xref: /petsc/src/mat/impls/baij/seq/baijfact.c (revision dc5cefde2fdcdb92db1db0a399b233ecd778849b)
1 #define PETSCMAT_DLL
2 
3 /*
4     Factorization code for BAIJ format.
5 */
6 #include "../src/mat/impls/baij/seq/baij.h"
7 #include "../src/mat/blockinvert.h"
8 
9 #undef __FUNCT__
10 #define __FUNCT__ "MatLUFactorNumeric_SeqBAIJ_2_newdatastruct"
11 PetscErrorCode MatLUFactorNumeric_SeqBAIJ_2_newdatastruct(Mat B,Mat A,const MatFactorInfo *info)
12 {
13   Mat            C=B;
14   Mat_SeqBAIJ    *a=(Mat_SeqBAIJ*)A->data,*b=(Mat_SeqBAIJ *)C->data;
15   IS             isrow = b->row,isicol = b->icol;
16   PetscErrorCode ierr;
17   const PetscInt *r,*ic,*ics;
18   PetscInt       i,j,k,n=a->mbs,*ai=a->i,*aj=a->j,*bi=b->i,*bj=b->j;
19   PetscInt       *ajtmp,*bjtmp,nz,nzL,row,*bdiag=b->diag,*pj;
20   MatScalar      *rtmp,*pc,*mwork,*v,*pv,*aa=a->a;
21   PetscInt       bs2 = a->bs2,flg;
22   PetscReal      shift = info->shiftinblocks;
23 
24   PetscFunctionBegin;
25   ierr = ISGetIndices(isrow,&r);CHKERRQ(ierr);
26   ierr = ISGetIndices(isicol,&ic);CHKERRQ(ierr);
27 
28   /* generate work space needed by the factorization */
29   ierr = PetscMalloc2(bs2*n,MatScalar,&rtmp,bs2,MatScalar,&mwork);CHKERRQ(ierr);
30   ierr = PetscMemzero(rtmp,bs2*n*sizeof(MatScalar));CHKERRQ(ierr);
31   ics  = ic;
32 
33   for (i=0; i<n; i++){
34     /* zero rtmp */
35     /* L part */
36     nz    = bi[i+1] - bi[i];
37     bjtmp = bj + bi[i];
38     for  (j=0; j<nz; j++){
39       ierr = PetscMemzero(rtmp+bs2*bjtmp[j],bs2*sizeof(MatScalar));CHKERRQ(ierr);
40     }
41 
42     /* U part */
43     nz = bdiag[i] - bdiag[i+1];
44     bjtmp = bj + bdiag[i+1]+1;
45     for  (j=0; j<nz; j++){
46       ierr = PetscMemzero(rtmp+bs2*bjtmp[j],bs2*sizeof(MatScalar));CHKERRQ(ierr);
47     }
48 
49     /* load in initial (unfactored row) */
50     nz    = ai[r[i]+1] - ai[r[i]];
51     ajtmp = aj + ai[r[i]];
52     v     = aa + bs2*ai[r[i]];
53     for (j=0; j<nz; j++) {
54       ierr = PetscMemcpy(rtmp+bs2*ic[ajtmp[j]],v+bs2*j,bs2*sizeof(MatScalar));CHKERRQ(ierr);
55     }
56 
57     /* elimination */
58     bjtmp = bj + bi[i];
59     nzL   = bi[i+1] - bi[i];
60     for(k=0;k < nzL;k++) {
61       row = bjtmp[k];
62       pc = rtmp + bs2*row;
63       for (flg=0,j=0; j<bs2; j++) { if (pc[j]!=0.0) { flg = 1; break; }}
64       if (flg) {
65         pv = b->a + bs2*bdiag[row];
66         /* Kernel_A_gets_A_times_B(bs,pc,pv,mwork); *pc = *pc * (*pv); */
67         ierr = Kernel_A_gets_A_times_B_2(pc,pv,mwork);CHKERRQ(ierr);
68 
69         pj = b->j + bdiag[row+1]+1; /* begining of U(row,:) */
70         pv = b->a + bs2*(bdiag[row+1]+1);
