197 lines
8.0 KiB
C
197 lines
8.0 KiB
C
/***************************************************************************
|
|
Copyright (c) 2020, The OpenBLAS Project
|
|
All rights reserved.
|
|
Redistribution and use in source and binary forms, with or without
|
|
modification, are permitted provided that the following conditions are
|
|
met:
|
|
1. Redistributions of source code must retain the above copyright
|
|
notice, this list of conditions and the following disclaimer.
|
|
2. 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.
|
|
3. Neither the name of the OpenBLAS project 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 OPENBLAS PROJECT 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.
|
|
*****************************************************************************/
|
|
|
|
#include "common.h"
|
|
|
|
#if !defined(DOUBLE)
|
|
#define RVV_EFLOAT RVV_E32
|
|
#define RVV_M RVV_M4
|
|
#define FLOAT_V_T float32xm4_t
|
|
#define VLEV_FLOAT vlev_float32xm4
|
|
#define VLSEV_FLOAT vlsev_float32xm4
|
|
#define VSEV_FLOAT vsev_float32xm4
|
|
#define VSSEV_FLOAT vssev_float32xm4
|
|
#define VFMACCVF_FLOAT vfmaccvf_float32xm4
|
|
#define VFMULVF_FLOAT vfmulvf_float32xm4
|
|
#define VFMSACVF_FLOAT vfmsacvf_float32xm4
|
|
#else
|
|
#define RVV_EFLOAT RVV_E64
|
|
#define RVV_M RVV_M4
|
|
#define FLOAT_V_T float64xm4_t
|
|
#define VLEV_FLOAT vlev_float64xm4
|
|
#define VLSEV_FLOAT vlsev_float64xm4
|
|
#define VSEV_FLOAT vsev_float64xm4
|
|
#define VSSEV_FLOAT vssev_float64xm4
|
|
#define VFMACCVF_FLOAT vfmaccvf_float64xm4
|
|
#define VFMULVF_FLOAT vfmulvf_float64xm4
|
|
#define VFMSACVF_FLOAT vfmsacvf_float64xm4
|
|
#endif
|
|
|
|
int CNAME(BLASLONG n, FLOAT *x, BLASLONG inc_x, FLOAT *y, BLASLONG inc_y, FLOAT c, FLOAT s)
|
|
{
|
|
BLASLONG i=0, j=0;
|
|
BLASLONG ix=0,iy=0;
|
|
|
|
if(n <= 0) return(0);
|
|
unsigned int gvl = 0;
|
|
FLOAT_V_T v0, v1, vx, vy;
|
|
|
|
if(inc_x == 1 && inc_y == 1){
|
|
gvl = vsetvli(n, RVV_EFLOAT, RVV_M);
|
|
for(i=0,j=0; i<n/gvl; i++){
|
|
vx = VLEV_FLOAT(&x[j], gvl);
|
|
vy = VLEV_FLOAT(&y[j], gvl);
|
|
|
|
v0 = VFMULVF_FLOAT(vx, c, gvl);
|
|
v0 = VFMACCVF_FLOAT(v0, s, vy, gvl);
|
|
VSEV_FLOAT(&x[j], v0, gvl);
|
|
|
|
v1 = VFMULVF_FLOAT(vx, s, gvl);
|
|
v1 = VFMSACVF_FLOAT(v1, c, vy, gvl);
|
|
VSEV_FLOAT(&y[j], v1, gvl);
|
|
|
|
j += gvl;
|
|
}
|
|
if(j<n){
|
|
gvl = vsetvli(n-j, RVV_EFLOAT, RVV_M);
|
|
vx = VLEV_FLOAT(&x[j], gvl);
|
|
vy = VLEV_FLOAT(&y[j], gvl);
|
|
|
|
v0 = VFMULVF_FLOAT(vx, c, gvl);
|
|
v0 = VFMACCVF_FLOAT(v0, s, vy, gvl);
|
|
VSEV_FLOAT(&x[j], v0, gvl);
|
|
|
|
v1 = VFMULVF_FLOAT(vx, s, gvl);
|
|
v1 = VFMSACVF_FLOAT(v1, c, vy, gvl);
|
|
VSEV_FLOAT(&y[j], v1, gvl);
|
|
}
|
|
}else if(inc_y == 1){
|
|
gvl = vsetvli(n, RVV_EFLOAT, RVV_M);
|
|
BLASLONG stride_x = inc_x * sizeof(FLOAT);
|
|
BLASLONG inc_xv = inc_x * gvl;
