/usr/include/fflas-ffpack/fflas/fflas_fgemm/schedule_winograd_acc_ip.inl is in fflas-ffpack-common 2.2.2-5.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 | /* -*- mode: C++; tab-width: 8; indent-tabs-mode: t; c-basic-offset: 8 -*- */
// vim:sts=8:sw=8:ts=8:noet:sr:cino=>s,f0,{0,g0,(0,\:0,t0,+0,=s
/* Copyright (C) 2014 the LinBox group
*
* Written by Brice Boyer (briceboyer) <boyer.brice@gmail.com>
*
*
* ========LICENCE========
* This file is part of the library FFLAS-FFPACK.
*
* FFLAS-FFPACK is free software: you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
* ========LICENCE========
*.
*/
/** @file fflas/fflas_fgemm/winograd_acc2.inl
* @ingroup MMalgos
* @brief Winograd implementation
* @bib ISSAC09 Scheduling
*/
#ifndef __FFLASFFPACK_fgemm_winograd_acc_ip_INL
#define __FFLASFFPACK_fgemm_winograd_acc_ip_INL
namespace FFLAS { namespace BLAS3 {
template < class Field, class FieldTrait >
inline void WinogradAcc_LR (const Field& F,
const FFLAS_TRANSPOSE ta,
const FFLAS_TRANSPOSE tb,
const size_t mr, const size_t nr, const size_t kr,
const typename Field::Element alpha,
typename Field::Element_ptr A,const size_t lda,
typename Field::Element_ptr B,const size_t ldb,
const typename Field::Element beta,
typename Field::Element_ptr C, const size_t ldc,
const MMHelper<Field, MMHelperAlgo::Winograd, FieldTrait > & WH
)
{
MMHelper<Field, MMHelperAlgo::Winograd, FieldTrait > H = WH ;
H.recLevel = H.recLevel - 1 ;
FFLASFFPACK_check(!F.isZero(beta));
typename Field::Element malpha ;
F.neg(malpha,alpha);
// A, B and c submatrices
typename Field::Element_ptr A11=A, A12, A21, A22;
typename Field::Element_ptr B11=B, B12, B21, B22;
typename Field::Element_ptr C11=C, C12=C+nr, C21=C+mr*ldc, C22=C21+nr;
typename Field::Element mbeta ;
F.neg(mbeta,beta);
size_t la, ca, lb, cb; // lines and columns in A,B sub matrices
// Three temporary submatrices are required
if (ta == FflasTrans) {
A21 = A + mr;
A12 = A + kr*lda;
A22 = A12 + mr;
la = kr ;
ca = mr ;
}
else { // ta == FflasNoTrans
A12 = A + kr;
A21 = A + mr*lda;
A22 = A21 + kr;
la = mr ;
ca = kr ;
}
if (tb == FflasTrans) {
B21 = B + kr;
B12 = B + nr*ldb;
B22 = B12 + kr;
lb = nr ;
cb = kr ;
}
else { // ta == FflasNoTrans
B12 = B + nr;
B21 = B + kr*ldb;
B22 = B21 + nr;
lb = kr ;
cb = nr ;
}
// Z1 = C22 - C12 in C22
fsubin(F,mr,nr,C12,ldc,C22,ldc);
// S1 = A21 + A22 in X
typename Field::Element_ptr X = fflas_new (F, std::max(std::max(mr*nr,kr*nr),mr*kr), 1);
fadd(F,la,ca,A21,lda,A22,lda,X,ca);
// T1 = B12 - B11 in Y
typename Field::Element_ptr Y = fflas_new (F, std::max(mr,kr), nr);
fsub(F,lb,cb,B12,ldb,B11,ldb,Y,cb);
// Z2 = C21 - Z1 in C21
fsubin(F,mr,nr,C22,ldc,C21,ldc);
// T3 = B22 - B12 in B12 ;
fsub(F,lb,cb,B22,ldb,B12,ldb,B12,ldb);
// S3 = A11 - A21 in A21
fsub(F,la,ca,A11,lda,A21,lda,A21,lda);
// P7 = a S3 T3 + b Z1 in C22