71         nz = bdiag[row] - bdiag[row+1] - 1; /* num of entries inU(row,:), excluding diag */
72         for (j=0; j<nz; j++) {
73           /* Kernel_A_gets_A_minus_B_times_C(bs,rtmp+bs2*pj[j],pc,pv+bs2*j); */
74           /* rtmp+bs2*pj[j] = rtmp+bs2*pj[j] - (*pc)*(pv+bs2*j) */
75           v    = rtmp + 4*pj[j];
76           ierr = Kernel_A_gets_A_minus_B_times_C_2(v,pc,pv);CHKERRQ(ierr);
77           pv  += 4;
78         }
79         ierr = PetscLogFlops(16*nz+12);CHKERRQ(ierr); /* flops = 2*bs^3*nz + 2*bs^3 - bs2) */
80       }
81     }
82 
83     /* finished row so stick it into b->a */
84     /* L part */
85     pv   = b->a + bs2*bi[i] ;
86     pj   = b->j + bi[i] ;
87     nz   = bi[i+1] - bi[i];
88     for (j=0; j<nz; j++) {
89       ierr = PetscMemcpy(pv+bs2*j,rtmp+bs2*pj[j],bs2*sizeof(MatScalar));CHKERRQ(ierr);
90     }
91 
92     /* Mark diagonal and invert diagonal for simplier triangular solves */
93     pv   = b->a + bs2*bdiag[i];
94     pj   = b->j + bdiag[i];
95     ierr = PetscMemcpy(pv,rtmp+bs2*pj[0],bs2*sizeof(MatScalar));CHKERRQ(ierr);
96     /* ierr = Kernel_A_gets_inverse_A(bs,pv,v_pivots,v_work);CHKERRQ(ierr); */
97     ierr = Kernel_A_gets_inverse_A_2(pv,shift);CHKERRQ(ierr);
98 
99     /* U part */
100     pv = b->a + bs2*(bdiag[i+1]+1);
101     pj = b->j + bdiag[i+1]+1;
102     nz = bdiag[i] - bdiag[i+1] - 1;
103     for (j=0; j<nz; j++){
104       ierr = PetscMemcpy(pv+bs2*j,rtmp+bs2*pj[j],bs2*sizeof(MatScalar));CHKERRQ(ierr);
105     }
106   }
107 
108   ierr = PetscFree2(rtmp,mwork);CHKERRQ(ierr);
109   ierr = ISRestoreIndices(isicol,&ic);CHKERRQ(ierr);
110   ierr = ISRestoreIndices(isrow,&r);CHKERRQ(ierr);
111   C->ops->solve          = MatSolve_SeqBAIJ_2_newdatastruct;
112   C->ops->solvetranspose = MatSolveTranspose_SeqBAIJ_2_newdatastruct;
113 
114   C->assembled = PETSC_TRUE;
115   ierr = PetscLogFlops(1.3333*bs2*n);CHKERRQ(ierr); /* from inverting diagonal blocks */
116   PetscFunctionReturn(0);
117 }
118 
119 #undef __FUNCT__
120 #define __FUNCT__ "MatLUFactorNumeric_SeqBAIJ_2_NaturalOrdering_newdatastruct"
121 PetscErrorCode MatLUFactorNumeric_SeqBAIJ_2_NaturalOrdering_newdatastruct(Mat B,Mat A,const MatFactorInfo *info)
122 {
123   Mat            C=B;
124   Mat_SeqBAIJ    *a=(Mat_SeqBAIJ*)A->data,*b=(Mat_SeqBAIJ *)C->data;
125   PetscErrorCode ierr;
126   PetscInt       i,j,k,n=a->mbs,*ai=a->i,*aj=a->j,*bi=b->i,*bj=b->j;
127   PetscInt       *ajtmp,*bjtmp,nz,nzL,row,*bdiag=b->diag,*pj;
128   MatScalar      *rtmp,*pc,*mwork,*v,*pv,*aa=a->a;
129   PetscInt       bs2 = a->bs2,flg;
130   PetscReal      shift = info->shiftinblocks;
131 
132   PetscFunctionBegin;
