xref: /petsc/src/mat/impls/aij/mpi/superlu_dist/superlu_dist.c (revision bebe2cf65d55febe21a5af8db2bd2e168caaa2e7)
1 
2 /*
3         Provides an interface to the SuperLU_DIST_2.2 sparse solver
4 */
5 
6 #include <../src/mat/impls/aij/seq/aij.h>
7 #include <../src/mat/impls/aij/mpi/mpiaij.h>
8 #if defined(PETSC_HAVE_STDLIB_H) /* This is to get around weird problem with SuperLU on cray */
9 #include <stdlib.h>
10 #endif
11 
12 #if defined(PETSC_USE_64BIT_INDICES)
13 /* ugly SuperLU_Dist variable telling it to use long long int */
14 #define _LONGINT
15 #endif
16 
17 EXTERN_C_BEGIN
18 #if defined(PETSC_USE_COMPLEX)
19 #include <superlu_zdefs.h>
20 #else
21 #include <superlu_ddefs.h>
22 #endif
23 EXTERN_C_END
24 
25 /*
26     GLOBAL - The sparse matrix and right hand side are all stored initially on process 0. Should be called centralized
27     DISTRIBUTED - The sparse matrix and right hand size are initially stored across the entire MPI communicator.
28 */
29 typedef enum {GLOBAL,DISTRIBUTED} SuperLU_MatInputMode;
30 const char *SuperLU_MatInputModes[] = {"GLOBAL","DISTRIBUTED","SuperLU_MatInputMode","PETSC_",0};
31 
32 typedef struct {
33   int_t                nprow,npcol,*row,*col;
34   gridinfo_t           grid;
35   superlu_options_t    options;
36   SuperMatrix          A_sup;
37   ScalePermstruct_t    ScalePermstruct;
38   LUstruct_t           LUstruct;
39   int                  StatPrint;
40   SuperLU_MatInputMode MatInputMode;
41   SOLVEstruct_t        SOLVEstruct;
42   fact_t               FactPattern;
43   MPI_Comm             comm_superlu;
44 #if defined(PETSC_USE_COMPLEX)
45   doublecomplex        *val;
46 #else
47   double               *val;
48 #endif
49   PetscBool            matsolve_iscalled,matmatsolve_iscalled;
50   PetscBool            CleanUpSuperLU_Dist;  /* Flag to clean up (non-global) SuperLU objects during Destroy */
51 } Mat_SuperLU_DIST;
52 
53 extern PetscErrorCode MatFactorInfo_SuperLU_DIST(Mat,PetscViewer);
54 extern PetscErrorCode MatLUFactorNumeric_SuperLU_DIST(Mat,Mat,const MatFactorInfo*);
55 extern PetscErrorCode MatDestroy_SuperLU_DIST(Mat);
56 extern PetscErrorCode MatView_SuperLU_DIST(Mat,PetscViewer);
57 extern PetscErrorCode MatSolve_SuperLU_DIST(Mat,Vec,Vec);
58 extern PetscErrorCode MatLUFactorSymbolic_SuperLU_DIST(Mat,Mat,IS,IS,const MatFactorInfo*);
59 extern PetscErrorCode MatDestroy_MPIAIJ(Mat);
60 
61 #undef __FUNCT__
62 #define __FUNCT__ "MatGetDiagonal_SuperLU_DIST"
63 PetscErrorCode MatGetDiagonal_SuperLU_DIST(Mat A,Vec v)
64 {
65   PetscFunctionBegin;
66   SETERRQ(PetscObjectComm((PetscObject)A),PETSC_ERR_SUP,"Mat type: SuperLU_DIST factor");
67   PetscFunctionReturn(0);
68 }
69 
70 #undef __FUNCT__
71 #define __FUNCT__ "MatDestroy_SuperLU_DIST"
72 PetscErrorCode MatDestroy_SuperLU_DIST(Mat A)
73 {
74   PetscErrorCode   ierr;
75   Mat_SuperLU_DIST *lu = (Mat_SuperLU_DIST*)A->spptr;
76   PetscBool        flg;
77 
78   PetscFunctionBegin;
79   if (lu && lu->CleanUpSuperLU_Dist) {
80     /* Deallocate SuperLU_DIST storage */
81     if (lu->MatInputMode == GLOBAL) {
82       PetscStackCall("SuperLU_DIST:Destroy_CompCol_Matrix_dist",Destroy_CompCol_Matrix_dist(&lu->A_sup));
83     } else {
84       PetscStackCall("SuperLU_DIST:Destroy_CompRowLoc_Matrix_dist",Destroy_CompRowLoc_Matrix_dist(&lu->A_sup));
85       if (lu->options.SolveInitialized) {
86 #if defined(PETSC_USE_COMPLEX)
87         PetscStackCall("SuperLU_DIST:zSolveFinalize",zSolveFinalize(&lu->options, &lu->SOLVEstruct));
88 #else
89         PetscStackCall("SuperLU_DIST:dSolveFinalize",dSolveFinalize(&lu->options, &lu->SOLVEstruct));
90 #endif
91       }
92     }
93     PetscStackCall("SuperLU_DIST:Destroy_LU",Destroy_LU(A->cmap->N, &lu->grid, &lu->LUstruct));
94     PetscStackCall("SuperLU_DIST:ScalePermstructFree",ScalePermstructFree(&lu->ScalePermstruct));
95     PetscStackCall("SuperLU_DIST:LUstructFree",LUstructFree(&lu->LUstruct));
96 
97     /* Release the SuperLU_DIST process grid. */
98     PetscStackCall("SuperLU_DIST:superlu_gridexit",superlu_gridexit(&lu->grid));
99     ierr = MPI_Comm_free(&(lu->comm_superlu));CHKERRQ(ierr);
100   }
101   ierr = PetscFree(A->spptr);CHKERRQ(ierr);
102 
103   ierr = PetscObjectTypeCompare((PetscObject)A,MATSEQAIJ,&flg);CHKERRQ(ierr);
104   if (flg) {
