]>
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 | #include <limits.h> | |
30 | #include "erasure_code.h" | |
31 | #include "types.h" | |
32 | ||
33 | void ec_init_tables(int k, int rows, unsigned char *a, unsigned char *g_tbls) | |
34 | { | |
35 | int i, j; | |
36 | ||
37 | for (i = 0; i < rows; i++) { | |
38 | for (j = 0; j < k; j++) { | |
39 | gf_vect_mul_init(*a++, g_tbls); | |
40 | g_tbls += 32; | |
41 | } | |
42 | } | |
43 | } | |
44 | ||
45 | void ec_encode_data_sse(int len, int k, int rows, unsigned char *g_tbls, unsigned char **data, | |
46 | unsigned char **coding) | |
47 | { | |
48 | ||
49 | if (len < 16) { | |
50 | ec_encode_data_base(len, k, rows, g_tbls, data, coding); | |
51 | return; | |
52 | } | |
53 | ||
54 | while (rows >= 4) { | |
55 | gf_4vect_dot_prod_sse(len, k, g_tbls, data, coding); | |
56 | g_tbls += 4 * k * 32; | |
57 | coding += 4; | |
58 | rows -= 4; | |
59 | } | |
60 | switch (rows) { | |
61 | case 3: | |
62 | gf_3vect_dot_prod_sse(len, k, g_tbls, data, coding); | |
63 | break; | |
64 | case 2: | |
65 | gf_2vect_dot_prod_sse(len, k, g_tbls, data, coding); | |
66 | break; | |
67 | case 1: | |
68 | gf_vect_dot_prod_sse(len, k, g_tbls, data, *coding); | |
69 | break; | |
70 | case 0: | |
71 | break; | |
72 | } | |
73 | ||
74 | } | |
75 | ||
76 | void ec_encode_data_avx(int len, int k, int rows, unsigned char *g_tbls, unsigned char **data, | |
77 | unsigned char **coding) | |
78 | { | |
79 | if (len < 16) { | |
80 | ec_encode_data_base(len, k, rows, g_tbls, data, coding); | |
81 | return; | |
82 | } | |
83 | ||
84 | while (rows >= 4) { | |
85 | gf_4vect_dot_prod_avx(len, k, g_tbls, data, coding); | |
86 | g_tbls += 4 * k * 32; | |
87 | coding += 4; | |
88 | rows -= 4; | |
89 | } | |
90 | switch (rows) { | |
91 | case 3: | |
92 | gf_3vect_dot_prod_avx(len, k, g_tbls, data, coding); | |
93 | break; | |
94 | case 2: | |
95 | gf_2vect_dot_prod_avx(len, k, g_tbls, data, coding); | |
96 | break; | |
97 | case 1: | |
98 | gf_vect_dot_prod_avx(len, k, g_tbls, data, *coding); | |
99 | break; | |
100 | case 0: | |
101 | break; | |
102 | } | |
103 | ||
104 | } | |
105 | ||
106 | void ec_encode_data_avx2(int len, int k, int rows, unsigned char *g_tbls, unsigned char **data, | |
107 | unsigned char **coding) | |
108 | { | |
109 | ||
110 | if (len < 32) { | |
111 | ec_encode_data_base(len, k, rows, g_tbls, data, coding); | |
112 | return; | |
113 | } | |
114 | ||
115 | while (rows >= 4) { | |
116 | gf_4vect_dot_prod_avx2(len, k, g_tbls, data, coding); | |
117 | g_tbls += 4 * k * 32; | |
118 | coding += 4; | |
119 | rows -= 4; | |
120 | } | |
121 | switch (rows) { | |
122 | case 3: | |
123 | gf_3vect_dot_prod_avx2(len, k, g_tbls, data, coding); | |
124 | break; | |
125 | case 2: | |
126 | gf_2vect_dot_prod_avx2(len, k, g_tbls, data, coding); | |
127 | break; | |
128 | case 1: | |
129 | gf_vect_dot_prod_avx2(len, k, g_tbls, data, *coding); | |
130 | break; | |
131 | case 0: | |
132 | break; | |
133 | } | |
134 | ||
135 | } | |
136 | ||
137 | #if __WORDSIZE == 64 || _WIN64 || __x86_64__ | |
138 | ||
139 | void ec_encode_data_update_sse(int len, int k, int rows, int vec_i, unsigned char *g_tbls, | |
140 | unsigned char *data, unsigned char **coding) | |
141 | { | |
142 | if (len < 16) { | |
143 | ec_encode_data_update_base(len, k, rows, vec_i, g_tbls, data, coding); | |
144 | return; | |
145 | } | |
146 | ||
147 | while (rows > 6) { | |
148 | gf_6vect_mad_sse(len, k, vec_i, g_tbls, data, coding); | |
149 | g_tbls += 6 * k * 32; | |
150 | coding += 6; | |
151 | rows -= 6; | |
152 | } | |