133   /* generate work space needed by the factorization */
134   ierr = PetscMalloc2(bs2*n,MatScalar,&rtmp,bs2,MatScalar,&mwork);CHKERRQ(ierr);
135   ierr = PetscMemzero(rtmp,bs2*n*sizeof(MatScalar));CHKERRQ(ierr);
136 
137   for (i=0; i<n; i++){
138     /* zero rtmp */
139     /* L part */
140     nz    = bi[i+1] - bi[i];
141     bjtmp = bj + bi[i];
142     for  (j=0; j<nz; j++){
143       ierr = PetscMemzero(rtmp+bs2*bjtmp[j],bs2*sizeof(MatScalar));CHKERRQ(ierr);
144     }
145 
146     /* U part */
147     nz = bdiag[i] - bdiag[i+1];
148     bjtmp = bj + bdiag[i+1]+1;
149     for  (j=0; j<nz; j++){
150       ierr = PetscMemzero(rtmp+bs2*bjtmp[j],bs2*sizeof(MatScalar));CHKERRQ(ierr);
151     }
152 
153     /* load in initial (unfactored row) */
154     nz    = ai[i+1] - ai[i];
155     ajtmp = aj + ai[i];
156     v     = aa + bs2*ai[i];
157     for (j=0; j<nz; j++) {
158       ierr = PetscMemcpy(rtmp+bs2*ajtmp[j],v+bs2*j,bs2*sizeof(MatScalar));CHKERRQ(ierr);
159     }
160 
161     /* elimination */
162     bjtmp = bj + bi[i];
163     nzL   = bi[i+1] - bi[i];
164     for(k=0;k < nzL;k++) {
165       row = bjtmp[k];
166       pc = rtmp + bs2*row;
167       for (flg=0,j=0; j<bs2; j++) { if (pc[j]!=0.0) { flg = 1; break; }}
168       if (flg) {
169         pv = b->a + bs2*bdiag[row];
170         /* Kernel_A_gets_A_times_B(bs,pc,pv,mwork); *pc = *pc * (*pv); */
171         ierr = Kernel_A_gets_A_times_B_2(pc,pv,mwork);CHKERRQ(ierr);
172 
173         pj = b->j + bdiag[row+1]+1; /* beginning of U(row,:) */
174 	pv = b->a + bs2*(bdiag[row+1]+1);
175 	nz = bdiag[row]-bdiag[row+1] - 1; /* num of entries in U(row,:) excluding diag */
176         for (j=0; j<nz; j++) {
177           /* Kernel_A_gets_A_minus_B_times_C(bs,rtmp+bs2*pj[j],pc,pv+bs2*j); */
178           /* rtmp+bs2*pj[j] = rtmp+bs2*pj[j] - (*pc)*(pv+bs2*j) */
179           v    = rtmp + 4*pj[j];
180           ierr = Kernel_A_gets_A_minus_B_times_C_2(v,pc,pv);CHKERRQ(ierr);
181           pv  += 4;
182         }
183         ierr = PetscLogFlops(16*nz+12);CHKERRQ(ierr); /* flops = 2*bs^3*nz + 2*bs^3 - bs2) */
184       }
185     }
186 
187     /* finished row so stick it into b->a */
188     /* L part */
189     pv   = b->a + bs2*bi[i] ;
190     pj   = b->j + bi[i] ;
191     nz   = bi[i+1] - bi[i];
192     for (j=0; j<nz; j++) {
193       ierr = PetscMemcpy(pv+bs2*j,rtmp+bs2*pj[j],bs2*sizeof(MatScalar));CHKERRQ(ierr);
194     }
195 
196     /* Mark diagonal and invert diagonal for simplier triangular solves */
197     pv   = b->a + bs2*bdiag[i];
198     pj   = b->j + bdiag[i];
199     ierr = PetscMemcpy(pv,rtmp+bs2*pj[0],bs2*sizeof(MatScalar));CHKERRQ(ierr);
