Skip to content

Commit 3e8f1c6

Browse files
author
wuanjun 00447568
committed
[OpenBlas]:Add benchmark tpmv.c and modify Makefile
[Description]:Solve the problem of missing tpmv.c benchmark file
1 parent 4e1c4e6 commit 3e8f1c6

2 files changed

Lines changed: 259 additions & 0 deletions

File tree

benchmark/Makefile

Lines changed: 87 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -80,6 +80,7 @@ goto :: slinpack.goto dlinpack.goto clinpack.goto zlinpack.goto \
8080
cher2k.goto zher2k.goto \
8181
sgemv.goto dgemv.goto cgemv.goto zgemv.goto \
8282
strmv.goto dtrmv.goto ctrmv.goto ztrmv.goto \
83+
stpmv.goto dtpmv.goto ctpmv.goto ztpmv.goto \
8384
strsv.goto dtrsv.goto ctrsv.goto ztrsv.goto \
8485
sgeev.goto dgeev.goto cgeev.goto zgeev.goto \
8586
sgesv.goto dgesv.goto cgesv.goto zgesv.goto \
@@ -117,6 +118,7 @@ acml :: slinpack.acml dlinpack.acml clinpack.acml zlinpack.acml \
117118
cher2k.acml zher2k.acml \
118119
sgemv.acml dgemv.acml cgemv.acml zgemv.acml \
119120
strmv.acml dtrmv.acml ctrmv.acml ztrmv.acml \
121+
stpmv.acml dtpmv.acml ctpmv.acml ztpmv.acml \
120122
strsv.acml dtrsv.acml ctrsv.acml ztrsv.acml \
121123
sgeev.acml dgeev.acml cgeev.acml zgeev.acml \
122124
sgesv.acml dgesv.acml cgesv.acml zgesv.acml \
@@ -155,6 +157,7 @@ atlas :: slinpack.atlas dlinpack.atlas clinpack.atlas zlinpack.atlas \
155157
cher2k.atlas zher2k.atlas \
156158
sgemv.atlas dgemv.atlas cgemv.atlas zgemv.atlas \
157159
strmv.atlas dtrmv.atlas ctrmv.atlas ztrmv.atlas \
160+
stpmv.atlas dtpmv.atlas ctpmv.atlas ztpmv.atlas \
158161
strsv.atlas dtrsv.atlas ctrsv.atlas ztrsv.atlas \
159162
sgeev.atlas dgeev.atlas cgeev.atlas zgeev.atlas \
160163
sgesv.atlas dgesv.atlas cgesv.atlas zgesv.atlas \
@@ -192,6 +195,7 @@ mkl :: slinpack.mkl dlinpack.mkl clinpack.mkl zlinpack.mkl \
192195
cher2k.mkl zher2k.mkl \
193196
sgemv.mkl dgemv.mkl cgemv.mkl zgemv.mkl \
194197
strmv.mkl dtrmv.mkl ctrmv.mkl ztrmv.mkl \
198+
stpmv.mkl dtpmv.mkl ctpmv.mkl ztpmv.mkl \
195199
strsv.mkl dtrsv.mkl ctrsv.mkl ztrsv.mkl \
196200
sgeev.mkl dgeev.mkl cgeev.mkl zgeev.mkl \
197201
sgesv.mkl dgesv.mkl cgesv.mkl zgesv.mkl \
@@ -229,6 +233,7 @@ goto :: sgemm.goto dgemm.goto cgemm.goto zgemm.goto \
229233
cher2k.goto zher2k.goto \
230234
sgemv.goto dgemv.goto cgemv.goto zgemv.goto \