105     ierr = MatDestroy_SeqAIJ(A);CHKERRQ(ierr);
106   } else {
107     ierr = MatDestroy_MPIAIJ(A);CHKERRQ(ierr);
108   }
109   PetscFunctionReturn(0);
110 }
111 
112 #undef __FUNCT__
113 #define __FUNCT__ "MatSolve_SuperLU_DIST"
114 PetscErrorCode MatSolve_SuperLU_DIST(Mat A,Vec b_mpi,Vec x)
115 {
116   Mat_SuperLU_DIST *lu = (Mat_SuperLU_DIST*)A->spptr;
117   PetscErrorCode   ierr;
118   PetscMPIInt      size;
119   PetscInt         m=A->rmap->n,M=A->rmap->N,N=A->cmap->N;
120   SuperLUStat_t    stat;
121   double           berr[1];
122   PetscScalar      *bptr;
123   PetscInt         nrhs=1;
124   Vec              x_seq;
125   IS               iden;
126   VecScatter       scat;
127   int              info; /* SuperLU_Dist info code is ALWAYS an int, even with long long indices */
128   static PetscBool cite = PETSC_FALSE;
129 
130   PetscFunctionBegin;
131   ierr = PetscCitationsRegister("@article{lidemmel03,\n  author = {Xiaoye S. Li and James W. Demmel},\n  title = {{SuperLU_DIST}: A Scalable Distributed-Memory Sparse Direct\n           Solver for Unsymmetric Linear Systems},\n  journal = {ACM Trans. Mathematical Software},\n  volume = {29},\n  number = {2},\n  pages = {110-140},\n  year = 2003\n}\n",&cite);CHKERRQ(ierr);
132   ierr = MPI_Comm_size(PetscObjectComm((PetscObject)A),&size);CHKERRQ(ierr);
133   if (size > 1 && lu->MatInputMode == GLOBAL) {
134     /* global mat input, convert b to x_seq */
135     ierr = VecCreateSeq(PETSC_COMM_SELF,N,&x_seq);CHKERRQ(ierr);
136     ierr = ISCreateStride(PETSC_COMM_SELF,N,0,1,&iden);CHKERRQ(ierr);
137     ierr = VecScatterCreate(b_mpi,iden,x_seq,iden,&scat);CHKERRQ(ierr);
138     ierr = ISDestroy(&iden);CHKERRQ(ierr);
139 
140     ierr = VecScatterBegin(scat,b_mpi,x_seq,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr);
141     ierr = VecScatterEnd(scat,b_mpi,x_seq,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr);
142     ierr = VecGetArray(x_seq,&bptr);CHKERRQ(ierr);
143   } else { /* size==1 || distributed mat input */
144     if (lu->options.SolveInitialized && !lu->matsolve_iscalled) {
145       /* see comments in MatMatSolve() */
146 #if defined(PETSC_USE_COMPLEX)
147       PetscStackCall("SuperLU_DIST:zSolveFinalize",zSolveFinalize(&lu->options, &lu->SOLVEstruct));
148 #else
149       PetscStackCall("SuperLU_DIST:dSolveFinalize",dSolveFinalize(&lu->options, &lu->SOLVEstruct));
150 #endif
151       lu->options.SolveInitialized = NO;
152     }
153     ierr = VecCopy(b_mpi,x);CHKERRQ(ierr);
154     ierr = VecGetArray(x,&bptr);CHKERRQ(ierr);
155   }
156 
157   if (lu->options.Fact != FACTORED) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"SuperLU_DIST options.Fact mush equal FACTORED");
158 
159   PetscStackCall("SuperLU_DIST:PStatInit",PStatInit(&stat));        /* Initialize the statistics variables. */
160   if (lu->MatInputMode == GLOBAL) {
161 #if defined(PETSC_USE_COMPLEX)
162     PetscStackCall("SuperLU_DIST:pzgssvx_ABglobal",pzgssvx_ABglobal(&lu->options,&lu->A_sup,&lu->ScalePermstruct,(doublecomplex*)bptr,M,nrhs,&lu->grid,&lu->LUstruct,berr,&stat,&info));
163 #else
164     PetscStackCall("SuperLU_DIST:pdgssvx_ABglobal",pdgssvx_ABglobal(&lu->options, &lu->A_sup, &lu->ScalePermstruct,bptr,M,nrhs,&lu->grid,&lu->LUstruct,berr,&stat,&info));
165 #endif
166   } else { /* distributed mat input */
167 #if defined(PETSC_USE_COMPLEX)
168     PetscStackCall("SuperLU_DIST:pzgssvx",pzgssvx(&lu->options,&lu->A_sup,&lu->ScalePermstruct,(doublecomplex*)bptr,m,nrhs,&lu->grid,&lu->LUstruct,&lu->SOLVEstruct,berr,&stat,&info));
169 #else
170     PetscStackCall("SuperLU_DIST:pdgssvx",pdgssvx(&lu->options,&lu->A_sup,&lu->ScalePermstruct,bptr,m,nrhs,&lu->grid,&lu->LUstruct,&lu->SOLVEstruct,berr,&stat,&info));
171 #endif
172   }
173   if (info) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_LIB,"pdgssvx fails, info: %d\n",info);
174 
175   if (lu->options.PrintStat) PStatPrint(&lu->options, &stat, &lu->grid);      /* Print the statistics. */
176   PetscStackCall("SuperLU_DIST:PStatFree",PStatFree(&stat));
177 
178   if (size > 1 && lu->MatInputMode == GLOBAL) {
179     /* convert seq x to mpi x */
180     ierr = VecRestoreArray(x_seq,&bptr);CHKERRQ(ierr);
181     ierr = VecScatterBegin(scat,x_seq,x,INSERT_VALUES,SCATTER_REVERSE);CHKERRQ(ierr);
