erasure_code: add AVX2 2vect mad with GFNI implementation

Signed-off-by: Marcel Cornu <marcel.d.cornu@intel.com>
This commit is contained in:
Marcel Cornu 2023-12-08 12:08:34 +00:00 committed by Pablo de Lara
parent 0052080f53
commit a0a149d674
4 changed files with 310 additions and 2 deletions

View File

@ -85,6 +85,8 @@ objs = \
bin\gf_5vect_mad_avx2.obj \
bin\gf_6vect_mad_avx2.obj \
bin\ec_multibinary.obj \
bin\gf_vect_mad_avx2_gfni.obj \
bin\gf_2vect_mad_avx2_gfni.obj \
bin\gf_vect_dot_prod_avx512.obj \
bin\gf_2vect_dot_prod_avx512.obj \
bin\gf_3vect_dot_prod_avx512.obj \
@ -107,7 +109,6 @@ objs = \
bin\gf_5vect_mad_avx512.obj \
bin\gf_6vect_mad_avx512.obj \
bin\gf_vect_mad_avx512_gfni.obj \
bin\gf_vect_mad_avx2_gfni.obj \
bin\gf_2vect_mad_avx512_gfni.obj \
bin\gf_3vect_mad_avx512_gfni.obj \
bin\gf_4vect_mad_avx512_gfni.obj \

View File

@ -78,6 +78,8 @@ lsrc_x86_64 += \
#if HAVE_AVX512
lsrc_x86_64 += \
erasure_code/gf_vect_mad_avx2_gfni.asm \
erasure_code/gf_2vect_mad_avx2_gfni.asm \
erasure_code/gf_vect_dot_prod_avx512.asm \
erasure_code/gf_2vect_dot_prod_avx512.asm \
erasure_code/gf_3vect_dot_prod_avx512.asm \
@ -100,7 +102,6 @@ lsrc_x86_64 += \
erasure_code/gf_5vect_mad_avx512.asm \
erasure_code/gf_6vect_mad_avx512.asm \
erasure_code/gf_vect_mad_avx512_gfni.asm \
erasure_code/gf_vect_mad_avx2_gfni.asm \
erasure_code/gf_2vect_mad_avx512_gfni.asm \
erasure_code/gf_3vect_mad_avx512_gfni.asm \
erasure_code/gf_4vect_mad_avx512_gfni.asm \

View File

@ -275,6 +275,8 @@ extern void gf_3vect_dot_prod_avx2_gfni(int len, int k, unsigned char *g_tbls,
unsigned char **data, unsigned char **coding);
extern void gf_vect_mad_avx2_gfni(int len, int vec, int vec_i, unsigned char *gftbls,
unsigned char *src, unsigned char *dest);
extern void gf_2vect_mad_avx2_gfni(int len, int vec, int vec_i, unsigned char *gftbls,
unsigned char *src, unsigned char **dest);
void ec_init_tables_gfni(int k, int rows, unsigned char *a, unsigned char *g_tbls)
{
@ -378,6 +380,12 @@ void ec_encode_data_update_avx2_gfni(int len, int k, int rows, int vec_i,
unsigned char *g_tbls, unsigned char *data,
unsigned char **coding)
{
while (rows >= 2) {
gf_2vect_mad_avx2_gfni(len, k, vec_i, g_tbls, data, coding);
g_tbls += 2 * k * 8;
coding += 2;
rows -= 2;
}
while (rows) {
gf_vect_mad_avx2_gfni(len, k, vec_i, g_tbls, data, *coding);
g_tbls += k * 8;