200     /* ierr = Kernel_A_gets_inverse_A(bs,pv,v_pivots,v_work);CHKERRQ(ierr); */
201     ierr = Kernel_A_gets_inverse_A_2(pv,shift);CHKERRQ(ierr);
202 
203     /* U part */
204     /*
205     pv = b->a + bs2*bi[2*n-i];
206     pj = b->j + bi[2*n-i];
207     nz = bi[2*n-i+1] - bi[2*n-i] - 1;
208     */
209     pv = b->a + bs2*(bdiag[i+1]+1);
210     pj = b->j + bdiag[i+1]+1;
211     nz = bdiag[i] - bdiag[i+1] - 1;
212     for (j=0; j<nz; j++){
213       ierr = PetscMemcpy(pv+bs2*j,rtmp+bs2*pj[j],bs2*sizeof(MatScalar));CHKERRQ(ierr);
214     }
215   }
216   ierr = PetscFree2(rtmp,mwork);CHKERRQ(ierr);
217 
218   C->ops->solve          = MatSolve_SeqBAIJ_2_NaturalOrdering_newdatastruct;
219   C->ops->solvetranspose = MatSolveTranspose_SeqBAIJ_2_NaturalOrdering_newdatastruct;
220   C->assembled = PETSC_TRUE;
221   ierr = PetscLogFlops(1.3333*bs2*n);CHKERRQ(ierr); /* from inverting diagonal blocks */
222   PetscFunctionReturn(0);
223 }
224 
225 #undef __FUNCT__
226 #define __FUNCT__ "MatLUFactorNumeric_SeqBAIJ_2"
227 PetscErrorCode MatLUFactorNumeric_SeqBAIJ_2(Mat B,Mat A,const MatFactorInfo *info)
228 {
229   Mat            C = B;
230   Mat_SeqBAIJ    *a = (Mat_SeqBAIJ*)A->data,*b = (Mat_SeqBAIJ *)C->data;
231   IS             isrow = b->row,isicol = b->icol;
232   PetscErrorCode ierr;
233   const PetscInt *r,*ic;
234   PetscInt       i,j,n = a->mbs,*bi = b->i,*bj = b->j;
235   PetscInt       *ajtmpold,*ajtmp,nz,row;
236   PetscInt       *diag_offset=b->diag,idx,*ai=a->i,*aj=a->j,*pj;
237   MatScalar      *pv,*v,*rtmp,m1,m2,m3,m4,*pc,*w,*x,x1,x2,x3,x4;
238   MatScalar      p1,p2,p3,p4;
239   MatScalar      *ba = b->a,*aa = a->a;
240   PetscReal      shift = info->shiftinblocks;
241 
242   PetscFunctionBegin;
243   ierr  = ISGetIndices(isrow,&r);CHKERRQ(ierr);
244   ierr  = ISGetIndices(isicol,&ic);CHKERRQ(ierr);
245   ierr  = PetscMalloc(4*(n+1)*sizeof(MatScalar),&rtmp);CHKERRQ(ierr);
246 
247   for (i=0; i<n; i++) {
248     nz    = bi[i+1] - bi[i];
249     ajtmp = bj + bi[i];
250     for  (j=0; j<nz; j++) {
251       x = rtmp+4*ajtmp[j]; x[0] = x[1] = x[2] = x[3] = 0.0;
252     }
253     /* load in initial (unfactored row) */
254     idx      = r[i];
255     nz       = ai[idx+1] - ai[idx];
256     ajtmpold = aj + ai[idx];
257     v        = aa + 4*ai[idx];
258     for (j=0; j<nz; j++) {
259       x    = rtmp+4*ic[ajtmpold[j]];
260       x[0] = v[0]; x[1] = v[1]; x[2] = v[2]; x[3] = v[3];
261       v    += 4;
262     }
263     row = *ajtmp++;
264     while (row < i) {
265       pc = rtmp + 4*row;
266       p1 = pc[0]; p2 = pc[1]; p3 = pc[2]; p4 = pc[3];
