Actual source code: mpisbaij.c
1: #define PETSCMAT_DLL
3: #include src/mat/impls/baij/mpi/mpibaij.h
4: #include mpisbaij.h
5: #include src/mat/impls/sbaij/seq/sbaij.h
7: EXTERN PetscErrorCode MatSetUpMultiply_MPISBAIJ(Mat);
8: EXTERN PetscErrorCode MatSetUpMultiply_MPISBAIJ_2comm(Mat);
9: EXTERN PetscErrorCode DisAssemble_MPISBAIJ(Mat);
10: EXTERN PetscErrorCode MatIncreaseOverlap_MPISBAIJ(Mat,PetscInt,IS[],PetscInt);
11: EXTERN PetscErrorCode MatGetValues_SeqSBAIJ(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],PetscScalar []);
12: EXTERN PetscErrorCode MatGetValues_SeqBAIJ(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],PetscScalar []);
13: EXTERN PetscErrorCode MatSetValues_SeqSBAIJ(Mat,PetscInt,const PetscInt [],PetscInt,const PetscInt [],const PetscScalar [],InsertMode);
14: EXTERN PetscErrorCode MatSetValuesBlocked_SeqSBAIJ(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const PetscScalar[],InsertMode);
15: EXTERN PetscErrorCode MatSetValuesBlocked_SeqBAIJ(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const PetscScalar[],InsertMode);
16: EXTERN PetscErrorCode MatGetRow_SeqSBAIJ(Mat,PetscInt,PetscInt*,PetscInt**,PetscScalar**);
17: EXTERN PetscErrorCode MatRestoreRow_SeqSBAIJ(Mat,PetscInt,PetscInt*,PetscInt**,PetscScalar**);
18: EXTERN PetscErrorCode MatZeroRows_SeqSBAIJ(Mat,IS,PetscScalar*);
19: EXTERN PetscErrorCode MatZeroRows_SeqBAIJ(Mat,IS,PetscScalar *);
20: EXTERN PetscErrorCode MatGetRowMaxAbs_MPISBAIJ(Mat,Vec,PetscInt[]);
21: EXTERN PetscErrorCode MatRelax_MPISBAIJ(Mat,Vec,PetscReal,MatSORType,PetscReal,PetscInt,PetscInt,Vec);
23: /* UGLY, ugly, ugly
24: When MatScalar == PetscScalar the function MatSetValuesBlocked_MPIBAIJ_MatScalar() does
25: not exist. Otherwise ..._MatScalar() takes matrix elements in single precision and
26: inserts them into the single precision data structure. The function MatSetValuesBlocked_MPIBAIJ()
27: converts the entries into single precision and then calls ..._MatScalar() to put them
28: into the single precision data structures.
29: */
30: #if defined(PETSC_USE_MAT_SINGLE)
31: EXTERN PetscErrorCode MatSetValuesBlocked_SeqSBAIJ_MatScalar(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const MatScalar[],InsertMode);
32: EXTERN PetscErrorCode MatSetValues_MPISBAIJ_MatScalar(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const MatScalar[],InsertMode);
33: EXTERN PetscErrorCode MatSetValuesBlocked_MPISBAIJ_MatScalar(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const MatScalar[],InsertMode);
34: EXTERN PetscErrorCode MatSetValues_MPISBAIJ_HT_MatScalar(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const MatScalar[],InsertMode);
35: EXTERN PetscErrorCode MatSetValuesBlocked_MPISBAIJ_HT_MatScalar(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const MatScalar[],InsertMode);
36: #else
37: #define MatSetValuesBlocked_SeqSBAIJ_MatScalar MatSetValuesBlocked_SeqSBAIJ
38: #define MatSetValues_MPISBAIJ_MatScalar MatSetValues_MPISBAIJ
39: #define MatSetValuesBlocked_MPISBAIJ_MatScalar MatSetValuesBlocked_MPISBAIJ
40: #define MatSetValues_MPISBAIJ_HT_MatScalar MatSetValues_MPISBAIJ_HT
41: #define MatSetValuesBlocked_MPISBAIJ_HT_MatScalar MatSetValuesBlocked_MPISBAIJ_HT
42: #endif
47: PetscErrorCode MatStoreValues_MPISBAIJ(Mat mat)
48: {
49: Mat_MPISBAIJ *aij = (Mat_MPISBAIJ *)mat->data;
53: MatStoreValues(aij->A);
54: MatStoreValues(aij->B);
55: return(0);
56: }
62: PetscErrorCode MatRetrieveValues_MPISBAIJ(Mat mat)
63: {
64: Mat_MPISBAIJ *aij = (Mat_MPISBAIJ *)mat->data;
68: MatRetrieveValues(aij->A);
69: MatRetrieveValues(aij->B);
70: return(0);
71: }
75: #define CHUNKSIZE 10
77: #define MatSetValues_SeqSBAIJ_A_Private(row,col,value,addv) \
78: { \
79: \
80: brow = row/bs; \
81: rp = aj + ai[brow]; ap = aa + bs2*ai[brow]; \
82: rmax = aimax[brow]; nrow = ailen[brow]; \
83: bcol = col/bs; \
84: ridx = row % bs; cidx = col % bs; \
85: low = 0; high = nrow; \
86: while (high-low > 3) { \
87: t = (low+high)/2; \
88: if (rp[t] > bcol) high = t; \
89: else low = t; \
90: } \
91: for (_i=low; _i<high; _i++) { \
92: if (rp[_i] > bcol) break; \
93: if (rp[_i] == bcol) { \
94: bap = ap + bs2*_i + bs*cidx + ridx; \
95: if (addv == ADD_VALUES) *bap += value; \
96: else *bap = value; \
97: goto a_noinsert; \
98: } \
99: } \
100: if (a->nonew == 1) goto a_noinsert; \
101: if (a->nonew == -1) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%D, %D) into matrix", row, col); \
102: MatSeqXAIJReallocateAIJ(A,a->mbs,bs2,nrow,brow,bcol,rmax,aa,ai,aj,rp,ap,aimax,a->nonew,MatScalar); \
103: N = nrow++ - 1; \
104: /* shift up all the later entries in this row */ \
105: for (ii=N; ii>=_i; ii--) { \
106: rp[ii+1] = rp[ii]; \
107: PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar)); \
108: } \
109: if (N>=_i) { PetscMemzero(ap+bs2*_i,bs2*sizeof(MatScalar)); } \
110: rp[_i] = bcol; \
111: ap[bs2*_i + bs*cidx + ridx] = value; \
112: a_noinsert:; \
113: ailen[brow] = nrow; \
114: }
115: #ifndef MatSetValues_SeqBAIJ_B_Private
116: #define MatSetValues_SeqSBAIJ_B_Private(row,col,value,addv) \
117: { \
118: brow = row/bs; \
119: rp = bj + bi[brow]; ap = ba + bs2*bi[brow]; \
120: rmax = bimax[brow]; nrow = bilen[brow]; \
121: bcol = col/bs; \
122: ridx = row % bs; cidx = col % bs; \
123: low = 0; high = nrow; \
124: while (high-low > 3) { \
125: t = (low+high)/2; \
126: if (rp[t] > bcol) high = t; \
127: else low = t; \
128: } \
129: for (_i=low; _i<high; _i++) { \
130: if (rp[_i] > bcol) break; \
131: if (rp[_i] == bcol) { \
132: bap = ap + bs2*_i + bs*cidx + ridx; \
133: if (addv == ADD_VALUES) *bap += value; \
134: else *bap = value; \
135: goto b_noinsert; \
136: } \
137: } \
138: if (b->nonew == 1) goto b_noinsert; \
139: if (b->nonew == -1) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%D, %D) into matrix", row, col); \
140: MatSeqXAIJReallocateAIJ(B,b->mbs,bs2,nrow,brow,bcol,rmax,ba,bi,bj,rp,ap,bimax,b->nonew,MatScalar); \
141: N = nrow++ - 1; \
142: /* shift up all the later entries in this row */ \
143: for (ii=N; ii>=_i; ii--) { \
144: rp[ii+1] = rp[ii]; \
145: PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar)); \
146: } \
147: if (N>=_i) { PetscMemzero(ap+bs2*_i,bs2*sizeof(MatScalar));} \
148: rp[_i] = bcol; \
149: ap[bs2*_i + bs*cidx + ridx] = value; \
150: b_noinsert:; \
151: bilen[brow] = nrow; \
152: }
153: #endif
155: #if defined(PETSC_USE_MAT_SINGLE)
158: PetscErrorCode MatSetValues_MPISBAIJ(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode addv)
159: {
160: Mat_MPISBAIJ *b = (Mat_MPISBAIJ*)mat->data;
162: PetscInt i,N = m*n;
163: MatScalar *vsingle;
166: if (N > b->setvalueslen) {
167: PetscFree(b->setvaluescopy);
168: PetscMalloc(N*sizeof(MatScalar),&b->setvaluescopy);
169: b->setvalueslen = N;
170: }
171: vsingle = b->setvaluescopy;
173: for (i=0; i<N; i++) {
174: vsingle[i] = v[i];
175: }
176: MatSetValues_MPISBAIJ_MatScalar(mat,m,im,n,in,vsingle,addv);
177: return(0);
178: }
182: PetscErrorCode MatSetValuesBlocked_MPISBAIJ(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode addv)
183: {
184: Mat_MPIBAIJ *b = (Mat_MPIBAIJ*)mat->data;
186: PetscInt i,N = m*n*b->bs2;
187: MatScalar *vsingle;
190: if (N > b->setvalueslen) {
191: PetscFree(b->setvaluescopy);
192: PetscMalloc(N*sizeof(MatScalar),&b->setvaluescopy);
193: b->setvalueslen = N;
194: }
195: vsingle = b->setvaluescopy;
196: for (i=0; i<N; i++) {
197: vsingle[i] = v[i];
198: }
199: MatSetValuesBlocked_MPISBAIJ_MatScalar(mat,m,im,n,in,vsingle,addv);
200: return(0);
201: }
202: #endif
204: /* Only add/insert a(i,j) with i<=j (blocks).
205: Any a(i,j) with i>j input by user is ingored.
