Mercurial > hg > sv-dependency-builds

comparison src/fftw-3.3.8/rdft/scalar/r2cb/r2cb_13.c @ 167:bd3cc4d1df30

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Add FFTW 3.3.8 source, and a Linux build
author Chris Cannam <cannam@all-day-breakfast.com>
date Tue, 19 Nov 2019 14:52:55 +0000
parents
children
comparison
equal deleted inserted replaced
166:cbd6d7e562c7 167:bd3cc4d1df30
1 /*
2 * Copyright (c) 2003, 2007-14 Matteo Frigo
3 * Copyright (c) 2003, 2007-14 Massachusetts Institute of Technology
4 *
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
9 *
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
14 *
15 * You should have received a copy of the GNU General Public License
16 * along with this program; if not, write to the Free Software
17 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
18 *
19 */
20
21 /* This file was automatically generated --- DO NOT EDIT */
22 /* Generated on Thu May 24 08:07:28 EDT 2018 */
23
24 #include "rdft/codelet-rdft.h"
25
26 #if defined(ARCH_PREFERS_FMA) || defined(ISA_EXTENSION_PREFERS_FMA)
27
28 /* Generated by: ../../../genfft/gen_r2cb.native -fma -compact -variables 4 -pipeline-latency 4 -sign 1 -n 13 -name r2cb_13 -include rdft/scalar/r2cb.h */
29
30 /*
31 * This function contains 76 FP additions, 58 FP multiplications,
32 * (or, 18 additions, 0 multiplications, 58 fused multiply/add),
33 * 63 stack variables, 26 constants, and 26 memory accesses
34 */
35 #include "rdft/scalar/r2cb.h"
36
37 static void r2cb_13(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
38 {
39 DK(KP875502302, +0.875502302409147941146295545768755143177842006);
40 DK(KP1_040057143, +1.040057143777729238234261000998465604986476278);
41 DK(KP968287244, +0.968287244361984016049539446938120421179794516);
42 DK(KP1_150281458, +1.150281458948006242736771094910906776922003215);
43 DK(KP1_200954543, +1.200954543865330565851538506669526018704025697);
44 DK(KP769338817, +0.769338817572980603471413688209101117038278899);
45 DK(KP686558370, +0.686558370781754340655719594850823015421401653);
46 DK(KP226109445, +0.226109445035782405468510155372505010481906348);
47 DK(KP1_033041561, +1.033041561246979445681802577138034271410067244);
48 DK(KP581704778, +0.581704778510515730456870384989698884939833902);
49 DK(KP1_007074065, +1.007074065727533254493747707736933954186697125);
50 DK(KP600925212, +0.600925212577331548853203544578415991041882762);
51 DK(KP859542535, +0.859542535098774820163672132761689612766401925);
52 DK(KP503537032, +0.503537032863766627246873853868466977093348562);
53 DK(KP522026385, +0.522026385161275033714027226654165028300441940);
54 DK(KP957805992, +0.957805992594665126462521754605754580515587217);
55 DK(KP853480001, +0.853480001859823990758994934970528322872359049);
56 DK(KP2_000000000, +2.000000000000000000000000000000000000000000000);
57 DK(KP514918778, +0.514918778086315755491789696138117261566051239);
58 DK(KP301479260, +0.301479260047709873958013540496673347309208464);
59 DK(KP166666666, +0.166666666666666666666666666666666666666666667);
60 DK(KP612264650, +0.612264650376756543746494474777125408779395514);