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A21, lda, B12, ldb, beta, C22, ldc, H);
// S2 = S1 - A11 in A21
fsub(F,la,ca,X,ca,A11,lda,A21,lda);
// T2 = B22 - T1 in B12
fsub(F,lb,cb,B22,ldb,Y,cb,B12,ldb);
// P5 = a S1 T1 + b C12 in C12
fgemm2 (F, ta, tb, mr, nr, kr, alpha, X, ca, Y, cb, beta, C12, ldc, H);
// T4 = T2 - B21 in X
fsub(F,lb,cb,B12,ldb,B21,ldb,X,cb);
// W1 = a A22 T4 in Y;
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A22, lda, X, cb, F.zero, Y, nr, H);
// P4 = W1 - b Z2 in C21
fadd(F,mr,nr,Y,nr,mbeta,C21,ldc,C21,ldc);
// S4 = A12 - S2 in A22
fsub(F,la,ca,A12,lda,A21,lda,A22,lda);
// P6 = a S2 T2 in X
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A21, lda, B12, ldb, F.zero, X, nr, H);
// W2 = a A12 B21 in Y
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A12, lda, B21, ldb, F.zero, Y, nr, H);
// P2 = W2 + beta C11 in C11
fadd(F,mr,nr,Y,nr,beta,C11,ldc,C11,ldc);
// P1 = a A11 B11 in Y
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A11, lda, B11, ldb, F.zero, Y, nr, H);
// U1 = P1 + P2 in C11
faddin(F,mr,nr,Y,nr,C11,ldc);
// U2 = P6 + P1 in X
faddin(F,mr,nr,Y,nr,X,nr);
fflas_delete (Y);
// U3 = U2 + P7 in C22
faddin(F,mr,nr,X,nr,C22,ldc);
// U4 = U2 + P5 in X
faddin(F,mr,nr,C12,ldc,X,nr);
// U6 = U3 - P4 in C21
fsub(F,mr,nr,C22,ldc,C21,ldc,C21,ldc);
// U7 = U3 + P5 in C22
faddin(F,mr,nr,C12,ldc,C22,ldc);
// P3 = a S4 B22 in C12
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A22, lda, B22, ldb, F.zero, C12, ldc, H);
// U5 = U4 + P3 in C12
faddin(F,mr,nr,X,nr,C12,ldc);
fflas_delete (X);
} // WinogradAccOld
template < class Field, class FieldTrait >
inline void WinogradAcc_R_S (const Field& F,
const FFLAS_TRANSPOSE ta,
const FFLAS_TRANSPOSE tb,
const size_t mr, const size_t nr, const size_t kr,
const typename Field::Element alpha,
const typename Field::Element_ptr A,const size_t lda,
typename Field::Element_ptr B,const size_t ldb,
const typename Field::Element beta,
typename Field::Element_ptr C, const size_t ldc,
const MMHelper<Field, MMHelperAlgo::Winograd,FieldTrait > & WH
)
{
MMHelper<Field, MMHelperAlgo::Winograd, FieldTrait > H = WH ;
H.recLevel = H.recLevel - 1 ;
FFLASFFPACK_check(!F.isZero(beta));
typename Field::Element malpha ;
F.neg(malpha,alpha);
// A, B and c submatrices
const typename Field::Element_ptr A11=A, A12, A21, A22;
typename Field::Element_ptr B11=B, B12, B21, B22;
typename Field::Element_ptr C11=C, C12=C+nr, C21=C+mr*ldc, C22=C21+nr;
typename Field::Element mbeta ;
F.neg(mbeta,beta);
size_t la, ca, lb, cb; // lines and columns in A,B sub matrices
// Three temporary submatrices are required
if (ta == FflasTrans) {
A21 = A + mr;
A12 = A + kr*lda;
A22 = A12 + mr;
la = kr ;
ca = mr ;
}
else { // ta == FflasNoTrans
A12 = A + kr;
A21 = A + mr*lda;
A22 = A21 + kr;
la = mr ;
ca = kr ;
}
if (tb == FflasTrans) {
B21 = B + kr;
B12 = B + nr*ldb;
B22 = B12 + kr;
lb = nr ;
cb = kr ;