206: */
209: PetscErrorCode MatSetValues_MPISBAIJ_MatScalar(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const MatScalar v[],InsertMode addv)
210: {
211: Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data;
212: MatScalar value;
213: PetscTruth roworiented = baij->roworiented;
215: PetscInt i,j,row,col;
216: PetscInt rstart_orig=mat->rmap.rstart;
217: PetscInt rend_orig=mat->rmap.rend,cstart_orig=mat->cmap.rstart;
218: PetscInt cend_orig=mat->cmap.rend,bs=mat->rmap.bs;
220: /* Some Variables required in the macro */
221: Mat A = baij->A;
222: Mat_SeqSBAIJ *a = (Mat_SeqSBAIJ*)(A)->data;
223: PetscInt *aimax=a->imax,*ai=a->i,*ailen=a->ilen,*aj=a->j;
224: MatScalar *aa=a->a;
226: Mat B = baij->B;
227: Mat_SeqBAIJ *b = (Mat_SeqBAIJ*)(B)->data;
228: PetscInt *bimax=b->imax,*bi=b->i,*bilen=b->ilen,*bj=b->j;
229: MatScalar *ba=b->a;
231: PetscInt *rp,ii,nrow,_i,rmax,N,brow,bcol;
232: PetscInt low,high,t,ridx,cidx,bs2=a->bs2;
233: MatScalar *ap,*bap;
235: /* for stash */
236: PetscInt n_loc, *in_loc = PETSC_NULL;
237: MatScalar *v_loc = PETSC_NULL;
241: if (!baij->donotstash){
242: if (n > baij->n_loc) {
243: PetscFree(baij->in_loc);
244: PetscFree(baij->v_loc);
245: PetscMalloc(n*sizeof(PetscInt),&baij->in_loc);
246: PetscMalloc(n*sizeof(MatScalar),&baij->v_loc);
247: baij->n_loc = n;
248: }
249: in_loc = baij->in_loc;
250: v_loc = baij->v_loc;
251: }
253: for (i=0; i<m; i++) {
254: if (im[i] < 0) continue;
255: #if defined(PETSC_USE_DEBUG)
256: if (im[i] >= mat->rmap.N) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",im[i],mat->rmap.N-1);
257: #endif
258: if (im[i] >= rstart_orig && im[i] < rend_orig) { /* this processor entry */
259: row = im[i] - rstart_orig; /* local row index */
260: for (j=0; j<n; j++) {
261: if (im[i]/bs > in[j]/bs){
262: if (a->ignore_ltriangular){
263: continue; /* ignore lower triangular blocks */
264: } else {
265: SETERRQ(PETSC_ERR_USER,"Lower triangular value cannot be set for sbaij format. Ignoring these values, run with -mat_ignore_lower_triangular or call MatSetOption(mat,MAT_IGNORE_LOWER_TRIANGULAR)");
266: }
267: }
268: if (in[j] >= cstart_orig && in[j] < cend_orig){ /* diag entry (A) */
269: col = in[j] - cstart_orig; /* local col index */
270: brow = row/bs; bcol = col/bs;
271: if (brow > bcol) continue; /* ignore lower triangular blocks of A */
272: if (roworiented) value = v[i*n+j]; else value = v[i+j*m];
273: MatSetValues_SeqSBAIJ_A_Private(row,col,value,addv);
274: /* MatSetValues_SeqBAIJ(baij->A,1,&row,1,&col,&value,addv); */
275: } else if (in[j] < 0) continue;
276: #if defined(PETSC_USE_DEBUG)
277: else if (in[j] >= mat->cmap.N) {SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %D max %D",in[j],mat->cmap.N-1);}
278: #endif
279: else { /* off-diag entry (B) */
280: if (mat->was_assembled) {
281: if (!baij->colmap) {
282: CreateColmap_MPIBAIJ_Private(mat);
283: }
284: #if defined (PETSC_USE_CTABLE)
285: PetscTableFind(baij->colmap,in[j]/bs + 1,&col);
286: col = col - 1;
287: #else
288: col = baij->colmap[in[j]/bs] - 1;
289: #endif
290: if (col < 0 && !((Mat_SeqSBAIJ*)(baij->A->data))->nonew) {
291: DisAssemble_MPISBAIJ(mat);
292: col = in[j];
293: /* Reinitialize the variables required by MatSetValues_SeqBAIJ_B_Private() */
294: B = baij->B;
295: b = (Mat_SeqBAIJ*)(B)->data;
296: bimax=b->imax;bi=b->i;bilen=b->ilen;bj=b->j;
297: ba=b->a;
298: } else col += in[j]%bs;
299: } else col = in[j];
300: if (roworiented) value = v[i*n+j]; else value = v[i+j*m];
301: MatSetValues_SeqSBAIJ_B_Private(row,col,value,addv);
302: /* MatSetValues_SeqBAIJ(baij->B,1,&row,1,&col,&value,addv); */
303: }
304: }
305: } else { /* off processor entry */
306: if (!baij->donotstash) {
307: n_loc = 0;
308: for (j=0; j<n; j++){
309: if (im[i]/bs > in[j]/bs) continue; /* ignore lower triangular blocks */
310: in_loc[n_loc] = in[j];
311: if (roworiented) {
312: v_loc[n_loc] = v[i*n+j];
313: } else {
314: v_loc[n_loc] = v[j*m+i];
315: }
316: n_loc++;
317: }
318: MatStashValuesRow_Private(&mat->stash,im[i],n_loc,in_loc,v_loc);
319: }
320: }
321: }
322: return(0);
323: }
327: PetscErrorCode MatSetValuesBlocked_MPISBAIJ_MatScalar(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const MatScalar v[],InsertMode addv)
328: {
329: Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data;
330: const MatScalar *value;
331: MatScalar *barray=baij->barray;
332: PetscTruth roworiented = baij->roworiented;
333: PetscErrorCode ierr;
334: PetscInt i,j,ii,jj,row,col,rstart=baij->rstartbs;
335: PetscInt rend=baij->rendbs,cstart=baij->rstartbs,stepval;
336: PetscInt cend=baij->rendbs,bs=mat->rmap.bs,bs2=baij->bs2;
339: if(!barray) {
340: PetscMalloc(bs2*sizeof(MatScalar),&barray);
341: baij->barray = barray;
342: }
344: if (roworiented) {
345: stepval = (n-1)*bs;
346: } else {
347: stepval = (m-1)*bs;
348: }
349: for (i=0; i<m; i++) {
350: if (im[i] < 0) continue;
351: #if defined(PETSC_USE_DEBUG)
352: if (im[i] >= baij->Mbs) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large, row %D max %D",im[i],baij->Mbs-1);
353: #endif
354: if (im[i] >= rstart && im[i] < rend) {
355: row = im[i] - rstart;
356: for (j=0; j<n; j++) {
357: /* If NumCol = 1 then a copy is not required */
358: if ((roworiented) && (n == 1)) {
359: barray = (MatScalar*) v + i*bs2;
360: } else if((!roworiented) && (m == 1)) {
361: barray = (MatScalar*) v + j*bs2;
362: } else { /* Here a copy is required */
363: if (roworiented) {
364: value = v + i*(stepval+bs)*bs + j*bs;
365: } else {
366: value = v + j*(stepval+bs)*bs + i*bs;
367: }
368: for (ii=0; ii<bs; ii++,value+=stepval) {
369: for (jj=0; jj<bs; jj++) {
370: *barray++ = *value++;
371: }
372: }
373: barray -=bs2;
374: }
375:
376: if (in[j] >= cstart && in[j] < cend){
377: col = in[j] - cstart;
378: MatSetValuesBlocked_SeqSBAIJ(baij->A,1,&row,1,&col,barray,addv);
379: }
380: else if (in[j] < 0) continue;
381: #if defined(PETSC_USE_DEBUG)
382: else if (in[j] >= baij->Nbs) {SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large, col %D max %D",in[j],baij->Nbs-1);}
383: #endif
384: else {
385: if (mat->was_assembled) {
386: if (!baij->colmap) {
387: CreateColmap_MPIBAIJ_Private(mat);
388: }
390: #if defined(PETSC_USE_DEBUG)
391: #if defined (PETSC_USE_CTABLE)
392: { PetscInt data;
393: PetscTableFind(baij->colmap,in[j]+1,&data);
394: if ((data - 1) % bs) SETERRQ(PETSC_ERR_PLIB,"Incorrect colmap");
395: }
396: #else
397: if ((baij->colmap[in[j]] - 1) % bs) SETERRQ(PETSC_ERR_PLIB,"Incorrect colmap");
398: #endif
399: #endif
400: #if defined (PETSC_USE_CTABLE)
401: PetscTableFind(baij->colmap,in[j]+1,&col);
402: col = (col - 1)/bs;
403: #else
404: col = (baij->colmap[in[j]] - 1)/bs;
405: #endif
406: if (col < 0 && !((Mat_SeqBAIJ*)(baij->A->data))->nonew) {
407: DisAssemble_MPISBAIJ(mat);
408: col = in[j];
409: }
410: }
411: else col = in[j];
412: MatSetValuesBlocked_SeqBAIJ(baij->B,1,&row,1,&col,barray,addv);
413: }
414: }
415: } else {
416: if (!baij->donotstash) {
417: if (roworiented) {
418: MatStashValuesRowBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);
419: } else {
420: MatStashValuesColBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);
421: }
422: }
423: }
424: }
425: return(0);
426: }
430: PetscErrorCode MatGetValues_MPISBAIJ(Mat mat,PetscInt m,const PetscInt idxm[],PetscInt n,const PetscInt idxn[],PetscScalar v[])
431: {
432: Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data;
434: PetscInt bs=mat->rmap.bs,i,j,bsrstart = mat->rmap.rstart,bsrend = mat->rmap.rend;
435: PetscInt bscstart = mat->cmap.rstart,bscend = mat->cmap.rend,row,col,data;
438: for (i=0; i<m; i++) {
439: if (idxm[i] < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Negative row: %D",idxm[i]);
440: if (idxm[i] >= mat->rmap.N) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",idxm[i],mat->rmap.N-1);
441: if (idxm[i] >= bsrstart && idxm[i] < bsrend) {
442: row = idxm[i] - bsrstart;
443: for (j=0; j<n; j++) {
444: if (idxn[j] < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Negative column %D",idxn[j]);
445: if (idxn[j] >= mat->cmap.N) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %D max %D",idxn[j],mat->cmap.N-1);
446: if (idxn[j] >= bscstart && idxn[j] < bscend){
447: col = idxn[j] - bscstart;
448: MatGetValues_SeqSBAIJ(baij->A,1,&row,1,&col,v+i*n+j);
449: } else {
450: if (!baij->colmap) {
451: CreateColmap_MPIBAIJ_Private(mat);
452: }
453: #if defined (PETSC_USE_CTABLE)
454: PetscTableFind(baij->colmap,idxn[j]/bs+1,&data);
455: data --;
456: #else
457: data = baij->colmap[idxn[j]/bs]-1;
458: #endif
459: if((data < 0) || (baij->garray[data/bs] != idxn[j]/bs)) *(v+i*n+j) = 0.0;
460: else {
461: col = data + idxn[j]%bs;
462: MatGetValues_SeqBAIJ(baij->B,1,&row,1,&col,v+i*n+j);
463: }
464: }
465: }
466: } else {
467: SETERRQ(PETSC_ERR_SUP,"Only local values currently supported");
468: }
469: }
470: return(0);
471: }
475: PetscErrorCode MatNorm_MPISBAIJ(Mat mat,NormType type,PetscReal *norm)
476: {
477: Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data;
479: PetscReal sum[2],*lnorm2;
482: if (baij->size == 1) {
483: MatNorm(baij->A,type,norm);
484: } else {
485: if (type == NORM_FROBENIUS) {
486: PetscMalloc(2*sizeof(PetscReal),&lnorm2);
487: MatNorm(baij->A,type,lnorm2);
488: *lnorm2 = (*lnorm2)*(*lnorm2); lnorm2++; /* squar power of norm(A) */
489: MatNorm(baij->B,type,lnorm2);
490: *lnorm2 = (*lnorm2)*(*lnorm2); lnorm2--; /* squar power of norm(B) */
491: MPI_Allreduce(lnorm2,&sum,2,MPIU_REAL,MPI_SUM,mat->comm);
492: *norm = sqrt(sum[0] + 2*sum[1]);
493: PetscFree(lnorm2);
494: } else if (type == NORM_INFINITY || type == NORM_1) { /* max row/column sum */
495: Mat_SeqSBAIJ *amat=(Mat_SeqSBAIJ*)baij->A->data;
496: Mat_SeqBAIJ *bmat=(Mat_SeqBAIJ*)baij->B->data;
497: PetscReal *rsum,*rsum2,vabs;
498: PetscInt *jj,*garray=baij->garray,rstart=baij->rstartbs,nz;
499: PetscInt brow,bcol,col,bs=baij->A->rmap.bs,row,grow,gcol,mbs=amat->mbs;
500: MatScalar *v;
502: PetscMalloc((2*mat->cmap.N+1)*sizeof(PetscReal),&rsum);
503: rsum2 = rsum + mat->cmap.N;
504: PetscMemzero(rsum,mat->cmap.N*sizeof(PetscReal));
505: /* Amat */
506: v = amat->a; jj = amat->j;
507: for (brow=0; brow<mbs; brow++) {
508: grow = bs*(rstart + brow);
509: nz = amat->i[brow+1] - amat->i[brow];
510: for (bcol=0; bcol<nz; bcol++){
511: gcol = bs*(rstart + *jj); jj++;
512: for (col=0; col<bs; col++){
513: for (row=0; row<bs; row++){
514: vabs = PetscAbsScalar(*v); v++;
515: rsum[gcol+col] += vabs;
516: /* non-diagonal block */
517: if (bcol > 0 && vabs > 0.0) rsum[grow+row] += vabs;
518: }
519: }
520: }
521: }
522: /* Bmat */
523: v = bmat->a; jj = bmat->j;
524: for (brow=0; brow<mbs; brow++) {
525: grow = bs*(rstart + brow);
526: nz = bmat->i[brow+1] - bmat->i[brow];
527: for (bcol=0; bcol<nz; bcol++){
528: gcol = bs*garray[*jj]; jj++;
529: for (col=0; col<bs; col++){
530: for (row=0; row<bs; row++){
531: vabs = PetscAbsScalar(*v); v++;
532: rsum[gcol+col] += vabs;
533: rsum[grow+row] += vabs;
534: }
535: }
536: }
537: }
538: MPI_Allreduce(rsum,rsum2,mat->cmap.N,MPIU_REAL,MPI_SUM,mat->comm);
539: *norm = 0.0;
540: for (col=0; col<mat->cmap.N; col++) {
541: if (rsum2[col] > *norm) *norm = rsum2[col];
542: }
543: PetscFree(rsum);
544: } else {
545: SETERRQ(PETSC_ERR_SUP,"No support for this norm yet");
546: }
547: }
548: return(0);
549: }
553: PetscErrorCode MatAssemblyBegin_MPISBAIJ(Mat mat,MatAssemblyType mode)
554: {
555: Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data;
557: PetscInt nstash,reallocs;
558: InsertMode addv;
561: if (baij->donotstash) {
562: return(0);
563: }
565: /* make sure all processors are either in INSERTMODE or ADDMODE */
566: MPI_Allreduce(&mat->insertmode,&addv,1,MPI_INT,MPI_BOR,mat->comm);
567: if (addv == (ADD_VALUES|INSERT_VALUES)) {
568: SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Some processors inserted others added");
569: }
570: mat->insertmode = addv; /* in case this processor had no cache */
572: MatStashScatterBegin_Private(&mat->stash,mat->rmap.range);
573: MatStashScatterBegin_Private(&mat->bstash,baij->rangebs);
574: MatStashGetInfo_Private(&mat->stash,&nstash,&reallocs);
575: PetscInfo2(0,"Stash has %D entries,uses %D mallocs.\n",nstash,reallocs);
576: MatStashGetInfo_Private(&mat->stash,&nstash,&reallocs);
577: PetscInfo2(0,"Block-Stash has %D entries, uses %D mallocs.\n",nstash,reallocs);
578: return(0);
579: }
583: PetscErrorCode MatAssemblyEnd_MPISBAIJ(Mat mat,MatAssemblyType mode)
584: {
585: Mat_MPISBAIJ *baij=(Mat_MPISBAIJ*)mat->data;
586: Mat_SeqSBAIJ *a=(Mat_SeqSBAIJ*)baij->A->data;
588: PetscInt i,j,rstart,ncols,flg,bs2=baij->bs2;
589: PetscInt *row,*col,other_disassembled;
590: PetscMPIInt n;
591: PetscTruth r1,r2,r3;
592: MatScalar *val;
593: InsertMode addv = mat->insertmode;
595: /* do not use 'b=(Mat_SeqBAIJ*)baij->B->data' as B can be reset in disassembly */
598: if (!baij->donotstash) {
599: while (1) {
600: MatStashScatterGetMesg_Private(&mat->stash,&n,&row,&col,&val,&flg);
601: if (!flg) break;
603: for (i=0; i<n;) {
604: /* Now identify the consecutive vals belonging to the same row */
605: for (j=i,rstart=row[j]; j<n; j++) { if (row[j] != rstart) break; }
606: if (j < n) ncols = j-i;
607: else ncols = n-i;
608: /* Now assemble all these values with a single function call */
609: MatSetValues_MPISBAIJ_MatScalar(mat,1,row+i,ncols,col+i,val+i,addv);
610: i = j;
611: }
612: }
613: MatStashScatterEnd_Private(&mat->stash);
614: /* Now process the block-stash. Since the values are stashed column-oriented,
615: set the roworiented flag to column oriented, and after MatSetValues()
616: restore the original flags */
