1 /* 2 Defines the basic matrix operations for the BAIJ (compressed row) 3 matrix storage format. 4 */ 5 #include "src/mat/impls/baij/seq/baij.h" 6 #include "src/inline/spops.h" 7 #include "petscsys.h" /*I "petscmat.h" I*/ 8 9 #include "src/inline/ilu.h" 10 11 #undef __FUNCT__ 12 #define __FUNCT__ "MatInvertBlockDiagonal_SeqBAIJ" 13 PetscErrorCode MatInvertBlockDiagonal_SeqBAIJ(Mat A) 14 { 15 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*) A->data; 16 PetscErrorCode ierr; 17 PetscInt *diag_offset,i,bs = A->bs,mbs = a->mbs; 18 PetscScalar *v = a->a,*odiag,*diag,*mdiag; 19 20 PetscFunctionBegin; 21 if (a->idiagvalid) PetscFunctionReturn(0); 22 ierr = MatMarkDiagonal_SeqBAIJ(A);CHKERRQ(ierr); 23 diag_offset = a->diag; 24 if (!a->idiag) { 25 ierr = PetscMalloc(2*bs*bs*mbs*sizeof(PetscScalar),&a->idiag);CHKERRQ(ierr); 26 } 27 diag = a->idiag; 28 mdiag = a->idiag+bs*bs*mbs; 29 /* factor and invert each block */ 30 switch (bs){ 31 case 2: 32 for (i=0; i<mbs; i++) { 33 odiag = v + 4*diag_offset[i]; 34 diag[0] = odiag[0]; diag[1] = odiag[1]; diag[2] = odiag[2]; diag[3] = odiag[3]; 35 mdiag[0] = odiag[0]; mdiag[1] = odiag[1]; mdiag[2] = odiag[2]; mdiag[3] = odiag[3]; 36 ierr = Kernel_A_gets_inverse_A_2(diag);CHKERRQ(ierr); 37 diag += 4; 38 mdiag += 4; 39 } 40 break; 41 case 3: 42 for (i=0; i<mbs; i++) { 43 odiag = v + 9*diag_offset[i]; 44 diag[0] = odiag[0]; diag[1] = odiag[1]; diag[2] = odiag[2]; diag[3] = odiag[3]; 45 diag[4] = odiag[4]; diag[5] = odiag[5]; diag[6] = odiag[6]; diag[7] = odiag[7]; 46 diag[8] = odiag[8]; 47 mdiag[0] = odiag[0]; mdiag[1] = odiag[1]; mdiag[2] = odiag[2]; mdiag[3] = odiag[3]; 48 mdiag[4] = odiag[4]; mdiag[5] = odiag[5]; mdiag[6] = odiag[6]; mdiag[7] = odiag[7]; 49 mdiag[8] = odiag[8]; 50 ierr = Kernel_A_gets_inverse_A_3(diag);CHKERRQ(ierr); 51 diag += 9; 52 mdiag += 9; 53 } 54 break; 55 case 4: 56 for (i=0; i<mbs; i++) { 57 odiag = v + 16*diag_offset[i]; 58 ierr = PetscMemcpy(diag,odiag,16*sizeof(PetscScalar));CHKERRQ(ierr); 59 ierr = PetscMemcpy(mdiag,odiag,16*sizeof(PetscScalar));CHKERRQ(ierr); 60 ierr = Kernel_A_gets_inverse_A_4(diag);CHKERRQ(ierr); 61 diag += 16; 62 mdiag += 16; 63 } 64 break; 65 case 5: 66 for (i=0; i<mbs; i++) { 67 odiag = v + 25*diag_offset[i]; 68 ierr = PetscMemcpy(diag,odiag,25*sizeof(PetscScalar));CHKERRQ(ierr); 69 ierr = PetscMemcpy(mdiag,odiag,25*sizeof(PetscScalar));CHKERRQ(ierr); 70 ierr = Kernel_A_gets_inverse_A_5(diag);CHKERRQ(ierr); 71 diag += 25; 72 mdiag += 25; 73 } 74 break; 75 default: 76 SETERRQ1(PETSC_ERR_SUP,"not supported for block size %D",bs); 77 } 78 a->idiagvalid = PETSC_TRUE; 79 PetscFunctionReturn(0); 80 } 81 82 #undef __FUNCT__ 83 #define __FUNCT__ "MatPBRelax_SeqBAIJ_2" 84 PetscErrorCode MatPBRelax_SeqBAIJ_2(Mat A,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,PetscInt its,PetscInt lits,Vec xx) 85 { 86 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 87 PetscScalar *x,x1,x2,s1,s2; 88 const PetscScalar *v,*aa = a->a, *b, *idiag,*mdiag; 89 PetscErrorCode ierr; 90 PetscInt m = a->mbs,i,i2,nz,idx; 91 const PetscInt *diag,*ai = a->i,*aj = a->j,*vi; 92 93 PetscFunctionBegin; 94 its = its*lits; 95 if (its <= 0) SETERRQ2(PETSC_ERR_ARG_WRONG,"Relaxation requires global its %D and local its %D both positive",its,lits); 96 if (fshift) SETERRQ(PETSC_ERR_SUP,"Sorry, no support for diagonal shift"); 97 if (omega != 1.0) SETERRQ(PETSC_ERR_SUP,"Sorry, no support for non-trivial relaxation factor"); 98 if ((flag & SOR_EISENSTAT) ||(flag & SOR_APPLY_UPPER) || (flag & SOR_APPLY_LOWER) ) SETERRQ(PETSC_ERR_SUP,"Sorry, no support for Eisenstat trick"); 99 if (its > 1) SETERRQ(PETSC_ERR_SUP,"Sorry, no support yet for multiple point block SOR iterations"); 100 101 if (!a->idiagvalid){ierr = MatInvertBlockDiagonal_SeqBAIJ(A);CHKERRQ(ierr);} 102 103 diag = a->diag; 104 idiag = a->idiag; 105 ierr = VecGetArray(xx,&x);CHKERRQ(ierr); 106 ierr = VecGetArray(bb,(PetscScalar**)&b);CHKERRQ(ierr); 107 108 if (flag & SOR_ZERO_INITIAL_GUESS) { 109 if (flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP){ 110 x[0] = b[0]*idiag[0] + b[1]*idiag[2]; 111 x[1] = b[0]*idiag[1] + b[1]*idiag[3]; 112 i2 = 2; 113 idiag += 4; 114 for (i=1; i<m; i++) { 115 v = aa + 4*ai[i]; 116 vi = aj + ai[i]; 117 nz = diag[i] - ai[i]; 118 s1 = b[i2]; s2 = b[i2+1]; 119 while (nz--) { 120 idx = 2*(*vi++); 121 x1 = x[idx]; x2 = x[1+idx]; 122 s1 -= v[0]*x1 + v[2]*x2; 123 s2 -= v[1]*x1 + v[3]*x2; 124 v += 4; 125 } 126 x[i2] = idiag[0]*s1 + idiag[2]*s2; 127 x[i2+1] = idiag[1]*s1 + idiag[3]*s2; 128 idiag += 4; 129 i2 += 2; 130 } 131 /* for logging purposes assume number of nonzero in lower half is 1/2 of total */ 132 PetscLogFlops(4*(a->nz)); 133 } 134 if ((flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP) && 135 (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP)) { 136 i2 = 0; 137 mdiag = a->idiag+4*a->mbs; 138 for (i=0; i<m; i++) { 139 x1 = x[i2]; x2 = x[i2+1]; 140 x[i2] = mdiag[0]*x1 + mdiag[2]*x2; 141 x[i2+1] = mdiag[1]*x1 + mdiag[3]*x2; 142 mdiag += 4; 143 i2 += 2; 144 } 145 PetscLogFlops(6*m); 146 } else if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP) { 147 ierr = PetscMemcpy(x,b,A->m*sizeof(PetscScalar));CHKERRQ(ierr); 148 } 149 if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP){ 150 idiag = a->idiag+4*a->mbs - 4; 151 i2 = 2*m - 2; 152 x1 = x[i2]; x2 = x[i2+1]; 153 x[i2] = idiag[0]*x1 + idiag[2]*x2; 154 x[i2+1] = idiag[1]*x1 + idiag[3]*x2; 155 idiag -= 4; 156 i2 -= 2; 157 for (i=m-2; i>=0; i--) { 158 v = aa + 4*(diag[i]+1); 159 vi = aj + diag[i] + 1; 160 nz = ai[i+1] - diag[i] - 1; 161 s1 = x[i2]; s2 = x[i2+1]; 162 while (nz--) { 163 idx = 2*(*vi++); 164 x1 = x[idx]; x2 = x[1+idx]; 165 s1 -= v[0]*x1 + v[2]*x2; 166 s2 -= v[1]*x1 + v[3]*x2; 167 v += 4; 168 } 169 x[i2] = idiag[0]*s1 + idiag[2]*s2; 170 x[i2+1] = idiag[1]*s1 + idiag[3]*s2; 171 idiag -= 4; 172 i2 -= 2; 173 } 174 PetscLogFlops(4*(a->nz)); 175 } 176 } else { 177 SETERRQ(PETSC_ERR_SUP,"Only supports point block SOR with zero initial guess"); 178 } 179 ierr = VecRestoreArray(xx,&x);CHKERRQ(ierr); 180 ierr = VecRestoreArray(bb,(PetscScalar**)&b);CHKERRQ(ierr); 181 PetscFunctionReturn(0); 182 } 183 184 #undef __FUNCT__ 185 #define __FUNCT__ "MatPBRelax_SeqBAIJ_3" 186 PetscErrorCode MatPBRelax_SeqBAIJ_3(Mat A,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,PetscInt its,PetscInt lits,Vec xx) 187 { 188 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 189 PetscScalar *x,x1,x2,x3,s1,s2,s3; 190 const PetscScalar *v,*aa = a->a, *b, *idiag,*mdiag; 191 PetscErrorCode ierr; 192 PetscInt m = a->mbs,i,i2,nz,idx; 193 const PetscInt *diag,*ai = a->i,*aj = a->j,*vi; 194 195 PetscFunctionBegin; 196 its = its*lits; 197 if (its <= 0) SETERRQ2(PETSC_ERR_ARG_WRONG,"Relaxation requires global its %D and local its %D both positive",its,lits); 198 if (fshift) SETERRQ(PETSC_ERR_SUP,"Sorry, no support for diagonal shift"); 199 if (omega != 1.0) SETERRQ(PETSC_ERR_SUP,"Sorry, no support for non-trivial relaxation factor"); 200 if ((flag & SOR_EISENSTAT) ||(flag & SOR_APPLY_UPPER) || (flag & SOR_APPLY_LOWER) ) SETERRQ(PETSC_ERR_SUP,"Sorry, no support for Eisenstat trick"); 201 if (its > 1) SETERRQ(PETSC_ERR_SUP,"Sorry, no support yet for multiple point block SOR iterations"); 202 203 if (!a->idiagvalid){ierr = MatInvertBlockDiagonal_SeqBAIJ(A);CHKERRQ(ierr);} 204 205 diag = a->diag; 206 idiag = a->idiag; 207 ierr = VecGetArray(xx,&x);CHKERRQ(ierr); 208 ierr = VecGetArray(bb,(PetscScalar**)&b);CHKERRQ(ierr); 209 210 if (flag & SOR_ZERO_INITIAL_GUESS) { 211 if (flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP){ 212 x[0] = b[0]*idiag[0] + b[1]*idiag[3] + b[2]*idiag[6]; 213 x[1] = b[0]*idiag[1] + b[1]*idiag[4] + b[2]*idiag[7]; 214 x[2] = b[0]*idiag[2] + b[1]*idiag[5] + b[2]*idiag[8]; 215 i2 = 3; 216 idiag += 9; 217 for (i=1; i<m; i++) { 218 v = aa + 9*ai[i]; 219 vi = aj + ai[i]; 220 nz = diag[i] - ai[i]; 221 s1 = b[i2]; s2 = b[i2+1]; s3 = b[i2+2]; 222 while (nz--) { 223 idx = 3*(*vi++); 224 x1 = x[idx]; x2 = x[1+idx];x3 = x[2+idx]; 225 s1 -= v[0]*x1 + v[3]*x2 + v[6]*x3; 226 s2 -= v[1]*x1 + v[4]*x2 + v[7]*x3; 227 s3 -= v[2]*x1 + v[5]*x2 + v[8]*x3; 228 v += 9; 229 } 230 x[i2] = idiag[0]*s1 + idiag[3]*s2 + idiag[6]*s3; 231 x[i2+1] = idiag[1]*s1 + idiag[4]*s2 + idiag[7]*s3; 232 x[i2+2] = idiag[2]*s1 + idiag[5]*s2 + idiag[8]*s3; 233 idiag += 9; 234 i2 += 3; 235 } 236 /* for logging purposes assume number of nonzero in lower half is 1/2 of total */ 237 PetscLogFlops(9*(a->nz)); 238 } 239 if ((flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP) && 240 (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP)) { 241 i2 = 0; 242 mdiag = a->idiag+9*a->mbs; 243 for (i=0; i<m; i++) { 244 x1 = x[i2]; x2 = x[i2+1]; x3 = x[i2+2]; 245 x[i2] = mdiag[0]*x1 + mdiag[3]*x2 + mdiag[6]*x3; 246 x[i2+1] = mdiag[1]*x1 + mdiag[4]*x2 + mdiag[7]*x3; 247 x[i2+2] = mdiag[2]*x1 + mdiag[5]*x2 + mdiag[8]*x3; 248 mdiag += 9; 249 i2 += 3; 250 } 251 PetscLogFlops(15*m); 252 } else if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP) { 253 ierr = PetscMemcpy(x,b,A->m*sizeof(PetscScalar));CHKERRQ(ierr); 254 } 255 if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP){ 256 idiag = a->idiag+9*a->mbs - 9; 257 i2 = 3*m - 3; 258 x1 = x[i2]; x2 = x[i2+1]; x3 = x[i2+2]; 259 x[i2] = idiag[0]*x1 + idiag[3]*x2 + idiag[6]*x3; 260 x[i2+1] = idiag[1]*x1 + idiag[4]*x2 + idiag[7]*x3; 261 x[i2+2] = idiag[2]*x1 + idiag[5]*x2 + idiag[8]*x3; 262 idiag -= 9; 263 i2 -= 3; 264 for (i=m-2; i>=0; i--) { 265 v = aa + 9*(diag[i]+1); 266 vi = aj + diag[i] + 1; 267 nz = ai[i+1] - diag[i] - 1; 268 s1 = x[i2]; s2 = x[i2+1]; s3 = x[i2+2]; 269 while (nz--) { 270 idx = 3*(*vi++); 271 x1 = x[idx]; x2 = x[1+idx]; x3 = x[2+idx]; 272 s1 -= v[0]*x1 + v[3]*x2 + v[6]*x3; 273 s2 -= v[1]*x1 + v[4]*x2 + v[7]*x3; 274 s3 -= v[2]*x1 + v[5]*x2 + v[8]*x3; 275 v += 9; 276 } 277 x[i2] = idiag[0]*s1 + idiag[3]*s2 + idiag[6]*s3; 278 x[i2+1] = idiag[1]*s1 + idiag[4]*s2 + idiag[7]*s3; 279 x[i2+2] = idiag[2]*s1 + idiag[5]*s2 + idiag[8]*s3; 280 idiag -= 9; 281 i2 -= 3; 282 } 283 PetscLogFlops(9*(a->nz)); 284 } 285 } else { 286 SETERRQ(PETSC_ERR_SUP,"Only supports point block SOR with zero initial guess"); 287 } 288 ierr = VecRestoreArray(xx,&x);CHKERRQ(ierr); 289 ierr = VecRestoreArray(bb,(PetscScalar**)&b);CHKERRQ(ierr); 290 PetscFunctionReturn(0); 291 } 292 293 #undef __FUNCT__ 294 #define __FUNCT__ "MatPBRelax_SeqBAIJ_4" 295 PetscErrorCode MatPBRelax_SeqBAIJ_4(Mat A,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,PetscInt its,PetscInt lits,Vec xx) 296 { 297 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 298 PetscScalar *x,x1,x2,x3,x4,s1,s2,s3,s4; 299 const PetscScalar *v,*aa = a->a, *b, *idiag,*mdiag; 300 PetscErrorCode ierr; 301 PetscInt m = a->mbs,i,i2,nz,idx; 302 const PetscInt *diag,*ai = a->i,*aj = a->j,*vi; 303 304 PetscFunctionBegin; 305 its = its*lits; 306 if (its <= 0) SETERRQ2(PETSC_ERR_ARG_WRONG,"Relaxation requires global its %D and local its %D both positive",its,lits); 307 if (fshift) SETERRQ(PETSC_ERR_SUP,"Sorry, no support for diagonal shift"); 308 if (omega != 1.0) SETERRQ(PETSC_ERR_SUP,"Sorry, no support for non-trivial relaxation factor"); 309 if ((flag & SOR_EISENSTAT) ||(flag & SOR_APPLY_UPPER) || (flag & SOR_APPLY_LOWER) ) SETERRQ(PETSC_ERR_SUP,"Sorry, no support for Eisenstat trick"); 310 if (its > 1) SETERRQ(PETSC_ERR_SUP,"Sorry, no support yet for multiple point block SOR iterations"); 311 312 if (!a->idiagvalid){ierr = MatInvertBlockDiagonal_SeqBAIJ(A);CHKERRQ(ierr);} 313 314 diag = a->diag; 315 idiag = a->idiag; 316 ierr = VecGetArray(xx,&x);CHKERRQ(ierr); 317 ierr = VecGetArray(bb,(PetscScalar**)&b);CHKERRQ(ierr); 318 319 if (flag & SOR_ZERO_INITIAL_GUESS) { 320 if (flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP){ 321 x[0] = b[0]*idiag[0] + b[1]*idiag[4] + b[2]*idiag[8] + b[3]*idiag[12]; 322 x[1] = b[0]*idiag[1] + b[1]*idiag[5] + b[2]*idiag[9] + b[3]*idiag[13]; 323 x[2] = b[0]*idiag[2] + b[1]*idiag[6] + b[2]*idiag[10] + b[3]*idiag[14]; 324 x[3] = b[0]*idiag[3] + b[1]*idiag[7] + b[2]*idiag[11] + b[3]*idiag[15]; 325 i2 = 