]>
Commit | Line | Data |
---|---|---|
7c673cae FG |
1 | /********************************************************************** |
2 | Copyright(c) 2011-2015 Intel Corporation All rights reserved. | |
3 | ||
4 | Redistribution and use in source and binary forms, with or without | |
5 | modification, are permitted provided that the following conditions | |
6 | are met: | |
7 | * Redistributions of source code must retain the above copyright | |
8 | notice, this list of conditions and the following disclaimer. | |
9 | * 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 | * Neither the name of Intel Corporation nor the names of its | |
14 | contributors may be used to endorse or promote products derived | |
15 | from this software without specific prior written permission. | |
16 | ||
17 | THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS | |
18 | "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT | |
19 | LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR | |
20 | A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT | |
21 | OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, | |
22 | SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT | |
23 | LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, | |
24 | DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY | |
25 | THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT | |
26 | (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE | |
27 | OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. | |
28 | **********************************************************************/ | |
29 | ||
30 | #include <stdio.h> | |
31 | #include <stdlib.h> | |
32 | #include <string.h> // for memset, memcmp | |
33 | #include "erasure_code.h" | |
34 | #include "test.h" | |
35 | ||
36 | #ifndef FUNCTION_UNDER_TEST | |
37 | # define FUNCTION_UNDER_TEST gf_2vect_dot_prod_sse | |
38 | #endif | |
39 | ||
40 | #define str(s) #s | |
41 | #define xstr(s) str(s) | |
42 | ||
43 | //#define CACHED_TEST | |
44 | #ifdef CACHED_TEST | |
45 | // Cached test, loop many times over small dataset | |
46 | # define TEST_SOURCES 10 | |
47 | # define TEST_LEN 8*1024 | |
48 | # define TEST_LOOPS 40000 | |
49 | # define TEST_TYPE_STR "_warm" | |
50 | #else | |
51 | # ifndef TEST_CUSTOM | |
52 | // Uncached test. Pull from large mem base. | |
53 | # define TEST_SOURCES 10 | |
54 | # define GT_L3_CACHE 32*1024*1024 /* some number > last level cache */ | |
55 | # define TEST_LEN ((GT_L3_CACHE / TEST_SOURCES) & ~(64-1)) | |
56 | # define TEST_LOOPS 100 | |
57 | # define TEST_TYPE_STR "_cold" | |
58 | # else | |
59 | # define TEST_TYPE_STR "_cus" | |
60 | # ifndef TEST_LOOPS | |
61 | # define TEST_LOOPS 1000 | |
62 | # endif | |
63 | # endif | |
64 | #endif | |
65 | ||
66 | typedef unsigned char u8; | |
67 | ||
68 | void dump(unsigned char *buf, int len) | |
69 | { | |
70 | int i; | |
71 | for (i = 0; i < len;) { | |
72 | printf(" %2x", 0xff & buf[i++]); | |
73 | if (i % 32 == 0) | |
74 | printf("\n"); | |
75 | } | |
76 | printf("\n"); | |
77 | } | |
78 | ||
79 | void dump_matrix(unsigned char **s, int k, int m) | |
80 | { | |
81 | int i, j; | |
82 | for (i = 0; i < k; i++) { | |
83 | for (j = 0; j < m; j++) { | |
84 | printf(" %2x", s[i][j]); | |
85 | } | |
86 | printf("\n"); | |
87 | } | |
88 | printf("\n"); | |
89 | } | |
90 | ||
91 | int main(int argc, char *argv[]) | |
92 | { | |
93 | int i, j; | |
94 | void *buf; | |
95 | u8 g1[TEST_SOURCES], g2[TEST_SOURCES], g_tbls[2 * TEST_SOURCES * 32]; | |
96 | u8 *dest1, *dest2, *dest_ref1, *dest_ref2, *dest_ptrs[2]; | |
97 | u8 *buffs[TEST_SOURCES]; | |
98 | struct perf start, stop; | |
99 | ||
100 | printf(xstr(FUNCTION_UNDER_TEST) ": %dx%d\n", TEST_SOURCES, TEST_LEN); | |
101 | ||
102 | // Allocate the arrays | |
103 | for (i = 0; i < TEST_SOURCES; i++) { | |