617: r1 = baij->roworiented;
618: r2 = a->roworiented;
619: r3 = ((Mat_SeqBAIJ*)baij->B->data)->roworiented;
620: baij->roworiented = PETSC_FALSE;
621: a->roworiented = PETSC_FALSE;
622: ((Mat_SeqBAIJ*)baij->B->data)->roworiented = PETSC_FALSE; /* b->roworinted */
623: while (1) {
624: MatStashScatterGetMesg_Private(&mat->bstash,&n,&row,&col,&val,&flg);
625: if (!flg) break;
626:
627: for (i=0; i<n;) {
628: /* Now identify the consecutive vals belonging to the same row */
629: for (j=i,rstart=row[j]; j<n; j++) { if (row[j] != rstart) break; }
630: if (j < n) ncols = j-i;
631: else ncols = n-i;
632: MatSetValuesBlocked_MPISBAIJ_MatScalar(mat,1,row+i,ncols,col+i,val+i*bs2,addv);
633: i = j;
634: }
635: }
636: MatStashScatterEnd_Private(&mat->bstash);
637: baij->roworiented = r1;
638: a->roworiented = r2;
639: ((Mat_SeqBAIJ*)baij->B->data)->roworiented = r3; /* b->roworinted */
640: }
642: MatAssemblyBegin(baij->A,mode);
643: MatAssemblyEnd(baij->A,mode);
645: /* determine if any processor has disassembled, if so we must
646: also disassemble ourselfs, in order that we may reassemble. */
647: /*
648: if nonzero structure of submatrix B cannot change then we know that
649: no processor disassembled thus we can skip this stuff
650: */
651: if (!((Mat_SeqBAIJ*)baij->B->data)->nonew) {
652: MPI_Allreduce(&mat->was_assembled,&other_disassembled,1,MPI_INT,MPI_PROD,mat->comm);
653: if (mat->was_assembled && !other_disassembled) {
654: DisAssemble_MPISBAIJ(mat);
655: }
656: }
658: if (!mat->was_assembled && mode == MAT_FINAL_ASSEMBLY) {
659: MatSetUpMultiply_MPISBAIJ(mat); /* setup Mvctx and sMvctx */
660: }
661: ((Mat_SeqBAIJ*)baij->B->data)->compressedrow.use = PETSC_TRUE; /* b->compressedrow.use */
662: MatAssemblyBegin(baij->B,mode);
663: MatAssemblyEnd(baij->B,mode);
664:
665: PetscFree(baij->rowvalues);
666: baij->rowvalues = 0;
668: return(0);
669: }
674: static PetscErrorCode MatView_MPISBAIJ_ASCIIorDraworSocket(Mat mat,PetscViewer viewer)
675: {
676: Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data;
677: PetscErrorCode ierr;
678: PetscInt bs = mat->rmap.bs;
679: PetscMPIInt size = baij->size,rank = baij->rank;
680: PetscTruth iascii,isdraw;
681: PetscViewer sviewer;
682: PetscViewerFormat format;
685: PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_ASCII,&iascii);
686: PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_DRAW,&isdraw);
687: if (iascii) {
688: PetscViewerGetFormat(viewer,&format);
689: if (format == PETSC_VIEWER_ASCII_INFO_DETAIL) {
690: MatInfo info;
691: MPI_Comm_rank(mat->comm,&rank);
692: MatGetInfo(mat,MAT_LOCAL,&info);
693: PetscViewerASCIISynchronizedPrintf(viewer,"[%d] Local rows %D nz %D nz alloced %D bs %D mem %D\n",
694: rank,mat->rmap.N,(PetscInt)info.nz_used*bs,(PetscInt)info.nz_allocated*bs,
695: mat->rmap.bs,(PetscInt)info.memory);
696: MatGetInfo(baij->A,MAT_LOCAL,&info);
697: PetscViewerASCIISynchronizedPrintf(viewer,"[%d] on-diagonal part: nz %D \n",rank,(PetscInt)info.nz_used*bs);
698: MatGetInfo(baij->B,MAT_LOCAL,&info);
699: PetscViewerASCIISynchronizedPrintf(viewer,"[%d] off-diagonal part: nz %D \n",rank,(PetscInt)info.nz_used*bs);
700: PetscViewerFlush(viewer);
701: PetscViewerASCIIPrintf(viewer,"Information on VecScatter used in matrix-vector product: \n");
702: VecScatterView(baij->Mvctx,viewer);
703: return(0);
704: } else if (format == PETSC_VIEWER_ASCII_INFO) {
705: PetscViewerASCIIPrintf(viewer," block size is %D\n",bs);
706: return(0);
707: } else if (format == PETSC_VIEWER_ASCII_FACTOR_INFO) {
708: return(0);
709: }
710: }
712: if (isdraw) {
713: PetscDraw draw;
714: PetscTruth isnull;
715: PetscViewerDrawGetDraw(viewer,0,&draw);
716: PetscDrawIsNull(draw,&isnull); if (isnull) return(0);
717: }
719: if (size == 1) {
720: PetscObjectSetName((PetscObject)baij->A,mat->name);
721: MatView(baij->A,viewer);
722: } else {
723: /* assemble the entire matrix onto first processor. */
724: Mat A;
725: Mat_SeqSBAIJ *Aloc;
726: Mat_SeqBAIJ *Bloc;
727: PetscInt M = mat->rmap.N,N = mat->cmap.N,*ai,*aj,col,i,j,k,*rvals,mbs = baij->mbs;
728: MatScalar *a;
730: /* Should this be the same type as mat? */
731: MatCreate(mat->comm,&A);
732: if (!rank) {
733: MatSetSizes(A,M,N,M,N);
734: } else {
735: MatSetSizes(A,0,0,M,N);
736: }
737: MatSetType(A,MATMPISBAIJ);
738: MatMPISBAIJSetPreallocation(A,mat->rmap.bs,0,PETSC_NULL,0,PETSC_NULL);
739: PetscLogObjectParent(mat,A);
741: /* copy over the A part */
742: Aloc = (Mat_SeqSBAIJ*)baij->A->data;
743: ai = Aloc->i; aj = Aloc->j; a = Aloc->a;
744: PetscMalloc(bs*sizeof(PetscInt),&rvals);
746: for (i=0; i<mbs; i++) {
747: rvals[0] = bs*(baij->rstartbs + i);
748: for (j=1; j<bs; j++) { rvals[j] = rvals[j-1] + 1; }
749: for (j=ai[i]; j<ai[i+1]; j++) {
750: col = (baij->cstartbs+aj[j])*bs;
751: for (k=0; k<bs; k++) {
752: MatSetValues_MPISBAIJ_MatScalar(A,bs,rvals,1,&col,a,INSERT_VALUES);
753: col++; a += bs;
754: }
755: }
756: }
757: /* copy over the B part */
758: Bloc = (Mat_SeqBAIJ*)baij->B->data;
759: ai = Bloc->i; aj = Bloc->j; a = Bloc->a;
760: for (i=0; i<mbs; i++) {
761:
762: rvals[0] = bs*(baij->rstartbs + i);
763: for (j=1; j<bs; j++) { rvals[j] = rvals[j-1] + 1; }
764: for (j=ai[i]; j<ai[i+1]; j++) {
765: col = baij->garray[aj[j]]*bs;
766: for (k=0; k<bs; k++) {
767: MatSetValues_MPIBAIJ(A,bs,rvals,1,&col,a,INSERT_VALUES);
768: col++; a += bs;
769: }
770: }
771: }
772: PetscFree(rvals);
773: MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);
774: MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);
775: /*
776: Everyone has to call to draw the matrix since the graphics waits are
777: synchronized across all processors that share the PetscDraw object
778: */
779: PetscViewerGetSingleton(viewer,&sviewer);
780: if (!rank) {
781: PetscObjectSetName((PetscObject)((Mat_MPISBAIJ*)(A->data))->A,mat->name);
782: MatView(((Mat_MPISBAIJ*)(A->data))->A,sviewer);
783: }
784: PetscViewerRestoreSingleton(viewer,&sviewer);
785: MatDestroy(A);
786: }
787: return(0);
788: }
792: PetscErrorCode MatView_MPISBAIJ(Mat mat,PetscViewer viewer)
793: {
795: PetscTruth iascii,isdraw,issocket,isbinary;
798: PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_ASCII,&iascii);
799: PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_DRAW,&isdraw);
800: PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_SOCKET,&issocket);
801: PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_BINARY,&isbinary);
802: if (iascii || isdraw || issocket || isbinary) {
803: MatView_MPISBAIJ_ASCIIorDraworSocket(mat,viewer);
804: } else {
805: SETERRQ1(PETSC_ERR_SUP,"Viewer type %s not supported by MPISBAIJ matrices",((PetscObject)viewer)->type_name);
806: }
807: return(0);
808: }
812: PetscErrorCode MatDestroy_MPISBAIJ(Mat mat)
813: {
814: Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data;
818: #if defined(PETSC_USE_LOG)
819: PetscLogObjectState((PetscObject)mat,"Rows=%D,Cols=%D",mat->rmap.N,mat->cmap.N);
820: #endif
821: MatStashDestroy_Private(&mat->stash);
822: MatStashDestroy_Private(&mat->bstash);
823: MatDestroy(baij->A);
824: MatDestroy(baij->B);
825: #if defined (PETSC_USE_CTABLE)
826: if (baij->colmap) {PetscTableDestroy(baij->colmap);}
827: #else
828: PetscFree(baij->colmap);
829: #endif
830: PetscFree(baij->garray);
831: if (baij->lvec) {VecDestroy(baij->lvec);}
832: if (baij->Mvctx) {VecScatterDestroy(baij->Mvctx);}
833: if (baij->slvec0) {
834: VecDestroy(baij->slvec0);
835: VecDestroy(baij->slvec0b);
836: }
837: if (baij->slvec1) {
838: VecDestroy(baij->slvec1);
839: VecDestroy(baij->slvec1a);
840: VecDestroy(baij->slvec1b);
841: }
842: if (baij->sMvctx) {VecScatterDestroy(baij->sMvctx);}
843: PetscFree(baij->rowvalues);
844: PetscFree(baij->barray);
845: PetscFree(baij->hd);
846: #if defined(PETSC_USE_MAT_SINGLE)
847: PetscFree(baij->setvaluescopy);
848: #endif
849: PetscFree(baij->in_loc);
850: PetscFree(baij->v_loc);
851: PetscFree(baij->rangebs);
852: PetscFree(baij);
854: PetscObjectChangeTypeName((PetscObject)mat,0);
855: PetscObjectComposeFunction((PetscObject)mat,"MatStoreValues_C","",PETSC_NULL);
856: PetscObjectComposeFunction((PetscObject)mat,"MatRetrieveValues_C","",PETSC_NULL);
857: PetscObjectComposeFunction((PetscObject)mat,"MatGetDiagonalBlock_C","",PETSC_NULL);
858: PetscObjectComposeFunction((PetscObject)mat,"MatMPISBAIJSetPreallocation_C","",PETSC_NULL);
859: return(0);
860: }
864: PetscErrorCode MatMult_MPISBAIJ(Mat A,Vec xx,Vec yy)
865: {
866: Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data;
868: PetscInt nt,mbs=a->mbs,bs=A->rmap.bs;
869: PetscScalar *x,*from,zero=0.0;
870:
872: VecGetLocalSize(xx,&nt);
873: if (nt != A->cmap.n) {
874: SETERRQ(PETSC_ERR_ARG_SIZ,"Incompatible partition of A and xx");
875: }
877: /* diagonal part */
878: (*a->A->ops->mult)(a->A,xx,a->slvec1a);
879: VecSet(a->slvec1b,zero);
881: /* subdiagonal part */
882: (*a->B->ops->multtranspose)(a->B,xx,a->slvec0b);
883: CHKMEMQ;
884: /* copy x into the vec slvec0 */
885: VecGetArray(a->slvec0,&from);
886: VecGetArray(xx,&x);
887: CHKMEMQ;
888: PetscMemcpy(from,x,bs*mbs*sizeof(MatScalar));
889: CHKMEMQ;
890: VecRestoreArray(a->slvec0,&from);
891:
892: CHKMEMQ;
893: VecScatterBegin(a->sMvctx,a->slvec0,a->slvec1,ADD_VALUES,SCATTER_FORWARD);
894: CHKMEMQ;
895: VecRestoreArray(xx,&x);
896: CHKMEMQ;
897: VecScatterEnd(a->sMvctx,a->slvec0,a->slvec1,ADD_VALUES,SCATTER_FORWARD);
898: CHKMEMQ;
899: /* supperdiagonal part */
900: (*a->B->ops->multadd)(a->B,a->slvec1b,a->slvec1a,yy);
901: CHKMEMQ;
902: return(0);
903: }
907: PetscErrorCode MatMult_MPISBAIJ_2comm(Mat A,Vec xx,Vec yy)
908: {
909: Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data;
911: PetscInt nt;
914: VecGetLocalSize(xx,&nt);
915: if (nt != A->cmap.n) {
916: SETERRQ(PETSC_ERR_ARG_SIZ,"Incompatible partition of A and xx");
917: }
918: VecGetLocalSize(yy,&nt);
919: if (nt != A->rmap.N) {
920: SETERRQ(PETSC_ERR_ARG_SIZ,"Incompatible parition of A and yy");
921: }
923: VecScatterBegin(a->Mvctx,xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD);
924: /* do diagonal part */
925: (*a->A->ops->mult)(a->A,xx,yy);
926: /* do supperdiagonal part */
927: VecScatterEnd(a->Mvctx,xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD);
928: (*a->B->ops->multadd)(a->B,a->lvec,yy,yy);
929: /* do subdiagonal part */
930: (*a->B->ops->multtranspose)(a->B,xx,a->lvec);
931: VecScatterBegin(a->Mvctx,a->lvec,yy,ADD_VALUES,SCATTER_REVERSE);
932: VecScatterEnd(a->Mvctx,a->lvec,yy,ADD_VALUES,SCATTER_REVERSE);
934: return(0);
935: }
939: PetscErrorCode MatMultAdd_MPISBAIJ(Mat A,Vec xx,Vec yy,Vec zz)
940: {
941: Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data;
943: PetscInt mbs=a->mbs,bs=A->rmap.bs;
944: PetscScalar *x,*from,zero=0.0;
945:
947: /*
948: PetscSynchronizedPrintf(A->comm," MatMultAdd is called ...\n");
949: PetscSynchronizedFlush(A->comm);
950: */
951: /* diagonal part */
952: (*a->A->ops->multadd)(a->A,xx,yy,a->slvec1a);
953: VecSet(a->slvec1b,zero);
955: /* subdiagonal part */
956: (*a->B->ops->multtranspose)(a->B,xx,a->slvec0b);
958: /* copy x into the vec slvec0 */
959: VecGetArray(a->slvec0,&from);
960: VecGetArray(xx,&x);
961: PetscMemcpy(from,x,bs*mbs*sizeof(MatScalar));
962: VecRestoreArray(a->slvec0,&from);
963:
964: VecScatterBegin(a->sMvctx,a->slvec0,a->slvec1,ADD_VALUES,SCATTER_FORWARD);
965: VecRestoreArray(xx,&x);
966: VecScatterEnd(a->sMvctx,a->slvec0,a->slvec1,ADD_VALUES,SCATTER_FORWARD);
967:
968: /* supperdiagonal part */
969: (*a->B->ops->multadd)(a->B,a->slvec1b,a->slvec1a,zz);
970:
971: return(0);
972: }
976: PetscErrorCode MatMultAdd_MPISBAIJ_2comm(Mat A,Vec xx,Vec yy,Vec zz)
977: {
978: Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data;
982: VecScatterBegin(a->Mvctx,xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD);
983: /* do diagonal part */
984: (*a->A->ops->multadd)(a->A,xx,yy,zz);
985: /* do supperdiagonal part */
986: VecScatterEnd(a->Mvctx,xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD);
987: (*a->B->ops->multadd)(a->B,a->lvec,zz,zz);
989: /* do subdiagonal part */
990: (*a->B->ops->multtranspose)(a->B,xx,a->lvec);
991: VecScatterBegin(a->Mvctx,a->lvec,zz,ADD_VALUES,SCATTER_REVERSE);
992: VecScatterEnd(a->Mvctx,a->lvec,zz,ADD_VALUES,SCATTER_REVERSE);
994: return(0);
995: }
997: /*
998: This only works correctly for square matrices where the subblock A->A is the
999: diagonal block
1000: */
1003: PetscErrorCode MatGetDiagonal_MPISBAIJ(Mat A,Vec v)
1004: {
1005: Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data;
1009: /* if (a->rmap.N != a->cmap.N) SETERRQ(PETSC_ERR_SUP,"Supports only square matrix where A->A is diag block"); */
1010: MatGetDiagonal(a->A,v);
1011: return(0);
1012: }
1016: PetscErrorCode MatScale_MPISBAIJ(Mat A,PetscScalar aa)
1017: {
1018: Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data;
1022: MatScale(a->A,aa);
1023: MatScale(a->B,aa);
1024: return(0);
1025: }
1029: PetscErrorCode MatGetRow_MPISBAIJ(Mat matin,PetscInt row,PetscInt *nz,PetscInt **idx,PetscScalar **v)
1030: {
1031: Mat_MPISBAIJ *mat = (Mat_MPISBAIJ*)matin->data;
1032: PetscScalar *vworkA,*vworkB,**pvA,**pvB,*v_p;
1034: PetscInt bs = matin->rmap.bs,bs2 = mat->bs2,i,*cworkA,*cworkB,**pcA,**pcB;
1035: PetscInt nztot,nzA,nzB,lrow,brstart = matin->rmap.rstart,brend = matin->rmap.rend;
1036: PetscInt *cmap,*idx_p,cstart = mat->rstartbs;
1039: if (mat->getrowactive == PETSC_TRUE) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Already active");
1040: mat->getrowactive = PETSC_TRUE;
1042: if (!mat->rowvalues && (idx || v)) {
1043: /*
1044: allocate enough space to hold information from the longest row.