4; 326 idiag += 16; 327 for (i=1; i<m; i++) { 328 v = aa + 16*ai[i]; 329 vi = aj + ai[i]; 330 nz = diag[i] - ai[i]; 331 s1 = b[i2]; s2 = b[i2+1]; s3 = b[i2+2]; s4 = b[i2+3]; 332 while (nz--) { 333 idx = 4*(*vi++); 334 x1 = x[idx]; x2 = x[1+idx]; x3 = x[2+idx]; x4 = x[3+idx]; 335 s1 -= v[0]*x1 + v[4]*x2 + v[8]*x3 + v[12]*x4; 336 s2 -= v[1]*x1 + v[5]*x2 + v[9]*x3 + v[13]*x4; 337 s3 -= v[2]*x1 + v[6]*x2 + v[10]*x3 + v[14]*x4; 338 s4 -= v[3]*x1 + v[7]*x2 + v[11]*x3 + v[15]*x4; 339 v += 16; 340 } 341 x[i2] = idiag[0]*s1 + idiag[4]*s2 + idiag[8]*s3 + idiag[12]*s4; 342 x[i2+1] = idiag[1]*s1 + idiag[5]*s2 + idiag[9]*s3 + idiag[13]*s4; 343 x[i2+2] = idiag[2]*s1 + idiag[6]*s2 + idiag[10]*s3 + idiag[14]*s4; 344 x[i2+3] = idiag[3]*s1 + idiag[7]*s2 + idiag[11]*s3 + idiag[15]*s4; 345 idiag += 16; 346 i2 += 4; 347 } 348 /* for logging purposes assume number of nonzero in lower half is 1/2 of total */ 349 PetscLogFlops(16*(a->nz)); 350 } 351 if ((flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP) && 352 (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP)) { 353 i2 = 0; 354 mdiag = a->idiag+16*a->mbs; 355 for (i=0; i<m; i++) { 356 x1 = x[i2]; x2 = x[i2+1]; x3 = x[i2+2]; x4 = x[i2+3]; 357 x[i2] = mdiag[0]*x1 + mdiag[4]*x2 + mdiag[8]*x3 + mdiag[12]*x4; 358 x[i2+1] = mdiag[1]*x1 + mdiag[5]*x2 + mdiag[9]*x3 + mdiag[13]*x4; 359 x[i2+2] = mdiag[2]*x1 + mdiag[6]*x2 + mdiag[10]*x3 + mdiag[14]*x4; 360 x[i2+3] = mdiag[3]*x1 + mdiag[7]*x2 + mdiag[11]*x3 + mdiag[15]*x4; 361 mdiag += 16; 362 i2 += 4; 363 } 364 PetscLogFlops(28*m); 365 } else if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP) { 366 ierr = PetscMemcpy(x,b,A->m*sizeof(PetscScalar));CHKERRQ(ierr); 367 } 368 if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP){ 369 idiag = a->idiag+16*a->mbs - 16; 370 i2 = 4*m - 4; 371 x1 = x[i2]; x2 = x[i2+1]; x3 = x[i2+2]; x4 = x[i2+3]; 372 x[i2] = idiag[0]*x1 + idiag[4]*x2 + idiag[8]*x3 + idiag[12]*x4; 373 x[i2+1] = idiag[1]*x1 + idiag[5]*x2 + idiag[9]*x3 + idiag[13]*x4; 374 x[i2+2] = idiag[2]*x1 + idiag[6]*x2 + idiag[10]*x3 + idiag[14]*x4; 375 x[i2+3] = idiag[3]*x1 + idiag[7]*x2 + idiag[11]*x3 + idiag[15]*x4; 376 idiag -= 16; 377 i2 -= 4; 378 for (i=m-2; i>=0; i--) { 379 v = aa + 16*(diag[i]+1); 380 vi = aj + diag[i] + 1; 381 nz = ai[i+1] - diag[i] - 1; 382 s1 = x[i2]; s2 = x[i2+1]; s3 = x[i2+2]; s4 = x[i2+3]; 383 while (nz--) { 384 idx = 4*(*vi++); 385 x1 = x[idx]; x2 = x[1+idx]; x3 = x[2+idx]; x4 = x[3+idx]; 386 s1 -= v[0]*x1 + v[4]*x2 + v[8]*x3 + v[12]*x4; 387 s2 -= v[1]*x1 + v[5]*x2 + v[9]*x3 + v[13]*x4; 388 s3 -= v[2]*x1 + v[6]*x2 + v[10]*x3 + v[14]*x4; 389 s4 -= v[3]*x1 + v[7]*x2 + v[11]*x3 + v[15]*x4; 390 v += 16; 391 } 392 x[i2] = idiag[0]*s1 + idiag[4]*s2 + idiag[8]*s3 + idiag[12]*s4; 393 x[i2+1] = idiag[1]*s1 + idiag[5]*s2 + idiag[9]*s3 + idiag[13]*s4; 394 x[i2+2] = idiag[2]*s1 + idiag[6]*s2 + idiag[10]*s3 + idiag[14]*s4; 395 x[i2+3] = idiag[3]*s1 + idiag[7]*s2 + idiag[11]*s3 + idiag[15]*s4; 396 idiag -= 16; 397 i2 -= 4; 398 } 399 PetscLogFlops(16*(a->nz)); 400 } 401 } else { 402 SETERRQ(PETSC_ERR_SUP,"Only supports point block SOR with zero initial guess"); 403 } 404 ierr = VecRestoreArray(xx,&x);CHKERRQ(ierr); 405 ierr = VecRestoreArray(bb,(PetscScalar**)&b);CHKERRQ(ierr); 406 PetscFunctionReturn(0); 407 } 408 409 #undef __FUNCT__ 410 #define __FUNCT__ "MatPBRelax_SeqBAIJ_5" 411 PetscErrorCode MatPBRelax_SeqBAIJ_5(Mat A,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,PetscInt its,PetscInt lits,Vec xx) 412 { 413 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 414 PetscScalar *x,x1,x2,x3,x4,x5,s1,s2,s3,s4,s5; 415 const PetscScalar *v,*aa = a->a, *b, *idiag,*mdiag; 416 PetscErrorCode ierr; 417 PetscInt m = a->mbs,i,i2,nz,idx; 418 const PetscInt *diag,*ai = a->i,*aj = a->j,*vi; 419 420 PetscFunctionBegin; 421 its = its*lits; 422 if (its <= 0) SETERRQ2(PETSC_ERR_ARG_WRONG,"Relaxation requires global its %D and local its %D both positive",its,lits); 423 if (fshift) SETERRQ(PETSC_ERR_SUP,"Sorry, no support for diagonal shift"); 424 if (omega != 1.0) SETERRQ(PETSC_ERR_SUP,"Sorry, no support for non-trivial relaxation factor"); 425 if ((flag & SOR_EISENSTAT) ||(flag & SOR_APPLY_UPPER) || (flag & SOR_APPLY_LOWER) ) SETERRQ(PETSC_ERR_SUP,"Sorry, no support for Eisenstat trick"); 426 if (its > 1) SETERRQ(PETSC_ERR_SUP,"Sorry, no support yet for multiple point block SOR iterations"); 427 428 if (!a->idiagvalid){ierr = MatInvertBlockDiagonal_SeqBAIJ(A);CHKERRQ(ierr);} 429 430 diag = a->diag; 431 idiag = a->idiag; 432 ierr = VecGetArray(xx,&x);CHKERRQ(ierr); 433 ierr = VecGetArray(bb,(PetscScalar**)&b);CHKERRQ(ierr); 434 435 if (flag & SOR_ZERO_INITIAL_GUESS) { 436 if (flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP){ 437 x[0] = b[0]*idiag[0] + b[1]*idiag[5] + b[2]*idiag[10] + b[3]*idiag[15] + b[4]*idiag[20]; 438 x[1] = b[0]*idiag[1] + b[1]*idiag[6] + b[2]*idiag[11] + b[3]*idiag[16] + b[4]*idiag[21]; 439 x[2] = b[0]*idiag[2] + b[1]*idiag[7] + b[2]*idiag[12] + b[3]*idiag[17] + b[4]*idiag[22]; 440 x[3] = b[0]*idiag[3] + b[1]*idiag[8] + b[2]*idiag[13] + b[3]*idiag[18] + b[4]*idiag[23]; 441 x[4] = b[0]*idiag[4] + b[1]*idiag[9] + b[2]*idiag[14] + b[3]*idiag[19] + b[4]*idiag[24]; 442 i2 = 5; 443 idiag += 25; 444 for (i=1; i<m; i++) { 445 v = aa + 25*ai[i]; 446 vi = aj + ai[i]; 447 nz = diag[i] - ai[i]; 448 s1 = b[i2]; s2 = b[i2+1]; s3 = b[i2+2]; s4 = b[i2+3]; s5 = b[i2+4]; 449 while (nz--) { 450 idx = 5*(*vi++); 451 x1 = x[idx]; x2 = x[1+idx]; x3 = x[2+idx]; x4 = x[3+idx]; x5 = x[4+idx]; 452 s1 -= v[0]*x1 + v[5]*x2 + v[10]*x3 + v[15]*x4 + v[20]*x5; 453 s2 -= v[1]*x1 + v[6]*x2 + v[11]*x3 + v[16]*x4 + v[21]*x5; 454 s3 -= v[2]*x1 + v[7]*x2 + v[12]*x3 + v[17]*x4 + v[22]*x5; 455 s4 -= v[3]*x1 + v[8]*x2 + v[13]*x3 + v[18]*x4 + v[23]*x5; 456 s5 -= v[4]*x1 + v[9]*x2 + v[14]*x3 + v[19]*x4 + v[24]*x5; 457 v += 25; 458 } 459 x[i2] = idiag[0]*s1 + idiag[5]*s2 + idiag[10]*s3 + idiag[15]*s4 + idiag[20]*s5; 460 x[i2+1] = idiag[1]*s1 + idiag[6]*s2 + idiag[11]*s3 + idiag[16]*s4 + idiag[21]*s5; 461 x[i2+2] = idiag[2]*s1 + idiag[7]*s2 + idiag[12]*s3 + idiag[17]*s4 + idiag[22]*s5; 462 x[i2+3] = idiag[3]*s1 + idiag[8]*s2 + idiag[13]*s3 + idiag[18]*s4 + idiag[23]*s5; 463 x[i2+4] = idiag[4]*s1 + idiag[9]*s2 + idiag[14]*s3 + idiag[19]*s4 + idiag[24]*s5; 464 idiag += 25; 465 i2 += 5; 466 } 467 /* for logging purposes assume number of nonzero in lower half is 1/2 of total */ 468 PetscLogFlops(25*(a->nz)); 469 } 470 if ((flag & SOR_FORWARD_SWEEP || flag & SOR_LOCAL_FORWARD_SWEEP) && 471 (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP)) { 472 i2 = 0; 473 mdiag = a->idiag+25*a->mbs; 474 for (i=0; i<m; i++) { 475 x1 = x[i2]; x2 = x[i2+1]; x3 = x[i2+2]; x4 = x[i2+3]; x5 = x[i2+4]; 476 x[i2] = mdiag[0]*x1 + mdiag[5]*x2 + mdiag[10]*x3 + mdiag[15]*x4 + mdiag[20]*x5; 477 x[i2+1] = mdiag[1]*x1 + mdiag[6]*x2 + mdiag[11]*x3 + mdiag[16]*x4 + mdiag[21]*x5; 478 x[i2+2] = mdiag[2]*x1 + mdiag[7]*x2 + mdiag[12]*x3 + mdiag[17]*x4 + mdiag[22]*x5; 479 x[i2+3] = mdiag[3]*x1 + mdiag[8]*x2 + mdiag[13]*x3 + mdiag[18]*x4 + mdiag[23]*x5; 480 x[i2+4] = mdiag[4]*x1 + mdiag[9]*x2 + mdiag[14]*x3 + mdiag[19]*x4 + mdiag[24]*x5; 481 mdiag += 25; 482 i2 += 5; 483 } 484 PetscLogFlops(45*m); 485 } else if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP) { 486 ierr = PetscMemcpy(x,b,A->m*sizeof(PetscScalar));CHKERRQ(ierr); 487 } 488 if (flag & SOR_BACKWARD_SWEEP || flag & SOR_LOCAL_BACKWARD_SWEEP){ 489 idiag = a->idiag+25*a->mbs - 25; 490 i2 = 5*m - 5; 491 x1 = x[i2]; x2 = x[i2+1]; x3 = x[i2+2]; x4 = x[i2+3]; x5 = x[i2+4]; 492 x[i2] = idiag[0]*x1 + idiag[5]*x2 + idiag[10]*x3 + idiag[15]*x4 + idiag[20]*x5; 493 x[i2+1] = idiag[1]*x1 + idiag[6]*x2 + idiag[11]*x3 + idiag[16]*x4 + idiag[21]*x5; 494 x[i2+2] = idiag[2]*x1 + idiag[7]*x2 + idiag[12]*x3 + idiag[17]*x4 + idiag[22]*x5; 495 x[i2+3] = idiag[3]*x1 + idiag[8]*x2 + idiag[13]*x3 + idiag[18]*x4 + idiag[23]*x5; 496 x[i2+4] = idiag[4]*x1 + idiag[9]*x2 + idiag[14]*x3 + idiag[19]*x4 + idiag[24]*x5; 497 idiag -= 25; 498 i2 -= 5; 499 for (i=m-2; i>=0; i--) { 500 v = aa + 25*(diag[i]+1); 501 vi = aj + diag[i] + 1; 502 nz = ai[i+1] - diag[i] - 1; 503 s1 = x[i2]; s2 = x[i2+1]; s3 = x[i2+2]; s4 = x[i2+3]; s5 = x[i2+4]; 504 while (nz--) { 505 idx = 5*(*vi++); 506 x1 = x[idx]; x2 = x[1+idx]; x3 = x[2+idx]; x4 = x[3+idx]; x5 = x[4+idx]; 507 s1 -= v[0]*x1 + v[5]*x2 + v[10]*x3 + v[15]*x4 + v[20]*x5; 508 s2 -= v[1]*x1 + v[6]*x2 + v[11]*x3 + v[16]*x4 + v[21]*x5; 509 s3 -= v[2]*x1 + v[7]*x2 + v[12]*x3 + v[17]*x4 + v[22]*x5; 510 s4 -= v[3]*x1 + v[8]*x2 + v[13]*x3 + v[18]*x4 + v[23]*x5; 511 s5 -= v[4]*x1 + v[9]*x2 + v[14]*x3 + v[19]*x4 + v[24]*x5; 512 v += 25; 513 } 514 x[i2] = idiag[0]*s1 + idiag[5]*s2 + idiag[10]*s3 + idiag[15]*s4 + idiag[20]*s5; 515 x[i2+1] = idiag[1]*s1 + idiag[6]*s2 + idiag[11]*s3 + idiag[16]*s4 + idiag[21]*s5; 516 x[i2+2] = idiag[2]*s1 + idiag[7]*s2 + idiag[12]*s3 + idiag[17]*s4 + idiag[22]*s5; 517 x[i2+3] = idiag[3]*s1 + idiag[8]*s2 + idiag[13]*s3 + idiag[18]*s4 + idiag[23]*s5; 518 x[i2+4] = idiag[4]*s1 + idiag[9]*s2 + idiag[14]*s3 + idiag[19]*s4 + idiag[24]*s5; 519 idiag -= 25; 520 i2 -= 5; 521 } 522 PetscLogFlops(25*(a->nz)); 523 } 524 } else { 525 SETERRQ(PETSC_ERR_SUP,"Only supports point block SOR with zero initial guess"); 526 } 527 ierr = VecRestoreArray(xx,&x);CHKERRQ(ierr); 528 ierr = VecRestoreArray(bb,(PetscScalar**)&b);CHKERRQ(ierr); 529 PetscFunctionReturn(0); 530 } 531 532 /* 533 Special version for Fun3d sequential benchmark 534 */ 535 #if defined(PETSC_HAVE_FORTRAN_CAPS) 536 #define matsetvaluesblocked4_ MATSETVALUESBLOCKED4 537 #elif !defined(PETSC_HAVE_FORTRAN_UNDERSCORE) 538 #define matsetvaluesblocked4_ matsetvaluesblocked4 539 #endif 540 541 EXTERN_C_BEGIN 542 #undef __FUNCT__ 543 #define __FUNCT__ "matsetvaluesblocked4_" 544 void matsetvaluesblocked4_(Mat *AA,PetscInt *mm,const PetscInt im[],PetscInt *nn,const PetscInt in[],const PetscScalar v[]) 545 { 546 Mat A = *AA; 547 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 548 PetscInt *rp,k,low,high,t,ii,jj,row,nrow,i,col,l,N,m = *mm,n = *nn; 549 PetscInt *ai=a->i,*ailen=a->ilen; 550 PetscInt *aj=a->j,stepval; 551 const PetscScalar *value = v; 552 MatScalar *ap,*aa = a->a,*bap; 553 554 PetscFunctionBegin; 555 stepval = (n-1)*4; 556 for (k=0; k<m; k++) { /* loop over added rows */ 557 row = im[k]; 558 rp = aj + ai[row]; 559 ap = aa + 16*ai[row]; 560 nrow = ailen[row]; 561 low = 0; 562 for (l=0; l<n; l++) { /* loop over added columns */ 563 col = in[l]; 564 value = v + k*(stepval+4)*4 + l*4; 565 low = 0; high = nrow; 566 while (high-low > 7) { 567 t = (low+high)/2; 568 if (rp[t] > col) high = t; 569 else low = t; 570 } 571 for (i=low; i<high; i++) { 572 if (rp[i] > col) break; 573 if (rp[i] == col) { 574 bap = ap + 16*i; 575 for (ii=0; ii<4; ii++,value+=stepval) { 576 for (jj=ii; jj<16; jj+=4) { 577 bap[jj] += *value++; 578 } 579 } 580 goto noinsert2; 581 } 582 } 583 N = nrow++ - 1; 584 /* shift up all the later entries in this row */ 585 for (ii=N; ii>=i; ii--) { 586 rp[ii+1] = rp[ii]; 587 PetscMemcpy(ap+16*(ii+1),ap+16*(ii),16*sizeof(MatScalar)); 588 } 589 if (N >= i) { 590 PetscMemzero(ap+16*i,16*sizeof(MatScalar)); 591 } 592 rp[i] = col; 593 bap = ap + 16*i; 594 for (ii=0; ii<4; ii++,value+=stepval) { 595 for (jj=ii; jj<16; jj+=4) { 596 bap[jj] = *value++; 597 } 598 } 599 noinsert2:; 600 low = i; 601 } 602 ailen[row] = nrow; 603 } 604 } 605 EXTERN_C_END 606 607 #if defined(PETSC_HAVE_FORTRAN_CAPS) 608 #define matsetvalues4_ MATSETVALUES4 609 #elif !defined(PETSC_HAVE_FORTRAN_UNDERSCORE) 610 #define matsetvalues4_ matsetvalues4 611 #endif 612 613 EXTERN_C_BEGIN 614 #undef __FUNCT__ 615 #define __FUNCT__ "MatSetValues4_" 616 void matsetvalues4_(Mat *AA,PetscInt *mm,PetscInt *im,PetscInt *nn,PetscInt *in,PetscScalar *v) 617 { 618 Mat A = *AA; 619 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 620 PetscInt *rp,k,low,high,t,ii,row,nrow,i,col,l,N,n = *nn,m = *mm; 621 PetscInt *ai=a->i,*ailen=a->ilen; 622 PetscInt *aj=a->j,brow,bcol; 623 PetscInt ridx,cidx; 624 MatScalar *ap,value,*aa=a->a,*bap; 625 626 PetscFunctionBegin; 627 for (k=0; k<m; k++) { /* loop over added rows */ 628 row = im[k]; brow = row/4; 629 rp = aj + ai[brow]; 630 ap = aa + 16*ai[brow]; 631 nrow = ailen[brow]; 632 low = 0; 633 for (l=0; l<n; l++) { /* loop over added columns */ 634 col = in[l]; bcol = col/4; 635 ridx = row % 4; cidx = col % 4; 636 value = v[l + k*n]; 637 low = 0; high = nrow; 638 while (high-low > 7) { 639 t = (low+high)/2; 640 if (rp[t] > bcol) high = t; 641 else low = t; 642 } 643 for (i=low; i<high; i++) { 644 if (rp[i] > bcol) break; 645 if (rp[i] == bcol) { 646 bap = ap + 16*i + 4*cidx + ridx; 647 *bap += value; 648 goto noinsert1; 649 } 650 } 651 N = nrow++ - 1; 652 /* shift up all the later entries in this row */ 653 for (ii=N; ii>=i; ii--) { 654 rp[ii+1] = rp[ii]; 655 PetscMemcpy(ap+16*(ii+1),ap+16*(ii),16*sizeof(MatScalar)); 656 } 657 if (N>=i) { 658 PetscMemzero(ap+16*i,16*sizeof(MatScalar)); 659 } 660 rp[i] = bcol; 661 ap[16*i + 4*cidx + ridx] = value; 662 noinsert1:; 663 low = i; 664 } 665 ailen[brow] = nrow; 666 } 667 } 668 EXTERN_C_END 669 670 /* UGLY, ugly, ugly 671 When MatScalar == PetscScalar the function MatSetValuesBlocked_SeqBAIJ_MatScalar() does 672 not exist. Otherwise ..._MatScalar() takes matrix dlements in single precision and 673 inserts them into the single precision data structure. The function MatSetValuesBlocked_SeqBAIJ() 674 converts the entries into single precision and then calls ..._MatScalar() to put them 675 into the single precision data structures. 