182     ierr = VecScatterEnd(scat,x_seq,x,INSERT_VALUES,SCATTER_REVERSE);CHKERRQ(ierr);
183     ierr = VecScatterDestroy(&scat);CHKERRQ(ierr);
184     ierr = VecDestroy(&x_seq);CHKERRQ(ierr);
185   } else {
186     ierr = VecRestoreArray(x,&bptr);CHKERRQ(ierr);
187 
188     lu->matsolve_iscalled    = PETSC_TRUE;
189     lu->matmatsolve_iscalled = PETSC_FALSE;
190   }
191   PetscFunctionReturn(0);
192 }
193 
194 #undef __FUNCT__
195 #define __FUNCT__ "MatMatSolve_SuperLU_DIST"
196 PetscErrorCode MatMatSolve_SuperLU_DIST(Mat A,Mat B_mpi,Mat X)
197 {
198   Mat_SuperLU_DIST *lu = (Mat_SuperLU_DIST*)A->spptr;
199   PetscErrorCode   ierr;
200   PetscMPIInt      size;
201   PetscInt         M=A->rmap->N,m=A->rmap->n,nrhs;
202   SuperLUStat_t    stat;
203   double           berr[1];
204   PetscScalar      *bptr;
205   int              info; /* SuperLU_Dist info code is ALWAYS an int, even with long long indices */
206   PetscBool        flg;
207 
208   PetscFunctionBegin;
209   if (lu->options.Fact != FACTORED) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"SuperLU_DIST options.Fact mush equal FACTORED");
210   ierr = PetscObjectTypeCompareAny((PetscObject)B_mpi,&flg,MATSEQDENSE,MATMPIDENSE,NULL);CHKERRQ(ierr);
211   if (!flg) SETERRQ(PetscObjectComm((PetscObject)A),PETSC_ERR_ARG_WRONG,"Matrix B must be MATDENSE matrix");
212   ierr = PetscObjectTypeCompareAny((PetscObject)X,&flg,MATSEQDENSE,MATMPIDENSE,NULL);CHKERRQ(ierr);
213   if (!flg) SETERRQ(PetscObjectComm((PetscObject)A),PETSC_ERR_ARG_WRONG,"Matrix X must be MATDENSE matrix");
214 
215   ierr = MPI_Comm_size(PetscObjectComm((PetscObject)A),&size);CHKERRQ(ierr);
216   if (size > 1 && lu->MatInputMode == GLOBAL) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"MatInputMode=GLOBAL for nproc %d>1 is not supported",size);
217   /* size==1 or distributed mat input */
218   if (lu->options.SolveInitialized && !lu->matmatsolve_iscalled) {
219     /* communication pattern of SOLVEstruct is unlikely created for matmatsolve,
220        thus destroy it and create a new SOLVEstruct.
221        Otherwise it may result in memory corruption or incorrect solution
222        See src/mat/examples/tests/ex125.c */
223 #if defined(PETSC_USE_COMPLEX)
224     PetscStackCall("SuperLU_DIST:zSolveFinalize",zSolveFinalize(&lu->options, &lu->SOLVEstruct));
225 #else
226     PetscStackCall("SuperLU_DIST:dSolveFinalize",dSolveFinalize(&lu->options, &lu->SOLVEstruct));
227 #endif
228     lu->options.SolveInitialized = NO;
229   }
230   ierr = MatCopy(B_mpi,X,SAME_NONZERO_PATTERN);CHKERRQ(ierr);
231 
232   ierr = MatGetSize(B_mpi,NULL,&nrhs);CHKERRQ(ierr);
233 
234   PetscStackCall("SuperLU_DIST:PStatInit",PStatInit(&stat));        /* Initialize the statistics variables. */
235   ierr = MatDenseGetArray(X,&bptr);CHKERRQ(ierr);
236   if (lu->MatInputMode == GLOBAL) { /* size == 1 */
237 #if defined(PETSC_USE_COMPLEX)
238     PetscStackCall("SuperLU_DIST:pzgssvx_ABglobal",pzgssvx_ABglobal(&lu->options, &lu->A_sup, &lu->ScalePermstruct,(doublecomplex*)bptr, M, nrhs,&lu->grid, &lu->LUstruct, berr, &stat, &info));
239 #else
240     PetscStackCall("SuperLU_DIST:pdgssvx_ABglobal",pdgssvx_ABglobal(&lu->options, &lu->A_sup, &lu->ScalePermstruct,bptr, M, nrhs, &lu->grid, &lu->LUstruct, berr, &stat, &info));
241 #endif
242   } else { /* distributed mat input */
243 #if defined(PETSC_USE_COMPLEX)
244     PetscStackCall("SuperLU_DIST:pzgssvx",pzgssvx(&lu->options,&lu->A_sup,&lu->ScalePermstruct,(doublecomplex*)bptr,m,nrhs,&lu->grid, &lu->LUstruct,&lu->SOLVEstruct,berr,&stat,&info));
245 #else
246     PetscStackCall("SuperLU_DIST:pdgssvx",pdgssvx(&lu->options,&lu->A_sup,&lu->ScalePermstruct,bptr,m,nrhs,&lu->grid,&lu->LUstruct,&lu->SOLVEstruct,berr,&stat,&info));
247 #endif
248   }
249   if (info) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_LIB,"pdgssvx fails, info: %d\n",info);
250   ierr = MatDenseRestoreArray(X,&bptr);CHKERRQ(ierr);
251 
252   if (lu->options.PrintStat) PStatPrint(&lu->options, &stat, &lu->grid); /* Print the statistics. */
253   PetscStackCall("SuperLU_DIST:PStatFree",PStatFree(&stat));
254   lu->matsolve_iscalled    = PETSC_FALSE;
255   lu->matmatsolve_iscalled = PETSC_TRUE;
256   PetscFunctionReturn(0);
257 }
258 
259 
260 #undef __FUNCT__
261 #define __FUNCT__ "MatLUFactorNumeric_SuperLU_DIST"