267       if (p1 != 0.0 || p2 != 0.0 || p3 != 0.0 || p4 != 0.0) {
268         pv = ba + 4*diag_offset[row];
269         pj = bj + diag_offset[row] + 1;
270         x1 = pv[0]; x2 = pv[1]; x3 = pv[2]; x4 = pv[3];
271         pc[0] = m1 = p1*x1 + p3*x2;
272         pc[1] = m2 = p2*x1 + p4*x2;
273         pc[2] = m3 = p1*x3 + p3*x4;
274         pc[3] = m4 = p2*x3 + p4*x4;
275         nz = bi[row+1] - diag_offset[row] - 1;
276         pv += 4;
277         for (j=0; j<nz; j++) {
278           x1   = pv[0]; x2 = pv[1]; x3 = pv[2]; x4 = pv[3];
279           x    = rtmp + 4*pj[j];
280           x[0] -= m1*x1 + m3*x2;
281           x[1] -= m2*x1 + m4*x2;
282           x[2] -= m1*x3 + m3*x4;
283           x[3] -= m2*x3 + m4*x4;
284           pv   += 4;
285         }
286         ierr = PetscLogFlops(16.0*nz+12.0);CHKERRQ(ierr);
287       }
288       row = *ajtmp++;
289     }
290     /* finished row so stick it into b->a */
291     pv = ba + 4*bi[i];
292     pj = bj + bi[i];
293     nz = bi[i+1] - bi[i];
294     for (j=0; j<nz; j++) {
295       x     = rtmp+4*pj[j];
296       pv[0] = x[0]; pv[1] = x[1]; pv[2] = x[2]; pv[3] = x[3];
297       pv   += 4;
298     }
299     /* invert diagonal block */
300     w = ba + 4*diag_offset[i];
301     ierr = Kernel_A_gets_inverse_A_2(w,shift);CHKERRQ(ierr);
302   }
303 
304   ierr = PetscFree(rtmp);CHKERRQ(ierr);
305   ierr = ISRestoreIndices(isicol,&ic);CHKERRQ(ierr);
306   ierr = ISRestoreIndices(isrow,&r);CHKERRQ(ierr);
307   C->ops->solve          = MatSolve_SeqBAIJ_2;
308   C->ops->solvetranspose = MatSolveTranspose_SeqBAIJ_2;
309   C->assembled = PETSC_TRUE;
310   ierr = PetscLogFlops(1.3333*8*b->mbs);CHKERRQ(ierr); /* from inverting diagonal blocks */
311   PetscFunctionReturn(0);
312 }
313 /*
314       Version for when blocks are 2 by 2 Using natural ordering
315 */
316 #undef __FUNCT__
317 #define __FUNCT__ "MatLUFactorNumeric_SeqBAIJ_2_NaturalOrdering"
318 PetscErrorCode MatLUFactorNumeric_SeqBAIJ_2_NaturalOrdering(Mat C,Mat A,const MatFactorInfo *info)
319 {
320   Mat_SeqBAIJ    *a = (Mat_SeqBAIJ*)A->data,*b = (Mat_SeqBAIJ *)C->data;
321   PetscErrorCode ierr;
322   PetscInt       i,j,n = a->mbs,*bi = b->i,*bj = b->j;
323   PetscInt       *ajtmpold,*ajtmp,nz,row;
324   PetscInt       *diag_offset = b->diag,*ai=a->i,*aj=a->j,*pj;
325   MatScalar      *pv,*v,*rtmp,*pc,*w,*x;
326   MatScalar      p1,p2,p3,p4,m1,m2,m3,m4,x1,x2,x3,x4;
327   MatScalar      *ba = b->a,*aa = a->a;
328   PetscReal      shift = info->shiftinblocks;
329 
330   PetscFunctionBegin;
331   ierr = PetscMalloc(4*(n+1)*sizeof(MatScalar),&rtmp);CHKERRQ(ierr);
332   for (i=0; i<n; i++) {