View File

@ -0,0 +1,298 @@
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; Copyright(c) 2023 Intel Corporation All rights reserved.
;
; Redistribution and use in source and binary forms, with or without
; modification, are permitted provided that the following conditions
; are met:
; * Redistributions of source code must retain the above copyright
; notice, this list of conditions and the following disclaimer.
; * Redistributions in binary form must reproduce the above copyright
; notice, this list of conditions and the following disclaimer in
; the documentation and/or other materials provided with the
; distribution.
; * Neither the name of Intel Corporation nor the names of its
; contributors may be used to endorse or promote products derived
; from this software without specific prior written permission.
;
; THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
; "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
; LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
; A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
; OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
; SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
; LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
; DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
; THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
; (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
; OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;
;;; gf_2vect_mad_avx2_gfni(len, vec, vec_i, mul_array, src, dest);
;;;
%include "reg_sizes.asm"
%include "gf_vect_gfni.inc"
%include "memcpy.asm"
%if AS_FEATURE_LEVEL >= 10
%ifidn __OUTPUT_FORMAT__, elf64
%define arg0 rdi
%define arg1 rsi
%define arg2 rdx
%define arg3 rcx
%define arg4 r8
%define arg5 r9
%define tmp r11
%define tmp2 r10
%define func(x) x: endbranch
%define FUNC_SAVE
%define FUNC_RESTORE
%endif
%ifidn __OUTPUT_FORMAT__, win64
%define arg0 rcx
%define arg1 rdx
%define arg2 r8
%define arg3 r9
%define arg4 r12 ; must be saved, loaded and restored
%define arg5 r13 ; must be saved and restored
%define tmp r11
%define tmp2 r10
%define stack_size 16*9 + 3*8
%define arg(x) [rsp + stack_size + 8 + 8*x]
%define func(x) proc_frame x
%macro FUNC_SAVE 0
sub rsp, stack_size
vmovdqa [rsp + 0*16], xmm6
vmovdqa [rsp + 1*16], xmm7
vmovdqa [rsp + 2*16], xmm8
vmovdqa [rsp + 3*16], xmm9
vmovdqa [rsp + 4*16], xmm10
vmovdqa [rsp + 5*16], xmm11
vmovdqa [rsp + 6*16], xmm12
vmovdqa [rsp + 7*16], xmm13
vmovdqa [rsp + 8*16], xmm14
mov [rsp + 9*16 + 0*8], r12
mov [rsp + 9*16 + 1*8], r13
end_prolog
mov arg4, arg(4)
mov arg5, arg(5)
%endmacro
%macro FUNC_RESTORE 0
vmovdqa xmm6, [rsp + 0*16]
vmovdqa xmm7, [rsp + 1*16]
vmovdqa xmm8, [rsp + 2*16]
vmovdqa xmm9, [rsp + 3*16]
vmovdqa xmm10, [rsp + 4*16]
vmovdqa xmm11, [rsp + 5*16]
vmovdqa xmm12, [rsp + 6*16]
vmovdqa xmm13, [rsp + 7*16]
vmovdqa xmm14, [rsp + 8*16]
mov r12, [rsp + 9*16 + 0*8]
mov r13, [rsp + 9*16 + 1*8]
add rsp, stack_size
%endmacro
%endif
%define len arg0
%define vec arg1
%define vec_i arg2
%define mul_array arg3
%define src arg4
%define dest1 arg5
%define pos rax
%define dest2 mul_array
%define dest3 vec_i
%ifndef EC_ALIGNED_ADDR
;;; Use Un-aligned load/store
%define XLDR vmovdqu
%define XSTR vmovdqu
%else
;;; Use Non-temporal load/stor
%ifdef NO_NT_LDST
%define XLDR vmovdqa
%define XSTR vmovdqa
%else
%define XLDR vmovntdqa
%define XSTR vmovntdq