262 PetscErrorCode MatLUFactorNumeric_SuperLU_DIST(Mat F,Mat A,const MatFactorInfo *info)
263 {
264   Mat              *tseq,A_seq = NULL;
265   Mat_SeqAIJ       *aa,*bb;
266   Mat_SuperLU_DIST *lu = (Mat_SuperLU_DIST*)(F)->spptr;
267   PetscErrorCode   ierr;
268   PetscInt         M=A->rmap->N,N=A->cmap->N,i,*ai,*aj,*bi,*bj,nz,rstart,*garray,
269                    m=A->rmap->n, colA_start,j,jcol,jB,countA,countB,*bjj,*ajj;
270   int              sinfo;   /* SuperLU_Dist info flag is always an int even with long long indices */
271   PetscMPIInt      size;
272   SuperLUStat_t    stat;
273   double           *berr=0;
274   IS               isrow;
275   Mat              F_diag=NULL;
276 #if defined(PETSC_USE_COMPLEX)
277   doublecomplex    *av, *bv;
278 #else
279   double           *av, *bv;
280 #endif
281 
282   PetscFunctionBegin;
283   ierr = MPI_Comm_size(PetscObjectComm((PetscObject)A),&size);CHKERRQ(ierr);
284 
285   if (lu->MatInputMode == GLOBAL) { /* global mat input */
286     if (size > 1) { /* convert mpi A to seq mat A */
287       ierr = ISCreateStride(PETSC_COMM_SELF,M,0,1,&isrow);CHKERRQ(ierr);
288       ierr = MatGetSubMatrices(A,1,&isrow,&isrow,MAT_INITIAL_MATRIX,&tseq);CHKERRQ(ierr);
289       ierr = ISDestroy(&isrow);CHKERRQ(ierr);
290 
291       A_seq = *tseq;
292       ierr  = PetscFree(tseq);CHKERRQ(ierr);
293       aa    = (Mat_SeqAIJ*)A_seq->data;
294     } else {
295       PetscBool flg;
296       ierr = PetscObjectTypeCompare((PetscObject)A,MATMPIAIJ,&flg);CHKERRQ(ierr);
297       if (flg) {
298         Mat_MPIAIJ *At = (Mat_MPIAIJ*)A->data;
299         A = At->A;
300       }
301       aa =  (Mat_SeqAIJ*)A->data;
302     }
303 
304     /* Convert Petsc NR matrix to SuperLU_DIST NC.
305        Note: memories of lu->val, col and row are allocated by CompRow_to_CompCol_dist()! */
306     if (lu->options.Fact != DOFACT) {/* successive numeric factorization, sparsity pattern is reused. */
307       PetscStackCall("SuperLU_DIST:Destroy_CompCol_Matrix_dist",Destroy_CompCol_Matrix_dist(&lu->A_sup));
308       if (lu->FactPattern == SamePattern_SameRowPerm) {
309         lu->options.Fact = SamePattern_SameRowPerm; /* matrix has similar numerical values */
310       } else { /* lu->FactPattern == SamePattern */
311         PetscStackCall("SuperLU_DIST:Destroy_LU",Destroy_LU(N, &lu->grid, &lu->LUstruct));
312         lu->options.Fact = SamePattern;
313       }
314     }
315 #if defined(PETSC_USE_COMPLEX)
316     PetscStackCall("SuperLU_DIST:zCompRow_to_CompCol_dist",zCompRow_to_CompCol_dist(M,N,aa->nz,(doublecomplex*)aa->a,(int_t*)aa->j,(int_t*)aa->i,&lu->val,&lu->col, &lu->row));
317 #else
318     PetscStackCall("SuperLU_DIST:dCompRow_to_CompCol_dist",dCompRow_to_CompCol_dist(M,N,aa->nz,aa->a,(int_t*)aa->j,(int_t*)aa->i,&lu->val, &lu->col, &lu->row));
319 #endif
320 
321     /* Create compressed column matrix A_sup. */
322 #if defined(PETSC_USE_COMPLEX)
323     PetscStackCall("SuperLU_DIST:zCreate_CompCol_Matrix_dist",zCreate_CompCol_Matrix_dist(&lu->A_sup, M, N, aa->nz, lu->val, lu->col, lu->row, SLU_NC, SLU_Z, SLU_GE));
324 #else
325     PetscStackCall("SuperLU_DIST:dCreate_CompCol_Matrix_dist",dCreate_CompCol_Matrix_dist(&lu->A_sup, M, N, aa->nz, lu->val, lu->col, lu->row, SLU_NC, SLU_D, SLU_GE));
326 #endif
327   } else { /* distributed mat input */
328     Mat_MPIAIJ *mat = (Mat_MPIAIJ*)A->data;
329     aa=(Mat_SeqAIJ*)(mat->A)->data;
330     bb=(Mat_SeqAIJ*)(mat->B)->data;
331     ai=aa->i; aj=aa->j;
332     bi=bb->i; bj=bb->j;
333 #if defined(PETSC_USE_COMPLEX)
334     av=(doublecomplex*)aa->a;
335     bv=(doublecomplex*)bb->a;
336 #else
337     av=aa->a;
338     bv=bb->a;
339 #endif
340     rstart = A->rmap->rstart;
341     nz     = aa->nz + bb->nz;
342     garray = mat->garray;
343 
344     if (lu->options.Fact == DOFACT) { /* first numeric factorization */
345 #if defined(PETSC_USE_COMPLEX)
346       PetscStackCall("SuperLU_DIST:zallocateA_dist",zallocateA_dist(m, nz, &lu->val, &lu->col, &lu->row));
347 #else
348       PetscStackCall("SuperLU_DIST:dallocateA_dist",dallocateA_dist(m, nz, &lu->val, &lu->col, &lu->row));
349 #endif
350     } else { /* successive numeric factorization, sparsity pattern and perm_c are reused. */
351       /* Destroy_CompRowLoc_Matrix_dist(&lu->A_sup); */ /* this leads to crash! However, see SuperLU_DIST_2.5/EXAMPLE/pzdrive2.c */