1045: */
1046: Mat_SeqSBAIJ *Aa = (Mat_SeqSBAIJ*)mat->A->data;
1047: Mat_SeqBAIJ *Ba = (Mat_SeqBAIJ*)mat->B->data;
1048: PetscInt max = 1,mbs = mat->mbs,tmp;
1049: for (i=0; i<mbs; i++) {
1050: tmp = Aa->i[i+1] - Aa->i[i] + Ba->i[i+1] - Ba->i[i]; /* row length */
1051: if (max < tmp) { max = tmp; }
1052: }
1053: PetscMalloc(max*bs2*(sizeof(PetscInt)+sizeof(PetscScalar)),&mat->rowvalues);
1054: mat->rowindices = (PetscInt*)(mat->rowvalues + max*bs2);
1055: }
1056:
1057: if (row < brstart || row >= brend) SETERRQ(PETSC_ERR_SUP,"Only local rows")
1058: lrow = row - brstart; /* local row index */
1060: pvA = &vworkA; pcA = &cworkA; pvB = &vworkB; pcB = &cworkB;
1061: if (!v) {pvA = 0; pvB = 0;}
1062: if (!idx) {pcA = 0; if (!v) pcB = 0;}
1063: (*mat->A->ops->getrow)(mat->A,lrow,&nzA,pcA,pvA);
1064: (*mat->B->ops->getrow)(mat->B,lrow,&nzB,pcB,pvB);
1065: nztot = nzA + nzB;
1067: cmap = mat->garray;
1068: if (v || idx) {
1069: if (nztot) {
1070: /* Sort by increasing column numbers, assuming A and B already sorted */
1071: PetscInt imark = -1;
1072: if (v) {
1073: *v = v_p = mat->rowvalues;
1074: for (i=0; i<nzB; i++) {
1075: if (cmap[cworkB[i]/bs] < cstart) v_p[i] = vworkB[i];
1076: else break;
1077: }
1078: imark = i;
1079: for (i=0; i<nzA; i++) v_p[imark+i] = vworkA[i];
1080: for (i=imark; i<nzB; i++) v_p[nzA+i] = vworkB[i];
1081: }
1082: if (idx) {
1083: *idx = idx_p = mat->rowindices;
1084: if (imark > -1) {
1085: for (i=0; i<imark; i++) {
1086: idx_p[i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs;
1087: }
1088: } else {
1089: for (i=0; i<nzB; i++) {
1090: if (cmap[cworkB[i]/bs] < cstart)
1091: idx_p[i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs ;
1092: else break;
1093: }
1094: imark = i;
1095: }
1096: for (i=0; i<nzA; i++) idx_p[imark+i] = cstart*bs + cworkA[i];
1097: for (i=imark; i<nzB; i++) idx_p[nzA+i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs ;
1098: }
1099: } else {
1100: if (idx) *idx = 0;
1101: if (v) *v = 0;
1102: }
1103: }
1104: *nz = nztot;
1105: (*mat->A->ops->restorerow)(mat->A,lrow,&nzA,pcA,pvA);
1106: (*mat->B->ops->restorerow)(mat->B,lrow,&nzB,pcB,pvB);
1107: return(0);
1108: }
1112: PetscErrorCode MatRestoreRow_MPISBAIJ(Mat mat,PetscInt row,PetscInt *nz,PetscInt **idx,PetscScalar **v)
1113: {
1114: Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data;
1117: if (!baij->getrowactive) {
1118: SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"MatGetRow() must be called first");
1119: }
1120: baij->getrowactive = PETSC_FALSE;
1121: return(0);
1122: }
1126: PetscErrorCode MatGetRowUpperTriangular_MPISBAIJ(Mat A)
1127: {
1128: Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data;
1129: Mat_SeqSBAIJ *aA = (Mat_SeqSBAIJ*)a->A->data;
1132: aA->getrow_utriangular = PETSC_TRUE;
1133: return(0);
1134: }
1137: PetscErrorCode MatRestoreRowUpperTriangular_MPISBAIJ(Mat A)
1138: {
1139: Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data;
1140: Mat_SeqSBAIJ *aA = (Mat_SeqSBAIJ*)a->A->data;
1143: aA->getrow_utriangular = PETSC_FALSE;
1144: return(0);
1145: }
1149: PetscErrorCode MatRealPart_MPISBAIJ(Mat A)
1150: {
1151: Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data;
1155: MatRealPart(a->A);
1156: MatRealPart(a->B);
1157: return(0);
1158: }
1162: PetscErrorCode MatImaginaryPart_MPISBAIJ(Mat A)
1163: {
1164: Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data;
1168: MatImaginaryPart(a->A);
1169: MatImaginaryPart(a->B);
1170: return(0);
1171: }
1175: PetscErrorCode MatZeroEntries_MPISBAIJ(Mat A)
1176: {
1177: Mat_MPISBAIJ *l = (Mat_MPISBAIJ*)A->data;
1181: MatZeroEntries(l->A);
1182: MatZeroEntries(l->B);
1183: return(0);
1184: }
1188: PetscErrorCode MatGetInfo_MPISBAIJ(Mat matin,MatInfoType flag,MatInfo *info)
1189: {
1190: Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)matin->data;
1191: Mat A = a->A,B = a->B;
1193: PetscReal isend[5],irecv[5];
1196: info->block_size = (PetscReal)matin->rmap.bs;
1197: MatGetInfo(A,MAT_LOCAL,info);
1198: isend[0] = info->nz_used; isend[1] = info->nz_allocated; isend[2] = info->nz_unneeded;
1199: isend[3] = info->memory; isend[4] = info->mallocs;
1200: MatGetInfo(B,MAT_LOCAL,info);
1201: isend[0] += info->nz_used; isend[1] += info->nz_allocated; isend[2] += info->nz_unneeded;
1202: isend[3] += info->memory; isend[4] += info->mallocs;
1203: if (flag == MAT_LOCAL) {
1204: info->nz_used = isend[0];
1205: info->nz_allocated = isend[1];
1206: info->nz_unneeded = isend[2];
1207: info->memory = isend[3];
1208: info->mallocs = isend[4];
1209: } else if (flag == MAT_GLOBAL_MAX) {
1210: MPI_Allreduce(isend,irecv,5,MPIU_REAL,MPI_MAX,matin->comm);
1211: info->nz_used = irecv[0];
1212: info->nz_allocated = irecv[1];
1213: info->nz_unneeded = irecv[2];
1214: info->memory = irecv[3];
1215: info->mallocs = irecv[4];
1216: } else if (flag == MAT_GLOBAL_SUM) {
1217: MPI_Allreduce(isend,irecv,5,MPIU_REAL,MPI_SUM,matin->comm);
1218: info->nz_used = irecv[0];
1219: info->nz_allocated = irecv[1];
1220: info->nz_unneeded = irecv[2];
1221: info->memory = irecv[3];
1222: info->mallocs = irecv[4];
1223: } else {
1224: SETERRQ1(PETSC_ERR_ARG_WRONG,"Unknown MatInfoType argument %d",(int)flag);
1225: }
1226: info->rows_global = (PetscReal)A->rmap.N;
1227: info->columns_global = (PetscReal)A->cmap.N;
1228: info->rows_local = (PetscReal)A->rmap.N;
1229: info->columns_local = (PetscReal)A->cmap.N;
1230: info->fill_ratio_given = 0; /* no parallel LU/ILU/Cholesky */
1231: info->fill_ratio_needed = 0;
1232: info->factor_mallocs = 0;
1233: return(0);
1234: }
1238: PetscErrorCode MatSetOption_MPISBAIJ(Mat A,MatOption op)
1239: {
1240: Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data;
1241: Mat_SeqSBAIJ *aA = (Mat_SeqSBAIJ*)a->A->data;
1245: switch (op) {
1246: case MAT_NO_NEW_NONZERO_LOCATIONS:
1247: case MAT_YES_NEW_NONZERO_LOCATIONS:
1248: case MAT_COLUMNS_UNSORTED:
1249: case MAT_COLUMNS_SORTED:
1250: case MAT_NEW_NONZERO_ALLOCATION_ERR:
1251: case MAT_KEEP_ZEROED_ROWS:
1252: case MAT_NEW_NONZERO_LOCATION_ERR:
1253: MatSetOption(a->A,op);
1254: MatSetOption(a->B,op);
1255: break;
1256: case MAT_ROW_ORIENTED:
1257: a->roworiented = PETSC_TRUE;
1258: MatSetOption(a->A,op);
1259: MatSetOption(a->B,op);
1260: break;
1261: case MAT_ROWS_SORTED:
1262: case MAT_ROWS_UNSORTED:
1263: case MAT_YES_NEW_DIAGONALS:
1264: PetscInfo1(A,"Option %s ignored\n",MatOptions[op]);
1265: break;
1266: case MAT_COLUMN_ORIENTED:
1267: a->roworiented = PETSC_FALSE;
1268: MatSetOption(a->A,op);
1269: MatSetOption(a->B,op);
1270: break;
1271: case MAT_IGNORE_OFF_PROC_ENTRIES:
1272: a->donotstash = PETSC_TRUE;
1273: break;
1274: case MAT_NO_NEW_DIAGONALS:
1275: SETERRQ(PETSC_ERR_SUP,"MAT_NO_NEW_DIAGONALS");
1276: case MAT_USE_HASH_TABLE:
1277: a->ht_flag = PETSC_TRUE;
1278: break;
1279: case MAT_NOT_SYMMETRIC:
1280: case MAT_NOT_STRUCTURALLY_SYMMETRIC:
1281: case MAT_HERMITIAN:
1282: SETERRQ(PETSC_ERR_SUP,"Matrix must be symmetric");
1283: case MAT_SYMMETRIC:
1284: case MAT_STRUCTURALLY_SYMMETRIC:
1285: case MAT_NOT_HERMITIAN:
1286: case MAT_SYMMETRY_ETERNAL:
1287: case MAT_NOT_SYMMETRY_ETERNAL:
1288: PetscInfo1(A,"Option %s ignored\n",MatOptions[op]);
1289: break;
1290: case MAT_IGNORE_LOWER_TRIANGULAR:
1291: aA->ignore_ltriangular = PETSC_TRUE;
1292: break;
1293: case MAT_ERROR_LOWER_TRIANGULAR:
1294: aA->ignore_ltriangular = PETSC_FALSE;
1295: break;
1296: case MAT_GETROW_UPPERTRIANGULAR:
1297: aA->getrow_utriangular = PETSC_TRUE;
1298: break;
1299: default:
1300: SETERRQ1(PETSC_ERR_SUP,"unknown option %d",op);
1301: }
1302: return(0);
1303: }
1307: PetscErrorCode MatTranspose_MPISBAIJ(Mat A,Mat *B)
1308: {
1311: MatDuplicate(A,MAT_COPY_VALUES,B);
1312: return(0);
1313: }
1317: PetscErrorCode MatDiagonalScale_MPISBAIJ(Mat mat,Vec ll,Vec rr)
1318: {
1319: Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data;
1320: Mat a=baij->A, b=baij->B;
1322: PetscInt nv,m,n;
1323: PetscTruth flg;
1326: if (ll != rr){
1327: VecEqual(ll,rr,&flg);
1328: if (!flg)
1329: SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"For symmetric format, left and right scaling vectors must be same\n");
1330: }
1331: if (!ll) return(0);
1333: MatGetLocalSize(mat,&m,&n);
1334: if (m != n) SETERRQ2(PETSC_ERR_ARG_SIZ,"For symmetric format, local size %d %d must be same",m,n);
1335:
1336: VecGetLocalSize(rr,&nv);
1337: if (nv!=n) SETERRQ(PETSC_ERR_ARG_SIZ,"Left and right vector non-conforming local size");