61 DK(KP302775637, +0.302775637731994646559610633735247973125648287);
62 DK(KP038632954, +0.038632954644348171955506895830342264440241080);
63 DK(KP866025403, +0.866025403784438646763723170752936183471402627);
64 DK(KP500000000, +0.500000000000000000000000000000000000000000000);
65 {
66 INT i;
67 for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(52, rs), MAKE_VOLATILE_STRIDE(52, csr), MAKE_VOLATILE_STRIDE(52, csi)) {
68 E TG, TU, TN, T16, TJ, TV, T1, Tp, Tc, Td, Tg, Tj, Tk, Tm, Tn;
69 E To;
70 {
71 E Ts, Tv, Tw, TE, TB, TC, Tz, TD, TA, TF;
72 {
73 E Tt, Tu, Tx, Ty;
74 Ts = Ci[WS(csi, 5)];
75 Tt = Ci[WS(csi, 2)];
76 Tu = Ci[WS(csi, 6)];
77 Tv = Tt + Tu;
78 Tw = FNMS(KP500000000, Tv, Ts);
79 TE = Tu - Tt;
80 TB = Ci[WS(csi, 1)];
81 Tx = Ci[WS(csi, 3)];
82 Ty = Ci[WS(csi, 4)];
83 TC = Tx - Ty;
84 Tz = Tx + Ty;
85 TD = FNMS(KP500000000, TC, TB);
86 }
87 TA = FMA(KP866025403, Tz, Tw);
88 TF = FMA(KP866025403, TE, TD);
89 TG = FNMS(KP038632954, TF, TA);
90 TU = FMA(KP038632954, TA, TF);
91 {
92 E TL, TM, TH, TI;
93 TL = Ts + Tv;
94 TM = TB + TC;
95 TN = FMA(KP302775637, TM, TL);
96 T16 = FNMS(KP302775637, TL, TM);
97 TH = FNMS(KP866025403, Tz, Tw);
98 TI = FNMS(KP866025403, TE, TD);
99 TJ = FNMS(KP612264650, TI, TH);
100 TV = FMA(KP612264650, TH, TI);
101 }
102 }
103 {
104 E Tb, Ti, Tf, T6, Th, Te;
105 T1 = Cr[0];
106 {
107 E T7, T8, T9, Ta;
108 T7 = Cr[WS(csr, 5)];
109 T8 = Cr[WS(csr, 2)];
110 T9 = Cr[WS(csr, 6)];
111 Ta = T8 + T9;
112 Tb = T7 + Ta;
113 Ti = FMS(KP500000000, Ta, T7);
114 Tf = T8 - T9;
115 }
116 {
117 E T2, T3, T4, T5;
118 T2 = Cr[WS(csr, 1)];
119 T3 = Cr[WS(csr, 3)];
120 T4 = Cr[WS(csr, 4)];
121 T5 = T3 + T4;
122 T6 = T2 + T5;
123 Th = FNMS(KP500000000, T5, T2);
124 Te = T3 - T4;
125 }
126 Tp = T6 - Tb;
127 Tc = T6 + Tb;
128 Td = FNMS(KP166666666, Tc, T1);
129 Tg = Te + Tf;
130 Tj = Th - Ti;
131 Tk = FMA(KP301479260, Tj, Tg);
132 Tm = Th + Ti;
133 Tn = Te - Tf;
134 To = FNMS(KP514918778, Tn, Tm);
135 }
136 R0[0] = FMA(KP2_000000000, Tc, T1);
137 {
138 E TW, T14, TO, TS, T18, T1e, TR, T13, Tr, T1d, TZ, T19;
139 {
140 E TK, T17, TP, TQ;
141 TW = FMA(KP853480001, TV, TU);
142 T14 = FMA(KP853480001, TJ, TG);
143 TK = FNMS(KP853480001, TJ, TG);
144 TO = FMA(KP957805992, TN, TK);
145 TS = FNMS(KP522026385, TK, TN);
146 T17 = FNMS(KP853480001, TV, TU);
147 T18 = FNMS(KP522026385, T17, T16);
148 T1e = FMA(KP957805992, T16, T17);
149 TP = FNMS(KP503537032, Tk, Td);
150 TQ = FNMS(KP859542535, To, Tp);
151 TR = FMA(KP600925212, TQ, TP);
152 T13 = FNMS(KP600925212, TQ, TP);
153 {
154 E Tl, Tq, TX, TY;
155 Tl = FMA(KP1_007074065, Tk, Td);
156 Tq = FMA(KP581704778, Tp, To);
157 Tr = FMA(KP1_033041561, Tq, Tl);
158 T1d = FNMS(KP1_033041561, Tq, Tl);
159 TX = FNMS(KP226109445, Tg, Tj);
160 TY = FMA(KP686558370, Tm, Tn);
161 TZ = FNMS(KP769338817, TY, TX);
162 T19 = FMA(KP769338817, TY, TX);
163 }
164 }
165 R1[0] = FNMS(KP1_200954543, TO, Tr);
166 R1[WS(rs, 2)] = FNMS(KP1_200954543, T1e, T1d);
167 R0[WS(rs, 4)] = FMA(KP1_200954543, T1e, T1d);
168 R0[WS(rs, 6)] = FMA(KP1_200954543, TO, Tr);