352       if (lu->FactPattern == SamePattern_SameRowPerm) {
353         lu->options.Fact = SamePattern_SameRowPerm; /* matrix has similar numerical values */
354       } else {
355         PetscStackCall("SuperLU_DIST:Destroy_LU",Destroy_LU(N, &lu->grid, &lu->LUstruct)); /* Deallocate storage associated with the L and U matrices. */
356         lu->options.Fact = SamePattern;
357       }
358     }
359     nz = 0;
360     for (i=0; i<m; i++) {
361       lu->row[i] = nz;
362       countA     = ai[i+1] - ai[i];
363       countB     = bi[i+1] - bi[i];
364       ajj        = aj + ai[i]; /* ptr to the beginning of this row */
365       bjj        = bj + bi[i];
366 
367       /* B part, smaller col index */
368       colA_start = rstart + ajj[0]; /* the smallest global col index of A */
369       jB         = 0;
370       for (j=0; j<countB; j++) {
371         jcol = garray[bjj[j]];
372         if (jcol > colA_start) {
373           jB = j;
374           break;
375         }
376         lu->col[nz]   = jcol;
377         lu->val[nz++] = *bv++;
378         if (j==countB-1) jB = countB;
379       }
380 
381       /* A part */
382       for (j=0; j<countA; j++) {
383         lu->col[nz]   = rstart + ajj[j];
384         lu->val[nz++] = *av++;
385       }
386 
387       /* B part, larger col index */
388       for (j=jB; j<countB; j++) {
389         lu->col[nz]   = garray[bjj[j]];
390         lu->val[nz++] = *bv++;
391       }
392     }
393     lu->row[m] = nz;
394 #if defined(PETSC_USE_COMPLEX)
395     PetscStackCall("SuperLU_DIST:zCreate_CompRowLoc_Matrix_dist",zCreate_CompRowLoc_Matrix_dist(&lu->A_sup, M, N, nz, m, rstart,lu->val, lu->col, lu->row, SLU_NR_loc, SLU_Z, SLU_GE));
396 #else
397     PetscStackCall("SuperLU_DIST:dCreate_CompRowLoc_Matrix_dist",dCreate_CompRowLoc_Matrix_dist(&lu->A_sup, M, N, nz, m, rstart,lu->val, lu->col, lu->row, SLU_NR_loc, SLU_D, SLU_GE));
398 #endif
399   }
400 
401   /* Factor the matrix. */
402   PetscStackCall("SuperLU_DIST:PStatInit",PStatInit(&stat));   /* Initialize the statistics variables. */
403   if (lu->MatInputMode == GLOBAL) { /* global mat input */
404 #if defined(PETSC_USE_COMPLEX)
405     PetscStackCall("SuperLU_DIST:pzgssvx_ABglobal",pzgssvx_ABglobal(&lu->options, &lu->A_sup, &lu->ScalePermstruct, 0, M, 0,&lu->grid, &lu->LUstruct, berr, &stat, &sinfo));
406 #else
407     PetscStackCall("SuperLU_DIST:pdgssvx_ABglobal",pdgssvx_ABglobal(&lu->options, &lu->A_sup, &lu->ScalePermstruct, 0, M, 0,&lu->grid, &lu->LUstruct, berr, &stat, &sinfo));
408 #endif
409   } else { /* distributed mat input */
410 #if defined(PETSC_USE_COMPLEX)
411     PetscStackCall("SuperLU_DIST:pzgssvx",pzgssvx(&lu->options, &lu->A_sup, &lu->ScalePermstruct, 0, m, 0, &lu->grid,&lu->LUstruct, &lu->SOLVEstruct, berr, &stat, &sinfo));
412     if (sinfo) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_LIB,"pzgssvx fails, info: %d\n",sinfo);
413 #else
414     PetscStackCall("SuperLU_DIST:pdgssvx",pdgssvx(&lu->options, &lu->A_sup, &lu->ScalePermstruct, 0, m, 0, &lu->grid,&lu->LUstruct, &lu->SOLVEstruct, berr, &stat, &sinfo));
415     if (sinfo) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_LIB,"pdgssvx fails, info: %d\n",sinfo);
416 #endif
417   }
418 
419   if (lu->MatInputMode == GLOBAL && size > 1) {
420     ierr = MatDestroy(&A_seq);CHKERRQ(ierr);
421   }
422 
423   if (lu->options.PrintStat) {
424     PStatPrint(&lu->options, &stat, &lu->grid);  /* Print the statistics. */
425   }
426   PStatFree(&stat);
427   if (size > 1) {
428     F_diag            = ((Mat_MPIAIJ*)(F)->data)->A;
429     F_diag->assembled = PETSC_TRUE;
430   }
431   (F)->assembled    = PETSC_TRUE;
432   (F)->preallocated = PETSC_TRUE;
433   lu->options.Fact  = FACTORED; /* The factored form of A is supplied. Local option used by this func. only */
434   PetscFunctionReturn(0);
435 }
436 
437 /* Note the Petsc r and c permutations are ignored */
438 #undef __FUNCT__
439 #define __FUNCT__ "MatLUFactorSymbolic_SuperLU_DIST"
440 PetscErrorCode MatLUFactorSymbolic_SuperLU_DIST(Mat F,Mat A,IS r,IS c,const MatFactorInfo *info)
441 {
442   Mat_SuperLU_DIST *lu = (Mat_SuperLU_DIST*)F->spptr;
443   PetscInt         M   = A->rmap->N,N=A->cmap->N;
444 
445   PetscFunctionBegin;
446   /* Initialize the SuperLU process grid. */