1339: VecScatterBegin(baij->Mvctx,rr,baij->lvec,INSERT_VALUES,SCATTER_FORWARD);
1340:
1341: /* left diagonalscale the off-diagonal part */
1342: (*b->ops->diagonalscale)(b,ll,PETSC_NULL);
1343:
1344: /* scale the diagonal part */
1345: (*a->ops->diagonalscale)(a,ll,rr);
1347: /* right diagonalscale the off-diagonal part */
1348: VecScatterEnd(baij->Mvctx,rr,baij->lvec,INSERT_VALUES,SCATTER_FORWARD);
1349: (*b->ops->diagonalscale)(b,PETSC_NULL,baij->lvec);
1350: return(0);
1351: }
1355: PetscErrorCode MatSetUnfactored_MPISBAIJ(Mat A)
1356: {
1357: Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data;
1361: MatSetUnfactored(a->A);
1362: return(0);
1363: }
1365: static PetscErrorCode MatDuplicate_MPISBAIJ(Mat,MatDuplicateOption,Mat *);
1369: PetscErrorCode MatEqual_MPISBAIJ(Mat A,Mat B,PetscTruth *flag)
1370: {
1371: Mat_MPISBAIJ *matB = (Mat_MPISBAIJ*)B->data,*matA = (Mat_MPISBAIJ*)A->data;
1372: Mat a,b,c,d;
1373: PetscTruth flg;
1377: a = matA->A; b = matA->B;
1378: c = matB->A; d = matB->B;
1380: MatEqual(a,c,&flg);
1381: if (flg) {
1382: MatEqual(b,d,&flg);
1383: }
1384: MPI_Allreduce(&flg,flag,1,MPI_INT,MPI_LAND,A->comm);
1385: return(0);
1386: }
1390: PetscErrorCode MatCopy_MPISBAIJ(Mat A,Mat B,MatStructure str)
1391: {
1393: Mat_MPISBAIJ *a = (Mat_MPISBAIJ *)A->data;
1394: Mat_MPISBAIJ *b = (Mat_MPISBAIJ *)B->data;
1397: /* If the two matrices don't have the same copy implementation, they aren't compatible for fast copy. */
1398: if ((str != SAME_NONZERO_PATTERN) || (A->ops->copy != B->ops->copy)) {
1399: MatGetRowUpperTriangular(A);
1400: MatCopy_Basic(A,B,str);
1401: MatRestoreRowUpperTriangular(A);
1402: } else {
1403: MatCopy(a->A,b->A,str);
1404: MatCopy(a->B,b->B,str);
1405: }
1406: return(0);
1407: }
1411: PetscErrorCode MatSetUpPreallocation_MPISBAIJ(Mat A)
1412: {
1416: MatMPISBAIJSetPreallocation(A,PetscMax(A->rmap.bs,1),PETSC_DEFAULT,0,PETSC_DEFAULT,0);
1417: return(0);
1418: }
1420: #include petscblaslapack.h
1423: PetscErrorCode MatAXPY_MPISBAIJ(Mat Y,PetscScalar a,Mat X,MatStructure str)
1424: {
1426: Mat_MPISBAIJ *xx=(Mat_MPISBAIJ *)X->data,*yy=(Mat_MPISBAIJ *)Y->data;
1427: PetscBLASInt bnz,one=1;
1428: Mat_SeqSBAIJ *xa,*ya;
1429: Mat_SeqBAIJ *xb,*yb;
1432: if (str == SAME_NONZERO_PATTERN) {
1433: PetscScalar alpha = a;
1434: xa = (Mat_SeqSBAIJ *)xx->A->data;
1435: ya = (Mat_SeqSBAIJ *)yy->A->data;
1436: bnz = (PetscBLASInt)xa->nz;
1437: BLASaxpy_(&bnz,&alpha,xa->a,&one,ya->a,&one);
1438: xb = (Mat_SeqBAIJ *)xx->B->data;
1439: yb = (Mat_SeqBAIJ *)yy->B->data;
1440: bnz = (PetscBLASInt)xb->nz;
1441: BLASaxpy_(&bnz,&alpha,xb->a,&one,yb->a,&one);
1442: } else {
1443: MatGetRowUpperTriangular(X);
1444: MatAXPY_Basic(Y,a,X,str);
1445: MatRestoreRowUpperTriangular(X);
1446: }
1447: return(0);
1448: }
1452: PetscErrorCode MatGetSubMatrices_MPISBAIJ(Mat A,PetscInt n,const IS irow[],const IS icol[],MatReuse scall,Mat *B[])
1453: {
1455: PetscInt i;
1456: PetscTruth flg;
1459: for (i=0; i<n; i++) {
1460: ISEqual(irow[i],icol[i],&flg);
1461: if (!flg) {
1462: SETERRQ(PETSC_ERR_SUP,"Can only get symmetric submatrix for MPISBAIJ matrices");
1463: }
1464: }
1465: MatGetSubMatrices_MPIBAIJ(A,n,irow,icol,scall,B);
1466: return(0);
1467: }
1468:
1470: /* -------------------------------------------------------------------*/
1471: static struct _MatOps MatOps_Values = {
1472: MatSetValues_MPISBAIJ,
1473: MatGetRow_MPISBAIJ,
1474: MatRestoreRow_MPISBAIJ,
1475: MatMult_MPISBAIJ,
1476: /* 4*/ MatMultAdd_MPISBAIJ,
1477: MatMult_MPISBAIJ, /* transpose versions are same as non-transpose */
1478: MatMultAdd_MPISBAIJ,
1479: 0,
1480: 0,
1481: 0,
1482: /*10*/ 0,
1483: 0,
1484: 0,
1485: MatRelax_MPISBAIJ,
1486: MatTranspose_MPISBAIJ,
1487: /*15*/ MatGetInfo_MPISBAIJ,
1488: MatEqual_MPISBAIJ,
1489: MatGetDiagonal_MPISBAIJ,
1490: MatDiagonalScale_MPISBAIJ,
1491: MatNorm_MPISBAIJ,
1492: /*20*/ MatAssemblyBegin_MPISBAIJ,
1493: MatAssemblyEnd_MPISBAIJ,
1494: 0,
1495: MatSetOption_MPISBAIJ,
1496: MatZeroEntries_MPISBAIJ,
1497: /*25*/ 0,
1498: 0,
1499: 0,
1500: 0,
1501: 0,
1502: /*30*/ MatSetUpPreallocation_MPISBAIJ,
1503: 0,
1504: 0,
1505: 0,
1506: 0,
1507: /*35*/ MatDuplicate_MPISBAIJ,
1508: 0,
1509: 0,
1510: 0,
1511: 0,
1512: /*40*/ MatAXPY_MPISBAIJ,
1513: MatGetSubMatrices_MPISBAIJ,
1514: MatIncreaseOverlap_MPISBAIJ,
1515: MatGetValues_MPISBAIJ,
1516: MatCopy_MPISBAIJ,
1517: /*45*/ 0,
1518: MatScale_MPISBAIJ,
1519: 0,
1520: 0,
1521: 0,
1522: /*50*/ 0,
1523: 0,
1524: 0,
1525: 0,
1526: 0,
1527: /*55*/ 0,
1528: 0,
1529: MatSetUnfactored_MPISBAIJ,
1530: 0,
1531: MatSetValuesBlocked_MPISBAIJ,
1532: /*60*/ 0,
1533: 0,
1534: 0,
1535: 0,
1536: 0,
1537: /*65*/ 0,
1538: 0,
1539: 0,
1540: 0,
1541: 0,
1542: /*70*/ MatGetRowMaxAbs_MPISBAIJ,
1543: 0,
1544: 0,
1545: 0,
1546: 0,
1547: /*75*/ 0,
1548: 0,
1549: 0,
1550: 0,
1551: 0,
1552: /*80*/ 0,
1553: 0,
1554: 0,
1555: 0,
1556: MatLoad_MPISBAIJ,
1557: /*85*/ 0,
1558: 0,
1559: 0,
1560: 0,
1561: 0,
1562: /*90*/ 0,
1563: 0,
1564: 0,
1565: 0,
1566: 0,
1567: /*95*/ 0,
1568: 0,
1569: 0,
1570: 0,
1571: 0,
1572: /*100*/0,
1573: 0,
1574: 0,
1575: 0,
1576: 0,
1577: /*105*/0,
1578: MatRealPart_MPISBAIJ,
1579: MatImaginaryPart_MPISBAIJ,
1580: MatGetRowUpperTriangular_MPISBAIJ,
1581: MatRestoreRowUpperTriangular_MPISBAIJ
1582: };
1588: PetscErrorCode MatGetDiagonalBlock_MPISBAIJ(Mat A,PetscTruth *iscopy,MatReuse reuse,Mat *a)
1589: {
1591: *a = ((Mat_MPISBAIJ *)A->data)->A;
1592: *iscopy = PETSC_FALSE;
1593: return(0);
1594: }
1600: PetscErrorCode MatMPISBAIJSetPreallocation_MPISBAIJ(Mat B,PetscInt bs,PetscInt d_nz,PetscInt *d_nnz,PetscInt o_nz,PetscInt *o_nnz)
1601: {
1602: Mat_MPISBAIJ *b;
1604: PetscInt i,mbs,Mbs;
1607: PetscOptionsBegin(B->comm,B->prefix,"Options for MPISBAIJ matrix","Mat");
1608: PetscOptionsInt("-mat_block_size","Set the blocksize used to store the matrix","MatMPIBAIJSetPreallocation",bs,&bs,PETSC_NULL);
1609: PetscOptionsEnd();
1611: if (bs < 1) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Invalid block size specified, must be positive");
1612: if (d_nz == PETSC_DECIDE || d_nz == PETSC_DEFAULT) d_nz = 3;
1613: if (o_nz == PETSC_DECIDE || o_nz == PETSC_DEFAULT) o_nz = 1;
1614: if (d_nz < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"d_nz cannot be less than 0: value %D",d_nz);
1615: if (o_nz < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"o_nz cannot be less than 0: value %D",o_nz);
1617: B->rmap.bs = B->cmap.bs = bs;
1618: PetscMapSetUp(&B->rmap);
1619: PetscMapSetUp(&B->cmap);
1621: if (d_nnz) {
1622: for (i=0; i<B->rmap.n/bs; i++) {
1623: if (d_nnz[i] < 0) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"d_nnz cannot be less than -1: local row %D value %D",i,d_nnz[i]);
1624: }
1625: }
1626: if (o_nnz) {
1627: for (i=0; i<B->rmap.n/bs; i++) {
1628: if (o_nnz[i] < 0) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"o_nnz cannot be less than -1: local row %D value %D",i,o_nnz[i]);
1629: }
1630: }
1631: B->preallocated = PETSC_TRUE;
1633: b = (Mat_MPISBAIJ*)B->data;
1634: mbs = B->rmap.n/bs;
1635: Mbs = B->rmap.N/bs;
1636: if (mbs*bs != B->rmap.n) {
1637: SETERRQ2(PETSC_ERR_ARG_SIZ,"No of local rows %D must be divisible by blocksize %D",B->rmap.N,bs);
1638: }
1640: B->rmap.bs = bs;
1641: b->bs2 = bs*bs;
1642: b->mbs = mbs;
1643: b->nbs = mbs;
1644: b->Mbs = Mbs;
1645: b->Nbs = Mbs;
1647: for (i=0; i<=b->size; i++) {
1648: b->rangebs[i] = B->rmap.range[i]/bs;
1649: }
1650: b->rstartbs = B->rmap.rstart/bs;
1651: b->rendbs = B->rmap.rend/bs;
1652:
1653: b->cstartbs = B->cmap.rstart/bs;
1654: b->cendbs = B->cmap.rend/bs;
1655:
1656: MatCreate(PETSC_COMM_SELF,&b->A);
1657: MatSetSizes(b->A,B->rmap.n,B->cmap.n,B->rmap.n,B->cmap.n);
1658: MatSetType(b->A,MATSEQSBAIJ);
1659: MatSeqSBAIJSetPreallocation(b->A,bs,d_nz,d_nnz);
1660: PetscLogObjectParent(B,b->A);
1662: MatCreate(PETSC_COMM_SELF,&b->B);
1663: MatSetSizes(b->B,B->rmap.n,B->cmap.N,B->rmap.n,B->cmap.N);
1664: MatSetType(b->B,MATSEQBAIJ);
1665: MatSeqBAIJSetPreallocation(b->B,bs,o_nz,o_nnz);
1666: PetscLogObjectParent(B,b->B);
1668: /* build cache for off array entries formed */
1669: MatStashCreate_Private(B->comm,bs,&B->bstash);
1671: return(0);
1672: }
1675: /*MC
1676: MATMPISBAIJ - MATMPISBAIJ = "mpisbaij" - A matrix type to be used for distributed symmetric sparse block matrices,
1677: based on block compressed sparse row format. Only the upper triangular portion of the matrix is stored.
