx:function
func(gf_vect_dot_prod_avx)
FUNC_SAVE
SLDR len, len_m
sub len, 16
@ -300,4 +294,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_VECT_DOT_PROD_AVX, 02, 04, 0061
slversion gf_vect_dot_prod_avx, 02, 04, 0061

View File

@ -31,12 +31,6 @@
;;; gf_vect_dot_prod_avx2(len, vec, *g_tbls, **buffs, *dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_VECT_DOT_PROD_AVX2 _gf_vect_dot_prod_avx2
%else
%define GF_VECT_DOT_PROD_AVX2 gf_vect_dot_prod_avx2
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -229,8 +223,8 @@ section .text
%define xp ymm2
align 16
global GF_VECT_DOT_PROD_AVX2:function
func(GF_VECT_DOT_PROD_AVX2)
global gf_vect_dot_prod_avx2:function
func(gf_vect_dot_prod_avx2)
FUNC_SAVE
SLDR len, len_m
sub len, 32
@ -312,4 +306,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_VECT_DOT_PROD_AVX2, 04, 04, 0190
slversion gf_vect_dot_prod_avx2, 04, 04, 0190

View File

@ -31,12 +31,6 @@
;;; gf_vect_dot_prod_sse(len, vec, *g_tbls, **buffs, *dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_VECT_DOT_PROD_SSE _gf_vect_dot_prod_sse
%else
%define GF_VECT_DOT_PROD_SSE gf_vect_dot_prod_sse
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -218,8 +212,8 @@ section .text
%define xp xmm2
align 16
global GF_VECT_DOT_PROD_SSE:function
func(GF_VECT_DOT_PROD_SSE)
global gf_vect_dot_prod_sse:function
func(gf_vect_dot_prod_sse)
FUNC_SAVE
SLDR len, len_m
sub len, 16
@ -300,4 +294,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_VECT_DOT_PROD_SSE, 00, 04, 0060
slversion gf_vect_dot_prod_sse, 00, 04, 0060

View File

@ -31,12 +31,6 @@
;;; gf_vect_mad_avx(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_VECT_MAD_AVX _gf_vect_mad_avx
%else
%define GF_VECT_MAD_AVX gf_vect_mad_avx
%endif
%ifidn __OUTPUT_FORMAT__, win64
%define arg0 rcx
%define arg0.w ecx
@ -89,6 +83,7 @@
%define func(x) x:
%define FUNC_SAVE
%define FUNC_RESTORE
%elifidn __OUTPUT_FORMAT__, macho64
%define arg0 rdi
%define arg0.w edi
@ -150,8 +145,8 @@ section .text
%define xtmpd xmm5
align 16
global GF_VECT_MAD_AVX:function
func(GF_VECT_MAD_AVX)
global gf_vect_mad_avx:function
func(gf_vect_mad_avx)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -220,4 +215,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_VECT_MAD_AVX, 02, 00, 0201
slversion gf_vect_mad_avx, 02, 00, 0201

View File

@ -31,12 +31,6 @@
;;; gf_vect_mad_avx2(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_VECT_MAD_AVX2 _gf_vect_mad_avx2
%else
%define GF_VECT_MAD_AVX2 gf_vect_mad_avx2
%endif
%ifidn __OUTPUT_FORMAT__, win64
%define arg0 rcx
%define arg0.w ecx
@ -95,6 +89,7 @@
%define func(x) x:
%define FUNC_SAVE
%define FUNC_RESTORE
%elifidn __OUTPUT_FORMAT__, macho64
%define arg0 rdi
%define arg0.w edi
@ -161,8 +156,8 @@ section .text
%define xtmpd ymm5
align 16
global GF_VECT_MAD_AVX2:function
func(GF_VECT_MAD_AVX2)
global gf_vect_mad_avx2:function
func(gf_vect_mad_avx2)
FUNC_SAVE
sub len, 32
jl .return_fail
@ -230,4 +225,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_VECT_MAD_AVX2, 04, 00, 0202
slversion gf_vect_mad_avx2, 04, 00, 0202

View File

@ -31,12 +31,6 @@
;;; gf_vect_mad_sse(len, vec, vec_i, mul_array, src, dest);
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_VECT_MAD_SSE _gf_vect_mad_sse
%else
%define GF_VECT_MAD_SSE gf_vect_mad_sse
%endif
%ifidn __OUTPUT_FORMAT__, win64
%define arg0 rcx
%define arg0.w ecx
@ -89,6 +83,7 @@
%define func(x) x:
%define FUNC_SAVE
%define FUNC_RESTORE
%elifidn __OUTPUT_FORMAT__, macho64
%define arg0 rdi
%define arg0.w edi
@ -150,8 +145,8 @@ section .text
align 16
global GF_VECT_MAD_SSE:function
func(GF_VECT_MAD_SSE)
global gf_vect_mad_sse:function
func(gf_vect_mad_sse)
FUNC_SAVE
sub len, 16
jl .return_fail
@ -221,4 +216,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_VECT_MAD_SSE, 00, 00, 0200
slversion gf_vect_mad_sse, 00, 00, 0200

View File

@ -31,12 +31,6 @@
;;; gf_vect_mul_avx(len, mul_array, src, dest)
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_VECT_MUL_AVX _gf_vect_mul_avx
%else
%define GF_VECT_MUL_AVX gf_vect_mul_avx
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -128,8 +122,8 @@ section .text
%define xtmp2c xmm7
align 16
global GF_VECT_MUL_AVX:function
func(GF_VECT_MUL_AVX)
global gf_vect_mul_avx:function
func(gf_vect_mul_avx)
FUNC_SAVE
mov pos, 0
vmovdqa xmask0f, [mask0f] ;Load mask of lower nibble in each byte
@ -186,4 +180,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_VECT_MUL_AVX, 01, 02, 0036
slversion gf_vect_mul_avx, 01, 02, 0036

View File

@ -31,12 +31,6 @@
;;; gf_vect_mul_sse(len, mul_array, src, dest)
;;;
%ifidn __OUTPUT_FORMAT__, macho64
%define GF_VECT_MUL_SSE _gf_vect_mul_sse
%else
%define GF_VECT_MUL_SSE gf_vect_mul_sse
%endif
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
@ -129,8 +123,8 @@ section .text
align 16
global GF_VECT_MUL_SSE:function
func(GF_VECT_MUL_SSE)
global gf_vect_mul_sse:function
func(gf_vect_mul_sse)
FUNC_SAVE
mov pos, 0
movdqa xmask0f, [mask0f] ;Load mask of lower nibble in each byte
@ -192,4 +186,4 @@ global %1_slver
db 0x%3, 0x%2
%endmacro
;;; func core, ver, snum
slversion GF_VECT_MUL_SSE, 00, 02, 0034
slversion gf_vect_mul_sse, 00, 02, 0034