447   PetscStackCall("SuperLU_DIST:superlu_gridinit",superlu_gridinit(lu->comm_superlu, lu->nprow, lu->npcol, &lu->grid));
448 
449   /* Initialize ScalePermstruct and LUstruct. */
450   PetscStackCall("SuperLU_DIST:ScalePermstructInit",ScalePermstructInit(M, N, &lu->ScalePermstruct));
451   PetscStackCall("SuperLU_DIST:LUstructInit",LUstructInit(N, &lu->LUstruct));
452   F->ops->lufactornumeric = MatLUFactorNumeric_SuperLU_DIST;
453   F->ops->solve           = MatSolve_SuperLU_DIST;
454   F->ops->matsolve        = MatMatSolve_SuperLU_DIST;
455   lu->CleanUpSuperLU_Dist = PETSC_TRUE;
456   PetscFunctionReturn(0);
457 }
458 
459 #undef __FUNCT__
460 #define __FUNCT__ "MatFactorGetSolverPackage_aij_superlu_dist"
461 PetscErrorCode MatFactorGetSolverPackage_aij_superlu_dist(Mat A,const MatSolverPackage *type)
462 {
463   PetscFunctionBegin;
464   *type = MATSOLVERSUPERLU_DIST;
465   PetscFunctionReturn(0);
466 }
467 
468 #undef __FUNCT__
469 #define __FUNCT__ "MatGetFactor_aij_superlu_dist"
470 PETSC_EXTERN PetscErrorCode MatGetFactor_aij_superlu_dist(Mat A,MatFactorType ftype,Mat *F)
471 {
472   Mat               B;
473   Mat_SuperLU_DIST  *lu;
474   PetscErrorCode    ierr;
475   PetscInt          M=A->rmap->N,N=A->cmap->N,indx;
476   PetscMPIInt       size;
477   superlu_options_t options;
478   PetscBool         flg;
479   const char        *colperm[]     = {"NATURAL","MMD_AT_PLUS_A","MMD_ATA","METIS_AT_PLUS_A","PARMETIS"};
480   const char        *rowperm[]     = {"LargeDiag","NATURAL"};
481   const char        *factPattern[] = {"SamePattern","SamePattern_SameRowPerm"};
482   PetscBool         set;
483 
484   PetscFunctionBegin;
485   /* Create the factorization matrix */
486   ierr = MatCreate(PetscObjectComm((PetscObject)A),&B);CHKERRQ(ierr);
487   ierr = MatSetSizes(B,A->rmap->n,A->cmap->n,M,N);CHKERRQ(ierr);
488   ierr = MatSetType(B,((PetscObject)A)->type_name);CHKERRQ(ierr);
489   ierr = MatSeqAIJSetPreallocation(B,0,NULL);
490   ierr = MatMPIAIJSetPreallocation(B,0,NULL,0,NULL);CHKERRQ(ierr);
491 
492   B->ops->lufactorsymbolic = MatLUFactorSymbolic_SuperLU_DIST;
493   B->ops->view             = MatView_SuperLU_DIST;
494   B->ops->destroy          = MatDestroy_SuperLU_DIST;
495   B->ops->getdiagonal      = MatGetDiagonal_SuperLU_DIST;
496 
497   ierr = PetscObjectComposeFunction((PetscObject)B,"MatFactorGetSolverPackage_C",MatFactorGetSolverPackage_aij_superlu_dist);CHKERRQ(ierr);
498 
499   B->factortype = MAT_FACTOR_LU;
500 
501   ierr     = PetscNewLog(B,&lu);CHKERRQ(ierr);
502   B->spptr = lu;
503 
504   /* Set the default input options:
505      options.Fact              = DOFACT;
506      options.Equil             = YES;
507      options.ParSymbFact       = NO;
508      options.ColPerm           = METIS_AT_PLUS_A;
509      options.RowPerm           = LargeDiag;
510      options.ReplaceTinyPivot  = YES;
511      options.IterRefine        = DOUBLE;
512      options.Trans             = NOTRANS;
513      options.SolveInitialized  = NO; -hold the communication pattern used MatSolve() and MatMatSolve()
514      options.RefineInitialized = NO;
515      options.PrintStat         = YES;
516   */
517   set_default_options_dist(&options);
518 
519   ierr = MPI_Comm_dup(PetscObjectComm((PetscObject)A),&(lu->comm_superlu));CHKERRQ(ierr);
520   ierr = MPI_Comm_size(PetscObjectComm((PetscObject)A),&size);CHKERRQ(ierr);
521   /* Default num of process columns and rows */
522   lu->npcol = (int_t) (0.5 + PetscSqrtReal((PetscReal)size));
523   if (!lu->npcol) lu->npcol = 1;
524   while (lu->npcol > 0) {
525     lu->nprow = (int_t) (size/lu->npcol);
526     if (size == lu->nprow * lu->npcol) break;
527     lu->npcol--;
528   }
529 
530   ierr = PetscOptionsBegin(PetscObjectComm((PetscObject)A),((PetscObject)A)->prefix,"SuperLU_Dist Options","Mat");CHKERRQ(ierr);
531   ierr = PetscOptionsInt("-mat_superlu_dist_r","Number rows in processor partition","None",lu->nprow,(PetscInt*)&lu->nprow,NULL);CHKERRQ(ierr);
532   ierr = PetscOptionsInt("-mat_superlu_dist_c","Number columns in processor partition","None",lu->npcol,(PetscInt*)&lu->npcol,NULL);CHKERRQ(ierr);
533   if (size != lu->nprow * lu->npcol) SETERRQ3(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Number of processes %d must equal to nprow %d * npcol %d",size,lu->nprow,lu->npcol);