231235
strmv.goto dtrmv.goto ctrmv.goto ztrmv.goto \
236+
stpmv.goto dtpmv.goto ctpmv.goto ztpmv.goto \
232237
strsv.goto dtrsv.goto ctrsv.goto ztrsv.goto \
233238
ssymm.goto dsymm.goto csymm.goto zsymm.goto \
234239
smallscaling \
@@ -272,6 +277,7 @@ acml :: slinpack.acml dlinpack.acml clinpack.acml zlinpack.acml \
272277
cher2k.acml zher2k.acml \
273278
sgemv.acml dgemv.acml cgemv.acml zgemv.acml \
274279
strmv.acml dtrmv.acml ctrmv.acml ztrmv.acml \
280+
stpmv.acml dtpmv.acml ctpmv.acml ztpmv.acml \
275281
strsv.acml dtrsv.acml ctrsv.acml ztrsv.acml \
276282
sgeev.acml dgeev.acml cgeev.acml zgeev.acml \
277283
sgesv.acml dgesv.acml cgesv.acml zgesv.acml \
@@ -310,6 +316,7 @@ atlas :: slinpack.atlas dlinpack.atlas clinpack.atlas zlinpack.atlas \
310316
cher2k.atlas zher2k.atlas \
311317
sgemv.atlas dgemv.atlas cgemv.atlas zgemv.atlas \
312318
strmv.atlas dtrmv.atlas ctrmv.atlas ztrmv.atlas \
319+
stpmv.atlas dtpmv.atlas ctpmv.atlas ztpmv.atlas \
313320
strsv.atlas dtrsv.atlas ctrsv.atlas ztrsv.atlas \
314321
sgeev.atlas dgeev.atlas cgeev.atlas zgeev.atlas \
315322
sgesv.atlas dgesv.atlas cgesv.atlas zgesv.atlas \
@@ -349,6 +356,7 @@ mkl :: slinpack.mkl dlinpack.mkl clinpack.mkl zlinpack.mkl \
349356
cher2k.mkl zher2k.mkl \
350357
sgemv.mkl dgemv.mkl cgemv.mkl zgemv.mkl \
351358
strmv.mkl dtrmv.mkl ctrmv.mkl ztrmv.mkl \
359+
stpmv.mkl dtpmv.mkl ctpmv.mkl ztpmv.mkl \
352360
strsv.mkl dtrsv.mkl ctrsv.mkl ztrsv.mkl \
353361
sgeev.mkl dgeev.mkl cgeev.mkl zgeev.mkl \
354362
sgesv.mkl dgesv.mkl cgesv.mkl zgesv.mkl \
@@ -397,6 +405,7 @@ veclib :: slinpack.veclib dlinpack.veclib clinpack.veclib zlinpack.veclib \
397405
cher2k.veclib zher2k.veclib \
398406
sgemv.veclib dgemv.veclib cgemv.veclib zgemv.veclib \
399407
strmv.veclib dtrmv.veclib ctrmv.veclib ztrmv.veclib \
408+
stpmv.veclib dtpmv.veclib ctpmv.veclib ztpmv.veclib \
400409
strsv.veclib dtrsv.veclib ctrsv.veclib ztrsv.veclib \
401410
sgeev.veclib dgeev.veclib cgeev.veclib zgeev.veclib \
402411
sgesv.veclib dgesv.veclib cgesv.veclib zgesv.veclib \
@@ -1363,6 +1372,72 @@ ztrmv.mkl : ztrmv.$(SUFFIX)
13631372
ztrmv.veclib : ztrmv.$(SUFFIX)
13641373
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBVECLIB) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
13651374