}
else { // ta == FflasNoTrans
B12 = B + nr;
B21 = B + kr*ldb;
B22 = B21 + nr;
lb = kr ;
cb = nr ;
}
FFLASFFPACK_check(mr == nr && kr == nr);
// Z1 = C22 - C12 in C22
fsubin(F,mr,nr,C12,ldc,C22,ldc);
// T1 = B12 - B11 in X
// typename Field::Element_ptr X = fflas_new (F, std::max(mr,kr)*nr];
typename Field::Element_ptr X = fflas_new (F, mr, nr);
fsub(F,lb,cb,B12,ldb,B11,ldb,X,cb);
// Z2 = C21 - Z1 in C21
fsubin(F,mr,nr,C22,ldc,C21,ldc);
// T3 = B22 - B12 in B12 ;
fsub(F,lb,cb,B22,ldb,B12,ldb,B12,ldb);
// S3 = A11 - A21 in Y
typename Field::Element_ptr Y = fflas_new (F, mr, kr);
fsub(F,la,ca,A11,lda,A21,lda,Y,ca);
// P7 = a S3 T3 + b Z1 in C22
fgemm2 (F, ta, tb, mr, nr, kr, alpha, Y, ca, B12, ldb, beta, C22, ldc, H);
// S1 = A21 + A22 in Y
fadd(F,la,ca,A21,lda,A22,lda,Y,ca);
// T2 = B22 - T1 in B12
fsub(F,lb,cb,B22,ldb,X,cb,B12,ldb);
// P5 = a S1 T1 + b C12 in C12
fgemm2 (F, ta, tb, mr, nr, kr, alpha, Y, ca, X, cb, beta, C12, ldc, H);
// T4 = T2 - B21 in X
fsub(F,lb,cb,B12,ldb,B21,ldb,X,cb);
// P4 = a A22 T4 - b Z2 in C21
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A22, lda, X, cb, mbeta, C21, ldc, H);
// W1 = a A12 B21 in X;
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A12, lda, B21, ldb, F.zero, X, nr, H);
// P2 = W1 + beta C11 in C11
fadd(F,mr,nr,X,nr,beta,C11,ldc,C11,ldc);
// S2 = S1 - A11 in Y
fsubin(F,la,ca,A11,lda,Y,ca);
// P6 = a S2 T2 in B21
fgemm2 (F, ta, tb, mr, nr, kr, alpha, Y, ca, B12, ldb, F.zero, B21, ldb, H);
// S4 = A12 - S2 in Y
fsub(F,la,ca,A12,lda,Y,ca,Y,ca);
// P1 = a A11 B11 in X
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A11, lda, B11, ldb, F.zero, X, nr, H);
// U2 = P6 + P1 in B21
faddin(F,mr,nr,X,nr,B21,ldb);
// U3 = U2 + P7 in C22
faddin(F,mr,nr,B21,ldb,C22,ldc);
// U4 = U2 + P5 in B21
faddin(F,mr,nr,C12,ldc,B21,ldb);
// U6 = U3 - P4 in C21
fsub(F,mr,nr,C22,ldc,C21,ldc,C21,ldc);
// U1 = P1 + P2 in C11
faddin(F,mr,nr,X,nr,C11,ldc);
fflas_delete (X);
// U7 = U3 + P5 in C22
faddin(F,mr,nr,C12,ldc,C22,ldc);
// P3 = a S4 B22 in C12
fgemm2 (F, ta, tb, mr, nr, kr, alpha, Y, ca, B22, ldb, F.zero, C12, ldc, H);
fflas_delete (Y);
// U5 = U4 + P3 in C12
faddin(F,mr,nr,B21,ldb,C12,ldc);
} // WinogradAccOld
template < class Field ,class FieldTrait>
inline void WinogradAcc_L_S (const Field& F,
const FFLAS_TRANSPOSE ta,
const FFLAS_TRANSPOSE tb,
const size_t mr, const size_t nr, const size_t kr,
const typename Field::Element alpha,
typename Field::Element_ptr A,const size_t lda,
const typename Field::Element_ptr B,const size_t ldb,
const typename Field::Element beta,
typename Field::Element_ptr C, const size_t ldc,
const MMHelper<Field, MMHelperAlgo::Winograd, FieldTrait > & WH
)
{
MMHelper<Field, MMHelperAlgo::Winograd, FieldTrait > H = WH ;
H.recLevel = H.recLevel - 1 ;
FFLASFFPACK_check(!F.isZero(beta));
typename Field::Element malpha ;
F.neg(malpha,alpha);