1679: Options Database Keys:
1680: . -mat_type mpisbaij - sets the matrix type to "mpisbaij" during a call to MatSetFromOptions()
1682: Level: beginner
1684: .seealso: MatCreateMPISBAIJ
1685: M*/
1690: PetscErrorCode MatCreate_MPISBAIJ(Mat B)
1691: {
1692: Mat_MPISBAIJ *b;
1694: PetscTruth flg;
1698: PetscNew(Mat_MPISBAIJ,&b);
1699: B->data = (void*)b;
1700: PetscMemcpy(B->ops,&MatOps_Values,sizeof(struct _MatOps));
1702: B->ops->destroy = MatDestroy_MPISBAIJ;
1703: B->ops->view = MatView_MPISBAIJ;
1704: B->mapping = 0;
1705: B->factor = 0;
1706: B->assembled = PETSC_FALSE;
1708: B->insertmode = NOT_SET_VALUES;
1709: MPI_Comm_rank(B->comm,&b->rank);
1710: MPI_Comm_size(B->comm,&b->size);
1712: /* build local table of row and column ownerships */
1713: PetscMalloc((b->size+2)*sizeof(PetscInt),&b->rangebs);
1715: /* build cache for off array entries formed */
1716: MatStashCreate_Private(B->comm,1,&B->stash);
1717: b->donotstash = PETSC_FALSE;
1718: b->colmap = PETSC_NULL;
1719: b->garray = PETSC_NULL;
1720: b->roworiented = PETSC_TRUE;
1722: #if defined(PETSC_USE_MAT_SINGLE)
1723: /* stuff for MatSetValues_XXX in single precision */
1724: b->setvalueslen = 0;
1725: b->setvaluescopy = PETSC_NULL;
1726: #endif
1728: /* stuff used in block assembly */
1729: b->barray = 0;
1731: /* stuff used for matrix vector multiply */
1732: b->lvec = 0;
1733: b->Mvctx = 0;
1734: b->slvec0 = 0;
1735: b->slvec0b = 0;
1736: b->slvec1 = 0;
1737: b->slvec1a = 0;
1738: b->slvec1b = 0;
1739: b->sMvctx = 0;
1741: /* stuff for MatGetRow() */
1742: b->rowindices = 0;
1743: b->rowvalues = 0;
1744: b->getrowactive = PETSC_FALSE;
1746: /* hash table stuff */
1747: b->ht = 0;
1748: b->hd = 0;
1749: b->ht_size = 0;
1750: b->ht_flag = PETSC_FALSE;
1751: b->ht_fact = 0;
1752: b->ht_total_ct = 0;
1753: b->ht_insert_ct = 0;
1755: b->in_loc = 0;
1756: b->v_loc = 0;
1757: b->n_loc = 0;
1758: PetscOptionsBegin(B->comm,PETSC_NULL,"Options for loading MPISBAIJ matrix 1","Mat");
1759: PetscOptionsTruth("-mat_use_hash_table","Use hash table to save memory in constructing matrix","MatSetOption",PETSC_FALSE,&flg,PETSC_NULL);
1760: if (flg) {
1761: PetscReal fact = 1.39;
1762: MatSetOption(B,MAT_USE_HASH_TABLE);
1763: PetscOptionsReal("-mat_use_hash_table","Use hash table factor","MatMPIBAIJSetHashTableFactor",fact,&fact,PETSC_NULL);
1764: if (fact <= 1.0) fact = 1.39;
1765: MatMPIBAIJSetHashTableFactor(B,fact);
1766: PetscInfo1(0,"Hash table Factor used %5.2f\n",fact);
1767: }
1768: PetscOptionsEnd();
1770: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatStoreValues_C",
1771: "MatStoreValues_MPISBAIJ",
1772: MatStoreValues_MPISBAIJ);
1773: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatRetrieveValues_C",
1774: "MatRetrieveValues_MPISBAIJ",
1775: MatRetrieveValues_MPISBAIJ);
1776: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatGetDiagonalBlock_C",
1777: "MatGetDiagonalBlock_MPISBAIJ",
1778: MatGetDiagonalBlock_MPISBAIJ);
1779: PetscObjectComposeFunctionDynamic((PetscObject)B,"MatMPISBAIJSetPreallocation_C",
1780: "MatMPISBAIJSetPreallocation_MPISBAIJ",
1781: MatMPISBAIJSetPreallocation_MPISBAIJ);
1782: B->symmetric = PETSC_TRUE;
1783: B->structurally_symmetric = PETSC_TRUE;
1784: B->symmetric_set = PETSC_TRUE;
1785: B->structurally_symmetric_set = PETSC_TRUE;
1786: PetscObjectChangeTypeName((PetscObject)B,MATMPISBAIJ);
1787: return(0);
1788: }
1791: /*MC
1792: MATSBAIJ - MATSBAIJ = "sbaij" - A matrix type to be used for symmetric block sparse matrices.
1794: This matrix type is identical to MATSEQSBAIJ when constructed with a single process communicator,
1795: and MATMPISBAIJ otherwise.
1797: Options Database Keys:
1798: . -mat_type sbaij - sets the matrix type to "sbaij" during a call to MatSetFromOptions()
1800: Level: beginner
1802: .seealso: MatCreateMPISBAIJ,MATSEQSBAIJ,MATMPISBAIJ
1803: M*/
1808: PetscErrorCode MatCreate_SBAIJ(Mat A)
1809: {
1811: PetscMPIInt size;
1814: MPI_Comm_size(A->comm,&size);
1815: if (size == 1) {
1816: MatSetType(A,MATSEQSBAIJ);
1817: } else {
1818: MatSetType(A,MATMPISBAIJ);
1819: }
1820: return(0);
1821: }
1826: /*@C
1827: MatMPISBAIJSetPreallocation - For good matrix assembly performance
1828: the user should preallocate the matrix storage by setting the parameters
1829: d_nz (or d_nnz) and o_nz (or o_nnz). By setting these parameters accurately,
1830: performance can be increased by more than a factor of 50.
1832: Collective on Mat
1834: Input Parameters:
1835: + A - the matrix
1836: . bs - size of blockk
1837: . d_nz - number of block nonzeros per block row in diagonal portion of local
1838: submatrix (same for all local rows)
1839: . d_nnz - array containing the number of block nonzeros in the various block rows
1840: in the upper triangular and diagonal part of the in diagonal portion of the local
1841: (possibly different for each block row) or PETSC_NULL. You must leave room
1842: for the diagonal entry even if it is zero.
1843: . o_nz - number of block nonzeros per block row in the off-diagonal portion of local
1844: submatrix (same for all local rows).
1845: - o_nnz - array containing the number of nonzeros in the various block rows of the
1846: off-diagonal portion of the local submatrix (possibly different for
1847: each block row) or PETSC_NULL.
1850: Options Database Keys:
1851: . -mat_no_unroll - uses code that does not unroll the loops in the
1852: block calculations (much slower)
1853: . -mat_block_size - size of the blocks to use
1855: Notes:
1857: If PETSC_DECIDE or PETSC_DETERMINE is used for a particular argument on one processor
1858: than it must be used on all processors that share the object for that argument.
1860: If the *_nnz parameter is given then the *_nz parameter is ignored
1862: Storage Information:
1863: For a square global matrix we define each processor's diagonal portion
1864: to be its local rows and the corresponding columns (a square submatrix);
1865: each processor's off-diagonal portion encompasses the remainder of the
1866: local matrix (a rectangular submatrix).
1868: The user can specify preallocated storage for the diagonal part of
1869: the local submatrix with either d_nz or d_nnz (not both). Set
1870: d_nz=PETSC_DEFAULT and d_nnz=PETSC_NULL for PETSc to control dynamic
1871: memory allocation. Likewise, specify preallocated storage for the
1872: off-diagonal part of the local submatrix with o_nz or o_nnz (not both).
1874: Consider a processor that owns rows 3, 4 and 5 of a parallel matrix. In
1875: the figure below we depict these three local rows and all columns (0-11).
1877: .vb
1878: 0 1 2 3 4 5 6 7 8 9 10 11
1879: -------------------
1880: row 3 | o o o d d d o o o o o o
1881: row 4 | o o o d d d o o o o o o
1882: row 5 | o o o d d d o o o o o o
1883: -------------------
1884: .ve
1885:
1886: Thus, any entries in the d locations are stored in the d (diagonal)
1887: submatrix, and any entries in the o locations are stored in the
1888: o (off-diagonal) submatrix. Note that the d matrix is stored in
1889: MatSeqSBAIJ format and the o submatrix in MATSEQBAIJ format.
1891: Now d_nz should indicate the number of block nonzeros per row in the upper triangular
1892: plus the diagonal part of the d matrix,
1893: and o_nz should indicate the number of block nonzeros per row in the o matrix.
1894: In general, for PDE problems in which most nonzeros are near the diagonal,
1895: one expects d_nz >> o_nz. For large problems you MUST preallocate memory
1896: or you will get TERRIBLE performance; see the users' manual chapter on
1897: matrices.
1899: Level: intermediate
1901: .keywords: matrix, block, aij, compressed row, sparse, parallel
1903: .seealso: MatCreate(), MatCreateSeqSBAIJ(), MatSetValues(), MatCreateMPIBAIJ()
1904: @*/
1905: PetscErrorCode MatMPISBAIJSetPreallocation(Mat B,PetscInt bs,PetscInt d_nz,const PetscInt d_nnz[],PetscInt o_nz,const PetscInt o_nnz[])
1906: {
1907: PetscErrorCode ierr,(*f)(Mat,PetscInt,PetscInt,const PetscInt[],PetscInt,const PetscInt[]);
1910: PetscObjectQueryFunction((PetscObject)B,"MatMPISBAIJSetPreallocation_C",(void (**)(void))&f);
1911: if (f) {
1912: (*f)(B,bs,d_nz,d_nnz,o_nz,o_nnz);
1913: }
1914: return(0);
1915: }
1919: /*@C
1920: MatCreateMPISBAIJ - Creates a sparse parallel matrix in symmetric block AIJ format
1921: (block compressed row). For good matrix assembly performance
1922: the user should preallocate the matrix storage by setting the parameters
1923: d_nz (or d_nnz) and o_nz (or o_nnz). By setting these parameters accurately,
1924: performance can be increased by more than a factor of 50.
1926: Collective on MPI_Comm
1928: Input Parameters:
1929: + comm - MPI communicator
1930: . bs - size of blockk
1931: . m - number of local rows (or PETSC_DECIDE to have calculated if M is given)
1932: This value should be the same as the local size used in creating the
1933: y vector for the matrix-vector product y = Ax.
1934: . n - number of local columns (or PETSC_DECIDE to have calculated if N is given)
1935: This value should be the same as the local size used in creating the
1936: x vector for the matrix-vector product y = Ax.
1937: . M - number of global rows (or PETSC_DETERMINE to have calculated if m is given)
1938: . N - number of global columns (or PETSC_DETERMINE to have calculated if n is given)
1939: . d_nz - number of block nonzeros per block row in diagonal portion of local
1940: submatrix (same for all local rows)
1941: . d_nnz - array containing the number of block nonzeros in the various block rows
1942: in the upper triangular portion of the in diagonal portion of the local
1943: (possibly different for each block block row) or PETSC_NULL.
1944: You must leave room for the diagonal entry even if it is zero.
1945: . o_nz - number of block nonzeros per block row in the off-diagonal portion of local
1946: submatrix (same for all local rows).
1947: - o_nnz - array containing the number of nonzeros in the various block rows of the
1948: off-diagonal portion of the local submatrix (possibly different for
1949: each block row) or PETSC_NULL.
1951: Output Parameter:
1952: . A - the matrix
1954: Options Database Keys:
1955: . -mat_no_unroll - uses code that does not unroll the loops in the
1956: block calculations (much slower)
1957: . -mat_block_size - size of the blocks to use
1958: . -mat_mpi - use the parallel matrix data structures even on one processor
1959: (defaults to using SeqBAIJ format on one processor)
1961: Notes:
1962: The number of rows and columns must be divisible by blocksize.
1963: This matrix type does not support complex Hermitian operation.
1965: The user MUST specify either the local or global matrix dimensions
1966: (possibly both).
1968: If PETSC_DECIDE or PETSC_DETERMINE is used for a particular argument on one processor
1969: than it must be used on all processors that share the object for that argument.
1971: If the *_nnz parameter is given then the *_nz parameter is ignored
1973: Storage Information:
1974: For a square global matrix we define each processor's diagonal portion
1975: to be its local rows and the corresponding columns (a square submatrix);
1976: each processor's off-diagonal portion encompasses the remainder of the
1977: local matrix (a rectangular submatrix).
1979: The user can specify preallocated storage for the diagonal part of
1980: the local submatrix with either d_nz or d_nnz (not both). Set
1981: d_nz=PETSC_DEFAULT and d_nnz=PETSC_NULL for PETSc to control dynamic
1982: memory allocation. Likewise, specify preallocated storage for the
1983: off-diagonal part of the local submatrix with o_nz or o_nnz (not both).
1985: Consider a processor that owns rows 3, 4 and 5 of a parallel matrix. In
1986: the figure below we depict these three local rows and all columns (0-11).
1988: .vb
1989: 0 1 2 3 4 5 6 7 8 9 10 11
1990: -------------------
1991: row 3 | o o o d d d o o o o o o
1992: row 4 | o o o d d d o o o o o o
1993: row 5 | o o o d d d o o o o o o
1994: -------------------
1995: .ve
1996:
1997: Thus, any entries in the d locations are stored in the d (diagonal)
1998: submatrix, and any entries in the o locations are stored in the
1999: o (off-diagonal) submatrix. Note that the d matrix is stored in
2000: MatSeqSBAIJ format and the o submatrix in MATSEQBAIJ format.
