1 #ifdef PETSC_RCS_HEADER 2 static char vcid[] = "$Id: baijfact.c,v 1.54 1997/12/05 19:10:22 balay Exp bsmith $"; 3 #endif 4 /* 5 Factorization code for BAIJ format. 6 */ 7 8 #include "src/mat/impls/baij/seq/baij.h" 9 #include "src/vec/vecimpl.h" 10 #include "src/inline/ilu.h" 11 12 13 /* 14 The symbolic factorization code is identical to that for AIJ format, 15 except for very small changes since this is now a SeqBAIJ datastructure. 16 NOT good code reuse. 17 */ 18 #undef __FUNC__ 19 #define __FUNC__ "MatLUFactorSymbolic_SeqBAIJ" 20 int MatLUFactorSymbolic_SeqBAIJ(Mat A,IS isrow,IS iscol,double f,Mat *B) 21 { 22 Mat_SeqBAIJ *a = (Mat_SeqBAIJ *) A->data, *b; 23 IS isicol; 24 int *r,*ic, ierr, i, n = a->mbs, *ai = a->i, *aj = a->j; 25 int *ainew,*ajnew, jmax,*fill, *ajtmp, nz, bs = a->bs, bs2=a->bs2; 26 int *idnew, idx, row,m,fm, nnz, nzi,realloc = 0,nzbd,*im; 27 28 PetscFunctionBegin; 29 PetscValidHeaderSpecific(isrow,IS_COOKIE); 30 PetscValidHeaderSpecific(iscol,IS_COOKIE); 31 ierr = ISInvertPermutation(iscol,&isicol); CHKERRQ(ierr); 32 ISGetIndices(isrow,&r); ISGetIndices(isicol,&ic); 33 34 /* get new row pointers */ 35 ainew = (int *) PetscMalloc( (n+1)*sizeof(int) ); CHKPTRQ(ainew); 36 ainew[0] = 0; 37 /* don't know how many column pointers are needed so estimate */ 38 jmax = (int) (f*ai[n] + 1); 39 ajnew = (int *) PetscMalloc( (jmax)*sizeof(int) ); CHKPTRQ(ajnew); 40 /* fill is a linked list of nonzeros in active row */ 41 fill = (int *) PetscMalloc( (2*n+1)*sizeof(int)); CHKPTRQ(fill); 42 im = fill + n + 1; 43 /* idnew is location of diagonal in factor */ 44 idnew = (int *) PetscMalloc( (n+1)*sizeof(int)); CHKPTRQ(idnew); 45 idnew[0] = 0; 46 47 for ( i=0; i<n; i++ ) { 48 /* first copy previous fill into linked list */ 49 nnz = nz = ai[r[i]+1] - ai[r[i]]; 50 if (!nz) SETERRQ(PETSC_ERR_MAT_LU_ZRPVT,1,"Empty row in matrix"); 51 ajtmp = aj + ai[r[i]]; 52 fill[n] = n; 53 while (nz--) { 54 fm = n; 55 idx = ic[*ajtmp++]; 56 do { 57 m = fm; 58 fm = fill[m]; 59 } while (fm < idx); 60 fill[m] = idx; 61 fill[idx] = fm; 62 } 63 row = fill[n]; 64 while ( row < i ) { 65 ajtmp = ajnew + idnew[row] + 1; 66 nzbd = 1 + idnew[row] - ainew[row]; 67 nz = im[row] - nzbd; 68 fm = row; 69 while (nz-- > 0) { 70 idx = *ajtmp++; 71 nzbd++; 72 if (idx == i) im[row] = nzbd; 73 do { 74 m = fm; 75 fm = fill[m]; 76 } while (fm < idx); 77 if (fm != idx) { 78 fill[m] = idx; 79 fill[idx] = fm; 80 fm = idx; 81 nnz++; 82 } 83 } 84 row = fill[row]; 85 } 86 /* copy new filled row into permanent storage */ 87 ainew[i+1] = ainew[i] + nnz; 88 if (ainew[i+1] > jmax) { 89 /* allocate a longer ajnew */ 90 int maxadd; 91 maxadd = (int) ((f*(ai[n]+1)*(n-i+5))/n); 92 if (maxadd < nnz) maxadd = (n-i)*(nnz+1); 93 jmax += maxadd; 94 ajtmp = (int *) PetscMalloc( jmax*sizeof(int) );CHKPTRQ(ajtmp); 95 PetscMemcpy(ajtmp,ajnew,ainew[i]*sizeof(int)); 96 PetscFree(ajnew); 97 ajnew = ajtmp; 98 realloc++; /* count how many times we realloc */ 99 } 100 ajtmp = ajnew + ainew[i]; 101 fm = fill[n]; 102 nzi = 0; 103 im[i] = nnz; 104 while (nnz--) { 105 if (fm < i) nzi++; 106 *ajtmp++ = fm; 107 fm = fill[fm]; 108 } 109 idnew[i] = ainew[i] + nzi; 110 } 111 112 if (ai[n] != 0) { 113 double af = ((double)ainew[n])/((double)ai[n]); 114 PLogInfo(A,"MatLUFactorSymbolic_SeqBAIJ:Reallocs %d Fill ratio:given %g needed %g\n", 115 realloc,f,af); 116 PLogInfo(A,"MatLUFactorSymbolic_SeqBAIJ:Run with -pc_lu_fill %g or use \n",af); 117 PLogInfo(A,"MatLUFactorSymbolic_SeqBAIJ:PCLUSetFill(pc,%g);\n",af); 118 PLogInfo(A,"MatLUFactorSymbolic_SeqBAIJ:for best performance.\n"); 119 } else { 120 PLogInfo(A,"MatLUFactorSymbolic_SeqBAIJ:Empty matrix.\n"); 121 } 122 123 ierr = ISRestoreIndices(isrow,&r); CHKERRQ(ierr); 124 ierr = ISRestoreIndices(isicol,&ic); CHKERRQ(ierr); 125 126 PetscFree(fill); 127 128 /* put together the new matrix */ 129 ierr = MatCreateSeqBAIJ(A->comm,bs,bs*n,bs*n,0,PETSC_NULL,B); CHKERRQ(ierr); 130 PLogObjectParent(*B,isicol); 131 ierr = ISDestroy(isicol); CHKERRQ(ierr); 132 b = (Mat_SeqBAIJ *) (*B)->data; 133 PetscFree(b->imax); 134 b->singlemalloc = 0; 135 /* the next line frees the default space generated by the Create() */ 136 PetscFree(b->a); PetscFree(b->ilen); 137 b->a = (Scalar *) PetscMalloc((ainew[n]+1)*sizeof(Scalar)*bs2);CHKPTRQ(b->a); 138 b->j = ajnew; 139 b->i = ainew; 140 b->diag = idnew; 141 b->ilen = 0; 142 b->imax = 0; 143 b->row = isrow; 144 b->col = iscol; 145 b->solve_work = (Scalar *) PetscMalloc( (bs*n+bs)*sizeof(Scalar));CHKPTRQ(b->solve_work); 146 /* In b structure: Free imax, ilen, old a, old j. 147 Allocate idnew, solve_work, new a, new j */ 148 