mirror of
https://github.com/intel/isa-l.git
synced 2025-09-07 14:40:53 +02:00
d414b2702a
The ISA-L EC code has been written using RVV vector instructions and the minimum multiplication table, resulting in a performance improvement of over 10 times compared to the existing implementation. Signed-off-by: Shuo Lv <lv.shuo@sanechips.com.cn>
189 lines
7.2 KiB
C
189 lines
7.2 KiB
C
/**************************************************************
|
|
Copyright (c) 2025 sanechips Technologies Co., Ltd.
|
|
|
|
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 sanechips 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.
|
|
**********************************************************************/
|
|
|
|
#if HAVE_RVV
|
|
#include "erasure_code.h"
|
|
|
|
/*external function*/
|
|
|
|
/* RVV */
|
|
extern void
|
|
gf_vect_dot_prod_rvv(int len, int vlen, unsigned char *gftbls, unsigned char **src,
|
|
unsigned char *dest);
|
|
extern void
|
|
gf_2vect_dot_prod_rvv(int len, int vlen, unsigned char *gftbls, unsigned char **src,
|
|
unsigned char **dest);
|
|
extern void
|
|
gf_3vect_dot_prod_rvv(int len, int vlen, unsigned char *gftbls, unsigned char **src,
|
|
unsigned char **dest);
|
|
extern void
|
|
gf_4vect_dot_prod_rvv(int len, int vlen, unsigned char *gftbls, unsigned char **src,
|
|
unsigned char **dest);
|
|
extern void
|
|
gf_5vect_dot_prod_rvv(int len, int vlen, unsigned char *gftbls, unsigned char **src,
|
|
unsigned char **dest);
|
|
extern void
|
|
gf_6vect_dot_prod_rvv(int len, int vlen, unsigned char *gftbls, unsigned char **src,
|
|
unsigned char **dest);
|
|
extern void
|
|
gf_7vect_dot_prod_rvv(int len, int vlen, unsigned char *gftbls, unsigned char **src,
|
|
unsigned char **dest);
|
|
extern void
|
|
gf_8vect_dot_prod_rvv(int len, int vlen, unsigned char *gftbls, unsigned char **src,
|
|
unsigned char **dest);
|
|
extern void
|
|
gf_vect_mad_rvv(int len, int vec, int vec_i, unsigned char *gftbls, unsigned char *src,
|
|
unsigned char *dest);
|
|
extern void
|
|
gf_2vect_mad_rvv(int len, int vec, int vec_i, unsigned char *gftbls, unsigned char *src,
|
|
unsigned char **dest);
|
|
extern void
|
|
gf_3vect_mad_rvv(int len, int vec, int vec_i, unsigned char *gftbls, unsigned char *src,
|
|
unsigned char **dest);
|
|
extern void
|
|
gf_4vect_mad_rvv(int len, int vec, int vec_i, unsigned char *gftbls, unsigned char *src,
|
|
unsigned char **dest);
|
|
extern void
|
|
gf_5vect_mad_rvv(int len, int vec, int vec_i, unsigned char *gftbls, unsigned char *src,
|
|
unsigned char **dest);
|
|
extern void
|
|
gf_6vect_mad_rvv(int len, int vec, int vec_i, unsigned char *gftbls, unsigned char *src,
|
|
unsigned char **dest);
|
|
|
|
void
|
|
ec_encode_data_rvv(int len, int k, int rows, unsigned char *g_tbls, unsigned char **data,
|
|
unsigned char **coding)
|
|
{
|
|
if (len < 16) {
|
|
ec_encode_data_base(len, k, rows, g_tbls, data, coding);
|
|
return;
|
|
}
|
|
|
|
while (rows > 11) {
|
|
gf_6vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
g_tbls += 6 * k * 32;
|
|
coding += 6;
|
|
rows -= 6;
|
|
}
|
|
|
|
switch (rows) {
|
|
case 11:
|
|
/* 7 + 4 */
|
|
gf_7vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
g_tbls += 7 * k * 32;
|
|
coding += 7;
|
|
gf_4vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
break;
|
|
case 10:
|
|
/* 6 + 4 */
|
|
gf_6vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
g_tbls += 6 * k * 32;
|
|
coding += 6;
|
|
gf_4vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
break;
|
|
case 9:
|
|
/* 5 + 4 */
|
|
gf_5vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
g_tbls += 5 * k * 32;
|
|
coding += 5;
|
|
gf_4vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
break;
|
|
case 8:
|
|
/* 4 + 4 */
|
|
gf_4vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
g_tbls += 4 * k * 32;
|
|
coding += 4;
|
|
gf_4vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
break;
|
|
case 7:
|
|
gf_7vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
break;
|
|
case 6:
|
|
gf_6vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
break;
|
|
case 5:
|
|
gf_5vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
break;
|
|
case 4:
|
|
gf_4vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
break;
|
|
case 3:
|
|
gf_3vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
break;
|
|
case 2:
|
|
gf_2vect_dot_prod_rvv(len, k, g_tbls, data, coding);
|
|
break;
|
|
case 1:
|
|
gf_vect_dot_prod_rvv(len, k, g_tbls, data, *coding);
|
|
break;
|
|
default:
|
|
break;
|
|
}
|
|
}
|
|
|
|
void
|
|
ec_encode_data_update_rvv(int len, int k, int rows, int vec_i, unsigned char *g_tbls,
|
|
unsigned char *data, unsigned char **coding)
|
|
{
|
|
if (len < 16) {
|
|
ec_encode_data_update_base(len, k, rows, vec_i, g_tbls, data, coding);
|
|
return;
|
|
}
|
|
while (rows > 6) {
|
|
gf_6vect_mad_rvv(len, k, vec_i, g_tbls, data, coding);
|
|
g_tbls += 6 * k * 32;
|
|
coding += 6;
|
|
rows -= 6;
|
|
}
|
|
switch (rows) {
|
|
case 6:
|
|
gf_6vect_mad_rvv(len, k, vec_i, g_tbls, data, coding);
|
|
break;
|
|
case 5:
|
|
gf_5vect_mad_rvv(len, k, vec_i, g_tbls, data, coding);
|
|
break;
|
|
case 4:
|
|
gf_4vect_mad_rvv(len, k, vec_i, g_tbls, data, coding);
|
|
break;
|
|
case 3:
|
|
gf_3vect_mad_rvv(len, k, vec_i, g_tbls, data, coding);
|
|
break;
|
|
case 2:
|
|
gf_2vect_mad_rvv(len, k, vec_i, g_tbls, data, coding);
|
|
break;
|
|
case 1:
|
|
gf_vect_mad_rvv(len, k, vec_i, g_tbls, data, *coding);
|
|
break;
|
|
default:
|
|
break;
|
|
}
|
|
}
|
|
|
|
#endif
|