153 | switch (rows) { | |
154 | case 6: | |
155 | gf_6vect_mad_sse(len, k, vec_i, g_tbls, data, coding); | |
156 | break; | |
157 | case 5: | |
158 | gf_5vect_mad_sse(len, k, vec_i, g_tbls, data, coding); | |
159 | break; | |
160 | case 4: | |
161 | gf_4vect_mad_sse(len, k, vec_i, g_tbls, data, coding); | |
162 | break; | |
163 | case 3: | |
164 | gf_3vect_mad_sse(len, k, vec_i, g_tbls, data, coding); | |
165 | break; | |
166 | case 2: | |
167 | gf_2vect_mad_sse(len, k, vec_i, g_tbls, data, coding); | |
168 | break; | |
169 | case 1: | |
170 | gf_vect_mad_sse(len, k, vec_i, g_tbls, data, *coding); | |
171 | break; | |
172 | case 0: | |
173 | break; | |
174 | } | |
175 | ||
176 | } | |
177 | ||
178 | void ec_encode_data_update_avx(int len, int k, int rows, int vec_i, unsigned char *g_tbls, | |
179 | unsigned char *data, unsigned char **coding) | |
180 | { | |
181 | if (len < 16) { | |
182 | ec_encode_data_update_base(len, k, rows, vec_i, g_tbls, data, coding); | |
183 | return; | |
184 | } | |
185 | while (rows > 6) { | |
186 | gf_6vect_mad_avx(len, k, vec_i, g_tbls, data, coding); | |
187 | g_tbls += 6 * k * 32; | |
188 | coding += 6; | |
189 | rows -= 6; | |
190 | } | |
191 | switch (rows) { | |
192 | case 6: | |
193 | gf_6vect_mad_avx(len, k, vec_i, g_tbls, data, coding); | |
194 | break; | |
195 | case 5: | |
196 | gf_5vect_mad_avx(len, k, vec_i, g_tbls, data, coding); | |
197 | break; | |
198 | case 4: | |
199 | gf_4vect_mad_avx(len, k, vec_i, g_tbls, data, coding); | |
200 | break; | |
201 | case 3: | |
202 | gf_3vect_mad_avx(len, k, vec_i, g_tbls, data, coding); | |
203 | break; | |
204 | case 2: | |
205 | gf_2vect_mad_avx(len, k, vec_i, g_tbls, data, coding); | |
206 | break; | |
207 | case 1: | |
208 | gf_vect_mad_avx(len, k, vec_i, g_tbls, data, *coding); | |
209 | break; | |
210 | case 0: | |
211 | break; | |
212 | } | |
213 | ||
214 | } | |
215 | ||
216 | void ec_encode_data_update_avx2(int len, int k, int rows, int vec_i, unsigned char *g_tbls, | |
217 | unsigned char *data, unsigned char **coding) | |
218 | { | |
219 | if (len < 32) { | |
220 | ec_encode_data_update_base(len, k, rows, vec_i, g_tbls, data, coding); | |
221 | return; | |
222 | } | |
223 | while (rows > 6) { | |
224 | gf_6vect_mad_avx2(len, k, vec_i, g_tbls, data, coding); | |
225 | g_tbls += 6 * k * 32; | |
226 | coding += 6; | |
227 | rows -= 6; | |
228 | } | |
229 | switch (rows) { | |
230 | case 6: | |
231 | gf_6vect_mad_avx2(len, k, vec_i, g_tbls, data, coding); | |
232 | break; | |
233 | case 5: | |
234 | gf_5vect_mad_avx2(len, k, vec_i, g_tbls, data, coding); | |
235 | break; | |
236 | case 4: | |
237 | gf_4vect_mad_avx2(len, k, vec_i, g_tbls, data, coding); | |
238 | break; | |
239 | case 3: | |
240 | gf_3vect_mad_avx2(len, k, vec_i, g_tbls, data, coding); | |
241 | break; | |
242 | case 2: | |
243 | gf_2vect_mad_avx2(len, k, vec_i, g_tbls, data, coding); | |
244 | break; | |
245 | case 1: | |
246 | gf_vect_mad_avx2(len, k, vec_i, g_tbls, data, *coding); | |
247 | break; | |
248 | case 0: | |
249 | break; | |
250 | } | |
251 | ||
252 | } | |
253 | ||
254 | #endif //__WORDSIZE == 64 || _WIN64 || __x86_64__ | |
255 | ||
256 | struct slver { | |
257 | UINT16 snum; | |
258 | UINT8 ver; | |
259 | UINT8 core; | |
260 | }; | |
261 | ||
262 | // Version info | |
263 | struct slver ec_init_tables_slver_00010068; | |
264 | struct slver ec_init_tables_slver = { 0x0068, 0x01, 0x00 }; | |
265 | ||
266 | struct slver ec_encode_data_sse_slver_00020069; | |
267 | struct slver ec_encode_data_sse_slver = { 0x0069, 0x02, 0x00 }; |