104 | if (posix_memalign(&buf, 64, TEST_LEN)) { | |
105 | printf("alloc error: Fail"); | |
106 | return -1; | |
107 | } | |
108 | buffs[i] = buf; | |
109 | } | |
110 | ||
111 | if (posix_memalign(&buf, 64, TEST_LEN)) { | |
112 | printf("alloc error: Fail"); | |
113 | return -1; | |
114 | } | |
115 | dest1 = buf; | |
116 | ||
117 | if (posix_memalign(&buf, 64, TEST_LEN)) { | |
118 | printf("alloc error: Fail"); | |
119 | return -1; | |
120 | } | |
121 | dest2 = buf; | |
122 | ||
123 | if (posix_memalign(&buf, 64, TEST_LEN)) { | |
124 | printf("alloc error: Fail"); | |
125 | return -1; | |
126 | } | |
127 | dest_ref1 = buf; | |
128 | ||
129 | if (posix_memalign(&buf, 64, TEST_LEN)) { | |
130 | printf("alloc error: Fail"); | |
131 | return -1; | |
132 | } | |
133 | dest_ref2 = buf; | |
134 | ||
135 | dest_ptrs[0] = dest1; | |
136 | dest_ptrs[1] = dest2; | |
137 | ||
138 | // Performance test | |
139 | for (i = 0; i < TEST_SOURCES; i++) | |
140 | for (j = 0; j < TEST_LEN; j++) | |
141 | buffs[i][j] = rand(); | |
142 | ||
143 | memset(dest1, 0, TEST_LEN); | |
144 | memset(dest2, 0, TEST_LEN); | |
145 | memset(dest_ref1, 0, TEST_LEN); | |
146 | memset(dest_ref2, 0, TEST_LEN); | |
147 | ||
148 | for (i = 0; i < TEST_SOURCES; i++) { | |
149 | g1[i] = rand(); | |
150 | g2[i] = rand(); | |
151 | } | |
152 | ||
153 | for (j = 0; j < TEST_SOURCES; j++) { | |
154 | gf_vect_mul_init(g1[j], &g_tbls[j * 32]); | |
155 | gf_vect_mul_init(g2[j], &g_tbls[(32 * TEST_SOURCES) + (j * 32)]); | |
156 | } | |
157 | ||
158 | gf_vect_dot_prod_base(TEST_LEN, TEST_SOURCES, &g_tbls[0], buffs, dest_ref1); | |
159 | gf_vect_dot_prod_base(TEST_LEN, TEST_SOURCES, &g_tbls[32 * TEST_SOURCES], buffs, | |
160 | dest_ref2); | |
161 | ||
162 | #ifdef DO_REF_PERF | |
163 | perf_start(&start); | |
164 | for (i = 0; i < TEST_LOOPS / 100; i++) { | |
165 | for (j = 0; j < TEST_SOURCES; j++) { | |
166 | gf_vect_mul_init(g1[j], &g_tbls[j * 32]); | |
167 | gf_vect_mul_init(g2[j], &g_tbls[(32 * TEST_SOURCES) + (j * 32)]); | |
168 | } | |
169 | ||
170 | gf_vect_dot_prod_base(TEST_LEN, TEST_SOURCES, &g_tbls[0], buffs, dest_ref1); | |
171 | gf_vect_dot_prod_base(TEST_LEN, TEST_SOURCES, &g_tbls[32 * TEST_SOURCES], | |
172 | buffs, dest_ref2); | |
173 | } | |
174 | perf_stop(&stop); | |
175 | printf("gf_2vect_dot_prod_base" TEST_TYPE_STR ": "); | |
176 | perf_print(stop, start, (long long)TEST_LEN * (TEST_SOURCES + 2) * i); | |
177 | #endif | |
178 | ||
179 | FUNCTION_UNDER_TEST(TEST_LEN, TEST_SOURCES, g_tbls, buffs, dest_ptrs); | |
180 | ||
181 | perf_start(&start); | |
182 | for (i = 0; i < TEST_LOOPS; i++) { | |
183 | for (j = 0; j < TEST_SOURCES; j++) { | |
184 | gf_vect_mul_init(g1[j], &g_tbls[j * 32]); | |
185 | gf_vect_mul_init(g2[j], &g_tbls[(32 * TEST_SOURCES) + (j * 32)]); | |
186 | } | |
187 | ||
188 | FUNCTION_UNDER_TEST(TEST_LEN, TEST_SOURCES, g_tbls, buffs, dest_ptrs); | |
189 | } | |
190 | perf_stop(&stop); | |
191 | printf(xstr(FUNCTION_UNDER_TEST) TEST_TYPE_STR ": "); | |
192 | perf_print(stop, start, (long long)TEST_LEN * (TEST_SOURCES + 2) * i); | |
193 | ||
194 | if (0 != memcmp(dest_ref1, dest1, TEST_LEN)) { | |
195 | printf("Fail perf " xstr(FUNCTION_UNDER_TEST) " test1\n"); | |
196 | dump_matrix(buffs, 5, TEST_SOURCES); | |
197 | printf("dprod_base:"); | |
198 | dump(dest_ref1, 25); | |
199 | printf("dprod_dut:"); | |
200 | dump(dest1, 25); | |
201 | return -1; | |
202 | } | |
203 | if (0 != memcmp(dest_ref2, dest2, TEST_LEN)) { | |
204 | printf("Fail perf " xstr(FUNCTION_UNDER_TEST) " test2\n"); | |
205 | dump_matrix(buffs, 5, TEST_SOURCES); | |
206 | printf("dprod_base:"); | |
207 | dump(dest_ref2, 25); | |
208 | printf("dprod_dut:"); | |
209 | dump(dest2, 25); | |
210 | return -1; | |
211 | } | |
212 | ||
213 | printf("pass perf check\n"); | |
214 | return 0; | |
215 | ||
216 | } |