2002: Now d_nz should indicate the number of block nonzeros per row in the upper triangular
2003: plus the diagonal part of the d matrix,
2004: and o_nz should indicate the number of block nonzeros per row in the o matrix.
2005: In general, for PDE problems in which most nonzeros are near the diagonal,
2006: one expects d_nz >> o_nz. For large problems you MUST preallocate memory
2007: or you will get TERRIBLE performance; see the users' manual chapter on
2008: matrices.
2010: Level: intermediate
2012: .keywords: matrix, block, aij, compressed row, sparse, parallel
2014: .seealso: MatCreate(), MatCreateSeqSBAIJ(), MatSetValues(), MatCreateMPIBAIJ()
2015: @*/
2017: PetscErrorCode MatCreateMPISBAIJ(MPI_Comm comm,PetscInt bs,PetscInt m,PetscInt n,PetscInt M,PetscInt N,PetscInt d_nz,const PetscInt d_nnz[],PetscInt o_nz,const PetscInt o_nnz[],Mat *A)
2018: {
2020: PetscMPIInt size;
2023: MatCreate(comm,A);
2024: MatSetSizes(*A,m,n,M,N);
2025: MPI_Comm_size(comm,&size);
2026: if (size > 1) {
2027: MatSetType(*A,MATMPISBAIJ);
2028: MatMPISBAIJSetPreallocation(*A,bs,d_nz,d_nnz,o_nz,o_nnz);
2029: } else {
2030: MatSetType(*A,MATSEQSBAIJ);
2031: MatSeqSBAIJSetPreallocation(*A,bs,d_nz,d_nnz);
2032: }
2033: return(0);
2034: }
2039: static PetscErrorCode MatDuplicate_MPISBAIJ(Mat matin,MatDuplicateOption cpvalues,Mat *newmat)
2040: {
2041: Mat mat;
2042: Mat_MPISBAIJ *a,*oldmat = (Mat_MPISBAIJ*)matin->data;
2044: PetscInt len=0,nt,bs=matin->rmap.bs,mbs=oldmat->mbs;
2045: PetscScalar *array;
2048: *newmat = 0;
2049: MatCreate(matin->comm,&mat);
2050: MatSetSizes(mat,matin->rmap.n,matin->cmap.n,matin->rmap.N,matin->cmap.N);
2051: MatSetType(mat,matin->type_name);
2052: PetscMemcpy(mat->ops,matin->ops,sizeof(struct _MatOps));
2053: PetscMapCopy(matin->comm,&matin->rmap,&mat->rmap);
2054: PetscMapCopy(matin->comm,&matin->cmap,&mat->cmap);
2055:
2056: mat->factor = matin->factor;
2057: mat->preallocated = PETSC_TRUE;
2058: mat->assembled = PETSC_TRUE;
2059: mat->insertmode = NOT_SET_VALUES;
2061: a = (Mat_MPISBAIJ*)mat->data;
2062: a->bs2 = oldmat->bs2;
2063: a->mbs = oldmat->mbs;
2064: a->nbs = oldmat->nbs;
2065: a->Mbs = oldmat->Mbs;
2066: a->Nbs = oldmat->Nbs;
2069: a->size = oldmat->size;
2070: a->rank = oldmat->rank;
2071: a->donotstash = oldmat->donotstash;
2072: a->roworiented = oldmat->roworiented;
2073: a->rowindices = 0;
2074: a->rowvalues = 0;
2075: a->getrowactive = PETSC_FALSE;
2076: a->barray = 0;
2077: a->rstartbs = oldmat->rstartbs;
2078: a->rendbs = oldmat->rendbs;
2079: a->cstartbs = oldmat->cstartbs;
2080: a->cendbs = oldmat->cendbs;
2082: /* hash table stuff */
2083: a->ht = 0;
2084: a->hd = 0;
2085: a->ht_size = 0;
2086: a->ht_flag = oldmat->ht_flag;
2087: a->ht_fact = oldmat->ht_fact;
2088: a->ht_total_ct = 0;
2089: a->ht_insert_ct = 0;
2090:
2091: PetscMemcpy(a->rangebs,oldmat->rangebs,(a->size+2)*sizeof(PetscInt));
2092: MatStashCreate_Private(matin->comm,1,&mat->stash);
2093: MatStashCreate_Private(matin->comm,matin->rmap.bs,&mat->bstash);
2094: if (oldmat->colmap) {
2095: #if defined (PETSC_USE_CTABLE)
2096: PetscTableCreateCopy(oldmat->colmap,&a->colmap);
2097: #else
2098: PetscMalloc((a->Nbs)*sizeof(PetscInt),&a->colmap);
2099: PetscLogObjectMemory(mat,(a->Nbs)*sizeof(PetscInt));
2100: PetscMemcpy(a->colmap,oldmat->colmap,(a->Nbs)*sizeof(PetscInt));
2101: #endif
2102: } else a->colmap = 0;
2104: if (oldmat->garray && (len = ((Mat_SeqBAIJ*)(oldmat->B->data))->nbs)) {
2105: PetscMalloc(len*sizeof(PetscInt),&a->garray);
2106: PetscLogObjectMemory(mat,len*sizeof(PetscInt));
2107: PetscMemcpy(a->garray,oldmat->garray,len*sizeof(PetscInt));
2108: } else a->garray = 0;
2109:
2110: VecDuplicate(oldmat->lvec,&a->lvec);
2111: PetscLogObjectParent(mat,a->lvec);
2112: VecScatterCopy(oldmat->Mvctx,&a->Mvctx);
2113: PetscLogObjectParent(mat,a->Mvctx);
2115: VecDuplicate(oldmat->slvec0,&a->slvec0);
2116: PetscLogObjectParent(mat,a->slvec0);
2117: VecDuplicate(oldmat->slvec1,&a->slvec1);
2118: PetscLogObjectParent(mat,a->slvec1);
2120: VecGetLocalSize(a->slvec1,&nt);
2121: VecGetArray(a->slvec1,&array);
2122: VecCreateSeqWithArray(PETSC_COMM_SELF,bs*mbs,array,&a->slvec1a);
2123: VecCreateSeqWithArray(PETSC_COMM_SELF,nt-bs*mbs,array+bs*mbs,&a->slvec1b);
2124: VecRestoreArray(a->slvec1,&array);
2125: VecGetArray(a->slvec0,&array);
2126: VecCreateSeqWithArray(PETSC_COMM_SELF,nt-bs*mbs,array+bs*mbs,&a->slvec0b);
2127: VecRestoreArray(a->slvec0,&array);
2128: PetscLogObjectParent(mat,a->slvec0);
2129: PetscLogObjectParent(mat,a->slvec1);
2130: PetscLogObjectParent(mat,a->slvec0b);
2131: PetscLogObjectParent(mat,a->slvec1a);
2132: PetscLogObjectParent(mat,a->slvec1b);
2134: /* VecScatterCopy(oldmat->sMvctx,&a->sMvctx); - not written yet, replaced by the lazy trick: */
2135: PetscObjectReference((PetscObject)oldmat->sMvctx);
2136: a->sMvctx = oldmat->sMvctx;
2137: PetscLogObjectParent(mat,a->sMvctx);
2139: MatDuplicate(oldmat->A,cpvalues,&a->A);
2140: PetscLogObjectParent(mat,a->A);
2141: MatDuplicate(oldmat->B,cpvalues,&a->B);
2142: PetscLogObjectParent(mat,a->B);
2143: PetscFListDuplicate(mat->qlist,&matin->qlist);
2144: *newmat = mat;
2145: return(0);
2146: }
2148: #include petscsys.h
2152: PetscErrorCode MatLoad_MPISBAIJ(PetscViewer viewer, MatType type,Mat *newmat)
2153: {
2154: Mat A;
2156: PetscInt i,nz,j,rstart,rend;
2157: PetscScalar *vals,*buf;
2158: MPI_Comm comm = ((PetscObject)viewer)->comm;
2159: MPI_Status status;
2160: PetscMPIInt rank,size,tag = ((PetscObject)viewer)->tag,*sndcounts = 0,*browners,maxnz,*rowners,*locrowlens;
2161: PetscInt header[4],*rowlengths = 0,M,N,m,*cols;
2162: PetscInt *procsnz = 0,jj,*mycols,*ibuf;
2163: PetscInt bs=1,Mbs,mbs,extra_rows;
2164: PetscInt *dlens,*odlens,*mask,*masked1,*masked2,rowcount,odcount;
2165: PetscInt dcount,kmax,k,nzcount,tmp;
2166: int fd;
2167:
2169: PetscOptionsBegin(comm,PETSC_NULL,"Options for loading MPISBAIJ matrix 2","Mat");
2170: PetscOptionsInt("-matload_block_size","Set the blocksize used to store the matrix","MatLoad",bs,&bs,PETSC_NULL);
2171: PetscOptionsEnd();
2173: MPI_Comm_size(comm,&size);
2174: MPI_Comm_rank(comm,&rank);
2175: if (!rank) {
2176: PetscViewerBinaryGetDescriptor(viewer,&fd);
2177: PetscBinaryRead(fd,(char *)header,4,PETSC_INT);
2178: if (header[0] != MAT_FILE_COOKIE) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"not matrix object");
2179: if (header[3] < 0) {
2180: SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"Matrix stored in special format, cannot load as MPISBAIJ");
2181: }
2182: }
2184: MPI_Bcast(header+1,3,MPIU_INT,0,comm);
2185: M = header[1]; N = header[2];
2187: if (M != N) SETERRQ(PETSC_ERR_SUP,"Can only do square matrices");
2189: /*
2190: This code adds extra rows to make sure the number of rows is
2191: divisible by the blocksize
2192: */
2193: Mbs = M/bs;
2194: extra_rows = bs - M + bs*(Mbs);
2195: if (extra_rows == bs) extra_rows = 0;
2196: else Mbs++;
2197: if (extra_rows &&!rank) {
2198: PetscInfo(0,"Padding loaded matrix to match blocksize\n");
2199: }
2201: /* determine ownership of all rows */
2202: mbs = Mbs/size + ((Mbs % size) > rank);
2203: m = mbs*bs;
2204: PetscMalloc(2*(size+2)*sizeof(PetscMPIInt),&rowners);
2205: browners = rowners + size + 1;
2206: MPI_Allgather(&mbs,1,MPI_INT,rowners+1,1,MPI_INT,comm);
2207: rowners[0] = 0;
2208: for (i=2; i<=size; i++) rowners[i] += rowners[i-1];
2209: for (i=0; i<=size; i++) browners[i] = rowners[i]*bs;
2210: rstart = rowners[rank];
2211: rend = rowners[rank+1];
2212:
2213: /* distribute row lengths to all processors */
2214: PetscMalloc((rend-rstart)*bs*sizeof(PetscMPIInt),&locrowlens);
2215: if (!rank) {
2216: PetscMalloc((M+extra_rows)*sizeof(PetscInt),&rowlengths);
2217: PetscBinaryRead(fd,rowlengths,M,PETSC_INT);
2218: for (i=0; i<extra_rows; i++) rowlengths[M+i] = 1;
2219: PetscMalloc(size*sizeof(PetscMPIInt),&sndcounts);
2220: for (i=0; i<size; i++) sndcounts[i] = browners[i+1] - browners[i];
2221: MPI_Scatterv(rowlengths,sndcounts,browners,MPIU_INT,locrowlens,(rend-rstart)*bs,MPIU_INT,0,comm);
2222: PetscFree(sndcounts);
2223: } else {
2224: MPI_Scatterv(0,0,0,MPIU_INT,locrowlens,(rend-rstart)*bs,MPIU_INT,0,comm);
2225: }
2226:
2227: if (!rank) { /* procs[0] */
2228: /* calculate the number of nonzeros on each processor */
2229: PetscMalloc(size*sizeof(PetscInt),&procsnz);
2230: PetscMemzero(procsnz,size*sizeof(PetscInt));
2231: for (i=0; i<size; i++) {
2232: for (j=rowners[i]*bs; j< rowners[i+1]*bs; j++) {
2233: procsnz[i] += rowlengths[j];
2234: }
2235: }
2236: PetscFree(rowlengths);
2237:
2238: /* determine max buffer needed and allocate it */
2239: maxnz = 0;
2240: for (i=0; i<size; i++) {
2241: maxnz = PetscMax(maxnz,procsnz[i]);
2242: }
2243: PetscMalloc(maxnz*sizeof(PetscInt),&cols);
2245: /* read in my part of the matrix column indices */
2246: nz = procsnz[0];
2247: PetscMalloc(nz*sizeof(PetscInt),&ibuf);
2248: mycols = ibuf;
2249: if (size == 1) nz -= extra_rows;
2250: PetscBinaryRead(fd,mycols,nz,PETSC_INT);
2251: if (size == 1) for (i=0; i< extra_rows; i++) { mycols[nz+i] = M+i; }
2253: /* read in every ones (except the last) and ship off */
2254: for (i=1; i<size-1; i++) {
2255: nz = procsnz[i];
2256: PetscBinaryRead(fd,cols,nz,PETSC_INT);
2257: MPI_Send(cols,nz,MPIU_INT,i,tag,comm);
2258: }
2259: /* read in the stuff for the last proc */
2260: if (size != 1) {
2261: nz = procsnz[size-1] - extra_rows; /* the extra rows are not on the disk */
2262: PetscBinaryRead(fd,cols,nz,PETSC_INT);
2263: for (i=0; i<extra_rows; i++) cols[nz+i] = M+i;
2264: MPI_Send(cols,nz+extra_rows,MPIU_INT,size-1,tag,comm);
2265: }
2266: PetscFree(cols);
2267: } else { /* procs[i], i>0 */
2268: /* determine buffer space needed for message */
2269: nz = 0;
2270: for (i=0; i<m; i++) {
2271: nz += locrowlens[i];
2272: }
2273: PetscMalloc(nz*sizeof(PetscInt),&ibuf);
2274: mycols = ibuf;
2275: /* receive message of column indices*/
2276: MPI_Recv(mycols,nz,MPIU_INT,0,tag,comm,&status);
2277: MPI_Get_count(&status,MPIU_INT,&maxnz);
2278: if (maxnz != nz) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"something is wrong with file");