|
|
for(i=0,j=0; i<n/gvl; i++){
|
|
vx = VLSEV_FLOAT(&x[ix], stride_x, gvl);
|
|
vy = VLEV_FLOAT(&y[j], gvl);
|
|
|
|
v0 = VFMULVF_FLOAT(vx, c, gvl);
|
|
v0 = VFMACCVF_FLOAT(v0, s, vy, gvl);
|
|
VSSEV_FLOAT(&x[ix], stride_x, v0, gvl);
|
|
|
|
v1 = VFMULVF_FLOAT(vx, s, gvl);
|
|
v1 = VFMSACVF_FLOAT(v1, c, vy, gvl);
|
|
VSEV_FLOAT(&y[j], v1, gvl);
|
|
|
|
j += gvl;
|
|
ix += inc_xv;
|
|
}
|
|
if(j<n){
|
|
gvl = vsetvli(n-j, RVV_EFLOAT, RVV_M);
|
|
vx = VLSEV_FLOAT(&x[j*inc_x], stride_x, gvl);
|
|
vy = VLEV_FLOAT(&y[j], gvl);
|
|
|
|
v0 = VFMULVF_FLOAT(vx, c, gvl);
|
|
v0 = VFMACCVF_FLOAT(v0, s, vy, gvl);
|
|
VSSEV_FLOAT(&x[j*inc_x], stride_x, v0, gvl);
|
|
|
|
v1 = VFMULVF_FLOAT(vx, s, gvl);
|
|
v1 = VFMSACVF_FLOAT(v1, c, vy, gvl);
|
|
VSEV_FLOAT(&y[j], v1, gvl);
|
|
}
|
|
}else if(inc_x == 1){
|
|
gvl = vsetvli(n, RVV_EFLOAT, RVV_M);
|
|
BLASLONG stride_y = inc_y * sizeof(FLOAT);
|
|
BLASLONG inc_yv = inc_y * gvl;
|
|
for(i=0,j=0; i<n/gvl; i++){
|
|
vx = VLEV_FLOAT(&x[j], gvl);
|
|
vy = VLSEV_FLOAT(&y[iy], stride_y, gvl);
|
|
|
|
v0 = VFMULVF_FLOAT(vx, c, gvl);
|
|
v0 = VFMACCVF_FLOAT(v0, s, vy, gvl);
|
|
VSEV_FLOAT(&x[j], v0, gvl);
|
|
|
|
v1 = VFMULVF_FLOAT(vx, s, gvl);
|
|
v1 = VFMSACVF_FLOAT(v1, c, vy, gvl);
|
|
VSSEV_FLOAT(&y[iy], stride_y, v1, gvl);
|
|
|
|
j += gvl;
|
|
iy += inc_yv;
|
|
}
|
|
if(j<n){
|
|
gvl = vsetvli(n-j, RVV_EFLOAT, RVV_M);
|
|
vx = VLEV_FLOAT(&x[j], gvl);
|
|
vy = VLSEV_FLOAT(&y[j*inc_y],stride_y, gvl);
|
|
|
|
v0 = VFMULVF_FLOAT(vx, c, gvl);
|
|
v0 = VFMACCVF_FLOAT(v0, s, vy, gvl);
|
|
VSEV_FLOAT(&x[j], v0, gvl);
|
|
|
|
v1 = VFMULVF_FLOAT(vx, s, gvl);
|
|
v1 = VFMSACVF_FLOAT(v1, c, vy, gvl);
|
|
VSSEV_FLOAT(&y[j*inc_y], stride_y, v1, gvl);
|
|
}
|
|
}else{
|
|
gvl = vsetvli(n, RVV_EFLOAT, RVV_M);
|
|
BLASLONG stride_x = inc_x * sizeof(FLOAT);
|
|
BLASLONG stride_y = inc_y * sizeof(FLOAT);
|
|
BLASLONG inc_xv = inc_x * gvl;
|
|
BLASLONG inc_yv = inc_y * gvl;
|
|
for(i=0,j=0; i<n/gvl; i++){
|
|
vx = VLSEV_FLOAT(&x[ix], stride_x, gvl);
|
|
vy = VLSEV_FLOAT(&y[iy], stride_y, gvl);
|
|
|
|
v0 = VFMULVF_FLOAT(vx, c, gvl);
|
|
v0 = VFMACCVF_FLOAT(v0, s, vy, gvl);
|
|
VSSEV_FLOAT(&x[ix], stride_x, v0, gvl);
|
|
|
|
v1 = VFMULVF_FLOAT(vx, s, gvl);
|
|
v1 = VFMSACVF_FLOAT(v1, c, vy, gvl);
|
|
VSSEV_FLOAT(&y[iy], stride_y, v1, gvl);
|
|
|
|
j += gvl;
|
|
ix += inc_xv;
|
|
iy += inc_yv;
|
|
}
|
|
if(j<n){
|
|
gvl = vsetvli(n-j, RVV_EFLOAT, RVV_M);
|
|
vx = VLSEV_FLOAT(&x[j*inc_x],stride_x, gvl);
|
|
vy = VLSEV_FLOAT(&y[j*inc_y],stride_y, gvl);
|
|
|
|
v0 = VFMULVF_FLOAT(vx, c, gvl);
|
|
v0 = VFMACCVF_FLOAT(v0, s, vy, gvl);
|
|
VSSEV_FLOAT(&x[j*inc_x], stride_x, v0, gvl);
|
|
|
|
v1 = VFMULVF_FLOAT(vx, s, gvl);
|
|
v1 = VFMSACVF_FLOAT(v1, c, vy, gvl);
|
|
VSSEV_FLOAT(&y[j*inc_y], stride_y, v1, gvl);
|
|
}
|
|
}
|
|
return(0);
|
|
|
|
}
|
|
|
|
|