676 */ 677 #if defined(PETSC_USE_MAT_SINGLE) 678 EXTERN PetscErrorCode MatSetValuesBlocked_SeqBAIJ_MatScalar(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const MatScalar[],InsertMode); 679 #else 680 #define MatSetValuesBlocked_SeqBAIJ_MatScalar MatSetValuesBlocked_SeqBAIJ 681 #endif 682 683 #define CHUNKSIZE 10 684 685 /* 686 Checks for missing diagonals 687 */ 688 #undef __FUNCT__ 689 #define __FUNCT__ "MatMissingDiagonal_SeqBAIJ" 690 PetscErrorCode MatMissingDiagonal_SeqBAIJ(Mat A) 691 { 692 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 693 PetscErrorCode ierr; 694 PetscInt *diag,*jj = a->j,i; 695 696 PetscFunctionBegin; 697 ierr = MatMarkDiagonal_SeqBAIJ(A);CHKERRQ(ierr); 698 diag = a->diag; 699 for (i=0; i<a->mbs; i++) { 700 if (jj[diag[i]] != i) { 701 SETERRQ1(PETSC_ERR_PLIB,"Matrix is missing diagonal number %D",i); 702 } 703 } 704 PetscFunctionReturn(0); 705 } 706 707 #undef __FUNCT__ 708 #define __FUNCT__ "MatMarkDiagonal_SeqBAIJ" 709 PetscErrorCode MatMarkDiagonal_SeqBAIJ(Mat A) 710 { 711 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 712 PetscErrorCode ierr; 713 PetscInt i,j,*diag,m = a->mbs; 714 715 PetscFunctionBegin; 716 if (a->diag) PetscFunctionReturn(0); 717 718 ierr = PetscMalloc((m+1)*sizeof(PetscInt),&diag);CHKERRQ(ierr); 719 ierr = PetscLogObjectMemory(A,(m+1)*sizeof(PetscInt));CHKERRQ(ierr); 720 for (i=0; i<m; i++) { 721 diag[i] = a->i[i+1]; 722 for (j=a->i[i]; j<a->i[i+1]; j++) { 723 if (a->j[j] == i) { 724 diag[i] = j; 725 break; 726 } 727 } 728 } 729 a->diag = diag; 730 PetscFunctionReturn(0); 731 } 732 733 734 EXTERN PetscErrorCode MatToSymmetricIJ_SeqAIJ(PetscInt,PetscInt*,PetscInt*,PetscInt,PetscInt,PetscInt**,PetscInt**); 735 736 #undef __FUNCT__ 737 #define __FUNCT__ "MatGetRowIJ_SeqBAIJ" 738 static PetscErrorCode MatGetRowIJ_SeqBAIJ(Mat A,PetscInt oshift,PetscTruth symmetric,PetscInt *nn,PetscInt *ia[],PetscInt *ja[],PetscTruth *done) 739 { 740 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 741 PetscErrorCode ierr; 742 PetscInt n = a->mbs,i; 743 744 PetscFunctionBegin; 745 *nn = n; 746 if (!ia) PetscFunctionReturn(0); 747 if (symmetric) { 748 ierr = MatToSymmetricIJ_SeqAIJ(n,a->i,a->j,0,oshift,ia,ja);CHKERRQ(ierr); 749 } else if (oshift == 1) { 750 /* temporarily add 1 to i and j indices */ 751 PetscInt nz = a->i[n]; 752 for (i=0; i<nz; i++) a->j[i]++; 753 for (i=0; i<n+1; i++) a->i[i]++; 754 *ia = a->i; *ja = a->j; 755 } else { 756 *ia = a->i; *ja = a->j; 757 } 758 759 PetscFunctionReturn(0); 760 } 761 762 #undef __FUNCT__ 763 #define __FUNCT__ "MatRestoreRowIJ_SeqBAIJ" 764 static PetscErrorCode MatRestoreRowIJ_SeqBAIJ(Mat A,PetscInt oshift,PetscTruth symmetric,PetscInt *nn,PetscInt *ia[],PetscInt *ja[],PetscTruth *done) 765 { 766 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 767 PetscErrorCode ierr; 768 PetscInt i,n = a->mbs; 769 770 PetscFunctionBegin; 771 if (!ia) PetscFunctionReturn(0); 772 if (symmetric) { 773 ierr = PetscFree(*ia);CHKERRQ(ierr); 774 ierr = PetscFree(*ja);CHKERRQ(ierr); 775 } else if (oshift == 1) { 776 PetscInt nz = a->i[n]-1; 777 for (i=0; i<nz; i++) a->j[i]--; 778 for (i=0; i<n+1; i++) a->i[i]--; 779 } 780 PetscFunctionReturn(0); 781 } 782 783 #undef __FUNCT__ 784 #define __FUNCT__ "MatDestroy_SeqBAIJ" 785 PetscErrorCode MatDestroy_SeqBAIJ(Mat A) 786 { 787 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 788 PetscErrorCode ierr; 789 790 PetscFunctionBegin; 791 #if defined(PETSC_USE_LOG) 792 PetscLogObjectState((PetscObject)A,"Rows=%D, Cols=%D, NZ=%D",A->m,A->n,a->nz); 793 #endif 794 ierr = PetscFree(a->a);CHKERRQ(ierr); 795 if (!a->singlemalloc) { 796 ierr = PetscFree(a->i);CHKERRQ(ierr); 797 ierr = PetscFree(a->j);CHKERRQ(ierr); 798 } 799 if (a->row) { 800 ierr = ISDestroy(a->row);CHKERRQ(ierr); 801 } 802 if (a->col) { 803 ierr = ISDestroy(a->col);CHKERRQ(ierr); 804 } 805 if (a->diag) {ierr = PetscFree(a->diag);CHKERRQ(ierr);} 806 if (a->idiag) {ierr = PetscFree(a->idiag);CHKERRQ(ierr);} 807 if (a->ilen) {ierr = PetscFree(a->ilen);CHKERRQ(ierr);} 808 if (a->imax) {ierr = PetscFree(a->imax);CHKERRQ(ierr);} 809 if (a->solve_work) {ierr = PetscFree(a->solve_work);CHKERRQ(ierr);} 810 if (a->mult_work) {ierr = PetscFree(a->mult_work);CHKERRQ(ierr);} 811 if (a->icol) {ierr = ISDestroy(a->icol);CHKERRQ(ierr);} 812 if (a->saved_values) {ierr = PetscFree(a->saved_values);CHKERRQ(ierr);} 813 #if defined(PETSC_USE_MAT_SINGLE) 814 if (a->setvaluescopy) {ierr = PetscFree(a->setvaluescopy);CHKERRQ(ierr);} 815 #endif 816 if (a->xtoy) {ierr = PetscFree(a->xtoy);CHKERRQ(ierr);} 817 if (a->compressedrow.use){ierr = PetscFree(a->compressedrow.i);} 818 819 ierr = PetscFree(a);CHKERRQ(ierr); 820 821 ierr = PetscObjectComposeFunction((PetscObject)A,"MatStoreValues_C","",PETSC_NULL);CHKERRQ(ierr); 822 ierr = PetscObjectComposeFunction((PetscObject)A,"MatRetrieveValues_C","",PETSC_NULL);CHKERRQ(ierr); 823 ierr = PetscObjectComposeFunction((PetscObject)A,"MatSeqBAIJSetColumnIndices_C","",PETSC_NULL);CHKERRQ(ierr); 824 ierr = PetscObjectComposeFunction((PetscObject)A,"MatConvert_seqbaij_seqaij_C","",PETSC_NULL);CHKERRQ(ierr); 825 ierr = PetscObjectComposeFunction((PetscObject)A,"MatConvert_seqbaij_seqsbaij_C","",PETSC_NULL);CHKERRQ(ierr); 826 ierr = PetscObjectComposeFunction((PetscObject)A,"MatSeqBAIJSetPreallocation_C","",PETSC_NULL);CHKERRQ(ierr); 827 PetscFunctionReturn(0); 828 } 829 830 #undef __FUNCT__ 831 #define __FUNCT__ "MatSetOption_SeqBAIJ" 832 PetscErrorCode MatSetOption_SeqBAIJ(Mat A,MatOption op) 833 { 834 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 835 836 PetscFunctionBegin; 837 switch (op) { 838 case MAT_ROW_ORIENTED: 839 a->roworiented = PETSC_TRUE; 840 break; 841 case MAT_COLUMN_ORIENTED: 842 a->roworiented = PETSC_FALSE; 843 break; 844 case MAT_COLUMNS_SORTED: 845 a->sorted = PETSC_TRUE; 846 break; 847 case MAT_COLUMNS_UNSORTED: 848 a->sorted = PETSC_FALSE; 849 break; 850 case MAT_KEEP_ZEROED_ROWS: 851 a->keepzeroedrows = PETSC_TRUE; 852 break; 853 case MAT_NO_NEW_NONZERO_LOCATIONS: 854 a->nonew = 1; 855 break; 856 case MAT_NEW_NONZERO_LOCATION_ERR: 857 a->nonew = -1; 858 break; 859 case MAT_NEW_NONZERO_ALLOCATION_ERR: 860 a->nonew = -2; 861 break; 862 case MAT_YES_NEW_NONZERO_LOCATIONS: 863 a->nonew = 0; 864 break; 865 case MAT_ROWS_SORTED: 866 case MAT_ROWS_UNSORTED: 867 case MAT_YES_NEW_DIAGONALS: 868 case MAT_IGNORE_OFF_PROC_ENTRIES: 869 case MAT_USE_HASH_TABLE: 870 PetscLogInfo(A,"MatSetOption_SeqBAIJ:Option ignored\n"); 871 break; 872 case MAT_NO_NEW_DIAGONALS: 873 SETERRQ(PETSC_ERR_SUP,"MAT_NO_NEW_DIAGONALS"); 874 case MAT_SYMMETRIC: 875 case MAT_STRUCTURALLY_SYMMETRIC: 876 case MAT_NOT_SYMMETRIC: 877 case MAT_NOT_STRUCTURALLY_SYMMETRIC: 878 case MAT_HERMITIAN: 879 case MAT_NOT_HERMITIAN: 880 case MAT_SYMMETRY_ETERNAL: 881 case MAT_NOT_SYMMETRY_ETERNAL: 882 break; 883 default: 884 SETERRQ(PETSC_ERR_SUP,"unknown option"); 885 } 886 PetscFunctionReturn(0); 887 } 888 889 #undef __FUNCT__ 890 #define __FUNCT__ "MatGetRow_SeqBAIJ" 891 PetscErrorCode MatGetRow_SeqBAIJ(Mat A,PetscInt row,PetscInt *nz,PetscInt **idx,PetscScalar **v) 892 { 893 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 894 PetscErrorCode ierr; 895 PetscInt itmp,i,j,k,M,*ai,*aj,bs,bn,bp,*idx_i,bs2; 896 MatScalar *aa,*aa_i; 897 PetscScalar *v_i; 898 899 PetscFunctionBegin; 900 bs = A->bs; 901 ai = a->i; 902 aj = a->j; 903 aa = a->a; 904 bs2 = a->bs2; 905 906 if (row < 0 || row >= A->m) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Row %D out of range", row); 907 908 bn = row/bs; /* Block number */ 909 bp = row % bs; /* Block Position */ 910 M = ai[bn+1] - ai[bn]; 911 *nz = bs*M; 912 913 if (v) { 914 *v = 0; 915 if (*nz) { 916 ierr = PetscMalloc((*nz)*sizeof(PetscScalar),v);CHKERRQ(ierr); 917 for (i=0; i<M; i++) { /* for each block in the block row */ 918 v_i = *v + i*bs; 919 aa_i = aa + bs2*(ai[bn] + i); 920 for (j=bp,k=0; j<bs2; j+=bs,k++) {v_i[k] = aa_i[j];} 921 } 922 } 923 } 924 925 if (idx) { 926 *idx = 0; 927 if (*nz) { 928 ierr = PetscMalloc((*nz)*sizeof(PetscInt),idx);CHKERRQ(ierr); 929 for (i=0; i<M; i++) { /* for each block in the block row */ 930 idx_i = *idx + i*bs; 931 itmp = bs*aj[ai[bn] + i]; 932 for (j=0; j<bs; j++) {idx_i[j] = itmp++;} 933 } 934 } 935 } 936 PetscFunctionReturn(0); 937 } 938 939 #undef __FUNCT__ 940 #define __FUNCT__ "MatRestoreRow_SeqBAIJ" 941 PetscErrorCode MatRestoreRow_SeqBAIJ(Mat A,PetscInt row,PetscInt *nz,PetscInt **idx,PetscScalar **v) 942 { 943 PetscErrorCode ierr; 944 945 PetscFunctionBegin; 946 if (idx) {if (*idx) {ierr = PetscFree(*idx);CHKERRQ(ierr);}} 947 if (v) {if (*v) {ierr = PetscFree(*v);CHKERRQ(ierr);}} 948 PetscFunctionReturn(0); 949 } 950 951 #undef __FUNCT__ 952 #define __FUNCT__ "MatTranspose_SeqBAIJ" 953 PetscErrorCode MatTranspose_SeqBAIJ(Mat A,Mat *B) 954 { 955 Mat_SeqBAIJ *a=(Mat_SeqBAIJ *)A->data; 956 Mat C; 957 PetscErrorCode ierr; 958 PetscInt i,j,k,*aj=a->j,*ai=a->i,bs=A->bs,mbs=a->mbs,nbs=a->nbs,len,*col; 959 PetscInt *rows,*cols,bs2=a->bs2; 960 PetscScalar *array; 961 962 PetscFunctionBegin; 963 if (!B && mbs!