PLogObjectMemory(*B,(ainew[n]-n)*(sizeof(int)+sizeof(Scalar))); 149 b->maxnz = b->nz = ainew[n]; 150 151 (*B)->info.factor_mallocs = realloc; 152 (*B)->info.fill_ratio_given = f; 153 if (ai[i] != 0) { 154 (*B)->info.fill_ratio_needed = ((double)ainew[n])/((double)ai[i]); 155 } else { 156 (*B)->info.fill_ratio_needed = 0.0; 157 } 158 159 160 PetscFunctionReturn(0); 161 } 162 163 /* ----------------------------------------------------------- */ 164 #undef __FUNC__ 165 #define __FUNC__ "MatLUFactorNumeric_SeqBAIJ_N" 166 int MatLUFactorNumeric_SeqBAIJ_N(Mat A,Mat *B) 167 { 168 Mat C = *B; 169 Mat_SeqBAIJ *a = (Mat_SeqBAIJ *) A->data,*b = (Mat_SeqBAIJ *)C->data; 170 IS iscol = b->col, isrow = b->row, isicol; 171 int *r,*ic, ierr, i, j, n = a->mbs, *bi = b->i, *bj = b->j; 172 int *ajtmpold, *ajtmp, nz, row, bslog,*ai=a->i,*aj=a->j,k,flg; 173 int *diag_offset=b->diag,diag,bs=a->bs,bs2 = a->bs2,*v_pivots; 174 register int *pj; 175 register Scalar *pv,*v,*rtmp,*multiplier,*v_work,*pc,*w; 176 Scalar *ba = b->a,*aa = a->a; 177 178 PetscFunctionBegin; 179 ierr = ISInvertPermutation(iscol,&isicol); CHKERRQ(ierr); 180 PLogObjectParent(*B,isicol); 181 ierr = ISGetIndices(isrow,&r); CHKERRQ(ierr); 182 ierr = ISGetIndices(isicol,&ic); CHKERRQ(ierr); 183 rtmp = (Scalar *) PetscMalloc(bs2*(n+1)*sizeof(Scalar));CHKPTRQ(rtmp); 184 PetscMemzero(rtmp,bs2*(n+1)*sizeof(Scalar)); 185 /* generate work space needed by dense LU factorization */ 186 v_work = (Scalar *) PetscMalloc(bs*sizeof(int) + (bs+bs2)*sizeof(Scalar)); 187 CHKPTRQ(v_work); 188 multiplier = v_work + bs; 189 v_pivots = (int *) (multiplier + bs2); 190 191 /* flops in while loop */ 192 bslog = 2*bs*bs2; 193 194 for ( i=0; i<n; i++ ) { 195 nz = bi[i+1] - bi[i]; 196 ajtmp = bj + bi[i]; 197 for ( j=0; j<nz; j++ ) { 198 PetscMemzero(rtmp+bs2*ajtmp[j],bs2*sizeof(Scalar)); 199 } 200 /* load in initial (unfactored row) */ 201 nz = ai[r[i]+1] - ai[r[i]]; 202 ajtmpold = aj + ai[r[i]]; 203 v = aa + bs2*ai[r[i]]; 204 for ( j=0; j<nz; j++ ) { 205 PetscMemcpy(rtmp+bs2*ic[ajtmpold[j]],v+bs2*j,bs2*sizeof(Scalar)); 206 } 207 row = *ajtmp++; 208 while (row < i) { 209 pc = rtmp + bs2*row; 210 /* if (*pc) { */ 211 for ( flg=0,k=0; k<bs2; k++ ) { if (pc[k]!=0.0) { flg =1; break; }} 212 if (flg) { 213 pv = ba + bs2*diag_offset[row]; 214 pj = bj + diag_offset[row] + 1; 215 Kernel_A_gets_A_times_B(bs,pc,pv,multiplier); 216 nz = bi[row+1] - diag_offset[row] - 1; 217 pv += bs2; 218 for (j=0; j<nz; j++) { 219 Kernel_A_gets_A_minus_B_times_C(bs,rtmp+bs2*pj[j],pc,pv+bs2*j); 220 } 221 PLogFlops(bslog*(nz+1)-bs); 222 } 223 row = *ajtmp++; 224 } 225 /* finished row so stick it into b->a */ 226 pv = ba + bs2*bi[i]; 227 pj = bj + bi[i]; 228 nz = bi[i+1] - bi[i]; 229 for ( j=0; j<nz; j++ ) { 230 PetscMemcpy(pv+bs2*j,rtmp+bs2*pj[j],bs2*sizeof(Scalar)); 231 } 232 diag = diag_offset[i] - bi[i]; 233 /* invert diagonal block */ 234 w = pv + bs2*diag; 235 Kernel_A_gets_inverse_A(bs,w,v_pivots,v_work); 236 } 237 238 PetscFree(rtmp); PetscFree(v_work); 239 ierr = ISRestoreIndices(isicol,&ic); CHKERRQ(ierr); 240 ierr = ISRestoreIndices(isrow,&r); CHKERRQ(ierr); 241 ierr = ISDestroy(isicol); CHKERRQ(ierr); 242 C->factor = FACTOR_LU; 243 C->assembled = PETSC_TRUE; 244 PLogFlops(1.3333*bs*bs2*b->mbs); /* from inverting diagonal blocks */ 245 PetscFunctionReturn(0); 246 } 247 /* ------------------------------------------------------------*/ 248 /* 249 Version for when blocks are 5 by 5 250 */ 251 #undef __FUNC__ 252 #define __FUNC__ "MatLUFactorNumeric_SeqBAIJ_5" 253 int MatLUFactorNumeric_SeqBAIJ_5(Mat A,Mat *B) 254 { 255 Mat C = *B; 256 Mat_SeqBAIJ *a = (Mat_SeqBAIJ *) A->data,*b = (Mat_SeqBAIJ *)C->data; 257 IS iscol = b->col, isrow = b->row, isicol; 258 int *r,*ic, ierr, i, j, n = a->mbs, *bi = b->i, *bj = b->j; 259 int *ajtmpold, *ajtmp, nz, row; 260 int *diag_offset = b->diag,idx,*ai=a->i,*aj=a->j; 261 register int *pj; 262 register Scalar *pv,*v,*rtmp,*pc,*w,*x; 263 Scalar p1,p2,p3,p4,m1,m2,m3,m4,m5,m6,m7,m8,m9,x1,x2,x3,x4; 264 Scalar p5,p6,p7,p8,p9,x5,x6,x7,x8,x9,x10,x11,x12,x13,x14,x15,x16; 265 Scalar x17,x18,x19,x20,x21,x22,x23,x24,x25,p10,p11,p12,p13,p14; 266 Scalar p15,p16,p17,p18,p19,p20,p21,p22,p23,p24,p25,m10,m11,m12; 267 Scalar m13,m14,m15,m16,m17,m18,m19,m20,m21,m22,m23,m24,m25; 268 Scalar *ba = b->a,*aa = a->a; 269 270 PetscFunctionBegin; 271 ierr = ISInvertPermutation(iscol,&isicol); CHKERRQ(ierr); 272 PLogObjectParent(*B,isicol); 273 ierr = ISGetIndices(isrow,&r); CHKERRQ(ierr); 274 ierr = ISGetIndices(isicol,&ic); CHKERRQ(ierr); 275 rtmp = (Scalar *) PetscMalloc(25*(n+1)*sizeof(Scalar));CHKPTRQ(rtmp); 276 277 for ( i=0; i<n; i++ ) { 278 nz = bi[i+1] - bi[i]; 279 ajtmp = bj + bi[i]; 280 for ( j=0; j<nz; j++ ) { 281 x = rtmp+25*ajtmp[j]; 282 x[0] = x[1] = x[2] = x[3] = x[4] = x[5] = x[6] = x[7] = x[8] = x[9] = 0.0; 283 x[10] = x[11] = x[12] = x[13] = x[14] = x[15] = x[16] = x[17] = 0.0; 284 x[18] = x[19] = x[20] = x[21] = x[22] = x[23] = x[24] = 0.0; 285 } 286 /* load in initial (unfactored row) */ 287 idx = r[i]; 288 nz = ai[idx+1] - ai[idx]; 289 ajtmpold = aj + ai[idx]; 290 v = aa + 25*ai[idx]; 291 for ( j=0; j<nz; j++ ) { 292 x = rtmp+25*ic[ajtmpold[j]]; 293 x[0] = v[0]; x[1] = v[1]; x[2] = v[2]; x[3] = v[3]; 294 x[4] = v[4]; x[5] = v[5]; x[6] = v[6]; x[7] = v[7]; x[8] = v[8]; 295 x[9] = v[9]; x[10] = v[10]; x[11] = v[11]; x[12] = v[12]; x[13] = v[13]; 296 x[14] = v[14]; x[15] = v[15]; x[16] = v[16]; x[17] = v[17]; 297 x[18] = v[18]; x[19] = v[19]; x[20] = v[20]; x[21] = v[21]; 298 x[22] = v[22]; x[23] = v[23]; x[24] = v[24]; 299 v += 25; 300 } 301 row = *ajtmp++; 302 while (row < i) { 303 pc = rtmp + 25*row; 304 p1 = pc[0]; p2 = pc[1]; p3 = pc[2]; p4 = pc[3]; 305 p5 = pc[4]; p6 = pc[5]; p7 = pc[6]; p8 = pc[7]; p9 = pc[8]; 306 p10 = pc[9]; p11 = pc[10]; p12 = pc[11]; p13 = pc[12]; p14 = pc[13]; 307 p15 = pc[14]; p16 = pc[15]; p17 = pc[16]; p18 = pc[17]; p19 = pc[18]; 308 p20 = pc[19]; p21 = pc[20]; p22 = pc[21]; p23 = pc[22]; p24 = pc[23]; 309 p25 = pc[24]; 310 if (p1 != 0.0 || p2 != 0.0 || p3 != 0.0 || p4 != 0.0 || p5 != 0.0 || 311 p6 != 0.0 || p7 != 0.0 || p8 != 0.0 || p9 != 0.0 || p10 != 0.0 || 312 p11 != 0.0 || p12 != 0.0 || p13 != 0.0 || p14 != 0.0 || p15 != 0.0 313 || p16 != 0.0 || p17 != 0.0 || p18 != 0.0 || p19 != 0.0 || 314 p20 != 0.0 || p21 != 0.0 || p22 != 0.0 || p23 != 0.0 || 315 p24 != 0.0 || p25 != 0.0) { 316 pv = ba + 25*diag_offset[row]; 317 pj = bj + diag_offset[row] + 1; 318 x1 = pv[0]; x2 = pv[1]; x3 = pv[2]; x4 = pv[3]; 319 x5 = pv[4]; x6 = pv[5]; x7 = pv[6]; x8 = pv[7]; x9 = pv[8]; 320 x10 = pv[9]; x11 = pv[10]; x12 = pv[11]; x13 = pv[12]; x14 = pv[13]; 321 x15 = pv[14]; x16 = pv[15]; x17 = pv[16]; x18 = pv[17]; 322 x19 = pv[18]; x20 = pv[19]; x21 = pv[20]; x22 = pv[21]; 323 x23 = pv[22]; x24 = pv[23]; x25 = pv[24]; 324 pc[0] = m1 = p1*x1 + p6*x2 + p11*x3 + p16*x4 + p21*x5; 325 pc[1] = m2 = p2*x1 + p7*x2 + p12*x3 + p17*x4 + p22*x5; 326 pc[2] = m3 = p3*x1 + p8*x2 + p13*x3 + p18*x4 + p23*x5; 327 pc[3] = m4 = p4*x1 + p9*x2 + p14*x3 + p19*x4 + p24*x5; 328 pc[4] = m5 = p5*x1 + p10*x2 + p15*x3 + p20*x4 + p25*x5; 329 330 pc[5] = m6 = p1*x6 + p6*x7 + p11*x8 + p16*x9 + p21*x10; 331 pc[6] = m7 = p2*x6 + p7*x7 + p12*x8 + p17*x9 + p22*x10; 332 pc[7] = m8 = p3*x6 + p8*x7 + p13*x8 + p18*x9 + p23*x10; 333 pc[8] = m9 = p4*x6 + p9*x7 + p14*x8 + p19*x9 + p24*x10; 334 pc[9] = m10 = p5*x6 + p10*x7 + p15*x8 + p20*x9 + p25*x10; 335 336 pc[10] = m11 = p1*x11 + p6*x12 + p11*x13 + p16*x14 + p21*x15; 337 pc[11] = m12 = p2*x11 + p7*x12 + p12*x13 + p17*x14 + p22*x15; 338 pc[12] = m13 = p3*x11 + p8*x12 + p13*x13 + p18*x14 + p23*x15; 339 pc[13] = m14 = p4*x11 + p9*x12 + p14*x13 + p19*x14 + p24*x15; 340 pc[14] = m15 = p5*x11 + p10*x12 + p15*x13 + p20*x14 + p25*x15; 341 342 pc[15] = m16 = p1*x16 + p6*x17 + p11*x18 + p16*x19 + p21*x20; 343 pc[16] = m17 = p2*x16 + p7*x17 + p12*x18 + p17*x19 + p22*x20; 344 pc[17] = m18 = p3*x16 + p8*x17 + p13*x18 + p18*x19 + p23*x20; 345 pc[18] = m19 = p4*x16 + p9*x17 + p14*x18 + p19*x19 + p24*x20; 346 pc[19] = m20 = p5*x16 + p10*x17 + p15*x18 + p20*x19 + p25*x20; 347 348 pc[20] = m21 = p1*x21 + p6*x22 + p11*x23 + p16*x24 + p21*x25; 349 pc[21] = m22 = p2*x21 + p7*x22 + p12*x23 + p17*x24 + p22*x25; 350 pc[22] = m23 = p3*x21 + p8*x22 + p13*x23 + p18*x24 + p23*x25; 351 pc[23] = m24 = p4*x21 + p9*x22 + p14*x23 + p19*x24 + p24*x25; 352 pc[24] = m25 = p5*x21 + p10*x22 + p15*x23 + p20*x24 + p25*x25; 353 354 nz = bi[row+1] - diag_offset[row] - 1; 355 pv += 25; 356 for (j=0; j<nz; j++) { 357 x1 = pv[0]; x2 = pv[1]; x3 = pv[2]; x4 = pv[3]; 358 x5 = pv[4]; x6 = pv[5]; x7 = pv[6]; x8 = pv[7]; x9 = pv[8]; 359 x10 = pv[9]; x11 = pv[10]; x12 = pv[11]; x13 = pv[12]; 360 x14 = pv[13]; x15 = pv[14]; x16 = pv[15]; x17 = pv[16]; 361 x18 = pv[17]; x19 = pv[18]; x20 = pv[19]; x21 = pv[20]; 362 x22 = pv[21]; x23 = pv[22]; x24 = pv[23]; x25 = pv[24]; 363 x = rtmp + 25*pj[j]; 364 x[0] -= m1*x1 + m6*x2 + m11*x3 + m16*x4 + m21*x5; 365 x[1] -= m2*x1 + m7*x2 + m12*x3 + m17*x4 + m22*x5; 366 x[2] -= m3*x1 + m8*x2 + m13*x3 + m18*x4 + m23*x5; 367 x[3] -= m4*x1 + m9*x2 + m14*x3 + m19*x4 + m24*x5; 368 x[4] -= m5*x1 + m10*x2 + m15*x3 + m20*x4 + m25*x5; 369 370 x[5] -= m1*x6 + m6*x7 + m11*x8 + m16*x9 + m21*x10; 371 x[6] -= m2*x6 + m7*x7 + m12*x8 + m17*x9 + m22*x10; 372 x[7] -= m3*x6 + m8*x7 + m13*x8 + m18*x9 + m23*x10; 373 x[8] -= m4*x6 + m9*x7 + m14*x8 + m19*x9 + m24*x10; 374 x[9] -= m5*x6 + m10*x7 + m15*x8 + m20*x9 + m25*x10; 375 376 x[10] -= m1*x11 + m6*x12 + m11*x13 + m16*x14 + m21*x15; 377 x[11] -= m2*x11 + m7*x12 + m12*x13 + m17*x14 + m22*x15; 378 x[12] -= m3*x11 + m8*x12 + m13*x13 + m18*x14 + m23*x15; 379 x[13] -= m4*x11 + m9*x12 + m14*x13 + m19*x14 + m24*x15; 380 x[14] -= m5*x11 + m10*x12 + m15*x13 + m20*x14 + m25*x15; 381 382 x[15] -= m1*x16 + m6*x17 + m11*x18 + m16*x19 + m21*x20; 383 x[16] -= m2*x16 + m7*x17 + m12*x18 + m17*x19 + m22*x20; 384 x[17] -= m3*x16 + m8*x17 + m13*x18 + m18*x19 + m23*x20; 385 x[18] -= m4*x16 + m9*x17 + m14*x18 + m19*x19 + m24*x20; 386 x[19] -= m5*x16 + m10*x17 + m15*x18 + m20*x19 + m25*x20; 387 388 x[20] -= m1*x21 + m6*x22 + m11*x23 + m16*x24 + m21*x25; 389 x[21] -= m2*x21 + m7*x22 + m12*x23 + m17*x24 + m22*x25; 390 x[22] -= m3*x21 + m8*x22 + m13*x23 + m18*x24 + m23*x25; 391 x[23] -= m4*x21 + m9*x22 + m14*x23 + m19*x24 + m24*x25; 392 x[24] -= m5*x21 + m10*x22 + m15*x23 + m20*x24 + m25*x25; 393 394 pv += 25; 395 } 396 PLogFlops(250*nz+225); 397 } 398 row = *ajtmp++; 399 } 400 /* finished row so stick it into b->a */ 401 pv = ba + 25*bi[i]; 402 pj = bj + bi[i]; 403 nz = bi[i+1] - bi[i]; 404 for ( j=0; j<nz; j++ ) { 405 x = rtmp+25*pj[j]; 406 pv[0] = x[0]; pv[1] = x[1]; pv[2] = x[2]; pv[3] = x[3]; 407 pv[4] = x[4]; pv[5] = x[5]; pv[6] = x[6]; pv[7] = x[7]; pv[8] = x[8]; 408 pv[9] = x[9]; pv[10] = x[10]; pv[11] = x[11]; pv[12] = x[12]; 409 pv[13] = x[13]; pv[14] = x[14]; pv[15] = x[15]; pv[16] = x[16]; 410 pv[17] = x[17]; pv[18] = x[18]; pv[19] = x[19]; pv[20] = x[20]; 411 pv[21] = x[21]; pv[22] = x[22]; pv[23] = x[23]; pv[24] = x[24]; 412 pv += 25; 413 } 414 /* invert diagonal block */ 415 w = ba + 25*diag_offset[i]; 416 ierr = Kernel_A_gets_inverse_A_5(w); CHKERRQ(ierr); 417 } 418 419 PetscFree(rtmp); 420 ierr = ISRestoreIndices(isicol,&ic); CHKERRQ(ierr); 421 ierr = ISRestoreIndices(isrow,&r); CHKERRQ(ierr); 422 ierr = ISDestroy(isicol); CHKERRQ(ierr); 423 C->factor = FACTOR_LU; 424 C->assembled = PETSC_TRUE; 425 PLogFlops(1.3333*125*b->mbs); /* from inverting diagonal blocks */ 426 PetscFunctionReturn(0); 427 } 428 429 /* ------------------------------------------------------------*/ 430 /* 431 Version for when blocks are 4 by 4 432 */ 433 #undef __FUNC__ 434 #define __FUNC__ "MatLUFactorNumeric_SeqBAIJ_4" 435 int MatLUFactorNumeric_SeqBAIJ_4(Mat A,Mat *B) 436 { 437 Mat C = *B; 438 Mat_SeqBAIJ *a = (Mat_SeqBAIJ *) A->data,*b = (Mat_SeqBAIJ *)C->data; 439 IS iscol = b->col, isrow = b->row, isicol; 440 int *r,*ic, ierr, i, j, n = a->mbs, *bi = b->i, *bj = b->j; 441 int *ajtmpold, *ajtmp, nz, row; 442 int *diag_offset = b->diag,idx,*ai=a->i,*aj=a->j; 443 register int *pj; 444 register Scalar *pv,*v,*rtmp,*pc,*w,*x; 445 Scalar p1,p2,p3,p4,m1,m2,m3,m4,m5,m6,m7,m8,m9,x1,x2,x3,x4; 446 Scalar p5,p6,p7,p8,p9,x5,x6,x7,x8,x9,x10,x11,x12,x13,x14,x15,x16; 447 Scalar p10,p11,p12,p13,p14,p15,p16,m10,m11,m12; 448 Scalar m13,m14,m15,m16; 449 Scalar *ba = b->a,*aa = a->a; 450 451 PetscFunctionBegin; 452 ierr = ISInvertPermutation(iscol,&isicol); CHKERRQ(ierr); 453 PLogObjectParent(*B,isicol); 454 ierr = ISGetIndices(isrow,&r); CHKERRQ(ierr); 455 ierr = ISGetIndices(isicol,&ic); CHKERRQ(ierr); 456 rtmp = (Scalar *) PetscMalloc(16*(n+1)*sizeof(Scalar));CHKPTRQ(rtmp); 457 458 for ( i=0; i<n; i++ ) { 459 nz = bi[i+1] - bi[i]; 460 ajtmp = bj + bi[i]; 461 for ( j=0; j<nz; j++ ) { 462 x = rtmp+16*ajtmp[j]; 463 x[0] = x[1] = x[2] = x[3] = x[4] = x[5] = x[6] = x[7] = x[8] = x[9] = 0.0; 464 x[10] = x[11] = x[12] = x[13] = x[14] = x[15] = 0.0; 465 } 466 /* load in initial (unfactored row) */ 467 idx = r[i]; 468 nz = ai[idx+1] - ai[idx]; 469 ajtmpold = aj + ai[idx]; 470 v = aa + 16*ai[idx]; 471 for ( j=0; j<nz; j++ ) { 472 x = rtmp+16*ic[ajtmpold[j]]; 473 x[0] = v[0]; x[1] = v[1]; x[2] = v[2]; x[3] = v[3]; 474 x[4] = v[4]; x[5] = v[5]; x[6] = v[6]; x[7] = v[7]; x[8] = v[8]; 475 x[9] = v[9]; x[10] = v[10]; x[11] = v[11]; x[12] = v[12]; x[13] = v[13]; 476 x[14] = v[14]; x[15] = v[15]; 477 v += 16; 478 } 479 row = *ajtmp++; 480 while (row < i) { 481 pc = rtmp + 16*row; 482 p1 = pc[0]; p2 = pc[1]; p3 = pc[2]; p4 = pc[3]; 483 p5 = pc[4]; p6 = pc[5]; p7 = pc[6]; p8 = pc[7]; p9 = pc[8]; 484 p10 = pc[9]; p11 = pc[10]; p12 = pc[11]; p13 = pc[12]; p14 = pc[13]; 485 p15 = pc[14]; p16 = pc[15]; 486 if (p1 != 0.0 || p2 != 0.0 || p3 != 0.0 || p4 != 0.0 || p5 != 0.0 || 487 p6 != 0.0 || p7 != 0.0 || p8 != 0.0 || p9 != 0.0 || p10 != 0.0 || 488 p11 != 0.0 || p12 != 0.0 || p13 != 0.0 || p14 != 0.0 || p15 != 0.0 489 || p16 != 0.0) { 490 pv = ba + 16*diag_offset[row]; 491 pj = bj + diag_offset[row] + 1; 492 x1 = pv[0]; x2 = pv[1]; x3 = pv[2]; x4 = pv[3]; 493 x5 = pv[4]; x6 = pv[5]; x7 = pv[6]; x8 = pv[7]; x9 = pv[8]; 494 x10 = pv[9]; x11 = pv[10]; x12 = pv[11]; x13 = pv[12]; x14 = pv[13]; 495 x15 = pv[14]; x16 = pv[15]; 496 pc[0] = m1 = p1*x1 + p5*x2 + p9*x3 + p13*x4; 497 pc[1] = m2 = p2*x1 + p6*x2 + p10*x3 + p14*x4; 498 pc[2] = m3 = p3*x1 + p7*x2 + p11*x3 + p15*x4; 499 pc[3] = m4 = p4*x1 + p8*x2 + p12*x3 + p16*x4; 500 501 pc[4] = m5 = p1*x5 + p5*x6 + p9*x7 + p13*x8; 502 pc[5] = m6 = p2*x5 + p6*x6 + p10*x7 + p14*x8; 503 pc[6] = m7 = p3*x5 + p7*x6 + p11*x7 + p15*x8; 504 pc[7] = m8 = p4*x5 + p8*x6 + p12*x7 + p16*x8; 505 506 pc[8] = m9 = p1*x9 + p5*x10 + p9*x11 + p13*x12; 507 pc[9] = m10 = p2*x9 + p6*x10 + p10*x11 + p14*x12; 508 pc[10] = m11 = p3*x9 + p7*x10 + p11*x11 + p15*x12; 509 pc[11] = m12 = p4*x9 + p8*x10 + p12*x11 + p16*x12; 510 511 pc[12] = m13 = p1*x13 + p5*x14 + p9*x15 + p13*x16; 512 pc[13] = m14 = p2*x13 + p6*x14 + p10*x15 + p14*x16; 513 pc[14] = m15 = p3*x13 + p7*x14 + p11*x15 + p15*x16; 514 pc[15] = m16 = p4*x13 + p8*x14 + p12*x15 + p16*x16; 515 516 nz = bi[row+1] - diag_offset[row] - 1; 517 pv += 16; 518 for (j=0; j<nz; j++) { 519 x1 = pv[0]; x2 = pv[1]; x3 = pv[2]; x4 = pv[3]; 520 x5 = pv[4]; x6 = pv[5]; x7 = pv[6]; x8 = pv[7]; x9 = pv[8]; 521 x10 = pv[9]; x11 = pv[10]; x12 = pv[11]; x13 = pv[12]; 522 x14 = pv[13]; x15 = pv[14]; x16 = pv[15]; 523 x = rtmp + 16*pj[j]; 524 x[0] -= m1*x1 + m5*x2 + m9*x3 + m13*x4; 525 x[1] -= m2*x1 + m6*x2 + m10*x3 + m14*x4; 526 x[2] -= m3*x1 + m7*x2 + m11*x3 + m15*x4; 527 x[3] -= m4*x1 + m8*x2 + m12*x3 + m16*x4; 528 529 x[4] -= m1*x5 + m5*x6 + m9*x7 + m13*x8; 530 x[5] -= m2*x5 + m6*x6 + m10*x7 + m14*x8; 531 x[6] -= m3*x5 + m7*x6 + m11*x7 + m15*x8; 532 x[7] -= m4*x5 + m8*x6 + m12*x7 + m16*x8; 533 534 x[8] -= m1*x9 + m5*x10 + m9*x11 + m13*x12; 535 x[9] -= m2*x9 + m6*x10 + m10*x11 + m14*x12; 536 x[10] -= m3*x9 + m7*x10 + m11*x11 + m15*x12; 537 x[11] -= m4*x9 + m8*x10 + m12*x11 + m16*x12; 538 539 x[12] -= m1*x13 + m5*x14 + m9*x15 + m13*x16; 540 x[13] -= m2*x13 + m6*x14 + m10*x15 + m14*x16; 541 x[14] -= m3*x13 + m7*x14 + m11*x15 + m15*x16; 542 x[15] -= m4*x13 + m8*x14 + m12*x15 + m16*x16; 543 544 pv += 16; 545 } 546 PLogFlops(128*nz+112); 547 } 548 row = *ajtmp++; 549 } 550 /* finished row so stick it into b->a */ 551 pv = ba + 16*bi[i]; 552 pj = bj + bi[i]; 553 nz = bi[i+1] - bi[i]; 554 for ( j=0; j<nz; j++ ) { 555 x = rtmp+16*pj[j]; 556 pv[0] = x[0]; pv[1] = x[1]; pv[2] = x[2]; pv[3] = x[3]; 557 pv[4] = x[4]; pv[5] = x[5]; pv[6] = x[6]; pv[7] = x[7]; pv[8] = x[8]; 558 pv[9] = x[9]; pv[10] = x[10]; pv[11] = x[11]; pv[12] = x[12]; 559 pv[13] = x[13]; pv[14] = x[14]; pv[15] = x[15]; 560 pv += 16; 561 } 562 /* invert diagonal block */ 563 w = ba + 16*diag_offset[i]; 564 ierr = Kernel_A_gets_inverse_A_4(w); CHKERRQ(ierr); 565 } 566 567 PetscFree(rtmp); 568 ierr = ISRestoreIndices(isicol,&ic); CHKERRQ(ierr); 569 ierr = ISRestoreIndices(isrow,&r); CHKERRQ(ierr); 570 ierr = ISDestroy(isicol); CHKERRQ(ierr); 571 C->factor = FACTOR_LU; 572 C->assembled = PETSC_TRUE; 573 PLogFlops(1.3333*64*b->mbs); /* from inverting diagonal blocks */ 574 PetscFunctionReturn(0); 575 } 576 /* 577 Version for when blocks are 4 by 4 Using natural ordering 578 */ 579 #undef __FUNC__ 580 #define __FUNC__ "MatLUFactorNumeric_SeqBAIJ_4_NaturalOrdering" 581 int MatLUFactorNumeric_SeqBAIJ_4_NaturalOrdering(Mat A,Mat *B) 582 { 583 Mat C = *B; 584 Mat_SeqBAIJ *a = (Mat_SeqBAIJ *) A->data,*b = (Mat_SeqBAIJ *)C->data; 585 int ierr, i, j, n = a->mbs, *bi = b->i, *bj = b->j; 586 int *ajtmpold, *ajtmp, nz, row; 587 int *diag_offset = b->diag,*ai=a->i,*aj=a->j; 588 register int *pj; 589 register Scalar *pv,*v,*rtmp,*pc,*w,*x; 590 Scalar p1,p2,p3,p4,m1,m2,m3,m4,m5,m6,m7,m8,m9,x1,x2,x3,x4; 591 Scalar p5,p6,p7,p8,p9,x5,x6,x7,x8,x9,x10,x11,x12,x13,x14,x15,x16; 592 Scalar p10,p11,p12,p13,p14,p15,p16,m10,m11,m12; 593 Scalar m13,m14,m15,m16; 594 Scalar *ba = b->a,*aa = a->a; 595 596 PetscFunctionBegin; 597 rtmp = (Scalar *) PetscMalloc(16*(n+1)*sizeof(Scalar));CHKPTRQ(rtmp); 598 599 for ( i=0; i<n; i++ ) { 600 nz = bi[i+1] - bi[i]; 601 ajtmp = bj + bi[i]; 602 for ( j=0; j<nz; j++ ) { 603 x = rtmp+16*ajtmp[j]; 604 x[0] = x[1] = x[2] = x[3] = x[4] = x[5] = x[6] = x[7] = x[8] = x[9] = 0.0; 605 x[10] = x[11] = x[12] = x[13] = x[14] = x[15] = 0.0; 606 } 607 /* load in initial (unfactored row) */ 608 nz = ai[i+1] - ai[i]; 609 ajtmpold = aj + ai[i]; 610 v = aa + 16*ai[i]; 611 for ( j=0; j<nz; j++ ) { 612 x = rtmp+16*ajtmpold[j]; 613 x[0] = v[0]; x[1] = v[1]; x[2] = v[2]; x[3] = v[3]; 614 