1375+
1376+
##################################### Stpmv ####################################################
1377+
stpmv.goto : stpmv.$(SUFFIX) ../$(LIBNAME)
1378+
$(CC) $(CFLAGS) -o $(@F) $^ $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB) -lm
1379+
1380+
stpmv.acml : stpmv.$(SUFFIX)
1381+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBACML) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1382+
1383+
stpmv.atlas : stpmv.$(SUFFIX)
1384+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBATLAS) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1385+
1386+
stpmv.mkl : stpmv.$(SUFFIX)
1387+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBMKL) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1388+
1389+
stpmv.veclib : stpmv.$(SUFFIX)
1390+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBVECLIB) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1391+
1392+
##################################### Dtpmv ####################################################
1393+
dtpmv.goto : dtpmv.$(SUFFIX) ../$(LIBNAME)
1394+
$(CC) $(CFLAGS) -o $(@F) $^ $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB) -lm
1395+
1396+
dtpmv.acml : dtpmv.$(SUFFIX)
1397+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBACML) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1398+
1399+
dtpmv.atlas : dtpmv.$(SUFFIX)
1400+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBATLAS) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1401+
1402+
dtpmv.mkl : dtpmv.$(SUFFIX)
1403+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBMKL) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1404+
1405+
dtpmv.veclib : dtpmv.$(SUFFIX)
1406+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBVECLIB) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1407+
1408+
##################################### Ctpmv ####################################################
1409+
1410+
ctpmv.goto : ctpmv.$(SUFFIX) ../$(LIBNAME)
1411+
$(CC) $(CFLAGS) -o $(@F) $^ $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB) -lm
1412+
1413+
ctpmv.acml : ctpmv.$(SUFFIX)
1414+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBACML) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1415+
1416+
ctpmv.atlas : ctpmv.$(SUFFIX)
1417+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBATLAS) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1418+
1419+
ctpmv.mkl : ctpmv.$(SUFFIX)
1420+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBMKL) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1421+
1422+
ctpmv.veclib : ctpmv.$(SUFFIX)
1423+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBVECLIB) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1424+
1425+
##################################### Ztpmv ####################################################
1426+
1427+
ztpmv.goto : ztpmv.$(SUFFIX) ../$(LIBNAME)
1428+
$(CC) $(CFLAGS) -o $(@F) $^ $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB) -lm
1429+
1430+
ztpmv.acml : ztpmv.$(SUFFIX)
1431+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBACML) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1432+
1433+
ztpmv.atlas : ztpmv.$(SUFFIX)
1434+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBATLAS) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1435+
1436+
ztpmv.mkl : ztpmv.$(SUFFIX)
1437+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBMKL) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
1438+
1439+
ztpmv.veclib : ztpmv.$(SUFFIX)
1440+
-$(CC) $(CFLAGS) -o $(@F) $^ $(LIBVECLIB) $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB)
13661441
##################################### Strsv ####################################################
13671442
strsv.goto : strsv.$(SUFFIX) ../$(LIBNAME)
13681443
$(CC) $(CFLAGS) -o $(@F) $^ $(CEXTRALIB) $(EXTRALIB) $(FEXTRALIB) -lm
@@ -2799,6 +2874,18 @@ ctrmv.$(SUFFIX) : trmv.c
27992874
ztrmv.$(SUFFIX) : trmv.c
28002875
$(CC) $(CFLAGS) -c -DCOMPLEX -DDOUBLE -o $(@F) $^
28012876