534 
535   lu->MatInputMode = DISTRIBUTED;
536 
537   ierr = PetscOptionsEnum("-mat_superlu_dist_matinput","Matrix input mode (global or distributed)","None",SuperLU_MatInputModes,(PetscEnum)lu->MatInputMode,(PetscEnum*)&lu->MatInputMode,NULL);CHKERRQ(ierr);
538   if (lu->MatInputMode == DISTRIBUTED && size == 1) lu->MatInputMode = GLOBAL;
539 
540   ierr = PetscOptionsBool("-mat_superlu_dist_equil","Equilibrate matrix","None",options.Equil ? PETSC_TRUE : PETSC_FALSE,&flg,&set);CHKERRQ(ierr);
541   if (set && !flg) options.Equil = NO;
542 
543   ierr = PetscOptionsEList("-mat_superlu_dist_rowperm","Row permutation","None",rowperm,2,rowperm[0],&indx,&flg);CHKERRQ(ierr);
544   if (flg) {
545     switch (indx) {
546     case 0:
547       options.RowPerm = LargeDiag;
548       break;
549     case 1:
550       options.RowPerm = NOROWPERM;
551       break;
552     }
553   }
554 
555   ierr = PetscOptionsEList("-mat_superlu_dist_colperm","Column permutation","None",colperm,5,colperm[3],&indx,&flg);CHKERRQ(ierr);
556   if (flg) {
557     switch (indx) {
558     case 0:
559       options.ColPerm = NATURAL;
560       break;
561     case 1:
562       options.ColPerm = MMD_AT_PLUS_A;
563       break;
564     case 2:
565       options.ColPerm = MMD_ATA;
566       break;
567     case 3:
568       options.ColPerm = METIS_AT_PLUS_A;
569       break;
570     case 4:
571       options.ColPerm = PARMETIS;   /* only works for np>1 */
572       break;
573     default:
574       SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Unknown column permutation");
575     }
576   }
577 
578   ierr = PetscOptionsBool("-mat_superlu_dist_replacetinypivot","Replace tiny pivots","None",options.ReplaceTinyPivot ? PETSC_TRUE : PETSC_FALSE,&flg,&set);CHKERRQ(ierr);
579   if (set && !flg) options.ReplaceTinyPivot = NO;
580 
581   options.ParSymbFact = NO;
582 
583   ierr = PetscOptionsBool("-mat_superlu_dist_parsymbfact","Parallel symbolic factorization","None",PETSC_FALSE,&flg,&set);CHKERRQ(ierr);
584   if (set && flg) {
585 #if defined(PETSC_HAVE_PARMETIS)
586     options.ParSymbFact = YES;
587     options.ColPerm     = PARMETIS;   /* in v2.2, PARMETIS is forced for ParSymbFact regardless of user ordering setting */
588 #else
589     printf("parsymbfact needs PARMETIS");
590 #endif
591   }
592 
593   lu->FactPattern = SamePattern_SameRowPerm;
594 
595   ierr = PetscOptionsEList("-mat_superlu_dist_fact","Sparsity pattern for repeated matrix factorization","None",factPattern,2,factPattern[1],&indx,&flg);CHKERRQ(ierr);
596   if (flg) {
597     switch (indx) {
598     case 0:
599       lu->FactPattern = SamePattern;
600       break;
601     case 1:
602       lu->FactPattern = SamePattern_SameRowPerm;
603       break;
604     }
605   }
606 
607   options.IterRefine = NOREFINE;
608   ierr               = PetscOptionsBool("-mat_superlu_dist_iterrefine","Use iterative refinement","None",options.IterRefine == NOREFINE ? PETSC_FALSE : PETSC_TRUE ,&flg,&set);CHKERRQ(ierr);
609   if (set) {
610     if (flg) options.IterRefine = SLU_DOUBLE;
611     else options.IterRefine = NOREFINE;
612   }
613 
614   if (PetscLogPrintInfo) options.PrintStat = YES;
615   else options.PrintStat = NO;
616   ierr = PetscOptionsBool("-mat_superlu_dist_statprint","Print factorization information","None",(PetscBool)options.PrintStat,(PetscBool*)&options.PrintStat,NULL);CHKERRQ(ierr);
617   PetscOptionsEnd();
618 
619   lu->options              = options;
620   lu->options.Fact         = DOFACT;
621   lu->matsolve_iscalled    = PETSC_FALSE;
622   lu->matmatsolve_iscalled = PETSC_FALSE;
623 
624   *F = B;
625   PetscFunctionReturn(0);
626 }
627 
628 #undef __FUNCT__
629 #define __FUNCT__ "MatSolverPackageRegister_SuperLU_DIST"
630 PETSC_EXTERN PetscErrorCode MatSolverPackageRegister_SuperLU_DIST(void)
631 {
632   PetscErrorCode ierr;
633   PetscFunctionBegin;
634   ierr = MatSolverPackageRegister(MATSOLVERSUPERLU_DIST,MATMPIAIJ,  MAT_FACTOR_LU,MatGetFactor_aij_superlu_dist);CHKERRQ(ierr);
635   ierr = MatSolverPackageRegister(MATSOLVERSUPERLU_DIST,MATSEQAIJ,  MAT_FACTOR_LU,MatGetFactor_aij_superlu_dist);CHKERRQ(ierr);
636   PetscFunctionReturn(0);
637 }
638 
639 #undef __FUNCT__
640 #define __FUNCT__ "MatFactorInfo_SuperLU_DIST"