x[4] = v[4]; x[5] = v[5]; x[6] = v[6]; x[7] = v[7]; x[8] = v[8]; 615 x[9] = v[9]; x[10] = v[10]; x[11] = v[11]; x[12] = v[12]; x[13] = v[13]; 616 x[14] = v[14]; x[15] = v[15]; 617 v += 16; 618 } 619 row = *ajtmp++; 620 while (row < i) { 621 pc = rtmp + 16*row; 622 p1 = pc[0]; p2 = pc[1]; p3 = pc[2]; p4 = pc[3]; 623 p5 = pc[4]; p6 = pc[5]; p7 = pc[6]; p8 = pc[7]; p9 = pc[8]; 624 p10 = pc[9]; p11 = pc[10]; p12 = pc[11]; p13 = pc[12]; p14 = pc[13]; 625 p15 = pc[14]; p16 = pc[15]; 626 if (p1 != 0.0 || p2 != 0.0 || p3 != 0.0 || p4 != 0.0 || p5 != 0.0 || 627 p6 != 0.0 || p7 != 0.0 || p8 != 0.0 || p9 != 0.0 || p10 != 0.0 || 628 p11 != 0.0 || p12 != 0.0 || p13 != 0.0 || p14 != 0.0 || p15 != 0.0 629 || p16 != 0.0) { 630 pv = ba + 16*diag_offset[row]; 631 pj = bj + diag_offset[row] + 1; 632 x1 = pv[0]; x2 = pv[1]; x3 = pv[2]; x4 = pv[3]; 633 x5 = pv[4]; x6 = pv[5]; x7 = pv[6]; x8 = pv[7]; x9 = pv[8]; 634 x10 = pv[9]; x11 = pv[10]; x12 = pv[11]; x13 = pv[12]; x14 = pv[13]; 635 x15 = pv[14]; x16 = pv[15]; 636 pc[0] = m1 = p1*x1 + p5*x2 + p9*x3 + p13*x4; 637 pc[1] = m2 = p2*x1 + p6*x2 + p10*x3 + p14*x4; 638 pc[2] = m3 = p3*x1 + p7*x2 + p11*x3 + p15*x4; 639 pc[3] = m4 = p4*x1 + p8*x2 + p12*x3 + p16*x4; 640 641 pc[4] = m5 = p1*x5 + p5*x6 + p9*x7 + p13*x8; 642 pc[5] = m6 = p2*x5 + p6*x6 + p10*x7 + p14*x8; 643 pc[6] = m7 = p3*x5 + p7*x6 + p11*x7 + p15*x8; 644 pc[7] = m8 = p4*x5 + p8*x6 + p12*x7 + p16*x8; 645 646 pc[8] = m9 = p1*x9 + p5*x10 + p9*x11 + p13*x12; 647 pc[9] = m10 = p2*x9 + p6*x10 + p10*x11 + p14*x12; 648 pc[10] = m11 = p3*x9 + p7*x10 + p11*x11 + p15*x12; 649 pc[11] = m12 = p4*x9 + p8*x10 + p12*x11 + p16*x12; 650 651 pc[12] = m13 = p1*x13 + p5*x14 + p9*x15 + p13*x16; 652 pc[13] = m14 = p2*x13 + p6*x14 + p10*x15 + p14*x16; 653 pc[14] = m15 = p3*x13 + p7*x14 + p11*x15 + p15*x16; 654 pc[15] = m16 = p4*x13 + p8*x14 + p12*x15 + p16*x16; 655 656 nz = bi[row+1] - diag_offset[row] - 1; 657 pv += 16; 658 for (j=0; j<nz; j++) { 659 x1 = pv[0]; x2 = pv[1]; x3 = pv[2]; x4 = pv[3]; 660 x5 = pv[4]; x6 = pv[5]; x7 = pv[6]; x8 = pv[7]; x9 = pv[8]; 661 x10 = pv[9]; x11 = pv[10]; x12 = pv[11]; x13 = pv[12]; 662 x14 = pv[13]; x15 = pv[14]; x16 = pv[15]; 663 x = rtmp + 16*pj[j]; 664 x[0] -= m1*x1 + m5*x2 + m9*x3 + m13*x4; 665 x[1] -= m2*x1 + m6*x2 + m10*x3 + m14*x4; 666 x[2] -= m3*x1 + m7*x2 + m11*x3 + m15*x4; 667 x[3] -= m4*x1 + m8*x2 + m12*x3 + m16*x4; 668 669 x[4] -= m1*x5 + m5*x6 + m9*x7 + m13*x8; 670 x[5] -= m2*x5 + m6*x6 + m10*x7 + m14*x8; 671 x[6] -= m3*x5 + m7*x6 + m11*x7 + m15*x8; 672 x[7] -= m4*x5 + m8*x6 + m12*x7 + m16*x8; 673 674 x[8] -= m1*x9 + m5*x10 + m9*x11 + m13*x12; 675 x[9] -= m2*x9 + m6*x10 + m10*x11 + m14*x12; 676 x[10] -= m3*x9 + m7*x10 + m11*x11 + m15*x12; 677 x[11] -= m4*x9 + m8*x10 + m12*x11 + m16*x12; 678 679 x[12] -= m1*x13 + m5*x14 + m9*x15 + m13*x16; 680 x[13] -= m2*x13 + m6*x14 + m10*x15 + m14*x16; 681 x[14] -= m3*x13 + m7*x14 + m11*x15 + m15*x16; 682 x[15] -= m4*x13 + m8*x14 + m12*x15 + m16*x16; 683 684 pv += 16; 685 } 686 PLogFlops(128*nz+112); 687 } 688 row = *ajtmp++; 689 } 690 /* finished row so stick it into b->a */ 691 pv = ba + 16*bi[i]; 692 pj = bj + bi[i]; 693 nz = bi[i+1] - bi[i]; 694 for ( j=0; j<nz; j++ ) { 695 x = rtmp+16*pj[j]; 696 pv[0] = x[0]; pv[1] = x[1]; pv[2] = x[2]; pv[3] = x[3]; 697 pv[4] = x[4]; pv[5] = x[5]; pv[6] = x[6]; pv[7] = x[7]; pv[8] = x[8]; 698 pv[9] = x[9]; pv[10] = x[10]; pv[11] = x[11]; pv[12] = x[12]; 699 pv[13] = x[13]; pv[14] = x[14]; pv[15] = x[15]; 700 pv += 16; 701 } 702 /* invert diagonal block */ 703 w = ba + 16*diag_offset[i]; 704 ierr = Kernel_A_gets_inverse_A_4(w); CHKERRQ(ierr); 705 } 706 707 PetscFree(rtmp); 708 C->factor = FACTOR_LU; 709 C->assembled = PETSC_TRUE; 710 PLogFlops(1.3333*64*b->mbs); /* from inverting diagonal blocks */ 711 PetscFunctionReturn(0); 712 } 713 714 /* ------------------------------------------------------------*/ 715 /* 716 Version for when blocks are 3 by 3 717 */ 718 #undef __FUNC__ 719 #define __FUNC__ "MatLUFactorNumeric_SeqBAIJ_3" 720 int MatLUFactorNumeric_SeqBAIJ_3(Mat A,Mat *B) 721 { 722 Mat C = *B; 723 Mat_SeqBAIJ *a = (Mat_SeqBAIJ *) A->data,*b = (Mat_SeqBAIJ *)C->data; 724 IS iscol = b->col, isrow = b->row, isicol; 725 int *r,*ic, ierr, i, j, n = a->mbs, *bi = b->i, *bj = b->j; 726 int *ajtmpold, *ajtmp, nz, row, *ai=a->i,*aj=a->j; 727 int *diag_offset = b->diag,idx; 728 register int *pj; 729 register Scalar *pv,*v,*rtmp,*pc,*w,*x; 730 Scalar p1,p2,p3,p4,m1,m2,m3,m4,m5,m6,m7,m8,m9,x1,x2,x3,x4; 731 Scalar p5,p6,p7,p8,p9,x5,x6,x7,x8,x9; 732 Scalar *ba = b->a,*aa = a->a; 733 734 PetscFunctionBegin; 735 ierr = ISInvertPermutation(iscol,&isicol); CHKERRQ(ierr); 