1 /***************************************************************************
2 Copyright (c) 2020, The OpenBLAS Project
3 All rights reserved.
4 Redistribution and use in source and binary forms, with or without
5 modification, are permitted provided that the following conditions are
6 met:
7 1. Redistributions of source code must retain the above copyright
8 notice, this list of conditions and the following disclaimer.
9 2. Redistributions in binary form must reproduce the above copyright
10 notice, this list of conditions and the following disclaimer in
11 the documentation and/or other materials provided with the
12 distribution.
13 3. Neither the name of the OpenBLAS project nor the names of
14 its contributors may be used to endorse or promote products
15 derived from this software without specific prior written permission.
16 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
17 AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
18 IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19 ARE DISCLAIMED. IN NO EVENT SHALL THE OPENBLAS PROJECT OR CONTRIBUTORS BE
20 LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
21 DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
22 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
23 CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
24 OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
25 USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
26 *****************************************************************************/
27
28 #include "common.h"
29
30
31 #if defined(__VEC__) || defined(__ALTIVEC__)
32 #include "zaxpy_microk_power10.c"
33 #endif
34
35
36 #ifndef HAVE_KERNEL_4
37
zaxpy_kernel_4(BLASLONG n,FLOAT * x,FLOAT * y,FLOAT da_r,FLOAT da_i)38 static void zaxpy_kernel_4(BLASLONG n, FLOAT *x, FLOAT *y, FLOAT da_r,FLOAT da_i)
39 {
40 BLASLONG register i = 0;
41 BLASLONG register ix = 0;
42
43
44
45 while(i < n)
46 {
47 #if !defined(CONJ)
48 y[ix] += ( da_r * x[ix] - da_i * x[ix+1] ) ;
49 y[ix+1] += ( da_r * x[ix+1] + da_i * x[ix] ) ;
50 y[ix+2] += ( da_r * x[ix+2] - da_i * x[ix+3] ) ;
51 y[ix+3] += ( da_r * x[ix+3] + da_i * x[ix+2] ) ;
52 #else
53 y[ix] += ( da_r * x[ix] + da_i * x[ix+1] ) ;
54 y[ix+1] -= ( da_r * x[ix+1] - da_i * x[ix] ) ;
55 y[ix+2] += ( da_r * x[ix+2] + da_i * x[ix+3] ) ;
56 y[ix+3] -= ( da_r * x[ix+3] - da_i * x[ix+2] ) ;
57 #endif
58
59 ix+=4 ;
60 i+=2 ;
61
62 }
63
64 }
65
66 #endif
67
CNAME(BLASLONG n,BLASLONG dummy0,BLASLONG dummy1,FLOAT da_r,FLOAT da_i,FLOAT * x,BLASLONG inc_x,FLOAT * y,BLASLONG inc_y,FLOAT * dummy,BLASLONG dummy2)68 int CNAME(BLASLONG n, BLASLONG dummy0, BLASLONG dummy1, FLOAT da_r, FLOAT da_i, FLOAT *x, BLASLONG inc_x, FLOAT *y, BLASLONG inc_y, FLOAT *dummy, BLASLONG dummy2)
69 {
70 BLASLONG i=0;
71 BLASLONG ix=0,iy=0;
72
73 if ( n <= 0 ) return(0);
74
75 if ( (inc_x == 1) && (inc_y == 1) )
76 {
77
78 BLASLONG n1 = n & -16;
79
80 if ( n1 )
81 {
82 zaxpy_kernel_4 (n1, x, y, da_r, da_i);
83 ix = 2 * n1;
84 }
85 i = n1;
86 while(i < n)
87 {
88 #if !defined(CONJ)
89 y[ix] += ( da_r * x[ix] - da_i * x[ix+1] ) ;
90 y[ix+1] += ( da_r * x[ix+1] + da_i * x[ix] ) ;
91 #else
92 y[ix] += ( da_r * x[ix] + da_i * x[ix+1] ) ;
93 y[ix+1] -= ( da_r * x[ix+1] - da_i * x[ix] ) ;
94 #endif
95 i++ ;
96 ix += 2;
97
98 }
99 return(0);
100
101
102 }
103
104 inc_x *=2;
105 inc_y *=2;
106
107 while(i < n)
108 {
109
110 #if !defined(CONJ)
111 y[iy] += ( da_r * x[ix] - da_i * x[ix+1] ) ;
112 y[iy+1] += ( da_r * x[ix+1] + da_i * x[ix] ) ;
113 #else
114 y[iy] += ( da_r * x[ix] + da_i * x[ix+1] ) ;
115 y[iy+1] -= ( da_r * x[ix+1] - da_i * x[ix] ) ;
116 #endif
117 ix += inc_x ;
118 iy += inc_y ;
119 i++ ;
120
121 }
122 return(0);
123
124 }
125
126
127