2877+
stpmv.$(SUFFIX) : tpmv.c
2878+
$(CC) $(CFLAGS) -c -UCOMPLEX -UDOUBLE -o $(@F) $^
2879+
2880+
dtpmv.$(SUFFIX) : tpmv.c
2881+
$(CC) $(CFLAGS) -c -UCOMPLEX -DDOUBLE -o $(@F) $^
2882+
2883+
ctpmv.$(SUFFIX) : tpmv.c
2884+
$(CC) $(CFLAGS) -c -DCOMPLEX -UDOUBLE -o $(@F) $^
2885+
2886+
ztpmv.$(SUFFIX) : tpmv.c
2887+
$(CC) $(CFLAGS) -c -DCOMPLEX -DDOUBLE -o $(@F) $^
2888+
28022889
strsv.$(SUFFIX) : trsv.c
28032890
$(CC) $(CFLAGS) -c -UCOMPLEX -UDOUBLE -o $(@F) $^
28042891

benchmark/tpmv.c

Lines changed: 172 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,172 @@
1+
/***************************************************************************
2+
Copyright (c) 2014, 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 <stdio.h>
29+
#include <stdlib.h>
30+
#ifdef __CYGWIN32__
31+
#include <sys/time.h>
32+
#endif
33+
#include "common.h"
34+
35+
#undef TPMV
36+
37+
#ifndef COMPLEX
38+
39+
#ifdef DOUBLE
40+
#define TPMV BLASFUNC(dtpmv)
41+
#else
42+
#define TPMV BLASFUNC(stpmv)
43+
#endif
44+
45+
#else
46+
47+
#ifdef DOUBLE
48+
#define TPMV BLASFUNC(ztpmv)
49+
#else
50+
#define TPMV BLASFUNC(ctpmv)
51+
#endif
52+
53+
#endif
54+
55+
#if !defined(__WIN32__) && !defined(__WIN64__) && !defined(__CYGWIN32__) && 0
56+
57+
static void *huge_malloc(BLASLONG size)
58+
{
59+
int shmid;
60+
void *address;
61+
62+
#ifndef SHM_HUGETLB
63+
#define SHM_HUGETLB 04000
64+
#endif
65+
66+
if ((shmid =shmget(IPC_PRIVATE,
67+
(size + HUGE_PAGESIZE) & ~(HUGE_PAGESIZE - 1),
68+
SHM_HUGETLB | IPC_CREAT |0600)) < 0) {
69+
printf( "Memory allocation failed(shmget).\n");
70+
exit(1);
71+
}
72+
73+
address = shmat(shmid, NULL, SHM_RND);
74+
75+
if ((BLASLONG)address == -1) {
76+
printf( "Memory allocation failed(shmat).\n");
77+
exit(1);
78+
}
79+
80+
shmctl(shmid, IPC_RMID, 0);
81+
82+
return address;
83+
}
84+
85+
#define malloc huge_malloc
86+
87+
#endif
88+
89+
int main(int argc, char *argv[])
90+
{
91+
92+
FLOAT *a, *x;
93+
char *p;
94+
95+
char uplo ='U';
96+
char trans='N';
97+
char diag ='U';
98+
99+
int loops = 1;
100+
int l;
101+
blasint inc_x=1;
102+
103+
if ((p = getenv("OPENBLAS_UPLO"))) uplo=*p;
104+
if ((p = getenv("OPENBLAS_TRANS"))) trans=*p;
105+
if ((p = getenv("OPENBLAS_DIAG"))) diag=*p;
106+
if ((p = getenv("OPENBLAS_LOOPS"))) loops = atoi(p);
107+
if ((p = getenv("OPENBLAS_INCX"))) inc_x = atoi(p);
108+
109+
blasint n, i, j;
110+
111+
int from = 1;
112+
int to = 200;
113+
int step = 1;
114+
115+
struct timespec start = { 0, 0 }, stop = { 0, 0 };
116+
double time1, timeg;
117+
118+
argc--;argv++;
119+
120+
if (argc > 0) { from = atol(*argv); argc--; argv++;}
121+
if (argc > 0) { to = MAX(atol(*argv), from); argc--; argv++;}
122+
if (argc > 0) { step = atol(*argv); argc--; argv++;}
123+
124+
fprintf(stderr, "From : %3d To : %3d Step = %3d Uplo = %c Trans = %c Diag = %c Loops=%d Inc_x=%d\n", from,
125+
to, step, uplo, trans, diag, loops, inc_x);
126+
127+
if (( a = (FLOAT *)malloc(sizeof(FLOAT) * to * to * COMPSIZE)) == NULL) {
128+
fprintf(stderr,"Out of Memory!!\n");exit(1);
129+
}
130+
131+
if (( x = (FLOAT *)malloc(sizeof(FLOAT) * to * abs(inc_x) * COMPSIZE)) == NULL){
132+
fprintf(stderr,"Out of Memory!!\n");exit(1);
133+
}
134+
135+
#ifdef linux
136+
srandom(getpid());
137+
#endif
138+
139+
fprintf(stderr, " SIZE Flops\n");
140+
141+
for(n = from; n <= to; n += step) {
142+
timeg=0;
143+
144+
fprintf(stderr, " %6d : ", (int)n);
145+
for(j = 0; j < n; j++) {
146+
for(i = 0; i < n * COMPSIZE; i++) {
147+
a[(long)i + (long)j * (long)n * COMPSIZE] = ((FLOAT) rand() / (FLOAT) RAND_MAX) - 0.5;
148+
}
149+
}
150+
151+
for (i = 0; i < n * COMPSIZE * abs(inc_x); i++) {
152+
x[i] = ((FLOAT) rand() / (FLOAT) RAND_MAX) - 0.5;
153+
}
154+
155+
for (l = 0; l < loops; l++) {
156+
clock_gettime(CLOCK_REALTIME, &start);
157+
TPMV (&uplo, &trans, &diag, &n, a, x, &inc_x);
158+
clock_gettime(CLOCK_REALTIME, &stop);
159+
160+
time1 = (double)(stop.tv_sec - start.tv_sec) + (double)((stop.tv_nsec - start.tv_nsec)) / 1.e9;
161+
timeg += time1;
162+
}
163+
164+
timeg /= loops;
165+
fprintf(stderr, " %10.2f MFlops %12.9f sec\n",
166+
COMPSIZE * COMPSIZE * 1. * (double)n * (double)n / timeg / 1.e6, timeg);
167+
}
168+
169+
return 0;
170+
}
171+
172+
// void main(int argc, char *argv[]) __attribute__((weak, alias("MAIN__")));

0 commit comments

Comments
 (0)