736 PLogObjectParent(*B,isicol); 737 ierr = ISGetIndices(isrow,&r); CHKERRQ(ierr); 738 ierr = ISGetIndices(isicol,&ic); CHKERRQ(ierr); 739 rtmp = (Scalar *) PetscMalloc(9*(n+1)*sizeof(Scalar));CHKPTRQ(rtmp); 740 741 for ( i=0; i<n; i++ ) { 742 nz = bi[i+1] - bi[i]; 743 ajtmp = bj + bi[i]; 744 for ( j=0; j<nz; j++ ) { 745 x = rtmp + 9*ajtmp[j]; 746 x[0] = x[1] = x[2] = x[3] = x[4] = x[5] = x[6] = x[7] = x[8] = x[9] = 0.0; 747 } 748 /* load in initial (unfactored row) */ 749 idx = r[i]; 750 nz = ai[idx+1] - ai[idx]; 751 ajtmpold = aj + ai[idx]; 752 v = aa + 9*ai[idx]; 753 for ( j=0; j<nz; j++ ) { 754 x = rtmp + 9*ic[ajtmpold[j]]; 755 x[0] = v[0]; x[1] = v[1]; x[2] = v[2]; x[3] = v[3]; 756 x[4] = v[4]; x[5] = v[5]; x[6] = v[6]; x[7] = v[7]; x[8] = v[8]; 757 v += 9; 758 } 759 row = *ajtmp++; 760 while (row < i) { 761 pc = rtmp + 9*row; 762 p1 = pc[0]; p2 = pc[1]; p3 = pc[2]; p4 = pc[3]; 763 p5 = pc[4]; p6 = pc[5]; p7 = pc[6]; p8 = pc[7]; p9 = pc[8]; 764 if (p1 != 0.0 || p2 != 0.0 || p3 != 0.0 || p4 != 0.0 || p5 != 0.0 || 765 p6 != 0.0 || p7 != 0.0 || p8 != 0.0 || p9 != 0.0) { 766 pv = ba + 9*diag_offset[row]; 767 pj = bj + diag_offset[row] + 1; 768 x1 = pv[0]; x2 = pv[1]; x3 = pv[2]; x4 = pv[3]; 769 x5 = pv[4]; x6 = pv[5]; x7 = pv[6]; x8 = pv[7]; x9 = pv[8]; 770 pc[0] = m1 = p1*x1 + p4*x2 + p7*x3; 771 pc[1] = m2 = p2*x1 + p5*x2 + p8*x3; 772 pc[2] = m3 = p3*x1 + p6*x2 + p9*x3; 773 774 pc[3] = m4 = p1*x4 + p4*x5 + p7*x6; 775 pc[4] = m5 = p2*x4 + p5*x5 + p8*x6; 776 pc[5] = m6 = p3*x4 + p6*x5 + p9*x6; 777 778 pc[6] = m7 = p1*x7 + p4*x8 + p7*x9; 779 pc[7] = m8 = p2*x7 + p5*x8 + p8*x9; 780 pc[8] = m9 = p3*x7 + p6*x8 + p9*x9; 781 nz = bi[row+1] - diag_offset[row] - 1; 782 pv += 9; 783 for (j=0; j<nz; j++) { 784 x1 = pv[0]; x2 = pv[1]; x3 = pv[2]; x4 = pv[3]; 785 x5 = pv[4]; x6 = pv[5]; x7 = pv[6]; x8 = pv[7]; x9 = pv[8]; 786 x = rtmp + 9*pj[j]; 787 x[0] -= m1*x1 + m4*x2 + m7*x3; 788 x[1] -= m2*x1 + m5*x2 + m8*x3; 789 x[2] -= m3*x1 + m6*x2 + m9*x3; 790 791 x[3] -= m1*x4 + m4*x5 + m7*x6; 792 x[4] -= m2*x4 + m5*x5 + m8*x6; 793 x[5] -= m3*x4 + m6*x5 + m9*x6; 794 795 x[6] -= m1*x7 + m4*x8 + m7*x9; 796 x[7] -= m2*x7 + m5*x8 + m8*x9; 797 x[8] -= m3*x7 + m6*x8 + m9*x9; 798 pv += 9; 799 } 800 PLogFlops(54*nz+36); 801 } 802 row = *ajtmp++; 803 } 804 /* finished row so stick it into b->a */ 805 pv = ba + 9*bi[i]; 806 pj = bj + bi[i]; 807 nz = bi[i+1] - bi[i]; 808 for ( j=0; j<nz; j++ ) { 809 x = rtmp + 9*pj[j]; 810 pv[0] = x[0]; pv[1] = x[1]; pv[2] = x[2]; pv[3] = x[3]; 811 pv[4] = x[4]; pv[5] = x[5]; pv[6] = x[6]; pv[7] = x[7]; pv[8] = x[8]; 812 pv += 9; 813 } 814 /* invert diagonal block */ 815 w = ba + 9*diag_offset[i]; 816 ierr = Kernel_A_gets_inverse_A_3(w); CHKERRQ(ierr); 817 } 818 819 PetscFree(rtmp); 820 ierr = ISRestoreIndices(isicol,&ic); CHKERRQ(ierr); 821 ierr = ISRestoreIndices(isrow,&r); CHKERRQ(ierr); 822 ierr = ISDestroy(isicol); CHKERRQ(ierr); 823 C->factor = FACTOR_LU; 824 C->assembled = PETSC_TRUE; 825 PLogFlops(1.3333*27*b->mbs); /* from inverting diagonal blocks */ 826 PetscFunctionReturn(0); 827 } 828 829 /* ------------------------------------------------------------*/ 830 /* 831 Version for when blocks are 2 by 2 832 */ 833 #undef __FUNC__ 834 #define __FUNC__ "MatLUFactorNumeric_SeqBAIJ_2" 835 int MatLUFactorNumeric_SeqBAIJ_2(Mat A,Mat *B) 836 { 837 Mat C = *B; 838 Mat_SeqBAIJ *a = (Mat_SeqBAIJ *) A->data,*b = (Mat_SeqBAIJ *)C->data; 839 IS iscol = b->col, isrow = b->row, isicol; 840 int *r,*ic, ierr, i, j, n = a->mbs, *bi = b->i, *bj = b->j; 841 int *ajtmpold, *ajtmp, nz, row, v_pivots[2]; 842 int *diag_offset=b->diag,bs = 2,idx,*ai=a->i,*aj=a->j; 843 register int *pj; 844 register Scalar *pv,*v,*rtmp,m1,m2,m3,m4,*pc,*w,*x,x1,x2,x3,x4; 845 Scalar p1,p2,p3,p4,v_work[2]; 846 Scalar *ba = b->a,*aa = a->a; 847 848 PetscFunctionBegin; 849 ierr = ISInvertPermutation(iscol,&isicol); CHKERRQ(ierr); 850 PLogObjectParent(*B,isicol); 851 ierr = ISGetIndices(isrow,&r); CHKERRQ(ierr); 852 ierr = ISGetIndices(isicol,&ic); CHKERRQ(ierr); 853 rtmp = (Scalar *) PetscMalloc(4*(n+1)*sizeof(Scalar));CHKPTRQ(rtmp); 854 855 for ( i=0; i<n; i++ ) { 856 nz = bi[i+1] - bi[i]; 857 ajtmp = bj + bi[i]; 858 for ( j=0; j<nz; j++ ) { 859 x = rtmp+4*ajtmp[j]; x[0] = x[1] = x[2] = x[3] = 0.0; 860 } 861 /* load in initial (unfactored row) */ 862 idx = r[i]; 863 nz = ai[idx+1] - ai[idx]; 864 ajtmpold = aj + ai[idx]; 865 v = aa + 4*ai[idx]; 866 for ( j=0; j<nz; j++ ) { 867 x = rtmp+4*ic[ajtmpold[j]]; 868 x[0] = v[0]; x[1] = v[1]; x[2] = v[2]; x[3] = v[3]; 869 v += 4; 870 } 871 row = *ajtmp++; 872 while (row < i) { 873 pc = rtmp + 4*row; 874 p1 = pc[0]; p2 = pc[1]; p3 = pc[2]; p4 = pc[3]; 875 if (p1 != 0.0 || p2 != 0.0 || p3 != 0.0 || p4 != 0.0) { 876 pv = ba + 4*diag_offset[row]; 877 pj = bj + diag_offset[row] + 1; 878 x1 = pv[0]; x2 = pv[1]; x3 = pv[2]; x4 = pv[3]; 879 pc[0] = m1 = p1*x1 + p3*x2; 880 pc[1] = m2 = p2*x1 + p4*x2; 881 pc[2] = m3 = p1*x3 + p3*x4; 882 pc[3] = m4 = p2*x3 + p4*x4; 883 nz = bi[row+1] - diag_offset[row] - 1; 884 pv += 4; 885 for (j=0; j<nz; j++) { 886 x1 = pv[0]; x2 = pv[1]; x3 = pv[2]; x4 = pv[3]; 887 x = rtmp + 4*pj[j]; 888 x[0] -= m1*x1 + m3*x2; 889 x[1] -= m2*x1 + m4*x2; 890 x[2] -= m1*x3 + m3*x4; 891 x[3] -= m2*x3 + m4*x4; 892 pv += 4; 893 } 894 PLogFlops(16*nz+12); 895 } 896 row = *ajtmp++; 897 } 898 /* finished row so stick it into b->a */ 899 pv = ba + 4*bi[i]; 900 pj = bj + bi[i]; 901 nz = bi[i+1] - bi[i]; 902 for ( j=0; j<nz; j++ ) { 903 x = rtmp+4*pj[j]; 904 pv[0] = x[0]; pv[1] = x[1]; pv[2] = x[2]; pv[3] = x[3]; 905 pv += 4; 906 } 907 /* invert diagonal block */ 908 w = ba + 4*diag_offset[i]; 909 Kernel_A_gets_inverse_A(bs,w,v_pivots,v_work); 910 } 911 912 PetscFree(rtmp); 913 ierr = ISRestoreIndices(isicol,&ic); CHKERRQ(ierr); 914 ierr = ISRestoreIndices(isrow,&r); CHKERRQ(ierr); 915 ierr = ISDestroy(isicol); CHKERRQ(ierr); 916 C->factor = FACTOR_LU; 917 C->assembled = PETSC_TRUE; 918 PLogFlops(1.3333*8*b->mbs); /* from inverting diagonal blocks */ 919 PetscFunctionReturn(0); 920 } 921 922 /* ----------------------------------------------------------- */ 923 /* 924 Version for when blocks are 1 by 1. 925 */ 926 #undef __FUNC__ 927 #define __FUNC__ "MatLUFactorNumeric_SeqBAIJ_1" 928 int MatLUFactorNumeric_SeqBAIJ_1(Mat A,Mat *B) 929 { 930 Mat C = *B; 931 Mat_SeqBAIJ *a = (Mat_SeqBAIJ *) A->data, *b = (Mat_SeqBAIJ *)C->data; 932 IS iscol = b->col, isrow = b->row, isicol; 933 int *r,*ic, ierr, i, j, n = a->mbs, *bi = b->i, *bj = b->j; 934 int *ajtmpold, *ajtmp, nz, row,*ai = a->i,*aj = a->j; 935 int *diag_offset = b->diag,diag; 936 register int *pj; 937 register Scalar *pv,*v,*rtmp,multiplier,*pc; 938 Scalar *ba = b->a,*aa = a->a; 939 940 PetscFunctionBegin; 941 ierr = ISInvertPermutation(iscol,&isicol); CHKERRQ(ierr); 942 PLogObjectParent(*B,isicol); 943 ierr = ISGetIndices(isrow,&r); CHKERRQ(ierr); 944 ierr = ISGetIndices(isicol,&ic); CHKERRQ(ierr); 945 rtmp = (Scalar *) PetscMalloc((n+1)*sizeof(Scalar));CHKPTRQ(rtmp); 946 947 for ( i=0; i<n; i++ ) { 948 nz = bi[i+1] - bi[i]; 949 ajtmp = bj + bi[i]; 950 for ( j=0; j<nz; j++ ) rtmp[ajtmp[j]] = 0.0; 951 952 /* load in initial (unfactored row) */ 953 nz = ai[r[i]+1] - ai[r[i]]; 954 ajtmpold = aj + ai[r[i]]; 955 v = aa + ai[r[i]]; 956 for ( j=0; j<nz; j++ ) rtmp[ic[ajtmpold[j]]] = v[j]; 957 958 row = *ajtmp++; 959 while (row < i) { 960 pc = rtmp + row; 961 if (*pc != 0.0) { 962 pv = ba + diag_offset[row]; 963 pj = bj + diag_offset[row] + 1; 964 multiplier = *pc * *pv++; 965 *pc = multiplier; 966 nz = bi[row+1] - diag_offset[row] - 1; 967 for (j=0; j<nz; j++) rtmp[pj[j]] -= multiplier * pv[j]; 968 PLogFlops(1+2*nz); 969 } 970 row = *ajtmp++; 971 } 972 /* finished row so stick it into b->a */ 973 pv = ba + bi[i]; 974 pj = bj + bi[i]; 975 nz = bi[i+1] - bi[i]; 976 for ( j=0; j<nz; j++ ) {pv[j] = rtmp[pj[j]];} 977 diag = diag_offset[i] - bi[i]; 978 /* check pivot entry for current row */ 979 if (pv[diag] == 0.0) { 980 SETERRQ(PETSC_ERR_MAT_LU_ZRPVT,0,"Zero pivot"); 981 } 982 pv[diag] = 1.0/pv[diag]; 983 } 984 985 PetscFree(rtmp); 986 ierr = ISRestoreIndices(isicol,&ic); CHKERRQ(ierr); 987 ierr = ISRestoreIndices(isrow,&r); CHKERRQ(ierr); 988 ierr = ISDestroy(isicol); CHKERRQ(ierr); 989 C->factor = FACTOR_LU; 990 C->assembled = PETSC_TRUE; 991 PLogFlops(b->n); 992 PetscFunctionReturn(0); 993 } 994 995 /* ----------------------------------------------------------- */ 996 #undef __FUNC__ 997 #define __FUNC__ "MatLUFactor_SeqBAIJ" 998 int MatLUFactor_SeqBAIJ(Mat A,IS row,IS col,double f) 999 { 1000 Mat_SeqBAIJ *mat = (Mat_SeqBAIJ *) A->data; 1001 int ierr; 1002 Mat C; 1003 1004 PetscFunctionBegin; 1005 ierr = MatLUFactorSymbolic(A,row,col,f,&C); CHKERRQ(ierr); 1006 ierr = MatLUFactorNumeric(A,&C); CHKERRQ(ierr); 1007 1008 /* free all the data structures from mat */ 1009 PetscFree(mat->a); 1010 if (!mat->singlemalloc) {PetscFree(mat->i); PetscFree(mat->j);} 1011 if (mat->diag) PetscFree(mat->diag); 1012 if (mat->ilen) PetscFree(mat->ilen); 1013 if (mat->imax) PetscFree(mat->imax); 1014 if (mat->solve_work) PetscFree(mat->solve_work); 1015 if (mat->mult_work) PetscFree(mat->mult_work); 1016 PetscFree(mat); 1017 1018 PetscMemcpy(A,C,sizeof(struct _p_Mat)); 1019 PetscHeaderDestroy(C); 1020 PetscFunctionReturn(0); 1021 } 1022