333     nz    = bi[i+1] - bi[i];
334     ajtmp = bj + bi[i];
335     for  (j=0; j<nz; j++) {
336       x = rtmp+4*ajtmp[j];
337       x[0]  = x[1]  = x[2]  = x[3]  = 0.0;
338     }
339     /* load in initial (unfactored row) */
340     nz       = ai[i+1] - ai[i];
341     ajtmpold = aj + ai[i];
342     v        = aa + 4*ai[i];
343     for (j=0; j<nz; j++) {
344       x    = rtmp+4*ajtmpold[j];
345       x[0]  = v[0];  x[1]  = v[1];  x[2]  = v[2];  x[3]  = v[3];
346       v    += 4;
347     }
348     row = *ajtmp++;
349     while (row < i) {
350       pc  = rtmp + 4*row;
351       p1  = pc[0];  p2  = pc[1];  p3  = pc[2];  p4  = pc[3];
352       if (p1 != 0.0 || p2 != 0.0 || p3 != 0.0 || p4 != 0.0) {
353         pv = ba + 4*diag_offset[row];
354         pj = bj + diag_offset[row] + 1;
355         x1  = pv[0];  x2  = pv[1];  x3  = pv[2];  x4  = pv[3];
356         pc[0] = m1 = p1*x1 + p3*x2;
357         pc[1] = m2 = p2*x1 + p4*x2;
358         pc[2] = m3 = p1*x3 + p3*x4;
359         pc[3] = m4 = p2*x3 + p4*x4;
360         nz = bi[row+1] - diag_offset[row] - 1;
361         pv += 4;
362         for (j=0; j<nz; j++) {
363           x1   = pv[0];  x2  = pv[1];   x3 = pv[2];  x4  = pv[3];
364           x    = rtmp + 4*pj[j];
365           x[0] -= m1*x1 + m3*x2;
366           x[1] -= m2*x1 + m4*x2;
367           x[2] -= m1*x3 + m3*x4;
368           x[3] -= m2*x3 + m4*x4;
369           pv   += 4;
370         }
371         ierr = PetscLogFlops(16.0*nz+12.0);CHKERRQ(ierr);
372       }
373       row = *ajtmp++;
374     }
375     /* finished row so stick it into b->a */
376     pv = ba + 4*bi[i];
377     pj = bj + bi[i];
378     nz = bi[i+1] - bi[i];
379     for (j=0; j<nz; j++) {
380       x      = rtmp+4*pj[j];
381       pv[0]  = x[0];  pv[1]  = x[1];  pv[2]  = x[2];  pv[3]  = x[3];
382       /*
383       printf(" col %d:",pj[j]);
384       PetscInt j1;
385       for (j1=0; j1<4; j1++) printf(" %g,",*(pv+j1));
386       printf("\n");
387       */
388       pv   += 4;
389     }
390     /* invert diagonal block */
391     w = ba + 4*diag_offset[i];
392     /*
393     printf(" \n%d -th: diag: ",i);
394     for (j=0; j<4; j++){
395       printf(" %g,",w[j]);
396     }
397     printf("\n----------------------------\n");
398     */
399     ierr = Kernel_A_gets_inverse_A_2(w,shift);CHKERRQ(ierr);
400   }
401 
402   ierr = PetscFree(rtmp);CHKERRQ(ierr);
403   C->ops->solve          = MatSolve_SeqBAIJ_2_NaturalOrdering;
404   C->ops->solvetranspose = MatSolveTranspose_SeqBAIJ_2_NaturalOrdering;
405   C->assembled = PETSC_TRUE;
406   ierr = PetscLogFlops(1.3333*8*b->mbs);CHKERRQ(ierr); /* from inverting diagonal blocks */