169 {
170 E TT, T10, T15, T1a;
171 TT = FNMS(KP1_150281458, TS, TR);
172 T10 = FNMS(KP968287244, TZ, TW);
173 R1[WS(rs, 1)] = FNMS(KP1_040057143, T10, TT);
174 R1[WS(rs, 4)] = FMA(KP1_040057143, T10, TT);
175 T15 = FMA(KP1_040057143, T14, T13);
176 T1a = FNMS(KP875502302, T19, T18);
177 R0[WS(rs, 1)] = FNMS(KP1_150281458, T1a, T15);
178 R1[WS(rs, 3)] = FMA(KP1_150281458, T1a, T15);
179 }
180 {
181 E T1b, T1c, T11, T12;
182 T1b = FNMS(KP1_040057143, T14, T13);
183 T1c = FMA(KP875502302, T19, T18);
184 R0[WS(rs, 3)] = FNMS(KP1_150281458, T1c, T1b);
185 R1[WS(rs, 5)] = FMA(KP1_150281458, T1c, T1b);
186 T11 = FMA(KP1_150281458, TS, TR);
187 T12 = FMA(KP968287244, TZ, TW);
188 R0[WS(rs, 2)] = FNMS(KP1_040057143, T12, T11);
189 R0[WS(rs, 5)] = FMA(KP1_040057143, T12, T11);
190 }
191 }
192 }
193 }
194 }
195
196 static const kr2c_desc desc = { 13, "r2cb_13", {18, 0, 58, 0}, &GENUS };
197
198 void X(codelet_r2cb_13) (planner *p) {
199 X(kr2c_register) (p, r2cb_13, &desc);
200 }
201
202 #else
203
204 /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 13 -name r2cb_13 -include rdft/scalar/r2cb.h */
205
206 /*
207 * This function contains 76 FP additions, 35 FP multiplications,
208 * (or, 56 additions, 15 multiplications, 20 fused multiply/add),
209 * 56 stack variables, 19 constants, and 26 memory accesses
210 */
211 #include "rdft/scalar/r2cb.h"
212
213 static void r2cb_13(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs)
214 {
215 DK(KP1_007074065, +1.007074065727533254493747707736933954186697125);
216 DK(KP227708958, +0.227708958111581597949308691735310621069285120);
217 DK(KP531932498, +0.531932498429674575175042127684371897596660533);
218 DK(KP774781170, +0.774781170935234584261351932853525703557550433);
219 DK(KP265966249, +0.265966249214837287587521063842185948798330267);
220 DK(KP516520780, +0.516520780623489722840901288569017135705033622);
221 DK(KP151805972, +0.151805972074387731966205794490207080712856746);
222 DK(KP503537032, +0.503537032863766627246873853868466977093348562);
223 DK(KP166666666, +0.166666666666666666666666666666666666666666667);
224 DK(KP600925212, +0.600925212577331548853203544578415991041882762);
225 DK(KP500000000, +0.500000000000000000000000000000000000000000000);
226 DK(KP256247671, +0.256247671582936600958684654061725059144125175);
227 DK(KP156891391, +0.156891391051584611046832726756003269660212636);
228 DK(KP348277202, +0.348277202304271810011321589858529485233929352);
229 DK(KP1_150281458, +1.150281458948006242736771094910906776922003215);
230 DK(KP300238635, +0.300238635966332641462884626667381504676006424);
231 DK(KP011599105, +0.011599105605768290721655456654083252189827041);
232 DK(KP1_732050807, +1.732050807568877293527446341505872366942805254);
233 DK(KP2_000000000, +2.000000000000000000000000000000000000000000000);
234 {
235 INT i;
236 for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(52, rs), MAKE_VOLATILE_STRIDE(52, csr), MAKE_VOLATILE_STRIDE(52, csi)) {
237 E TG, TS, TR, T15, TJ, TT, T1, Tm, Tc, Td, Tg, Tj, Tk, Tn, To;
238 E Tp;
239 {
240 E Ts, Tv, Tw, TE, TC, TB, Tz, TD, TA, TF;