%endif
%endif
default rel
[bits 64]
section .text
%define x0l ymm0
%define x0h ymm1
%define x0x ymm2
%define xgft1 ymm3
%define xgft2 ymm4
%define xd1l ymm5
%define xd1h ymm6
%define xd1x ymm7
%define xd2l ymm8
%define xd2h ymm9
%define xd2x ymm10
%define xret1l ymm11
%define xret1h ymm12
%define xret2l ymm13
%define xret2h ymm14
%define x0 x0l
%define xd1 xd1l
%define xd2 xd2l
%define xret1 xret1l
%define xret2 xret2l
;;
;; Encodes 96 bytes of a single source into 2x 96 bytes (parity disks)
;;
%macro ENCODE_96B_2 0
;Get next source vector
XLDR x0l, [src + pos]
XLDR x0h, [src + pos + 32]
XLDR x0x, [src + pos + 64]
;Get next dest vectors
XLDR xd1l, [dest1 + pos]
XLDR xd1h, [dest1 + pos + 32]
XLDR xd1x, [dest1 + pos + 64]
XLDR xd2l, [dest2 + pos]
XLDR xd2h, [dest2 + pos + 32]
XLDR xd2x, [dest2 + pos + 64]
GF_MUL_XOR VEX, x0l, xgft1, xret1l, xd1l, xgft2, xret2l, xd2l
GF_MUL_XOR VEX, x0h, xgft1, xret1h, xd1h, xgft2, xret2h, xd2h
GF_MUL_XOR VEX, x0x, xgft1, xret1l, xd1x, xgft2, xret2l, xd2x
XSTR [dest1 + pos], xd1l
XSTR [dest1 + pos + 32], xd1h
XSTR [dest1 + pos + 64], xd1x
XSTR [dest2 + pos], xd2l
XSTR [dest2 + pos + 32], xd2h
XSTR [dest2 + pos + 64], xd2x
%endmacro
;;
;; Encodes 64 bytes of a single source into 2x 64 bytes (parity disks)
;;
%macro ENCODE_64B_2 0
;Get next source vector
XLDR x0l, [src + pos]
XLDR x0h, [src + pos + 32]
;Get next dest vectors
XLDR xd1l, [dest1 + pos]
XLDR xd1h, [dest1 + pos + 32]
XLDR xd2l, [dest2 + pos]
XLDR xd2h, [dest2 + pos + 32]
GF_MUL_XOR VEX, x0l, xgft1, xret1l, xd1l, xgft2, xret2l, xd2l
GF_MUL_XOR VEX, x0h, xgft1, xret1h, xd1h, xgft2, xret2h, xd2h
XSTR [dest1 + pos], xd1l
XSTR [dest1 + pos + 32], xd1h
XSTR [dest2 + pos], xd2l
XSTR [dest2 + pos + 32], xd2h
%endmacro
;;
;; Encodes 32 bytes of a single source into 2x 32 bytes (parity disks)
;;
%macro ENCODE_32B_2 0
;Get next source vector
XLDR x0, [src + pos]
;Get next dest vectors
XLDR xd1, [dest1 + pos]
XLDR xd2, [dest2 + pos]
GF_MUL_XOR VEX, x0, xgft1, xret1, xd1, xgft2, xret2, xd2
XSTR [dest1 + pos], xd1
XSTR [dest2 + pos], xd2
%endmacro
;;
;; Encodes less than 32 bytes of a single source into 2x parity disks
;;
%macro ENCODE_LT_32B_2 1
%define %%LEN %1
;Get next source vector
simd_load_avx2 x0, src + pos, %%LEN, tmp, tmp2
;Get next dest vectors
simd_load_avx2 xd1, dest1 + pos, %%LEN, tmp, tmp2
simd_load_avx2 xd2, dest2 + pos, %%LEN, tmp, tmp2
GF_MUL_XOR VEX, x0, xgft1, xret1, xd1, xgft2, xret2, xd2
lea dest1, [dest1 + pos]
simd_store_avx2 dest1, xd1, %%LEN, tmp, tmp2
lea dest2, [dest2 + pos]
simd_store_avx2 dest2, xd2, %%LEN, tmp, tmp2
%endmacro
align 16
mk_global gf_2vect_mad_avx2_gfni, function
func(gf_2vect_mad_avx2_gfni)
FUNC_SAVE
xor pos, pos
shl vec_i, 3 ;Multiply by 8
shl vec, 3 ;Multiply by 8
lea tmp, [mul_array + vec_i]
vbroadcastsd xgft1, [tmp]
vbroadcastsd xgft2, [tmp + vec]
mov dest2, [dest1 + 8] ; reuse mul_array
mov dest1, [dest1]
cmp len, 96
jl .len_lt_96
.loop96:
ENCODE_96B_2
add pos, 96 ;; loop on 96 bytes at a time
sub len, 96
cmp len, 96
jge .loop96
.len_lt_96:
cmp len, 64
jl .len_lt_64
ENCODE_64B_2 ;; encode next 64 bytes
add pos, 64
sub len, 64
.len_lt_64:
cmp len, 32
jl .len_lt_32
ENCODE_32B_2 ;; encode next 32 bytes
add pos, 32
sub len, 32
.len_lt_32:
cmp len, 0
jle .exit
ENCODE_LT_32B_2 len ;; encode final bytes
.exit:
vzeroupper
FUNC_RESTORE
ret
endproc_frame
%endif ; if AS_FEATURE_LEVEL >= 10