407   PetscFunctionReturn(0);
408 }
409 
410 /* ----------------------------------------------------------- */
411 /*
412      Version for when blocks are 1 by 1.
413 */
414 #undef __FUNCT__
415 #define __FUNCT__ "MatLUFactorNumeric_SeqBAIJ_1"
416 PetscErrorCode MatLUFactorNumeric_SeqBAIJ_1(Mat C,Mat A,const MatFactorInfo *info)
417 {
418   Mat_SeqBAIJ    *a = (Mat_SeqBAIJ*)A->data,*b = (Mat_SeqBAIJ *)C->data;
419   IS             isrow = b->row,isicol = b->icol;
420   PetscErrorCode ierr;
421   const PetscInt *r,*ic;
422   PetscInt       i,j,n = a->mbs,*bi = b->i,*bj = b->j;
423   PetscInt       *ajtmpold,*ajtmp,nz,row,*ai = a->i,*aj = a->j;
424   PetscInt       *diag_offset = b->diag,diag,*pj;
425   MatScalar      *pv,*v,*rtmp,multiplier,*pc;
426   MatScalar      *ba = b->a,*aa = a->a;
427   PetscTruth     row_identity, col_identity;
428 
429   PetscFunctionBegin;
430   ierr  = ISGetIndices(isrow,&r);CHKERRQ(ierr);
431   ierr  = ISGetIndices(isicol,&ic);CHKERRQ(ierr);
432   ierr  = PetscMalloc((n+1)*sizeof(MatScalar),&rtmp);CHKERRQ(ierr);
433 
434   for (i=0; i<n; i++) {
435     nz    = bi[i+1] - bi[i];
436     ajtmp = bj + bi[i];
437     for  (j=0; j<nz; j++) rtmp[ajtmp[j]] = 0.0;
438 
439     /* load in initial (unfactored row) */
440     nz       = ai[r[i]+1] - ai[r[i]];
441     ajtmpold = aj + ai[r[i]];
442     v        = aa + ai[r[i]];
443     for (j=0; j<nz; j++) rtmp[ic[ajtmpold[j]]] =  v[j];
444 
445     row = *ajtmp++;
446     while (row < i) {
447       pc = rtmp + row;
448       if (*pc != 0.0) {
449         pv         = ba + diag_offset[row];
450         pj         = bj + diag_offset[row] + 1;
451         multiplier = *pc * *pv++;
452         *pc        = multiplier;
453         nz         = bi[row+1] - diag_offset[row] - 1;
454         for (j=0; j<nz; j++) rtmp[pj[j]] -= multiplier * pv[j];
455         ierr = PetscLogFlops(1.0+2.0*nz);CHKERRQ(ierr);
456       }
457       row = *ajtmp++;
458     }
459     /* finished row so stick it into b->a */
460     pv = ba + bi[i];
461     pj = bj + bi[i];
462     nz = bi[i+1] - bi[i];
463     for (j=0; j<nz; j++) {pv[j] = rtmp[pj[j]];}
464     diag = diag_offset[i] - bi[i];
465     /* check pivot entry for current row */
466     if (pv[diag] == 0.0) {
467       SETERRQ2(PETSC_ERR_MAT_LU_ZRPVT,"Zero pivot: row in original ordering %D in permuted ordering %D",r[i],i);
468     }
469     pv[diag] = 1.0/pv[diag];
470   }
471 
472   ierr = PetscFree(rtmp);CHKERRQ(ierr);
473   ierr = ISRestoreIndices(isicol,&ic);CHKERRQ(ierr);
474   ierr = ISRestoreIndices(isrow,&r);CHKERRQ(ierr);
475   ierr = ISIdentity(isrow,&row_identity);CHKERRQ(ierr);
476   ierr = ISIdentity(isicol,&col_identity);CHKERRQ(ierr);
477   if (row_identity && col_identity) {