// A, B and c submatrices
typename Field::Element_ptr A11=A, A12, A21, A22;
const typename Field::Element_ptr B11=B, B12, B21, B22;
typename Field::Element_ptr C11=C, C12=C+nr, C21=C+mr*ldc, C22=C21+nr;
typename Field::Element mbeta ;
F.neg(mbeta,beta);
size_t la, ca, lb, cb; // lines and columns in A,B sub matrices
// Three temporary submatrices are required
if (ta == FflasTrans) {
A21 = A + mr;
A12 = A + kr*lda;
A22 = A12 + mr;
la = kr ;
ca = mr ;
}
else { // ta == FflasNoTrans
A12 = A + kr;
A21 = A + mr*lda;
A22 = A21 + kr;
la = mr ;
ca = kr ;
}
if (tb == FflasTrans) {
B21 = B + kr;
B12 = B + nr*ldb;
B22 = B12 + kr;
lb = nr ;
cb = kr ;
}
else { // ta == FflasNoTrans
B12 = B + nr;
B21 = B + kr*ldb;
B22 = B21 + nr;
lb = kr ;
cb = nr ;
}
FFLASFFPACK_check(mr == nr && kr == nr);
// Z1 = C22 - C12 in C22
fsubin(F,mr,nr,C12,ldc,C22,ldc);
// Z2 = C21 - Z1 in C21
fsubin(F,mr,nr,C22,ldc,C21,ldc);
// S3 = A11 - A21 in X
typename Field::Element_ptr X = fflas_new (F, mr, nr);
fsub(F,la,ca,A11,lda,A21,lda,X,ca);
// S1 = A21 + A22 in A21
faddin(F,la,ca,A22,lda,A21,lda);
// T3 = B22 - B12 in Y ;
typename Field::Element_ptr Y = fflas_new (F, mr, kr);
fsub(F,lb,cb,B22,ldb,B12,ldb,Y,cb);
// P7 = a S3 T3 + b Z1 in C22
fgemm2 (F, ta, tb, mr, nr, kr, alpha, X, ca, Y, cb, beta, C22, ldc, H);
// T1 = B12 - B11 in X
fsub(F,lb,cb,B12,ldb,B11,ldb,X,cb);
// T2 = B22 - T1 in Y
fsub(F,lb,cb,B22,ldb,X,cb,Y,cb);
// P5 = a S1 T1 + b C12 in C12
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A21, lda, X, cb, beta, C12, ldc, H);
// S2 = S1 - A11 in A21
fsubin(F,la,ca,A11,lda,A21,lda);
// P1 = a A11 B11 in X
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A11, lda, B11, ldb, F.zero, X, nr, H);
// S4 = A12 - S2 in A11
fsub(F,la,ca,A12,lda,A21,lda,A11,lda);
// P2 = a A12 B21 + b C11 in C11;
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A12, lda, B21, ldb, beta, C11, ldc, H);
// U1 = P1 + P2 in C11
faddin(F,mr,nr,X,nr,C11,ldc);
// P6 = a S2 T2 in A12
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A21, lda, Y, cb, F.zero, A12, lda, H);
// T4 = T2 - B21 in Y
fsubin(F,lb,cb,B21,ldb,Y,cb);
// W2 = a A22 T4 in A21
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A22, lda, Y, cb, F.zero, A21, lda, H);
// P4 = W2 - beta Z2 in C21
fadd(F,mr,nr,A21,lda,mbeta,C21,ldc,C21,ldc);
// U2 = P6 + P1 in X
faddin(F,mr,nr,A12,lda,X,nr);
// U3 = U2 + P7 in C22
faddin(F,mr,nr,X,nr,C22,ldc);
// U6 = U3 - P4 in C21
fsub(F,mr,nr,C22,ldc,C21,ldc,C21,ldc);
// U7 = U3 + P5 in C22
faddin(F,mr,nr,C12,ldc,C22,ldc);
// U4 = U2 + P5 in C12
faddin(F,mr,nr,X,nr,C12,ldc);
fflas_delete (X);
// W3 = a S4 B22 in Y
fgemm2 (F, ta, tb, mr, nr, kr, alpha, A11, lda, B22, ldb, F.zero, Y, nr, H);
// U5 = U4 + W3 in C12
faddin(F,mr,nr,Y,nr,C12,ldc);
fflas_delete (Y);
} // WinogradAccOld
} // BLAS3
} // FFLAS
#endif // __FFLASFFPACK_fgemm_winograd_acc_ip_INL
|