641 PetscErrorCode MatFactorInfo_SuperLU_DIST(Mat A,PetscViewer viewer)
642 {
643   Mat_SuperLU_DIST  *lu=(Mat_SuperLU_DIST*)A->spptr;
644   superlu_options_t options;
645   PetscErrorCode    ierr;
646 
647   PetscFunctionBegin;
648   /* check if matrix is superlu_dist type */
649   if (A->ops->solve != MatSolve_SuperLU_DIST) PetscFunctionReturn(0);
650 
651   options = lu->options;
652   ierr    = PetscViewerASCIIPrintf(viewer,"SuperLU_DIST run parameters:\n");CHKERRQ(ierr);
653   ierr    = PetscViewerASCIIPrintf(viewer,"  Process grid nprow %D x npcol %D \n",lu->nprow,lu->npcol);CHKERRQ(ierr);
654   ierr    = PetscViewerASCIIPrintf(viewer,"  Equilibrate matrix %s \n",PetscBools[options.Equil != NO]);CHKERRQ(ierr);
655   ierr    = PetscViewerASCIIPrintf(viewer,"  Matrix input mode %d \n",lu->MatInputMode);CHKERRQ(ierr);
656   ierr    = PetscViewerASCIIPrintf(viewer,"  Replace tiny pivots %s \n",PetscBools[options.ReplaceTinyPivot != NO]);CHKERRQ(ierr);
657   ierr    = PetscViewerASCIIPrintf(viewer,"  Use iterative refinement %s \n",PetscBools[options.IterRefine == SLU_DOUBLE]);CHKERRQ(ierr);
658   ierr    = PetscViewerASCIIPrintf(viewer,"  Processors in row %d col partition %d \n",lu->nprow,lu->npcol);CHKERRQ(ierr);
659   ierr    = PetscViewerASCIIPrintf(viewer,"  Row permutation %s \n",(options.RowPerm == NOROWPERM) ? "NATURAL" : "LargeDiag");CHKERRQ(ierr);
660 
661   switch (options.ColPerm) {
662   case NATURAL:
663     ierr = PetscViewerASCIIPrintf(viewer,"  Column permutation NATURAL\n");CHKERRQ(ierr);
664     break;
665   case MMD_AT_PLUS_A:
666     ierr = PetscViewerASCIIPrintf(viewer,"  Column permutation MMD_AT_PLUS_A\n");CHKERRQ(ierr);
667     break;
668   case MMD_ATA:
669     ierr = PetscViewerASCIIPrintf(viewer,"  Column permutation MMD_ATA\n");CHKERRQ(ierr);
670     break;
671   case METIS_AT_PLUS_A:
672     ierr = PetscViewerASCIIPrintf(viewer,"  Column permutation METIS_AT_PLUS_A\n");CHKERRQ(ierr);
673     break;
674   case PARMETIS:
675     ierr = PetscViewerASCIIPrintf(viewer,"  Column permutation PARMETIS\n");CHKERRQ(ierr);
676     break;
677   default:
678     SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Unknown column permutation");
679   }
680 
681   ierr = PetscViewerASCIIPrintf(viewer,"  Parallel symbolic factorization %s \n",PetscBools[options.ParSymbFact != NO]);CHKERRQ(ierr);
682 
683   if (lu->FactPattern == SamePattern) {
684     ierr = PetscViewerASCIIPrintf(viewer,"  Repeated factorization SamePattern\n");CHKERRQ(ierr);
685   } else {
686     ierr = PetscViewerASCIIPrintf(viewer,"  Repeated factorization SamePattern_SameRowPerm\n");CHKERRQ(ierr);
687   }
688   PetscFunctionReturn(0);
689 }
690 
691 #undef __FUNCT__
692 #define __FUNCT__ "MatView_SuperLU_DIST"
693 PetscErrorCode MatView_SuperLU_DIST(Mat A,PetscViewer viewer)
694 {
695   PetscErrorCode    ierr;
696   PetscBool         iascii;
697   PetscViewerFormat format;
698 
699   PetscFunctionBegin;
700   ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERASCII,&iascii);CHKERRQ(ierr);
701   if (iascii) {
702     ierr = PetscViewerGetFormat(viewer,&format);CHKERRQ(ierr);
703     if (format == PETSC_VIEWER_ASCII_INFO) {
704       ierr = MatFactorInfo_SuperLU_DIST(A,viewer);CHKERRQ(ierr);
705     }
706   }
707   PetscFunctionReturn(0);
708 }
709 
710 
711 /*MC
712   MATSOLVERSUPERLU_DIST - Parallel direct solver package for LU factorization
713 
714    Works with AIJ matrices
715 
716   Options Database Keys:
717 + -mat_superlu_dist_r <n> - number of rows in processor partition
718 . -mat_superlu_dist_c <n> - number of columns in processor partition
719 . -mat_superlu_dist_matinput <0,1> - matrix input mode; 0=global, 1=distributed
720 . -mat_superlu_dist_equil - equilibrate the matrix
721 . -mat_superlu_dist_rowperm <LargeDiag,NATURAL> - row permutation
722 . -mat_superlu_dist_colperm <MMD_AT_PLUS_A,MMD_ATA,NATURAL> - column permutation
723 . -mat_superlu_dist_replacetinypivot - replace tiny pivots
724 . -mat_superlu_dist_fact <SamePattern> - (choose one of) SamePattern SamePattern_SameRowPerm
725 . -mat_superlu_dist_iterrefine - use iterative refinement
726 - -mat_superlu_dist_statprint - print factorization information
727 
728    Level: beginner
729 
730 .seealso: PCLU
731 
732 .seealso: PCFactorSetMatSolverPackage(), MatSolverPackage
733 
734 M*/
735 
736