=nbs) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Square matrix only for in-place"); 964 ierr = PetscMalloc((1+nbs)*sizeof(PetscInt),&col);CHKERRQ(ierr); 965 ierr = PetscMemzero(col,(1+nbs)*sizeof(PetscInt));CHKERRQ(ierr); 966 967 #if defined(PETSC_USE_MAT_SINGLE) 968 ierr = PetscMalloc(a->bs2*a->nz*sizeof(PetscScalar),&array);CHKERRQ(ierr); 969 for (i=0; i<a->bs2*a->nz; i++) array[i] = (PetscScalar)a->a[i]; 970 #else 971 array = a->a; 972 #endif 973 974 for (i=0; i<ai[mbs]; i++) col[aj[i]] += 1; 975 ierr = MatCreate(A->comm,A->n,A->m,A->n,A->m,&C);CHKERRQ(ierr); 976 ierr = MatSetType(C,A->type_name);CHKERRQ(ierr); 977 ierr = MatSeqBAIJSetPreallocation(C,bs,PETSC_NULL,col);CHKERRQ(ierr); 978 ierr = PetscFree(col);CHKERRQ(ierr); 979 ierr = PetscMalloc(2*bs*sizeof(PetscInt),&rows);CHKERRQ(ierr); 980 cols = rows + bs; 981 for (i=0; i<mbs; i++) { 982 cols[0] = i*bs; 983 for (k=1; k<bs; k++) cols[k] = cols[k-1] + 1; 984 len = ai[i+1] - ai[i]; 985 for (j=0; j<len; j++) { 986 rows[0] = (*aj++)*bs; 987 for (k=1; k<bs; k++) rows[k] = rows[k-1] + 1; 988 ierr = MatSetValues(C,bs,rows,bs,cols,array,INSERT_VALUES);CHKERRQ(ierr); 989 array += bs2; 990 } 991 } 992 ierr = PetscFree(rows);CHKERRQ(ierr); 993 #if defined(PETSC_USE_MAT_SINGLE) 994 ierr = PetscFree(array); 995 #endif 996 997 ierr = MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 998 ierr = MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 999 1000 if (B) { 1001 *B = C; 1002 } else { 1003 ierr = MatHeaderCopy(A,C);CHKERRQ(ierr); 1004 } 1005 PetscFunctionReturn(0); 1006 } 1007 1008 #undef __FUNCT__ 1009 #define __FUNCT__ "MatView_SeqBAIJ_Binary" 1010 static PetscErrorCode MatView_SeqBAIJ_Binary(Mat A,PetscViewer viewer) 1011 { 1012 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 1013 PetscErrorCode ierr; 1014 PetscInt i,*col_lens,bs = A->bs,count,*jj,j,k,l,bs2=a->bs2; 1015 int fd; 1016 PetscScalar *aa; 1017 FILE *file; 1018 1019 PetscFunctionBegin; 1020 ierr = PetscViewerBinaryGetDescriptor(viewer,&fd);CHKERRQ(ierr); 1021 ierr = PetscMalloc((4+A->m)*sizeof(PetscInt),&col_lens);CHKERRQ(ierr); 1022 col_lens[0] = MAT_FILE_COOKIE; 1023 1024 col_lens[1] = A->m; 1025 col_lens[2] = A->n; 1026 col_lens[3] = a->nz*bs2; 1027 1028 /* store lengths of each row and write (including header) to file */ 1029 count = 0; 1030 for (i=0; i<a->mbs; i++) { 1031 for (j=0; j<bs; j++) { 1032 col_lens[4+count++] = bs*(a->i[i+1] - a->i[i]); 1033 } 1034 } 1035 ierr = PetscBinaryWrite(fd,col_lens,4+A->m,PETSC_INT,PETSC_TRUE);CHKERRQ(ierr); 1036 ierr = PetscFree(col_lens);CHKERRQ(ierr); 1037 1038 /* store column indices (zero start index) */ 1039 ierr = PetscMalloc((a->nz+1)*bs2*sizeof(PetscInt),&jj);CHKERRQ(ierr); 1040 count = 0; 1041 for (i=0; i<a->mbs; i++) { 1042 for (j=0; j<bs; j++) { 1043 for (k=a->i[i]; k<a->i[i+1]; k++) { 1044 for (l=0; l<bs; l++) { 1045 jj[count++] = bs*a->j[k] + l; 1046 } 1047 } 1048 } 1049 } 1050 ierr = PetscBinaryWrite(fd,jj,bs2*a->nz,PETSC_INT,PETSC_FALSE);CHKERRQ(ierr); 1051 ierr = PetscFree(jj);CHKERRQ(ierr); 1052 1053 /* store nonzero values */ 1054 ierr = PetscMalloc((a->nz+1)*bs2*sizeof(PetscScalar),&aa);CHKERRQ(ierr); 1055 count = 0; 1056 for (i=0; i<a->mbs; i++) { 1057 for (j=0; j<bs; j++) { 1058 for (k=a->i[i]; k<a->i[i+1]; k++) { 1059 for (l=0; l<bs; l++) { 1060 aa[count++] = a->a[bs2*k + l*bs + j]; 1061 } 1062 } 1063 } 1064 } 1065 ierr = PetscBinaryWrite(fd,aa,bs2*a->nz,PETSC_SCALAR,PETSC_FALSE);CHKERRQ(ierr); 1066 ierr = PetscFree(aa);CHKERRQ(ierr); 1067 1068 ierr = PetscViewerBinaryGetInfoPointer(viewer,&file);CHKERRQ(ierr); 1069 if (file) { 1070 fprintf(file,"-matload_block_size %d\n",(int)A->bs); 1071 } 1072 PetscFunctionReturn(0); 1073 } 1074 1075 #undef __FUNCT__ 1076 #define __FUNCT__ "MatView_SeqBAIJ_ASCII" 1077 static PetscErrorCode MatView_SeqBAIJ_ASCII(Mat A,PetscViewer viewer) 1078 { 1079 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 1080 PetscErrorCode ierr; 1081 PetscInt i,j,bs = A->bs,k,l,bs2=a->bs2; 1082 PetscViewerFormat format; 1083 1084 PetscFunctionBegin; 1085 ierr = PetscViewerGetFormat(viewer,&format);CHKERRQ(ierr); 1086 if (format == PETSC_VIEWER_ASCII_INFO || format == PETSC_VIEWER_ASCII_INFO_DETAIL) { 1087 ierr = PetscViewerASCIIPrintf(viewer," block size is %D\n",bs);CHKERRQ(ierr); 1088 } else if (format == PETSC_VIEWER_ASCII_MATLAB) { 1089 Mat aij; 1090 ierr = MatConvert(A,MATSEQAIJ,&aij);CHKERRQ(ierr); 1091 ierr = MatView(aij,viewer);CHKERRQ(ierr); 1092 ierr = MatDestroy(aij);CHKERRQ(ierr); 1093 } else if (format == PETSC_VIEWER_ASCII_FACTOR_INFO) { 1094 PetscFunctionReturn(0); 1095 } else if (format == PETSC_VIEWER_ASCII_COMMON) { 1096 ierr = PetscViewerASCIIUseTabs(viewer,PETSC_NO);CHKERRQ(ierr); 1097 for (i=0; i<a->mbs; i++) { 1098 for (j=0; j<bs; j++) { 1099 ierr = PetscViewerASCIIPrintf(viewer,"row %D:",i*bs+j);CHKERRQ(ierr); 1100 for (k=a->i[i]; k<a->i[i+1]; k++) { 1101 for (l=0; l<bs; l++) { 1102 #if defined(PETSC_USE_COMPLEX) 1103 if (PetscImaginaryPart(a->a[bs2*k + l*bs + j]) > 0.0 && PetscRealPart(a->a[bs2*k + l*bs + j]) != 0.0) { 1104 ierr = PetscViewerASCIIPrintf(viewer," (%D, %g + %gi) ",bs*a->j[k]+l, 1105 PetscRealPart(a->a[bs2*k + l*bs + j]),PetscImaginaryPart(a->a[bs2*k + l*bs + j]));CHKERRQ(ierr); 1106 } else if (PetscImaginaryPart(a->a[bs2*k + l*bs + j]) < 0.0 && PetscRealPart(a->a[bs2*k + l*bs + j]) != 0.0) { 1107 ierr = PetscViewerASCIIPrintf(viewer," (%D, %g - %gi) ",bs*a->j[k]+l, 1108 PetscRealPart(a->a[bs2*k + l*bs + j]),-PetscImaginaryPart(a->a[bs2*k + l*bs + j]));CHKERRQ(ierr); 1109 } else if (PetscRealPart(a->a[bs2*k + l*bs + j]) != 0.0) { 1110 ierr = PetscViewerASCIIPrintf(viewer," (%D, %g) ",bs*a->j[k]+l,PetscRealPart(a->a[bs2*k + l*bs + j]));CHKERRQ(ierr); 1111 } 1112 #else 1113 if (a->a[bs2*k + l*bs + j] != 0.0) { 1114 ierr = PetscViewerASCIIPrintf(viewer," (%D, %g) ",bs*a->j[k]+l,a->a[bs2*k + l*bs + j]);CHKERRQ(ierr); 1115 } 1116 #endif 1117 } 1118 } 1119 ierr = PetscViewerASCIIPrintf(viewer,"\n");CHKERRQ(ierr); 1120 } 1121 } 1122 ierr = PetscViewerASCIIUseTabs(viewer,PETSC_YES);CHKERRQ(ierr); 1123 } else { 1124 ierr = PetscViewerASCIIUseTabs(viewer,PETSC_NO);CHKERRQ(ierr); 1125 for (i=0; i<a->mbs; i++) { 1126 for (j=0; j<bs; j++) { 1127 ierr = PetscViewerASCIIPrintf(viewer,"row %D:",i*bs+j);CHKERRQ(ierr); 1128 for (k=a->i[i]; k<a->i[i+1]; k++) { 1129 for (l=0; l<bs; l++) { 1130 #if defined(PETSC_USE_COMPLEX) 1131 if (PetscImaginaryPart(a->a[bs2*k + l*bs + j]) > 0.0) { 1132 ierr = PetscViewerASCIIPrintf(viewer," (%D, %g + %g i) ",bs*a->j[k]+l, 1133 PetscRealPart(a->a[bs2*k + l*bs + j]),PetscImaginaryPart(a->a[bs2*k + l*bs + j]));CHKERRQ(ierr); 1134 } else if (PetscImaginaryPart(a->a[bs2*k + l*bs + j]) < 0.0) { 1135 ierr = PetscViewerASCIIPrintf(viewer," (%D, %g - %g i) ",bs*a->j[k]+l, 1136 PetscRealPart(a->a[bs2*k + l*bs + j]),-PetscImaginaryPart(a->a[bs2*k + l*bs + j]));CHKERRQ(ierr); 1137 } else { 1138 ierr = PetscViewerASCIIPrintf(viewer," (%D, %g) ",bs*a->j[k]+l,PetscRealPart(a->a[bs2*k + l*bs + j]));CHKERRQ(ierr); 1139 } 1140 #else 1141 ierr = PetscViewerASCIIPrintf(viewer," (%D, %g) ",bs*a->j[k]+l,a->a[bs2*k + l*bs + j]);CHKERRQ(ierr); 1142 #endif 1143 } 1144 } 1145 ierr = PetscViewerASCIIPrintf(viewer,"\n");CHKERRQ(ierr); 1146 } 1147 } 1148 ierr = PetscViewerASCIIUseTabs(viewer,PETSC_YES);CHKERRQ(ierr); 1149 } 1150 ierr = PetscViewerFlush(viewer);CHKERRQ(ierr); 1151 PetscFunctionReturn(0); 1152 } 1153 1154 #undef __FUNCT__ 1155 #define __FUNCT__ "MatView_SeqBAIJ_Draw_Zoom" 1156 static PetscErrorCode MatView_SeqBAIJ_Draw_Zoom(PetscDraw draw,void *Aa) 1157 { 1158 Mat A = (Mat) Aa; 1159 Mat_SeqBAIJ *a=(Mat_SeqBAIJ*)A->data; 1160 PetscErrorCode ierr; 1161 PetscInt row,i,j,k,l,mbs=a->mbs,color,bs=A->bs,bs2=a->bs2; 1162 PetscReal xl,yl,xr,yr,x_l,x_r,y_l,y_r; 1163 MatScalar *aa; 1164 PetscViewer viewer; 1165 1166 PetscFunctionBegin; 1167 1168 /* still need to add support for contour plot of nonzeros; see MatView_SeqAIJ_Draw_Zoom()*/ 1169 ierr = PetscObjectQuery((PetscObject)A,"Zoomviewer",(PetscObject*)&viewer);CHKERRQ(ierr); 1170 1171 ierr = PetscDrawGetCoordinates(draw,&xl,&yl,&xr,&yr);CHKERRQ(ierr); 1172 1173 /* loop over matrix elements drawing boxes */ 1174 color = PETSC_DRAW_BLUE; 1175 for (i=0,row=0; i<mbs; i++,row+=bs) { 1176 for (j=a->i[i]; j<a->i[i+1]; j++) { 1177 y_l = A->m - row - 1.0; y_r = y_l + 1.0; 1178 x_l = a->j[j]*bs; x_r = x_l + 1.0; 1179 aa = a->a + j*bs2; 1180 for (k=0; k<bs; k++) { 1181 for (l=0; l<bs; l++) { 1182 if (PetscRealPart(*aa++) >= 0.) continue; 1183 ierr = PetscDrawRectangle(draw,x_l+k,y_l-l,x_r+k,y_r-l,color,color,color,color);CHKERRQ(ierr); 1184 } 1185 } 1186 } 1187 } 1188 color = PETSC_DRAW_CYAN; 1189 for (i=0,row=0; i<mbs; i++,row+=bs) { 1190 for (j=a->i[i]; j<a->i[i+1]; j++) { 1191 y_l = A->m - row - 1.0; y_r = y_l + 1.0; 1192 x_l = a->j[j]*bs; x_r = x_l + 1.0; 1193 aa = a->a + j*bs2; 1194 for (k=0; k<bs; k++) { 1195 for (l=0; l<bs; l++) { 1196 if (PetscRealPart(*aa++) != 0.) continue; 1197 ierr = PetscDrawRectangle(draw,x_l+k,y_l-l,x_r+k,y_r-l,color,color,color,color);CHKERRQ(ierr); 1198 } 1199 } 1200 } 1201 } 1202 1203 color = PETSC_DRAW_RED; 1204 for (i=0,row=0; i<mbs; i++,row+=bs) { 1205 for (j=a->i[i]; j<a->i[i+1]; j++) { 1206 y_l = A->m - row - 1.0; y_r = y_l + 1.0; 1207 x_l = a->j[j]*bs; x_r = x_l + 1.0; 1208 aa = a->a + j*bs2; 1209 for (k=0; k<bs; k++) { 1210 for (l=0; l<bs; l++) { 1211 if (PetscRealPart(*aa++) <= 0.) continue; 1212 ierr = PetscDrawRectangle(draw,x_l+k,y_l-l,x_r+k,y_r-l,color,color,color,color);CHKERRQ(ierr); 1213 } 1214 } 1215 } 1216 } 1217 PetscFunctionReturn(0); 1218 } 1219 1220 #undef __FUNCT__ 1221 #define __FUNCT__ "MatView_SeqBAIJ_Draw" 1222 static PetscErrorCode MatView_SeqBAIJ_Draw(Mat A,PetscViewer viewer) 1223 { 1224 PetscErrorCode ierr; 1225 PetscReal xl,yl,xr,yr,w,h; 1226 PetscDraw draw; 1227 PetscTruth isnull; 1228 1229 PetscFunctionBegin; 1230 1231 ierr = PetscViewerDrawGetDraw(viewer,0,&draw);CHKERRQ(ierr); 1232 ierr = PetscDrawIsNull(draw,&isnull);CHKERRQ(ierr); if (isnull) PetscFunctionReturn(0); 1233 1234 ierr = PetscObjectCompose((PetscObject)A,"Zoomviewer",(PetscObject)viewer);CHKERRQ(ierr); 1235 xr = A->n; yr = A->m; h = yr/10.0; w = xr/10.0; 1236 xr += w; yr += h; xl = -w; yl = -h; 1237 ierr = PetscDrawSetCoordinates(draw,xl,yl,xr,yr);CHKERRQ(ierr); 1238 ierr = PetscDrawZoom(draw,MatView_SeqBAIJ_Draw_Zoom,A);CHKERRQ(ierr); 1239 ierr = PetscObjectCompose((PetscObject)A,"Zoomviewer",PETSC_NULL);CHKERRQ(ierr); 1240 PetscFunctionReturn(0); 1241 } 1242 1243 #undef __FUNCT__ 1244 #define __FUNCT__ "MatView_SeqBAIJ" 1245 PetscErrorCode MatView_SeqBAIJ(Mat A,PetscViewer viewer) 1246 { 1247 PetscErrorCode ierr; 1248 PetscTruth iascii,isbinary,isdraw; 1249 1250 PetscFunctionBegin; 1251 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_ASCII,&iascii);CHKERRQ(ierr); 1252 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_BINARY,&isbinary);CHKERRQ(ierr); 1253 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_DRAW,&isdraw);CHKERRQ(ierr); 1254 if (iascii){ 1255 ierr = MatView_SeqBAIJ_ASCII(A,viewer);CHKERRQ(ierr); 1256 } else if (isbinary) { 1257 ierr = MatView_SeqBAIJ_Binary(A,viewer);CHKERRQ(ierr); 1258 } else if (isdraw) { 1259 ierr = MatView_SeqBAIJ_Draw(A,viewer);CHKERRQ(ierr); 1260 } else { 1261 Mat B; 1262 ierr = MatConvert(A,MATSEQAIJ,&B);CHKERRQ(ierr); 1263 ierr = MatView(B,viewer);CHKERRQ(ierr); 1264 ierr = MatDestroy(B);CHKERRQ(ierr); 1265 } 1266 PetscFunctionReturn(0); 1267 } 1268 1269 1270 #undef __FUNCT__ 1271 #define __FUNCT__ "MatGetValues_SeqBAIJ" 1272 PetscErrorCode MatGetValues_SeqBAIJ(Mat A,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],PetscScalar v[]) 1273 { 1274 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 1275 PetscInt *rp,k,low,high,t,row,nrow,i,col,l,*aj = a->j; 1276 PetscInt *ai = a->i,*ailen = a->ilen; 1277 PetscInt brow,bcol,ridx,cidx,bs=A->bs,bs2=a->bs2; 1278 MatScalar *ap,*aa = a->a,zero = 0.0; 1279 1280 PetscFunctionBegin; 1281 for (k=0; k<m; k++) { /* loop over rows */ 1282 row = im[k]; brow = row/bs; 1283 if (row < 0) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Negative row"); 1284 if (row >= A->m) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Row %D too large", row); 1285 rp = aj + ai[brow] ; ap = aa + bs2*ai[brow] ; 1286 nrow = ailen[brow]; 1287 for (l=0; l<n; l++) { /* loop over columns */ 1288 if (in[l] < 0) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Negative column"); 1289 if (in[l] >= A->n) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Column %D too large", in[l]); 1290 col = in[l] ; 1291 bcol = col/bs; 1292 cidx = col%bs; 1293 ridx = row%bs; 1294 high = nrow; 1295 low = 0; /* assume unsorted */ 1296 while (high-low > 5) { 1297 t = (low+high)/2; 1298 if (rp[t] > bcol) high = t; 1299 else low = t; 1300 } 1301 for (i=low; i<high; i++) { 1302 if (rp[i] > bcol) break; 1303 if (rp[i] == bcol) { 1304 *v++ = ap[bs2*i+bs*cidx+ridx]; 1305 goto finished; 1306 } 1307 } 1308 *v++ = zero; 1309 finished:; 1310 } 1311 } 1312 PetscFunctionReturn(0); 1313 } 1314 1315 #if defined(PETSC_USE_MAT_SINGLE) 1316 #undef __FUNCT__ 1317 #define __FUNCT__ "MatSetValuesBlocked_SeqBAIJ" 1318 PetscErrorCode MatSetValuesBlocked_SeqBAIJ(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode addv) 1319 { 1320 Mat_SeqBAIJ *b = (Mat_SeqBAIJ*)mat->data; 1321 PetscErrorCode ierr; 1322 PetscInt i,N = m*n*b->bs2; 1323 MatScalar *vsingle; 1324 1325 PetscFunctionBegin; 1326 if (N > b->setvalueslen) { 1327 if (b->setvaluescopy) {ierr = PetscFree(b->setvaluescopy);CHKERRQ(ierr);} 1328 ierr = PetscMalloc(N*sizeof(MatScalar),&b->setvaluescopy);CHKERRQ(ierr); 1329 b->setvalueslen = N; 1330 } 1331 vsingle = b->setvaluescopy; 1332 for (i=0; i<N; i++) { 1333 vsingle[i] = v[i]; 1334 } 1335 ierr = MatSetValuesBlocked_SeqBAIJ_MatScalar(mat,m,im,n,in,vsingle,addv);CHKERRQ(ierr); 1336 PetscFunctionReturn(0); 1337 } 1338 #endif 1339 1340 1341 #undef __FUNCT__ 1342 #define __FUNCT__ "MatSetValuesBlocked_SeqBAIJ" 1343 PetscErrorCode MatSetValuesBlocked_SeqBAIJ_MatScalar(Mat A,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const MatScalar v[],InsertMode is) 1344 { 1345 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 1346 PetscInt *rp,k,low,high,t,ii,jj,row,nrow,i,col,l,rmax,N,lastcol = -1; 1347 PetscInt *imax=a->imax,*ai=a->i,*ailen=a->ilen; 1348 PetscErrorCode ierr; 1349 PetscInt *aj=a->j,nonew=a->nonew,bs2=a->bs2,bs=A->bs,stepval; 1350 PetscTruth roworiented=a->roworiented; 1351 const MatScalar *value = v; 1352 MatScalar *ap,*aa = a->a,*bap; 1353 1354 PetscFunctionBegin; 1355 if (roworiented) { 1356 stepval = (n-1)*bs; 1357 } else { 1358 stepval = (m-1)*bs; 1359 } 1360 for (k=0; k<m; k++) { /* loop over added rows */ 1361 row = im[k]; 1362 if (row < 0) continue; 1363 #if defined(PETSC_USE_DEBUG) 1364 if (row >= a->mbs) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",row,a->mbs-1); 1365 #endif 1366 rp = aj + ai[row]; 1367 ap = aa + bs2*ai[row]; 1368 rmax = imax[row]; 1369 nrow = ailen[row]; 1370 low = 0; 1371 for (l=0; l<n; l++) { /* loop over added columns */ 1372 if (in[l] < 0) continue; 1373 #if defined(PETSC_USE_DEBUG) 1374 if (in[l] >= a->nbs) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %D max %D",in[l],a->nbs-1); 1375 #endif 1376 col = in[l]; 1377 if (roworiented) { 1378 value = v + k*(stepval+bs)*bs + l*bs; 1379 } else { 1380 value = v + l*(stepval+bs)*bs + k*bs; 1381 } 1382 if (col < lastcol) low = 0; high = nrow; 1383 lastcol = col; 1384 while (high-low > 7) { 1385 t = (low+high)/2; 1386 if (rp[t] > col) high = t; 1387 else low = t; 1388 } 1389 for (i=low; i<high; i++) { 1390 if (rp[i] > col) break; 1391 if (rp[i] == col) { 1392 bap = ap + bs2*i; 1393 if (roworiented) { 1394 if (is == ADD_VALUES) { 1395 for (ii=0; ii<bs; ii++,value+=stepval) { 1396 for (jj=ii; jj<bs2; jj+=bs) { 1397 bap[jj] += *value++; 1398 } 1399 } 1400 } else { 1401 for (ii=0; ii<bs; ii++,value+=stepval) { 1402 for (jj=ii; jj<bs2; jj+=bs) { 1403 bap[jj] = *value++; 1404 } 1405 } 1406 } 1407 } else { 1408 if (is == ADD_VALUES) { 1409 for (ii=0; ii<bs; ii++,value+=stepval) { 1410 for (jj=0; jj<bs; jj++) { 1411 *bap++ += *value++; 1412 } 1413 } 1414 } else { 1415 for (ii=0; ii<bs; ii++,value+=stepval) { 1416 for (jj=0; jj<bs; jj++) { 1417 *bap++ = *value++; 1418 } 1419 } 1420 } 1421 } 1422 goto noinsert2; 1423 } 1424 } 1425 if (nonew == 1) goto noinsert2; 1426 else if (nonew == -1) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%D, %D) in the matrix", row, col); 1427 if (nrow >= rmax) { 1428 /* there is no extra room in row, therefore enlarge */ 1429 PetscInt new_nz = ai[a->mbs] + CHUNKSIZE,len,*new_i,*new_j; 1430 MatScalar *new_a; 1431 1432 if (nonew == -2) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%D, %D) in the matrix", row, col); 1433 1434 /* malloc new storage space */ 1435 len = new_nz*(sizeof(PetscInt)+bs2*sizeof(MatScalar))+(a->mbs+1)*sizeof(PetscInt); 1436 ierr = PetscMalloc(len,&new_a);CHKERRQ(ierr); 1437 new_j = (PetscInt*)(new_a + bs2*new_nz); 1438 new_i = new_j + new_nz; 1439 1440 /* copy over old data into new slots */ 1441 for (ii=0; ii<row+1; ii++) {new_i[ii] = ai[ii];} 1442 for (ii=row+1; ii<a->mbs+1; ii++) {new_i[ii] = ai[ii]+CHUNKSIZE;} 1443 ierr = PetscMemcpy(new_j,aj,(ai[row]+nrow)*sizeof(PetscInt));CHKERRQ(ierr); 1444 len = (new_nz - CHUNKSIZE - ai[row] - nrow); 1445 ierr = PetscMemcpy(new_j+ai[row]+nrow+CHUNKSIZE,aj+ai[row]+nrow,len*sizeof(PetscInt));CHKERRQ(ierr); 1446 ierr = PetscMemcpy(new_a,aa,(ai[row]+nrow)*bs2*sizeof(MatScalar));CHKERRQ(ierr); 1447 ierr = PetscMemzero(new_a+bs2*(ai[row]+nrow),bs2*CHUNKSIZE*sizeof(MatScalar));CHKERRQ(ierr); 1448 ierr = PetscMemcpy(new_a+bs2*(ai[row]+nrow+CHUNKSIZE),aa+bs2*(ai[row]+nrow),bs2*len*sizeof(MatScalar));CHKERRQ(ierr); 1449 /* free up old matrix storage */ 1450 ierr = PetscFree(a->a);CHKERRQ(ierr); 1451 if (!a->singlemalloc) { 1452 ierr = PetscFree(a->i);CHKERRQ(ierr); 1453 ierr = PetscFree(a->j);CHKERRQ(ierr); 1454 } 1455 aa = a->a = new_a; ai = a->i = new_i; aj = a->j = new_j; 1456 a->singlemalloc = PETSC_TRUE; 1457 1458 rp = aj + ai[row]; ap = aa + bs2*ai[row]; 1459 rmax = imax[row] = imax[row] + CHUNKSIZE; 1460 ierr = PetscLogObjectMemory(A,CHUNKSIZE*(sizeof(PetscInt) + bs2*sizeof(MatScalar)));CHKERRQ(ierr); 1461 a->maxnz += bs2*CHUNKSIZE; 1462 a->reallocs++; 1463 a->nz++; 1464 } 1465 N = nrow++ - 1; 1466 /* shift up all the later entries in this row */ 1467 for (ii=N; ii>=i; ii--) { 1468 rp[ii+1] = rp[ii]; 1469 ierr = PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar));CHKERRQ(ierr); 1470 } 1471 if (N >= i) { 1472 ierr = PetscMemzero(ap+bs2*i,bs2*sizeof(MatScalar));CHKERRQ(ierr); 1473 } 1474 rp[i] = col; 1475 bap = ap + bs2*i; 1476 if (roworiented) { 1477 for (ii=0; ii<bs; ii++,value+=stepval) { 1478 for (jj=ii; jj<bs2; jj+=bs) { 1479 bap[jj] = *value++; 1480 } 1481 } 1482 } else { 1483 for (ii=0; ii<bs; ii++,value+=stepval) { 1484 for (jj=0; jj<bs; jj++) { 1485 *bap++ = *value++; 1486 } 1487 } 1488 } 1489 noinsert2:; 1490 low = i; 1491 } 1492 ailen[row] = nrow; 1493 } 1494 PetscFunctionReturn(0); 1495 } 1496 1497 #undef __FUNCT__ 1498 #define __FUNCT__ "MatAssemblyEnd_SeqBAIJ" 1499 PetscErrorCode MatAssemblyEnd_SeqBAIJ(Mat A,MatAssemblyType mode) 1500 { 1501 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 1502 PetscInt fshift = 0,i,j,*ai = a->i,*aj = a->j,*imax = a->imax; 1503 PetscInt m = A->m,*ip,N,*ailen = a->ilen; 1504 PetscErrorCode ierr; 1505 PetscInt mbs = a->mbs,bs2 = a->bs2,rmax = 0; 1506 MatScalar *aa = a->a,*ap; 1507 PetscReal ratio=0.6; 1508 1509 PetscFunctionBegin; 1510 if (mode == MAT_FLUSH_ASSEMBLY) PetscFunctionReturn(0); 1511 1512 if (m) rmax = ailen[0]; 1513 for (i=1; i<mbs; i++) { 1514 /* move each row back by the amount of empty slots (fshift) before it*/ 1515 fshift += imax[i-1] - ailen[i-1]; 1516 rmax = PetscMax(rmax,ailen[i]); 1517 if (fshift) { 1518 ip = aj + ai[i]; ap = aa + bs2*ai[i]; 1519 N = ailen[i]; 1520 for (j=0; j<N; j++) { 1521 ip[j-fshift] = ip[j]; 1522 ierr = PetscMemcpy(ap+(j-fshift)*bs2,ap+j*bs2,bs2*sizeof(MatScalar));CHKERRQ(ierr); 1523 } 1524 } 1525 ai[i] = ai[i-1] + ailen[i-1]; 1526 } 1527 if (mbs) { 1528 fshift += imax[mbs-1] - ailen[mbs-1]; 1529 ai[mbs] = ai[mbs-1] + ailen[mbs-1]; 1530 } 1531 /* reset ilen and imax for each row */ 1532 for (i=0; i<mbs; i++) { 1533 ailen[i] = imax[i] = ai[i+1] - ai[i]; 1534 } 1535 a->nz = ai[mbs]; 1536 1537 /* diagonals may have moved, so kill the diagonal pointers */ 1538 a->idiagvalid = PETSC_FALSE; 1539 if (fshift && a->diag) { 1540 ierr = PetscFree(a->diag);CHKERRQ(ierr); 1541 ierr = PetscLogObjectMemory(A,-(mbs+1)*sizeof(PetscInt));CHKERRQ(ierr); 1542 a->diag = 0; 1543 } 1544 PetscLogInfo(A,"MatAssemblyEnd_SeqBAIJ:Matrix size: %D X %D, block size %D; storage space: %D unneeded, %D used\n",m,A->n,A->bs,fshift*bs2,a->nz*bs2); 1545 PetscLogInfo(A,"MatAssemblyEnd_SeqBAIJ:Number of mallocs during MatSetValues is %D\n",a->reallocs); 1546 PetscLogInfo(A,"MatAssemblyEnd_SeqBAIJ:Most nonzeros blocks in any row is %D\n",rmax); 1547 a->reallocs = 0; 1548 A->info.nz_unneeded = (PetscReal)fshift*bs2; 1549 1550 /* check for zero rows. If found a large number of zero rows, use CompressedRow functions */ 1551 if (a->compressedrow.use){ 1552 ierr = Mat_CheckCompressedRow(A,&a->compressedrow,a->i,mbs,ratio);CHKERRQ(ierr); 1553 } 1554 1555 A->same_nonzero = PETSC_TRUE; 1556 PetscFunctionReturn(0); 1557 } 1558 1559 /* 1560 This function returns an array of flags which indicate the locations of contiguous 1561 blocks that should be zeroed. for eg: if bs = 3 and is = [0,1,2,3,5,6,7,8,9] 1562 then the resulting sizes = [3,1,1,3,1] correspondig to sets [(0,1,2),(3),(5),(6,7,8),(9)] 1563 Assume: sizes should be long enough to hold all the values. 1564 */ 1565 #undef __FUNCT__ 1566 #define __FUNCT__ "MatZeroRows_SeqBAIJ_Check_Blocks" 1567 static PetscErrorCode MatZeroRows_SeqBAIJ_Check_Blocks(PetscInt idx[],PetscInt n,PetscInt bs,PetscInt sizes[], PetscInt *bs_max) 1568 { 1569 PetscInt i,j,k,row; 1570 PetscTruth flg; 1571 1572 PetscFunctionBegin; 1573 for (i=0,j=0; i<n; j++) { 1574 row = idx[i]; 1575 if (row%bs!=0) { /* Not the begining of a block */ 1576 sizes[j] = 1; 1577 i++; 1578 } else if (i+bs > n) { /* complete block doesn't exist (at idx end) */ 1579 sizes[j] = 1; /* Also makes sure atleast 'bs' values exist for next else */ 1580 i++; 1581 } else { /* Begining of the block, so check if the complete block exists */ 1582 flg = PETSC_TRUE; 1583 for (k=1; k<bs; k++) { 1584 if (row+k != idx[i+k]) { /* break in the block */ 1585 flg = PETSC_FALSE; 1586 break; 1587 } 1588 } 1589 if (flg) { /* No break in the bs */ 1590 sizes[j] = bs; 1591 i+= bs; 1592 } else { 1593 sizes[j] = 1; 1594 i++; 1595 } 1596 } 1597 } 1598 *bs_max = j; 1599 PetscFunctionReturn(0); 1600 } 1601 1602 #undef __FUNCT__ 1603 #define __FUNCT__ "MatZeroRows_SeqBAIJ" 1604 PetscErrorCode MatZeroRows_SeqBAIJ(Mat A,IS is,const PetscScalar *diag) 1605 { 1606 Mat_SeqBAIJ *baij=(Mat_SeqBAIJ*)A->data; 1607 PetscErrorCode ierr; 1608 PetscInt i,j,k,count,is_n,*is_idx,*rows; 1609 PetscInt bs=A->bs,bs2=baij->bs2,*sizes,row,bs_max; 1610 PetscScalar zero = 0.0; 1611 MatScalar *aa; 1612 1613 PetscFunctionBegin; 1614 /* Make a copy of the IS and sort it */ 1615 ierr = ISGetLocalSize(is,&is_n);CHKERRQ(ierr); 1616 ierr = ISGetIndices(is,&is_idx);CHKERRQ(ierr); 1617 1618 /* allocate memory for rows,sizes */ 1619 ierr = PetscMalloc((3*is_n+1)*sizeof(PetscInt),&rows);CHKERRQ(ierr); 1620 sizes = rows + is_n; 1621 1622 /* copy IS values to rows, and sort them */ 1623 for (i=0; i<is_n; i++) { rows[i] = is_idx[i]; } 1624 ierr = PetscSortInt(is_n,rows);CHKERRQ(ierr); 1625 if (baij->keepzeroedrows) { 1626 for (i=0; i<is_n; i++) { sizes[i] = 1; } 1627 bs_max = is_n; 1628 A->same_nonzero = PETSC_TRUE; 1629 } else { 1630 ierr = MatZeroRows_SeqBAIJ_Check_Blocks(rows,is_n,bs,sizes,&bs_max);CHKERRQ(ierr); 1631 A->same_nonzero = PETSC_FALSE; 1632 } 1633 ierr = ISRestoreIndices(is,&is_idx);CHKERRQ(ierr); 1634 1635 for (i=0,j=0; i<bs_max; j+=sizes[i],i++) { 1636 row = rows[j]; 1637 if (row < 0 || row > A->m) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"row %D out of range",row); 1638 count = (baij->i[row/bs +1] - baij->i[row/bs])*bs; 1639 aa = baij->a + baij->i[row/bs]*bs2 + (row%bs); 1640 if (sizes[i] == bs && !baij->keepzeroedrows) { 1641 if (diag) { 1642 if (baij->ilen[row/bs] > 0) { 1643 baij->ilen[row/bs] = 1; 1644 baij->j[baij->i[row/bs]] = row/bs; 1645 ierr = PetscMemzero(aa,count*bs*sizeof(MatScalar));CHKERRQ(ierr); 1646 } 1647 /* Now insert all the diagonal values for this bs */ 1648 for (k=0; k<bs; k++) { 1649 ierr = (*A->ops->setvalues)(A,1,rows+j+k,1,rows+j+k,diag,INSERT_VALUES);CHKERRQ(ierr); 1650 } 1651 } else { /* (!diag) */ 1652 baij->ilen[row/bs] = 0; 1653 } /* end (!diag) */ 1654 } else { /* (sizes[i] != bs) */ 1655 #if defined (PETSC_USE_DEBUG) 1656 if (sizes[i] != 1) SETERRQ(PETSC_ERR_PLIB,"Internal Error. Value should be 1"); 1657 #endif 1658 for (k=0; k<count; k++) { 1659 aa[0] = zero; 1660 aa += bs; 1661 } 1662 if (diag) { 1663 ierr = (*A->ops->setvalues)(A,1,rows+j,1,rows+j,diag,INSERT_VALUES);CHKERRQ(ierr); 1664 } 1665 } 1666 } 1667 1668 ierr = PetscFree(rows);CHKERRQ(ierr); 1669 ierr = MatAssemblyEnd_SeqBAIJ(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1670 PetscFunctionReturn(0); 1671 } 1672 1673 #undef __FUNCT__ 1674 #define __FUNCT__ "MatSetValues_SeqBAIJ" 1675 PetscErrorCode MatSetValues_SeqBAIJ(Mat A,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode is) 1676 { 1677 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 1678 PetscInt *rp,k,low,high,t,ii,row,nrow,i,col,l,rmax,N,lastcol = -1; 1679 PetscInt *imax=a->imax,*ai=a->i,*ailen=a->ilen; 1680 PetscInt *aj=a->j,nonew=a->nonew,bs=A->bs,brow,bcol; 1681 PetscErrorCode ierr; 1682 PetscInt ridx,cidx,bs2=a->bs2; 1683 PetscTruth roworiented=a->roworiented; 1684 MatScalar *ap,value,*aa=a->a,*bap; 1685 1686 PetscFunctionBegin; 1687 for (k=0; k<m; k++) { /* loop over added rows */ 1688 row = im[k]; brow = row/bs; 1689 if (row < 0) continue; 1690 #if defined(PETSC_USE_DEBUG) 1691 if (row >= A->m) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",row,A->m-1); 1692 #endif 1693 rp = aj + ai[brow]; 1694 ap = aa + bs2*ai[brow]; 1695 rmax = imax[brow]; 1696 nrow = ailen[brow]; 1697 low = 0; 1698 for (l=0; l<n; l++) { /* loop over added columns */ 1699 if (in[l] < 0) continue; 1700 #if defined(PETSC_USE_DEBUG) 1701 if (in[l] >= A->n) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %D max %D",in[l],A->n-1); 1702 #endif 1703 col = in[l]; bcol = col/bs; 1704 ridx = row % bs; cidx = col % bs; 1705 if (roworiented) { 1706 value = v[l + k*n]; 1707 } else { 1708 value = v[k + l*m]; 1709 } 1710 if (col < lastcol) low = 0; high = nrow; 1711 lastcol = col; 1712 while (high-low > 7) { 1713 t = (low+high)/2; 1714 if (rp[t] > bcol) high = t; 1715 else low = t; 1716 } 1717 for (i=low; i<high; i++) { 1718 if (rp[i] > bcol) break; 1719 if (rp[i] == bcol) { 1720 bap = ap + bs2*i + bs*cidx + ridx; 1721 if (is == ADD_VALUES) *bap += value; 1722 else *bap = value; 1723 goto noinsert1; 1724 } 1725 } 1726 if (nonew == 1) goto noinsert1; 1727 else if (nonew == -1) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%D, %D) in the matrix", row, col); 1728 if (nrow >= rmax) { 1729 /* there is no extra room in row, therefore enlarge */ 1730 PetscInt new_nz = ai[a->mbs] + CHUNKSIZE,len,*new_i,*new_j; 1731 MatScalar *new_a; 1732 1733 if (nonew == -2) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%D, %D) in the matrix", row, col); 1734 1735 /* Malloc new storage space */ 1736 len = new_nz*(sizeof(PetscInt)+bs2*sizeof(MatScalar))+(a->mbs+1)*sizeof(PetscInt); 1737 ierr = PetscMalloc(len,&new_a);CHKERRQ(ierr); 1738 new_j = (PetscInt*)(new_a + bs2*new_nz); 1739 new_i = new_j + new_nz; 1740 1741 /* copy over old data into new slots */ 1742 for (ii=0; ii<brow+1; ii++) {new_i[ii] = ai[ii];} 1743 for (ii=brow+1; ii<a->mbs+1; ii++) {new_i[ii] = ai[ii]+CHUNKSIZE;} 1744 ierr = PetscMemcpy(new_j,aj,(ai[brow]+nrow)*sizeof(PetscInt));CHKERRQ(ierr); 1745 len = (new_nz - CHUNKSIZE - ai[brow] - nrow); 1746 ierr = PetscMemcpy(new_j+ai[brow]+nrow+CHUNKSIZE,aj+ai[brow]+nrow,len*sizeof(PetscInt));CHKERRQ(ierr); 1747 ierr = PetscMemcpy(new_a,aa,(ai[brow]+nrow)*bs2*sizeof(MatScalar));CHKERRQ(ierr); 1748 ierr = PetscMemzero(new_a+bs2*(ai[brow]+nrow),bs2*CHUNKSIZE*sizeof(MatScalar));CHKERRQ(ierr); 1749 ierr = PetscMemcpy(new_a+bs2*(ai[brow]+nrow+CHUNKSIZE),aa+bs2*(ai[brow]+nrow),bs2*len*sizeof(MatScalar));CHKERRQ(ierr); 1750 /* free up old matrix storage */ 1751 ierr = PetscFree(a->a);CHKERRQ(ierr); 1752 if (!a->singlemalloc) { 1753 ierr = PetscFree(a->i);CHKERRQ(ierr); 1754 ierr = PetscFree(a->j);CHKERRQ(ierr); 1755 } 1756 aa = a->a = new_a; ai = a->i = new_i; aj = a->j = new_j; 1757 a->singlemalloc = PETSC_TRUE; 1758 1759 rp = aj + ai[brow]; ap = aa + bs2*ai[brow]; 1760 rmax = imax[brow] = imax[brow] + CHUNKSIZE; 1761 ierr = PetscLogObjectMemory(A,CHUNKSIZE*(sizeof(PetscInt) + bs2*sizeof(MatScalar)));CHKERRQ(ierr); 1762 a->maxnz += bs2*CHUNKSIZE; 1763 a->reallocs++; 1764 a->nz++; 1765 } 1766 N = nrow++ - 1; 1767 /* shift up all the later entries in this row */ 1768 for (ii=N; ii>=i; ii--) { 1769 rp[ii+1] = rp[ii]; 1770 ierr = PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar));CHKERRQ(ierr); 1771 } 1772 if (N>=i) { 1773 ierr = PetscMemzero(ap+bs2*i,bs2*sizeof(MatScalar));CHKERRQ(ierr); 1774 } 1775 rp[i] = bcol; 1776 ap[bs2*i + bs*cidx + ridx] = value; 1777 noinsert1:; 1778 low = i; 1779 } 1780 ailen[brow] = nrow; 1781 } 1782 A->same_nonzero = PETSC_FALSE; 1783 PetscFunctionReturn(0); 1784 } 1785 1786 1787 #undef __FUNCT__ 1788 #define __FUNCT__ "MatILUFactor_SeqBAIJ" 1789 PetscErrorCode MatILUFactor_SeqBAIJ(Mat inA,IS row,IS col,MatFactorInfo *info) 1790 { 1791 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)inA->data; 1792 Mat outA; 1793 PetscErrorCode ierr; 1794 PetscTruth row_identity,col_identity; 1795 1796 PetscFunctionBegin; 1797 if (info->levels != 0) SETERRQ(PETSC_ERR_SUP,"Only levels = 0 supported for in-place ILU"); 1798 ierr = ISIdentity(row,&row_identity);CHKERRQ(ierr); 1799 ierr = ISIdentity(col,&col_identity);CHKERRQ(ierr); 1800 if (!row_identity || !col_identity) { 1801 SETERRQ(PETSC_ERR_ARG_WRONG,"Row and column permutations must be identity for in-place ILU"); 1802 } 1803 1804 outA = inA; 1805 inA->factor = FACTOR_LU; 1806 1807 if (!a->diag) { 1808 ierr = MatMarkDiagonal_SeqBAIJ(inA);CHKERRQ(ierr); 1809 } 1810 1811 a->row = row; 1812 a->col = col; 1813 ierr = PetscObjectReference((PetscObject)row);CHKERRQ(ierr); 1814 ierr = PetscObjectReference((PetscObject)col);CHKERRQ(ierr); 1815 1816 /* Create the invert permutation so that it can be used in MatLUFactorNumeric() */ 1817 ierr = ISInvertPermutation(col,PETSC_DECIDE,&a->icol);CHKERRQ(ierr); 1818 ierr = PetscLogObjectParent(inA,a->icol);CHKERRQ(ierr); 1819 1820 /* 1821 Blocksize 2, 3, 4, 5, 6 and 7 have a special faster factorization/solver 1822 for ILU(0) factorization with natural ordering 1823 */ 1824 if (inA->bs < 8) { 1825 ierr = MatSeqBAIJ_UpdateFactorNumeric_NaturalOrdering(inA);CHKERRQ(ierr); 1826 } else { 1827 if (!a->solve_work) { 1828 ierr = PetscMalloc((inA->m+inA->bs)*sizeof(PetscScalar),&a->solve_work);CHKERRQ(ierr); 1829 ierr = PetscLogObjectMemory(inA,(inA->m+inA->bs)*sizeof(PetscScalar));CHKERRQ(ierr); 1830 } 1831 } 1832 1833 ierr = MatLUFactorNumeric(inA,info,&outA);CHKERRQ(ierr); 1834 1835 PetscFunctionReturn(0); 1836 } 1837 #undef __FUNCT__ 1838 #define __FUNCT__ "MatPrintHelp_SeqBAIJ" 1839 PetscErrorCode MatPrintHelp_SeqBAIJ(Mat A) 1840 { 1841 static PetscTruth called = PETSC_FALSE; 1842 MPI_Comm comm = A->comm; 1843 PetscErrorCode ierr; 1844 1845 PetscFunctionBegin; 1846 if (called) {PetscFunctionReturn(0);} else called = PETSC_TRUE; 1847 ierr = (*PetscHelpPrintf)(comm," Options for MATSEQBAIJ and MATMPIBAIJ matrix formats (the defaults):\n");CHKERRQ(ierr); 1848 ierr = (*PetscHelpPrintf)(comm," -mat_block_size <block_size>\n");CHKERRQ(ierr); 1849 PetscFunctionReturn(0); 1850 } 1851 1852 EXTERN_C_BEGIN 1853 #undef __FUNCT__ 1854 #define __FUNCT__ "MatSeqBAIJSetColumnIndices_SeqBAIJ" 1855 PetscErrorCode MatSeqBAIJSetColumnIndices_SeqBAIJ(Mat mat,PetscInt *indices) 1856 { 1857 Mat_SeqBAIJ *baij = (Mat_SeqBAIJ *)mat->data; 1858 PetscInt i,nz,nbs; 1859 1860 PetscFunctionBegin; 1861 nz = baij->maxnz/baij->bs2; 1862 nbs = baij->nbs; 1863 for (i=0; i<nz; i++) { 1864 baij->j[i] = indices[i]; 1865 } 1866 baij->nz = nz; 1867 for (i=0; i<nbs; i++) { 1868 baij->ilen[i] = baij->imax[i]; 1869 } 1870 1871 PetscFunctionReturn(0); 1872 } 1873 EXTERN_C_END 1874 1875 #undef __FUNCT__ 1876 #define __FUNCT__ "MatSeqBAIJSetColumnIndices" 1877 /*@ 1878 MatSeqBAIJSetColumnIndices - Set the column indices for all the rows 1879 in the matrix. 1880 1881 Input Parameters: 1882 + mat - the SeqBAIJ matrix 1883 - indices - the column indices 1884 1885 Level: advanced 1886 1887 Notes: 1888 This can be called if you have precomputed the nonzero structure of the 1889 matrix and want to provide it to the matrix object to improve the performance 1890 of the MatSetValues() operation. 1891 1892 You MUST have set the correct numbers of nonzeros per row in the call to 1893 MatCreateSeqBAIJ(). 1894 1895 MUST be called before any calls to MatSetValues(); 1896 1897 @*/ 1898 PetscErrorCode MatSeqBAIJSetColumnIndices(Mat mat,PetscInt *indices) 1899 { 1900 PetscErrorCode ierr,(*f)(Mat,PetscInt *); 1901 1902 PetscFunctionBegin; 1903 PetscValidHeaderSpecific(mat,MAT_COOKIE,1); 1904 PetscValidPointer(indices,2); 1905 ierr = PetscObjectQueryFunction((PetscObject)mat,"MatSeqBAIJSetColumnIndices_C",(void (**)(void))&f);CHKERRQ(ierr); 1906 if (f) { 1907 ierr = (*f)(mat,indices);CHKERRQ(ierr); 1908 } else { 1909 SETERRQ(PETSC_ERR_ARG_WRONG,"Wrong type of matrix to set column indices"); 1910 } 1911 PetscFunctionReturn(0); 1912 } 1913 1914 #undef __FUNCT__ 1915 #define __FUNCT__ "MatGetRowMax_SeqBAIJ" 1916 PetscErrorCode MatGetRowMax_SeqBAIJ(Mat A,Vec v) 1917 { 1918 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 1919 PetscErrorCode ierr; 1920 PetscInt i,j,n,row,bs,*ai,*aj,mbs; 1921 PetscReal atmp; 1922 PetscScalar *x,zero = 0.0; 1923 MatScalar *aa; 1924 PetscInt ncols,brow,krow,kcol; 1925 1926 PetscFunctionBegin; 1927 if (A->factor) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 1928 bs = A->bs; 1929 aa = a->a; 1930 ai = a->i; 1931 aj = a->j; 1932 mbs = a->mbs; 1933 1934 ierr = VecSet(&zero,v);CHKERRQ(ierr); 1935 ierr = VecGetArray(v,&x);CHKERRQ(ierr); 1936 ierr = VecGetLocalSize(v,&n);CHKERRQ(ierr); 1937 if (n != A->m) SETERRQ(PETSC_ERR_ARG_SIZ,"Nonconforming matrix and vector"); 1938 for (i=0; i<mbs; i++) { 1939 ncols = ai[1] - ai[0]; ai++; 1940 brow = bs*i; 1941 for (j=0; j<ncols; j++){ 1942 /* bcol = bs*(*aj); */ 1943 for (kcol=0; kcol<bs; kcol++){ 1944 for (krow=0; krow<bs; krow++){ 1945 atmp = PetscAbsScalar(*aa); aa++; 1946 row = brow + krow; /* row index */ 1947 /* printf("val[%d,%d]: %g\n",row,bcol+kcol,atmp); */ 1948 if (PetscAbsScalar(x[row]) < atmp) x[row] = atmp; 1949 } 1950 } 1951 aj++; 1952 } 1953 } 1954 ierr = VecRestoreArray(v,&x);CHKERRQ(ierr); 1955 PetscFunctionReturn(0); 1956 } 1957 1958 #undef __FUNCT__ 1959 #define __FUNCT__ "MatSetUpPreallocation_SeqBAIJ" 1960 PetscErrorCode MatSetUpPreallocation_SeqBAIJ(Mat A) 1961 { 1962 PetscErrorCode ierr; 1963 1964 PetscFunctionBegin; 1965 ierr = MatSeqBAIJSetPreallocation(A,1,PETSC_DEFAULT,0);CHKERRQ(ierr); 1966 PetscFunctionReturn(0); 1967 } 1968 1969 #undef __FUNCT__ 1970 #define __FUNCT__ "MatGetArray_SeqBAIJ" 1971 PetscErrorCode MatGetArray_SeqBAIJ(Mat A,PetscScalar *array[]) 1972 { 1973 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 1974 PetscFunctionBegin; 1975 *array = a->a; 1976 PetscFunctionReturn(0); 1977 } 1978 1979 #undef __FUNCT__ 1980 #define __FUNCT__ "MatRestoreArray_SeqBAIJ" 1981 PetscErrorCode MatRestoreArray_SeqBAIJ(Mat A,PetscScalar *array[]) 1982 { 1983 PetscFunctionBegin; 1984 PetscFunctionReturn(0); 1985 } 1986 1987 #include "petscblaslapack.h" 1988 #undef __FUNCT__ 1989 #define __FUNCT__ "MatAXPY_SeqBAIJ" 1990 PetscErrorCode MatAXPY_SeqBAIJ(const PetscScalar *a,Mat X,Mat Y,MatStructure str) 1991 { 1992 Mat_SeqBAIJ *x = (Mat_SeqBAIJ *)X->data,*y = (Mat_SeqBAIJ *)Y->data; 1993 PetscErrorCode ierr; 1994 PetscInt i,bs=Y->bs,j,bs2; 1995 PetscBLASInt one=1,bnz = (PetscBLASInt)x->nz; 1996 1997 PetscFunctionBegin; 1998 if (str == SAME_NONZERO_PATTERN) { 1999 BLASaxpy_(&bnz,(PetscScalar*)a,x->a,&one,y->a,&one); 2000 } else if (str == SUBSET_NONZERO_PATTERN) { /* nonzeros of X is a subset of Y's */ 2001 if (y->xtoy && y->XtoY != X) { 2002 ierr = PetscFree(y->xtoy);CHKERRQ(ierr); 2003 ierr = MatDestroy(y->XtoY);CHKERRQ(ierr); 2004 } 2005 if (!y->xtoy) { /* get xtoy */ 2006 ierr = MatAXPYGetxtoy_Private(x->mbs,x->i,x->j,PETSC_NULL, y->i,y->j,PETSC_NULL, &y->xtoy);CHKERRQ(ierr); 2007 y->XtoY = X; 2008 } 2009 bs2 = bs*bs; 2010 for (i=0; i<x->nz; i++) { 2011 j = 0; 2012 while (j < bs2){ 2013 y->a[bs2*y->xtoy[i]+j] += (*a)*(x->a[bs2*i+j]); 2014 j++; 2015 } 2016 } 2017 PetscLogInfo(0,"MatAXPY_SeqBAIJ: ratio of nnz(X)/nnz(Y): %D/%D = %g\n",bs2*x->nz,bs2*y->nz,(PetscReal)(bs2*x->nz)/(bs2*y->nz)); 2018 } else { 2019 ierr = MatAXPY_Basic(a,X,Y,str);CHKERRQ(ierr); 2020 } 2021 PetscFunctionReturn(0); 2022 } 2023 2024 /* -------------------------------------------------------------------*/ 2025 static struct _MatOps MatOps_Values = {MatSetValues_SeqBAIJ, 2026 MatGetRow_SeqBAIJ, 2027 MatRestoreRow_SeqBAIJ, 2028 MatMult_SeqBAIJ_N, 2029 /* 4*/ MatMultAdd_SeqBAIJ_N, 2030 MatMultTranspose_SeqBAIJ, 2031 MatMultTransposeAdd_SeqBAIJ, 2032 