2279: }
2281: /* loop over local rows, determining number of off diagonal entries */
2282: PetscMalloc(2*(rend-rstart+1)*sizeof(PetscInt),&dlens);
2283: odlens = dlens + (rend-rstart);
2284: PetscMalloc(3*Mbs*sizeof(PetscInt),&mask);
2285: PetscMemzero(mask,3*Mbs*sizeof(PetscInt));
2286: masked1 = mask + Mbs;
2287: masked2 = masked1 + Mbs;
2288: rowcount = 0; nzcount = 0;
2289: for (i=0; i<mbs; i++) {
2290: dcount = 0;
2291: odcount = 0;
2292: for (j=0; j<bs; j++) {
2293: kmax = locrowlens[rowcount];
2294: for (k=0; k<kmax; k++) {
2295: tmp = mycols[nzcount++]/bs; /* block col. index */
2296: if (!mask[tmp]) {
2297: mask[tmp] = 1;
2298: if (tmp < rstart || tmp >= rend) masked2[odcount++] = tmp; /* entry in off-diag portion */
2299: else masked1[dcount++] = tmp; /* entry in diag portion */
2300: }
2301: }
2302: rowcount++;
2303: }
2304:
2305: dlens[i] = dcount; /* d_nzz[i] */
2306: odlens[i] = odcount; /* o_nzz[i] */
2308: /* zero out the mask elements we set */
2309: for (j=0; j<dcount; j++) mask[masked1[j]] = 0;
2310: for (j=0; j<odcount; j++) mask[masked2[j]] = 0;
2311: }
2312:
2313: /* create our matrix */
2314: MatCreate(comm,&A);
2315: MatSetSizes(A,m,m,PETSC_DETERMINE,PETSC_DETERMINE);
2316: MatSetType(A,type);
2317: MatMPISBAIJSetPreallocation(A,bs,0,dlens,0,odlens);
2318: MatSetOption(A,MAT_COLUMNS_SORTED);
2319:
2320: if (!rank) {
2321: PetscMalloc(maxnz*sizeof(PetscScalar),&buf);
2322: /* read in my part of the matrix numerical values */
2323: nz = procsnz[0];
2324: vals = buf;
2325: mycols = ibuf;
2326: if (size == 1) nz -= extra_rows;
2327: PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);
2328: if (size == 1) for (i=0; i< extra_rows; i++) { vals[nz+i] = 1.0; }
2330: /* insert into matrix */
2331: jj = rstart*bs;
2332: for (i=0; i<m; i++) {
2333: MatSetValues(A,1,&jj,locrowlens[i],mycols,vals,INSERT_VALUES);
2334: mycols += locrowlens[i];
2335: vals += locrowlens[i];
2336: jj++;
2337: }
2339: /* read in other processors (except the last one) and ship out */
2340: for (i=1; i<size-1; i++) {
2341: nz = procsnz[i];
2342: vals = buf;
2343: PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);
2344: MPI_Send(vals,nz,MPIU_SCALAR,i,A->tag,comm);
2345: }
2346: /* the last proc */
2347: if (size != 1){
2348: nz = procsnz[i] - extra_rows;
2349: vals = buf;
2350: PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);
2351: for (i=0; i<extra_rows; i++) vals[nz+i] = 1.0;
2352: MPI_Send(vals,nz+extra_rows,MPIU_SCALAR,size-1,A->tag,comm);
2353: }
2354: PetscFree(procsnz);
2356: } else {
2357: /* receive numeric values */
2358: PetscMalloc(nz*sizeof(PetscScalar),&buf);
2360: /* receive message of values*/
2361: vals = buf;
2362: mycols = ibuf;
2363: MPI_Recv(vals,nz,MPIU_SCALAR,0,A->tag,comm,&status);
2364: MPI_Get_count(&status,MPIU_SCALAR,&maxnz);
2365: if (maxnz != nz) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"something is wrong with file");
2367: /* insert into matrix */
2368: jj = rstart*bs;
2369: for (i=0; i<m; i++) {
2370: MatSetValues_MPISBAIJ(A,1,&jj,locrowlens[i],mycols,vals,INSERT_VALUES);
2371: mycols += locrowlens[i];
2372: vals += locrowlens[i];
2373: jj++;
2374: }
2375: }
2377: PetscFree(locrowlens);
2378: PetscFree(buf);
2379: PetscFree(ibuf);
2380: PetscFree(rowners);
2381: PetscFree(dlens);
2382: PetscFree(mask);
2383: MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);
2384: MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);
2385: *newmat = A;
2386: return(0);
2387: }
2391: /*XXXXX@
2392: MatMPISBAIJSetHashTableFactor - Sets the factor required to compute the size of the HashTable.
2394: Input Parameters:
2395: . mat - the matrix
2396: . fact - factor
2398: Collective on Mat
2400: Level: advanced
2402: Notes:
2403: This can also be set by the command line option: -mat_use_hash_table fact
2405: .keywords: matrix, hashtable, factor, HT
2407: .seealso: MatSetOption()
2408: @XXXXX*/
2413: PetscErrorCode MatGetRowMaxAbs_MPISBAIJ(Mat A,Vec v,PetscInt idx[])
2414: {
2415: Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data;
2416: Mat_SeqBAIJ *b = (Mat_SeqBAIJ*)(a->B)->data;
2417: PetscReal atmp;
2418: PetscReal *work,*svalues,*rvalues;
2420: PetscInt i,bs,mbs,*bi,*bj,brow,j,ncols,krow,kcol,col,row,Mbs,bcol;
2421: PetscMPIInt rank,size;
2422: PetscInt *rowners_bs,dest,count,source;
2423: PetscScalar *va;
2424: MatScalar *ba;
2425: MPI_Status stat;
2428: if (idx) SETERRQ(PETSC_ERR_SUP,"Send email to petsc-maint@mcs.anl.gov");
2429: MatGetRowMaxAbs(a->A,v,PETSC_NULL);
2430: VecGetArray(v,&va);
2432: MPI_Comm_size(A->comm,&size);
2433: MPI_Comm_rank(A->comm,&rank);
2435: bs = A->rmap.bs;
2436: mbs = a->mbs;
2437: Mbs = a->Mbs;
2438: ba = b->a;
2439: bi = b->i;
2440: bj = b->j;
2442: /* find ownerships */
2443: rowners_bs = A->rmap.range;
2445: /* each proc creates an array to be distributed */
2446: PetscMalloc(bs*Mbs*sizeof(PetscReal),&work);
2447: PetscMemzero(work,bs*Mbs*sizeof(PetscReal));
2449: /* row_max for B */
2450: if (rank != size-1){
2451: for (i=0; i<mbs; i++) {
2452: ncols = bi[1] - bi[0]; bi++;
2453: brow = bs*i;
2454: for (j=0; j<ncols; j++){
2455: bcol = bs*(*bj);
2456: for (kcol=0; kcol<bs; kcol++){
2457: col = bcol + kcol; /* local col index */
2458: col += rowners_bs[rank+1]; /* global col index */
2459: for (krow=0; krow<bs; krow++){
2460: atmp = PetscAbsScalar(*ba); ba++;
2461: row = brow + krow; /* local row index */
2462: if (PetscRealPart(va[row]) < atmp) va[row] = atmp;
2463: if (work[col] < atmp) work[col] = atmp;
2464: }
2465: }
2466: bj++;
2467: }
2468: }
2470: /* send values to its owners */
2471: for (dest=rank+1; dest<size; dest++){
2472: svalues = work + rowners_bs[dest];
2473: count = rowners_bs[dest+1]-rowners_bs[dest];
2474: MPI_Send(svalues,count,MPIU_REAL,dest,rank,A->comm);
2475: }
2476: }
2477:
2478: /* receive values */
2479: if (rank){
2480: rvalues = work;
2481: count = rowners_bs[rank+1]-rowners_bs[rank];
2482: for (source=0; source<rank; source++){
2483: MPI_Recv(rvalues,count,MPIU_REAL,MPI_ANY_SOURCE,MPI_ANY_TAG,A->comm,&stat);
2484: /* process values */
2485: for (i=0; i<count; i++){
2486: if (PetscRealPart(va[i]) < rvalues[i]) va[i] = rvalues[i];
2487: }
2488: }
2489: }
2491: VecRestoreArray(v,&va);
2492: PetscFree(work);
2493: return(0);
2494: }
2498: PetscErrorCode MatRelax_MPISBAIJ(Mat matin,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,PetscInt its,PetscInt lits,Vec xx)
2499: {
2500: Mat_MPISBAIJ *mat = (Mat_MPISBAIJ*)matin->data;
2502: PetscInt mbs=mat->mbs,bs=matin->rmap.bs;
2503: PetscScalar *x,*b,*ptr,zero=0.0;
2504: Vec bb1;
2505:
2507: if (its <= 0 || lits <= 0) SETERRQ2(PETSC_ERR_ARG_WRONG,"Relaxation requires global its %D and local its %D both positive",its,lits);
2508: if (bs > 1)
2509: SETERRQ(PETSC_ERR_SUP,"SSOR for block size > 1 is not yet implemented");
2511: if ((flag & SOR_LOCAL_SYMMETRIC_SWEEP) == SOR_LOCAL_SYMMETRIC_SWEEP){
2512: if ( flag & SOR_ZERO_INITIAL_GUESS ) {
2513: (*mat->A->ops->relax)(mat->A,bb,omega,flag,fshift,lits,lits,xx);
2514: its--;
2515: }
2517: VecDuplicate(bb,&bb1);
2518: while (its--){
2519:
2520: /* lower triangular part: slvec0b = - B^T*xx */
2521: (*mat->B->ops->multtranspose)(mat->B,xx,mat->slvec0b);
2522:
2523: /* copy xx into slvec0a */
2524: VecGetArray(mat->slvec0,&ptr);
2525: VecGetArray(xx,&x);
2526: PetscMemcpy(ptr,x,bs*mbs*sizeof(MatScalar));
2527: VecRestoreArray(mat->slvec0,&ptr);
2529: VecScale(mat->slvec0,-1.0);
2531: /* copy bb into slvec1a */
2532: VecGetArray(mat->slvec1,&ptr);
2533: VecGetArray(bb,&b);
2534: PetscMemcpy(ptr,b,bs*mbs*sizeof(MatScalar));
2535: VecRestoreArray(mat->slvec1,&ptr);
2537: /* set slvec1b = 0 */
2538: VecSet(mat->slvec1b,zero);
2540: VecScatterBegin(mat->sMvctx,mat->slvec0,mat->slvec1,ADD_VALUES,SCATTER_FORWARD);
2541: VecRestoreArray(xx,&x);
2542: VecRestoreArray(bb,&b);
2543: VecScatterEnd(mat->sMvctx,mat->slvec0,mat->slvec1,ADD_VALUES,SCATTER_FORWARD);
2545: /* upper triangular part: bb1 = bb1 - B*x */
2546: (*mat->B->ops->multadd)(mat->B,mat->slvec1b,mat->slvec1a,bb1);
2547:
2548: /* local diagonal sweep */
2549: (*mat->A->ops->relax)(mat->A,bb1,omega,SOR_SYMMETRIC_SWEEP,fshift,lits,lits,xx);
2550: }
2551: VecDestroy(bb1);
2552: } else {
2553: SETERRQ(PETSC_ERR_SUP,"MatSORType is not supported for SBAIJ matrix format");
2554: }
2555: return(0);
2556: }
2560: PetscErrorCode MatRelax_MPISBAIJ_2comm(Mat matin,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,PetscInt its,PetscInt lits,Vec xx)
2561: {
2562: Mat_MPISBAIJ *mat = (Mat_MPISBAIJ*)matin->data;
2564: Vec lvec1,bb1;
2565:
2567: if (its <= 0 || lits <= 0) SETERRQ2(PETSC_ERR_ARG_WRONG,"Relaxation requires global its %D and local its %D both positive",its,lits);
2568: if (matin->rmap.bs > 1)
2569: SETERRQ(PETSC_ERR_SUP,"SSOR for block size > 1 is not yet implemented");
2571: if ((flag & SOR_LOCAL_SYMMETRIC_SWEEP) == SOR_LOCAL_SYMMETRIC_SWEEP){
2572: if ( flag & SOR_ZERO_INITIAL_GUESS ) {
2573: (*mat->A->ops->relax)(mat->A,bb,omega,flag,fshift,lits,lits,xx);
2574: its--;
2575: }
2577: VecDuplicate(mat->lvec,&lvec1);
2578: VecDuplicate(bb,&bb1);
2579: while (its--){
2580: VecScatterBegin(mat->Mvctx,xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD);
2581:
2582: /* lower diagonal part: bb1 = bb - B^T*xx */
2583: (*mat->B->ops->multtranspose)(mat->B,xx,lvec1);
2584: VecScale(lvec1,-1.0);
2586: VecScatterEnd(mat->Mvctx,xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD);
2587: VecCopy(bb,bb1);
2588: VecScatterBegin(mat->Mvctx,lvec1,bb1,ADD_VALUES,SCATTER_REVERSE);
2590: /* upper diagonal part: bb1 = bb1 - B*x */
2591: VecScale(mat->lvec,-1.0);
2592: (*mat->B->ops->multadd)(mat->B,mat->lvec,bb1,bb1);
2594: VecScatterEnd(mat->Mvctx,lvec1,bb1,ADD_VALUES,SCATTER_REVERSE);
2595:
2596: /* diagonal sweep */
2597: (*mat->A->ops->relax)(mat->A,bb1,omega,SOR_SYMMETRIC_SWEEP,fshift,lits,lits,xx);
2598: }
2599: VecDestroy(lvec1);
2600: VecDestroy(bb1);
2601: } else {
2602: SETERRQ(PETSC_ERR_SUP,"MatSORType is not supported for SBAIJ matrix format");
2603: }
2604: return(0);
2605: }