241 {
242 E Tt, Tu, Tx, Ty;
243 Ts = Ci[WS(csi, 1)];
244 Tt = Ci[WS(csi, 3)];
245 Tu = Ci[WS(csi, 4)];
246 Tv = Tt - Tu;
247 Tw = FMS(KP2_000000000, Ts, Tv);
248 TE = KP1_732050807 * (Tt + Tu);
249 TC = Ci[WS(csi, 5)];
250 Tx = Ci[WS(csi, 6)];
251 Ty = Ci[WS(csi, 2)];
252 TB = Tx + Ty;
253 Tz = KP1_732050807 * (Tx - Ty);
254 TD = FNMS(KP2_000000000, TC, TB);
255 }
256 TA = Tw + Tz;
257 TF = TD - TE;
258 TG = FMA(KP011599105, TA, KP300238635 * TF);
259 TS = FNMS(KP011599105, TF, KP300238635 * TA);
260 {
261 E TP, TQ, TH, TI;
262 TP = Ts + Tv;
263 TQ = TB + TC;
264 TR = FNMS(KP348277202, TQ, KP1_150281458 * TP);
265 T15 = FMA(KP348277202, TP, KP1_150281458 * TQ);
266 TH = Tw - Tz;
267 TI = TE + TD;
268 TJ = FMA(KP156891391, TH, KP256247671 * TI);
269 TT = FNMS(KP256247671, TH, KP156891391 * TI);
270 }
271 }
272 {
273 E Tb, Ti, Tf, T6, Th, Te;
274 T1 = Cr[0];
275 {
276 E T7, T8, T9, Ta;
277 T7 = Cr[WS(csr, 5)];
278 T8 = Cr[WS(csr, 2)];
279 T9 = Cr[WS(csr, 6)];
280 Ta = T8 + T9;
281 Tb = T7 + Ta;
282 Ti = FNMS(KP500000000, Ta, T7);
283 Tf = T8 - T9;
284 }
285 {
286 E T2, T3, T4, T5;
287 T2 = Cr[WS(csr, 1)];
288 T3 = Cr[WS(csr, 3)];
289 T4 = Cr[WS(csr, 4)];
290 T5 = T3 + T4;
291 T6 = T2 + T5;
292 Th = FNMS(KP500000000, T5, T2);
293 Te = T3 - T4;
294 }
295 Tm = KP600925212 * (T6 - Tb);
296 Tc = T6 + Tb;
297 Td = FNMS(KP166666666, Tc, T1);
298 Tg = Te + Tf;
299 Tj = Th + Ti;
300 Tk = FMA(KP503537032, Tg, KP151805972 * Tj);
301 Tn = Th - Ti;
302 To = Te - Tf;
303 Tp = FNMS(KP265966249, To, KP516520780 * Tn);
304 }
305 R0[0] = FMA(KP2_000000000, Tc, T1);
306 {
307 E TK, T1b, TV, T12, T16, T18, TO, T1a, Tr, T17, T11, T13;
308 {
309 E TU, T14, TM, TN;
310 TK = KP1_732050807 * (TG + TJ);
311 T1b = KP1_732050807 * (TS - TT);
312 TU = TS + TT;
313 TV = TR - TU;
314 T12 = FMA(KP2_000000000, TU, TR);
315 T14 = TG - TJ;
316 T16 = FMS(KP2_000000000, T14, T15);
317 T18 = T14 + T15;
318 TM = FMA(KP774781170, To, KP531932498 * Tn);
319 TN = FNMS(KP1_007074065, Tj, KP227708958 * Tg);
320 TO = TM - TN;
321 T1a = TM + TN;
322 {
323 E Tl, Tq, TZ, T10;
324 Tl = Td - Tk;
325 Tq = Tm - Tp;
326 Tr = Tl - Tq;
327 T17 = Tq + Tl;
328 TZ = FMA(KP2_000000000, Tk, Td);
329 T10 = FMA(KP2_000000000, Tp, Tm);
330 T11 = TZ - T10;
331 T13 = T10 + TZ;
332 }
333 }
334 R1[WS(rs, 2)] = T11 - T12;
335 R0[WS(rs, 6)] = T13 - T16;
336 R1[0] = T13 + T16;
337 R0[WS(rs, 4)] = T11 + T12;
338 {
339 E TL, TW, T19, T1c;
340 TL = Tr - TK;
341 TW = TO - TV;
342 R1[WS(rs, 3)] = TL - TW;
343 R0[WS(rs, 1)] = TL + TW;
344 T19 = T17 - T18;
345 T1c = T1a + T1b;
346 R1[WS(rs, 1)] = T19 - T1c;
347 R1[WS(rs, 4)] = T1c + T19;
348 }
349 {
350 E T1d, T1e, TX, TY;
351 T1d = T1a - T1b;
352 T1e = T17 + T18;
353 R0[WS(rs, 2)] = T1d + T1e;
354 R0[WS(rs, 5)] = T1e - T1d;
355 TX = Tr + TK;
356 TY = TO + TV;
357 R0[WS(rs, 3)] = TX - TY;
358 R1[WS(rs, 5)] = TX + TY;
359 }
360 }
361 }
362 }
363 }
364
365 static const kr2c_desc desc = { 13, "r2cb_13", {56, 15, 20, 0}, &GENUS };
366
367 void X(codelet_r2cb_13) (planner *p) {
368 X(kr2c_register) (p, r2cb_13, &desc);
369 }
370
371 #endif