MatSolve_SeqBAIJ_N, 2033 0, 2034 0, 2035 /*10*/ 0, 2036 MatLUFactor_SeqBAIJ, 2037 0, 2038 0, 2039 MatTranspose_SeqBAIJ, 2040 /*15*/ MatGetInfo_SeqBAIJ, 2041 MatEqual_SeqBAIJ, 2042 MatGetDiagonal_SeqBAIJ, 2043 MatDiagonalScale_SeqBAIJ, 2044 MatNorm_SeqBAIJ, 2045 /*20*/ 0, 2046 MatAssemblyEnd_SeqBAIJ, 2047 0, 2048 MatSetOption_SeqBAIJ, 2049 MatZeroEntries_SeqBAIJ, 2050 /*25*/ MatZeroRows_SeqBAIJ, 2051 MatLUFactorSymbolic_SeqBAIJ, 2052 MatLUFactorNumeric_SeqBAIJ_N, 2053 MatCholeskyFactorSymbolic_SeqBAIJ, 2054 MatCholeskyFactorNumeric_SeqBAIJ_N, 2055 /*30*/ MatSetUpPreallocation_SeqBAIJ, 2056 MatILUFactorSymbolic_SeqBAIJ, 2057 MatICCFactorSymbolic_SeqBAIJ, 2058 MatGetArray_SeqBAIJ, 2059 MatRestoreArray_SeqBAIJ, 2060 /*35*/ MatDuplicate_SeqBAIJ, 2061 0, 2062 0, 2063 MatILUFactor_SeqBAIJ, 2064 0, 2065 /*40*/ MatAXPY_SeqBAIJ, 2066 MatGetSubMatrices_SeqBAIJ, 2067 MatIncreaseOverlap_SeqBAIJ, 2068 MatGetValues_SeqBAIJ, 2069 0, 2070 /*45*/ MatPrintHelp_SeqBAIJ, 2071 MatScale_SeqBAIJ, 2072 0, 2073 0, 2074 0, 2075 /*50*/ 0, 2076 MatGetRowIJ_SeqBAIJ, 2077 MatRestoreRowIJ_SeqBAIJ, 2078 0, 2079 0, 2080 /*55*/ 0, 2081 0, 2082 0, 2083 0, 2084 MatSetValuesBlocked_SeqBAIJ, 2085 /*60*/ MatGetSubMatrix_SeqBAIJ, 2086 MatDestroy_SeqBAIJ, 2087 MatView_SeqBAIJ, 2088 MatGetPetscMaps_Petsc, 2089 0, 2090 /*65*/ 0, 2091 0, 2092 0, 2093 0, 2094 0, 2095 /*70*/ MatGetRowMax_SeqBAIJ, 2096 MatConvert_Basic, 2097 0, 2098 0, 2099 0, 2100 /*75*/ 0, 2101 0, 2102 0, 2103 0, 2104 0, 2105 /*80*/ 0, 2106 0, 2107 0, 2108 0, 2109 MatLoad_SeqBAIJ, 2110 /*85*/ 0, 2111 0, 2112 0, 2113 0, 2114 0, 2115 /*90*/ 0, 2116 0, 2117 0, 2118 0, 2119 0, 2120 /*95*/ 0, 2121 0, 2122 0, 2123 0}; 2124 2125 EXTERN_C_BEGIN 2126 #undef __FUNCT__ 2127 #define __FUNCT__ "MatStoreValues_SeqBAIJ" 2128 PetscErrorCode MatStoreValues_SeqBAIJ(Mat mat) 2129 { 2130 Mat_SeqBAIJ *aij = (Mat_SeqBAIJ *)mat->data; 2131 PetscInt nz = aij->i[mat->m]*mat->bs*aij->bs2; 2132 PetscErrorCode ierr; 2133 2134 PetscFunctionBegin; 2135 if (aij->nonew != 1) { 2136 SETERRQ(PETSC_ERR_ORDER,"Must call MatSetOption(A,MAT_NO_NEW_NONZERO_LOCATIONS);first"); 2137 } 2138 2139 /* allocate space for values if not already there */ 2140 if (!aij->saved_values) { 2141 ierr = PetscMalloc((nz+1)*sizeof(PetscScalar),&aij->saved_values);CHKERRQ(ierr); 2142 } 2143 2144 /* copy values over */ 2145 ierr = PetscMemcpy(aij->saved_values,aij->a,nz*sizeof(PetscScalar));CHKERRQ(ierr); 2146 PetscFunctionReturn(0); 2147 } 2148 EXTERN_C_END 2149 2150 EXTERN_C_BEGIN 2151 #undef __FUNCT__ 2152 #define __FUNCT__ "MatRetrieveValues_SeqBAIJ" 2153 PetscErrorCode MatRetrieveValues_SeqBAIJ(Mat mat) 2154 { 2155 Mat_SeqBAIJ *aij = (Mat_SeqBAIJ *)mat->data; 2156 PetscErrorCode ierr; 2157 PetscInt nz = aij->i[mat->m]*mat->bs*aij->bs2; 2158 2159 PetscFunctionBegin; 2160 if (aij->nonew != 1) { 2161 SETERRQ(PETSC_ERR_ORDER,"Must call MatSetOption(A,MAT_NO_NEW_NONZERO_LOCATIONS);first"); 2162 } 2163 if (!aij->saved_values) { 2164 SETERRQ(PETSC_ERR_ORDER,"Must call MatStoreValues(A);first"); 2165 } 2166 2167 /* copy values over */ 2168 ierr = PetscMemcpy(aij->a,aij->saved_values,nz*sizeof(PetscScalar));CHKERRQ(ierr); 2169 PetscFunctionReturn(0); 2170 } 2171 EXTERN_C_END 2172 2173 EXTERN_C_BEGIN 2174 extern PetscErrorCode MatConvert_SeqBAIJ_SeqAIJ(Mat,const MatType,Mat*); 2175 extern PetscErrorCode MatConvert_SeqBAIJ_SeqSBAIJ(Mat,const MatType,Mat*); 2176 EXTERN_C_END 2177 2178 EXTERN_C_BEGIN 2179 #undef __FUNCT__ 2180 #define __FUNCT__ "MatSeqBAIJSetPreallocation_SeqBAIJ" 2181 PetscErrorCode MatSeqBAIJSetPreallocation_SeqBAIJ(Mat B,PetscInt bs,PetscInt nz,PetscInt *nnz) 2182 { 2183 Mat_SeqBAIJ *b; 2184 PetscErrorCode ierr; 2185 PetscInt i,len,mbs,nbs,bs2,newbs = bs; 2186 PetscTruth flg; 2187 2188 PetscFunctionBegin; 2189 2190 B->preallocated = PETSC_TRUE; 2191 ierr = PetscOptionsGetInt(B->prefix,"-mat_block_size",&newbs,PETSC_NULL);CHKERRQ(ierr); 2192 if (nnz && newbs != bs) { 2193 SETERRQ(PETSC_ERR_ARG_WRONG,"Cannot change blocksize from command line if setting nnz"); 2194 } 2195 bs = newbs; 2196 2197 mbs = B->m/bs; 2198 nbs = B->n/bs; 2199 bs2 = bs*bs; 2200 2201 if (mbs*bs!=B->m || nbs*bs!=B->n) { 2202 SETERRQ3(PETSC_ERR_ARG_SIZ,"Number rows %D, cols %D must be divisible by blocksize %D",B->m,B->n,bs); 2203 } 2204 2205 if (nz == PETSC_DEFAULT || nz == PETSC_DECIDE) nz = 5; 2206 if (nz < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"nz cannot be less than 0: value %D",nz); 2207 if (nnz) { 2208 for (i=0; i<mbs; i++) { 2209 if (nnz[i] < 0) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"nnz cannot be less than 0: local row %D value %D",i,nnz[i]); 2210 if (nnz[i] > nbs) SETERRQ3(PETSC_ERR_ARG_OUTOFRANGE,"nnz cannot be greater than block row length: local row %D value %D rowlength %D",i,nnz[i],nbs); 2211 } 2212 } 2213 2214 b = (Mat_SeqBAIJ*)B->data; 2215 ierr = PetscOptionsHasName(PETSC_NULL,"-mat_no_unroll",&flg);CHKERRQ(ierr); 2216 B->ops->solve = MatSolve_SeqBAIJ_Update; 2217 B->ops->solvetranspose = MatSolveTranspose_SeqBAIJ_Update; 2218 if (!flg) { 2219 switch (bs) { 2220 case 1: 2221 B->ops->lufactornumeric = MatLUFactorNumeric_SeqBAIJ_1; 2222 B->ops->mult = MatMult_SeqBAIJ_1; 2223 B->ops->multadd = MatMultAdd_SeqBAIJ_1; 2224 break; 2225 case 2: 2226 B->ops->lufactornumeric = MatLUFactorNumeric_SeqBAIJ_2; 2227 B->ops->mult = MatMult_SeqBAIJ_2; 2228 B->ops->multadd = MatMultAdd_SeqBAIJ_2; 2229 B->ops->pbrelax = MatPBRelax_SeqBAIJ_2; 2230 break; 2231 case 3: 2232 B->ops->lufactornumeric = MatLUFactorNumeric_SeqBAIJ_3; 2233 B->ops->mult = MatMult_SeqBAIJ_3; 2234 B->ops->multadd = MatMultAdd_SeqBAIJ_3; 2235 B->ops->pbrelax = MatPBRelax_SeqBAIJ_3; 2236 break; 2237 case 4: 2238 B->ops->lufactornumeric = MatLUFactorNumeric_SeqBAIJ_4; 2239 B->ops->mult = MatMult_SeqBAIJ_4; 2240 B->ops->multadd = MatMultAdd_SeqBAIJ_4; 2241 B->ops->pbrelax = MatPBRelax_SeqBAIJ_4; 2242 break; 2243 case 5: 2244 B->ops->lufactornumeric = MatLUFactorNumeric_SeqBAIJ_5; 2245 B->ops->mult = MatMult_SeqBAIJ_5; 2246 B->ops->multadd = MatMultAdd_SeqBAIJ_5; 2247 B->ops->pbrelax = MatPBRelax_SeqBAIJ_5; 2248 break; 2249 case 6: 2250 B->ops->lufactornumeric = MatLUFactorNumeric_SeqBAIJ_6; 2251 B->ops->mult = MatMult_SeqBAIJ_6; 2252 B->ops->multadd = MatMultAdd_SeqBAIJ_6; 2253 break; 2254 case 7: 2255 B->ops->lufactornumeric = MatLUFactorNumeric_SeqBAIJ_7; 2256 B->ops->mult = MatMult_SeqBAIJ_7; 2257 B->ops->multadd = MatMultAdd_SeqBAIJ_7; 2258 break; 2259 default: 2260 B->ops->lufactornumeric = MatLUFactorNumeric_SeqBAIJ_N; 2261 B->ops->mult = MatMult_SeqBAIJ_N; 2262 B->ops->multadd = MatMultAdd_SeqBAIJ_N; 2263 break; 2264 } 2265 } 2266 B->bs = bs; 2267 b->mbs = mbs; 2268 b->nbs = nbs; 2269 ierr = PetscMalloc((mbs+1)*sizeof(PetscInt),&b->imax);CHKERRQ(ierr); 2270 if (!nnz) { 2271 if (nz == PETSC_DEFAULT || nz == PETSC_DECIDE) nz = 5; 2272 else if (nz <= 0) nz = 1; 2273 for (i=0; i<mbs; i++) b->imax[i] = nz; 2274 nz = nz*mbs; 2275 } else { 2276 nz = 0; 2277 for (i=0; i<mbs; i++) {b->imax[i] = nnz[i]; nz += nnz[i];} 2278 } 2279 2280 /* allocate the matrix space */ 2281 len = nz*sizeof(PetscInt) + nz*bs2*sizeof(MatScalar) + (B->m+1)*sizeof(PetscInt); 2282 ierr = PetscMalloc(len,&b->a);CHKERRQ(ierr); 2283 ierr = PetscMemzero(b->a,nz*bs2*sizeof(MatScalar));CHKERRQ(ierr); 2284 b->j = (PetscInt*)(b->a + nz*bs2); 2285 ierr = PetscMemzero(b->j,nz*sizeof(PetscInt));CHKERRQ(ierr); 2286 b->i = b->j + nz; 2287 b->singlemalloc = PETSC_TRUE; 2288 2289 b->i[0] = 0; 2290 for (i=1; i<mbs+1; i++) { 2291 b->i[i] = b->i[i-1] + b->imax[i-1]; 2292 } 2293 2294 /* b->ilen will count nonzeros in each block row so far. */ 2295 ierr = PetscMalloc((mbs+1)*sizeof(PetscInt),&b->ilen);CHKERRQ(ierr); 2296 ierr = PetscLogObjectMemory(B,len+2*(mbs+1)*sizeof(PetscInt)+sizeof(struct _p_Mat)+sizeof(Mat_SeqBAIJ));CHKERRQ(ierr); 2297 for (i=0; i<mbs; i++) { b->ilen[i] = 0;} 2298 2299 B->bs = bs; 2300 b->bs2 = bs2; 2301 b->mbs = mbs; 2302 b->nz = 0; 2303 b->maxnz = nz*bs2; 2304 B->info.nz_unneeded = (PetscReal)b->maxnz; 2305 PetscFunctionReturn(0); 2306 } 2307 EXTERN_C_END 2308 2309 /*MC 2310 MATSEQBAIJ - MATSEQBAIJ = "seqbaij" - A matrix type to be used for sequential block sparse matrices, based on 2311 block sparse compressed row format. 2312 2313 Options Database Keys: 2314 . -mat_type seqbaij - sets the matrix type to "seqbaij" during a call to MatSetFromOptions() 2315 2316 Level: beginner 2317 2318 .seealso: MatCreateSeqBAIJ 2319 M*/ 2320 2321 EXTERN_C_BEGIN 2322 #undef __FUNCT__ 2323 #define __FUNCT__ "MatCreate_SeqBAIJ" 2324 PetscErrorCode MatCreate_SeqBAIJ(Mat B) 2325 { 2326 PetscErrorCode ierr; 2327 PetscMPIInt size; 2328 Mat_SeqBAIJ *b; 2329 2330 PetscFunctionBegin; 2331 ierr = MPI_Comm_size(B->comm,&size);CHKERRQ(ierr); 2332 if (size > 1) SETERRQ(PETSC_ERR_ARG_WRONG,"Comm must be of size 1"); 2333 2334 B->m = B->M = PetscMax(B->m,B->M); 2335 B->n = B->N = PetscMax(B->n,B->N); 2336 ierr = PetscNew(Mat_SeqBAIJ,&b);CHKERRQ(ierr); 2337 B->data = (void*)b; 2338 ierr = PetscMemcpy(B->ops,&MatOps_Values,sizeof(struct _MatOps));CHKERRQ(ierr); 2339 B->factor = 0; 2340 B->lupivotthreshold = 1.0; 2341 B->mapping = 0; 2342 b->row = 0; 2343 b->col = 0; 2344 b->icol = 0; 2345 b->reallocs = 0; 2346 b->saved_values = 0; 2347 #if defined(PETSC_USE_MAT_SINGLE) 2348 b->setvalueslen = 0; 2349 b->setvaluescopy = PETSC_NULL; 2350 #endif 2351 2352 ierr = PetscMapCreateMPI(B->comm,B->m,B->m,&B->rmap);CHKERRQ(ierr); 2353 ierr = PetscMapCreateMPI(B->comm,B->n,B->n,&B->cmap);CHKERRQ(ierr); 2354 2355 b->sorted = PETSC_FALSE; 2356 b->roworiented = PETSC_TRUE; 2357 b->nonew = 0; 2358 b->diag = 0; 2359 b->solve_work = 0; 2360 b->mult_work = 0; 2361 B->spptr = 0; 2362 B->info.nz_unneeded = (PetscReal)b->maxnz; 2363 b->keepzeroedrows = PETSC_FALSE; 2364 b->xtoy = 0; 2365 b->XtoY = 0; 2366 b->compressedrow.use = PETSC_FALSE; 2367 b->compressedrow.nrows = 0; 2368 b->compressedrow.i = PETSC_NULL; 2369 b->compressedrow.rindex = PETSC_NULL; 2370 b->compressedrow.checked = PETSC_FALSE; 2371 B->same_nonzero = PETSC_FALSE; 2372 2373 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatStoreValues_C", 2374 "MatStoreValues_SeqBAIJ", 2375 MatStoreValues_SeqBAIJ);CHKERRQ(ierr); 2376 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatRetrieveValues_C", 2377 "MatRetrieveValues_SeqBAIJ", 2378 MatRetrieveValues_SeqBAIJ);CHKERRQ(ierr); 2379 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatSeqBAIJSetColumnIndices_C", 2380 "MatSeqBAIJSetColumnIndices_SeqBAIJ", 2381 MatSeqBAIJSetColumnIndices_SeqBAIJ);CHKERRQ(ierr); 2382 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatConvert_seqbaij_seqaij_C", 2383 "MatConvert_SeqBAIJ_SeqAIJ", 2384 MatConvert_SeqBAIJ_SeqAIJ);CHKERRQ(ierr); 2385 #if !defined(PETSC_USE_64BIT_INT) 2386 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatConvert_seqbaij_seqsbaij_C", 2387 "MatConvert_SeqBAIJ_SeqSBAIJ", 2388 MatConvert_SeqBAIJ_SeqSBAIJ);CHKERRQ(ierr); 2389 #endif 2390 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatSeqBAIJSetPreallocation_C", 2391 "MatSeqBAIJSetPreallocation_SeqBAIJ", 2392 MatSeqBAIJSetPreallocation_SeqBAIJ);CHKERRQ(ierr); 2393 PetscFunctionReturn(0); 2394 } 2395 EXTERN_C_END 2396 2397 #undef __FUNCT__ 2398 #define __FUNCT__ "MatDuplicate_SeqBAIJ" 2399 PetscErrorCode MatDuplicate_SeqBAIJ(Mat A,MatDuplicateOption cpvalues,Mat *B) 2400 { 2401 Mat C; 2402 Mat_SeqBAIJ *c,*a = (Mat_SeqBAIJ*)A->data; 2403 PetscErrorCode ierr; 2404 PetscInt i,len,mbs = a->mbs,nz = a->nz,bs2 = a->bs2; 2405 2406 PetscFunctionBegin; 2407 if (a->i[mbs] != nz) SETERRQ(PETSC_ERR_PLIB,"Corrupt matrix"); 2408 2409 *B = 0; 2410 ierr = MatCreate(A->comm,A->m,A->n,A->m,A->n,&C);CHKERRQ(ierr); 2411 ierr = MatSetType(C,A->type_name);CHKERRQ(ierr); 2412 ierr = PetscMemcpy(C->ops,A->ops,sizeof(struct _MatOps));CHKERRQ(ierr); 2413 c = (Mat_SeqBAIJ*)C->data; 2414 2415 C->M = A->M; 2416 C->N = A->N; 2417 C->bs = A->bs; 2418 c->bs2 = a->bs2; 2419 c->mbs = a->mbs; 2420 c->nbs = a->nbs; 2421 2422 ierr = PetscMalloc((mbs+1)*sizeof(PetscInt),&c->imax);CHKERRQ(ierr); 2423 ierr = PetscMalloc((mbs+1)*sizeof(PetscInt),&c->ilen);CHKERRQ(ierr); 2424 for (i=0; i<mbs; i++) { 2425 c->imax[i] = a->imax[i]; 2426 c->ilen[i] = a->ilen[i]; 2427 } 2428 2429 /* allocate the matrix space */ 2430 c->singlemalloc = PETSC_TRUE; 2431 len = (mbs+1)*sizeof(PetscInt) + nz*(bs2*sizeof(MatScalar) + sizeof(PetscInt)); 2432 ierr = PetscMalloc(len,&c->a);CHKERRQ(ierr); 2433 c->j = (PetscInt*)(c->a + nz*bs2); 2434 c->i = c->j + nz; 2435 ierr = PetscMemcpy(c->i,a->i,(mbs+1)*sizeof(PetscInt));CHKERRQ(ierr); 2436 if (mbs > 0) { 2437 ierr = PetscMemcpy(c->j,a->j,nz*sizeof(PetscInt));CHKERRQ(ierr); 2438 if (cpvalues == MAT_COPY_VALUES) { 2439 ierr = PetscMemcpy(c->a,a->a,bs2*nz*sizeof(MatScalar));CHKERRQ(ierr); 2440 } else { 2441 ierr = PetscMemzero(c->a,bs2*nz*sizeof(MatScalar));CHKERRQ(ierr); 2442 } 2443 } 2444 2445 ierr = PetscLogObjectMemory(C,len+2*(mbs+1)*sizeof(PetscInt)+sizeof(struct _p_Mat)+sizeof(Mat_SeqBAIJ));CHKERRQ(ierr); 2446 c->sorted = a->sorted; 2447 c->roworiented = a->roworiented; 2448 c->nonew = a->nonew; 2449 2450 if (a->diag) { 2451 ierr = PetscMalloc((mbs+1)*sizeof(PetscInt),&c->diag);CHKERRQ(ierr); 2452 ierr = PetscLogObjectMemory(C,(mbs+1)*sizeof(PetscInt));CHKERRQ(ierr); 2453 for (i=0; i<mbs; i++) { 2454 c->diag[i] = a->diag[i]; 2455 } 2456 } else c->diag = 0; 2457 c->nz = a->nz; 2458 c->maxnz = a->maxnz; 2459 c->solve_work = 0; 2460 c->mult_work = 0; 2461 C->preallocated = PETSC_TRUE; 2462 C->assembled = PETSC_TRUE; 2463 2464 c->compressedrow.use = a->compressedrow.use; 2465 c->compressedrow.nrows = a->compressedrow.nrows; 2466 c->compressedrow.checked = a->compressedrow.checked; 2467 if ( a->compressedrow.checked && a->compressedrow.use){ 2468 i = a->compressedrow.nrows; 2469 ierr = PetscMalloc((2*i+1)*sizeof(PetscInt),&c->compressedrow.i);CHKERRQ(ierr); 2470 c->compressedrow.rindex = c->compressedrow.i + i + 1; 2471 ierr = PetscMemcpy(c->compressedrow.i,a->compressedrow.i,(i+1)*sizeof(PetscInt));CHKERRQ(ierr); 2472 ierr = PetscMemcpy(c->compressedrow.rindex,a->compressedrow.rindex,i*sizeof(PetscInt));CHKERRQ(ierr); 2473 } else { 2474 c->compressedrow.use = PETSC_FALSE; 2475 c->compressedrow.i = PETSC_NULL; 2476 c->compressedrow.rindex = PETSC_NULL; 2477 } 2478 C->same_nonzero = A->same_nonzero; 2479 *B = C; 2480 ierr = PetscFListDuplicate(A->qlist,&C->qlist);CHKERRQ(ierr); 2481 PetscFunctionReturn(0); 2482 } 2483 2484 #undef __FUNCT__ 2485 #define __FUNCT__ "MatLoad_SeqBAIJ" 2486 PetscErrorCode MatLoad_SeqBAIJ(PetscViewer viewer,const MatType type,Mat *A) 2487 { 2488 Mat_SeqBAIJ *a; 2489 Mat B; 2490 PetscErrorCode ierr; 2491 PetscInt i,nz,header[4],*rowlengths=0,M,N,bs=1; 2492 PetscInt *mask,mbs,*jj,j,rowcount,nzcount,k,*browlengths,maskcount; 2493 PetscInt kmax,jcount,block,idx,point,nzcountb,extra_rows; 2494 PetscInt *masked,nmask,tmp,bs2,ishift; 2495 PetscMPIInt size; 2496 int fd; 2497 PetscScalar *aa; 2498 MPI_Comm comm = ((PetscObject)viewer)->comm; 2499 2500 PetscFunctionBegin; 2501 ierr = PetscOptionsGetInt(PETSC_NULL,"-matload_block_size",&bs,PETSC_NULL);CHKERRQ(ierr); 2502 bs2 = bs*bs; 2503 2504 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 2505 if (size > 1) SETERRQ(PETSC_ERR_ARG_WRONG,"view must have one processor"); 2506 ierr = PetscViewerBinaryGetDescriptor(viewer,&fd);CHKERRQ(ierr); 2507 ierr = PetscBinaryRead(fd,header,4,PETSC_INT);CHKERRQ(ierr); 2508 if (header[0] != MAT_FILE_COOKIE) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"not Mat object"); 2509 M = header[1]; N = header[2]; nz = header[3]; 2510 2511 if (header[3] < 0) { 2512 SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"Matrix stored in special format, cannot load as SeqBAIJ"); 2513 } 2514 2515 if (M != N) SETERRQ(PETSC_ERR_SUP,"Can only do square matrices"); 2516 2517 /* 2518 This code adds extra rows to make sure the number of rows is 2519 divisible by the blocksize 2520 */ 2521 mbs = M/bs; 2522 extra_rows = bs - M + bs*(mbs); 2523 if (extra_rows == bs) extra_rows = 0; 2524 else mbs++; 2525 if (extra_rows) { 2526 PetscLogInfo(0,"MatLoad_SeqBAIJ:Padding loaded matrix to match blocksize\n"); 2527 } 2528 2529 /* read in row lengths */ 2530 ierr = PetscMalloc((M+extra_rows)*sizeof(PetscInt),&rowlengths);CHKERRQ(ierr); 2531 ierr = PetscBinaryRead(fd,rowlengths,M,PETSC_INT);CHKERRQ(ierr); 2532 for (i=0; i<extra_rows; i++) rowlengths[M+i] = 1; 2533 2534 /* read in column indices */ 2535 ierr = PetscMalloc((nz+extra_rows)*sizeof(PetscInt),&jj);CHKERRQ(ierr); 2536 ierr = PetscBinaryRead(fd,jj,nz,PETSC_INT);CHKERRQ(ierr); 2537 for (i=0; i<extra_rows; i++) jj[nz+i] = M+i; 2538 2539 /* loop over row lengths determining block row lengths */ 2540 ierr = PetscMalloc(mbs*sizeof(PetscInt),&browlengths);CHKERRQ(ierr); 2541 ierr = PetscMemzero(browlengths,mbs*sizeof(PetscInt));CHKERRQ(ierr); 2542 ierr = PetscMalloc(2*mbs*sizeof(PetscInt),&mask);CHKERRQ(ierr); 2543 ierr = PetscMemzero(mask,mbs*sizeof(PetscInt));CHKERRQ(ierr); 2544 masked = mask + mbs; 2545 rowcount = 0; nzcount = 0; 2546 for (i=0; i<mbs; i++) { 2547 nmask = 0; 2548 for (j=0; j<bs; j++) { 2549 kmax = rowlengths[rowcount]; 2550 for (k=0; k<kmax; k++) { 2551 tmp = jj[nzcount++]/bs; 2552 if (!mask[tmp]) {masked[nmask++] = tmp; mask[tmp] = 1;} 2553 } 2554 rowcount++; 2555 } 2556 browlengths[i] += nmask; 2557 /* zero out the mask elements we set */ 2558 for (j=0; j<nmask; j++) mask[masked[j]] = 0; 2559 } 2560 2561 /* create our matrix */ 2562 ierr = MatCreate(comm,PETSC_DECIDE,PETSC_DECIDE,M+extra_rows,N+extra_rows,&B); 2563 ierr = MatSetType(B,type);CHKERRQ(ierr); 2564 ierr = MatSeqBAIJSetPreallocation(B,bs,0,browlengths);CHKERRQ(ierr); 2565 a = (Mat_SeqBAIJ*)B->data; 2566 2567 /* set matrix "i" values */ 2568 a->i[0] = 0; 2569 for (i=1; i<= mbs; i++) { 2570 a->i[i] = a->i[i-1] + browlengths[i-1]; 2571 a->ilen[i-1] = browlengths[i-1]; 2572 } 2573 a->nz = 0; 2574 for (i=0; i<mbs; i++) a->nz += browlengths[i]; 2575 2576 /* read in nonzero values */ 2577 ierr = PetscMalloc((nz+extra_rows)*sizeof(PetscScalar),&aa);CHKERRQ(ierr); 2578 ierr = PetscBinaryRead(fd,aa,nz,PETSC_SCALAR);CHKERRQ(ierr); 2579 for (i=0; i<extra_rows; i++) aa[nz+i] = 1.0; 2580 2581 /* set "a" and "j" values into matrix */ 2582 nzcount = 0; jcount = 0; 2583 for (i=0; i<mbs; i++) { 2584 nzcountb = nzcount; 2585 nmask = 0; 2586 for (j=0; j<bs; j++) { 2587 kmax = rowlengths[i*bs+j]; 2588 for (k=0; k<kmax; k++) { 2589 tmp = jj[nzcount++]/bs; 2590 if (!mask[tmp]) { masked[nmask++] = tmp; mask[tmp] = 1;} 2591 } 2592 } 2593 /* sort the masked values */ 2594 ierr = PetscSortInt(nmask,masked);CHKERRQ(ierr); 2595 2596 /* set "j" values into matrix */ 2597 maskcount = 1; 2598 for (j=0; j<nmask; j++) { 2599 a->j[jcount++] = masked[j]; 2600 mask[masked[j]] = maskcount++; 2601 } 2602 /* set "a" values into matrix */ 2603 ishift = bs2*a->i[i]; 2604 for (j=0; j<bs; j++) { 2605 kmax = rowlengths[i*bs+j]; 2606 for (k=0; k<kmax; k++) { 2607 tmp = jj[nzcountb]/bs ; 2608 block = mask[tmp] - 1; 2609 point = jj[nzcountb] - bs*tmp; 2610 idx = ishift + bs2*block + j + bs*point; 2611 a->a[idx] = (MatScalar)aa[nzcountb++]; 2612 } 2613 } 2614 /* zero out the mask elements we set */ 2615 for (j=0; j<nmask; j++) mask[masked[j]] = 0; 2616 } 2617 if (jcount != a->nz) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"Bad binary matrix"); 2618 2619 ierr = PetscFree(rowlengths);CHKERRQ(ierr); 2620 ierr = PetscFree(browlengths);CHKERRQ(ierr); 2621 ierr = PetscFree(aa);CHKERRQ(ierr); 2622 ierr = PetscFree(jj);CHKERRQ(ierr); 2623 ierr = PetscFree(mask);CHKERRQ(ierr); 2624 2625 ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2626 ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2627 ierr = MatView_Private(B);CHKERRQ(ierr); 2628 2629 *A = B; 2630 PetscFunctionReturn(0); 2631 } 2632 2633 #undef __FUNCT__ 2634 #define __FUNCT__ "MatCreateSeqBAIJ" 2635 /*@C 2636 MatCreateSeqBAIJ - Creates a sparse matrix in block AIJ (block 2637 compressed row) format. For good matrix assembly performance the 2638 user should preallocate the matrix storage by setting the parameter nz 2639 (or the array nnz). By setting these parameters accurately, performance 2640 during matrix assembly can be increased by more than a factor of 50. 2641 2642 Collective on MPI_Comm 2643 2644 Input Parameters: 2645 + comm - MPI communicator, set to PETSC_COMM_SELF 2646 . bs - size of block 2647 . m - number of rows 2648 . n - number of columns 2649 . nz - number of nonzero blocks per block row (same for all rows) 2650 - nnz - array containing the number of nonzero blocks in the various block rows 2651 (possibly different for each block row) or PETSC_NULL 2652 2653 Output Parameter: 2654 . A - the matrix 2655 2656 Options Database Keys: 2657 . -mat_no_unroll - uses code that does not unroll the loops in the 2658 block calculations (much slower) 2659 . -mat_block_size - size of the blocks to use 2660 2661 Level: intermediate 2662 2663 Notes: 2664 If the nnz parameter is given then the nz parameter is ignored 2665 2666 A nonzero block is any block that as 1 or more nonzeros in it 2667 2668 The block AIJ format is fully compatible with standard Fortran 77 2669 storage. That is, the stored row and column indices can begin at 2670 either one (as in Fortran) or zero. See the users' manual for details. 2671 2672 Specify the preallocated storage with either nz or nnz (not both). 2673 Set nz=PETSC_DEFAULT and nnz=PETSC_NULL for PETSc to control dynamic memory 2674 allocation. For additional details, see the users manual chapter on 2675 matrices. 2676 2677 .seealso: MatCreate(), MatCreateSeqAIJ(), MatSetValues(), MatCreateMPIBAIJ() 2678 @*/ 2679 PetscErrorCode MatCreateSeqBAIJ(MPI_Comm comm,PetscInt bs,PetscInt m,PetscInt n,PetscInt nz,const PetscInt nnz[],Mat *A) 2680 { 2681 PetscErrorCode ierr; 2682 2683 PetscFunctionBegin; 2684 ierr = MatCreate(comm,m,n,m,n,A);CHKERRQ(ierr); 2685 ierr = MatSetType(*A,MATSEQBAIJ);CHKERRQ(ierr); 2686 ierr = MatSeqBAIJSetPreallocation(*A,bs,nz,nnz);CHKERRQ(ierr); 2687 PetscFunctionReturn(0); 2688 } 2689 2690 #undef __FUNCT__ 2691 #define __FUNCT__ "MatSeqBAIJSetPreallocation" 2692 /*@C 2693 MatSeqBAIJSetPreallocation - Sets the block size and expected nonzeros 2694 per row in the matrix. For good matrix assembly performance the 2695 user should preallocate the matrix storage by setting the parameter nz 2696 (or the array nnz). By setting these parameters accurately, performance 2697 during matrix assembly can be increased by more than a factor of 50. 2698 2699 Collective on MPI_Comm 2700 2701 Input Parameters: 2702 + A - the matrix 2703 . bs - size of block 2704 . nz - number of block nonzeros per block row (same for all rows) 2705 - nnz - array containing the number of block nonzeros in the various block rows 2706 (possibly different for each block row) or PETSC_NULL 2707 2708 Options Database Keys: 2709 . -mat_no_unroll - uses code that does not unroll the loops in the 2710 block calculations (much slower) 2711 . -mat_block_size - size of the blocks to use 2712 2713 Level: intermediate 2714 2715 Notes: 2716 If the nnz parameter is given then the nz parameter is ignored 2717 2718 The block AIJ format is fully compatible with standard Fortran 77 2719 storage. That is, the stored row and column indices can begin at 2720 either one (as in Fortran) or zero. See the users' manual for details. 2721 2722 Specify the preallocated storage with either nz or nnz (not both). 2723 Set nz=PETSC_DEFAULT and nnz=PETSC_NULL for PETSc to control dynamic memory 2724 allocation. For additional details, see the users manual chapter on 2725 matrices. 2726 2727 .seealso: MatCreate(), MatCreateSeqAIJ(), MatSetValues(), MatCreateMPIBAIJ() 2728 @*/ 2729 PetscErrorCode MatSeqBAIJSetPreallocation(Mat B,PetscInt bs,PetscInt nz,const PetscInt nnz[]) 2730 { 2731 PetscErrorCode ierr,(*f)(Mat,PetscInt,PetscInt,const PetscInt[]); 2732 2733 PetscFunctionBegin; 2734 ierr = PetscObjectQueryFunction((PetscObject)B,"MatSeqBAIJSetPreallocation_C",(void (**)(void))&f);CHKERRQ(ierr); 2735 if (f) { 2736 ierr = (*f)(B,bs,nz,nnz);CHKERRQ(ierr); 2737 } 2738 PetscFunctionReturn(0); 2739 } 2740 2741