478     C->ops->solve          = MatSolve_SeqBAIJ_1_NaturalOrdering;
479     C->ops->solvetranspose = MatSolveTranspose_SeqBAIJ_1_NaturalOrdering;
480   } else {
481     C->ops->solve          = MatSolve_SeqBAIJ_1;
482     C->ops->solvetranspose = MatSolveTranspose_SeqBAIJ_1;
483   }
484   C->assembled = PETSC_TRUE;
485   ierr = PetscLogFlops(C->cmap->n);CHKERRQ(ierr);
486   PetscFunctionReturn(0);
487 }
488 
489 EXTERN_C_BEGIN
490 #undef __FUNCT__
491 #define __FUNCT__ "MatGetFactor_seqbaij_petsc"
492 PetscErrorCode MatGetFactor_seqbaij_petsc(Mat A,MatFactorType ftype,Mat *B)
493 {
494   PetscInt           n = A->rmap->n;
495   PetscErrorCode     ierr;
496 
497   PetscFunctionBegin;
498   ierr = MatCreate(((PetscObject)A)->comm,B);CHKERRQ(ierr);
499   ierr = MatSetSizes(*B,n,n,n,n);CHKERRQ(ierr);
500   if (ftype == MAT_FACTOR_LU || ftype == MAT_FACTOR_ILU || ftype == MAT_FACTOR_ILUDT) {
501     ierr = MatSetType(*B,MATSEQBAIJ);CHKERRQ(ierr);
502     (*B)->ops->lufactorsymbolic  = MatLUFactorSymbolic_SeqBAIJ;
503     (*B)->ops->ilufactorsymbolic = MatILUFactorSymbolic_SeqBAIJ;
504     (*B)->ops->iludtfactor       = MatILUDTFactor_SeqBAIJ;
505   } else if (ftype == MAT_FACTOR_CHOLESKY || ftype == MAT_FACTOR_ICC) {
506     ierr = MatSetType(*B,MATSEQSBAIJ);CHKERRQ(ierr);
507     ierr = MatSeqSBAIJSetPreallocation(*B,1,MAT_SKIP_ALLOCATION,PETSC_NULL);CHKERRQ(ierr);
508     (*B)->ops->iccfactorsymbolic      = MatICCFactorSymbolic_SeqBAIJ;
509     (*B)->ops->choleskyfactorsymbolic = MatCholeskyFactorSymbolic_SeqBAIJ;
510   } else SETERRQ(PETSC_ERR_SUP,"Factor type not supported");
511   (*B)->factor = ftype;
512   PetscFunctionReturn(0);
513 }
514 EXTERN_C_END
515 
516 EXTERN_C_BEGIN
517 #undef __FUNCT__
518 #define __FUNCT__ "MatGetFactorAvailable_seqaij_petsc"
519 PetscErrorCode MatGetFactorAvailable_seqbaij_petsc(Mat A,MatFactorType ftype,PetscTruth *flg)
520 {
521   PetscFunctionBegin;
522   *flg = PETSC_TRUE;
523   PetscFunctionReturn(0);
524 }
525 EXTERN_C_END
526 
527 /* ----------------------------------------------------------- */
528 #undef __FUNCT__
529 #define __FUNCT__ "MatLUFactor_SeqBAIJ"
530 PetscErrorCode MatLUFactor_SeqBAIJ(Mat A,IS row,IS col,const MatFactorInfo *info)
531 {
532   PetscErrorCode ierr;
533   Mat            C;
534 
535   PetscFunctionBegin;
536   ierr = MatGetFactor(A,MAT_SOLVER_PETSC,MAT_FACTOR_LU,&C);CHKERRQ(ierr);
537   ierr = MatLUFactorSymbolic(C,A,row,col,info);CHKERRQ(ierr);
538   ierr = MatLUFactorNumeric(C,A,info);CHKERRQ(ierr);
539   A->ops->solve            = C->ops->solve;
540   A->ops->solvetranspose   = C->ops->solvetranspose;
541   ierr = MatHeaderCopy(A,C);CHKERRQ(ierr);
542   ierr = PetscLogObjectParent(A,((Mat_SeqBAIJ*)(A->data))->icol);CHKERRQ(ierr);
543   PetscFunctionReturn(0);
544 }
545 
546 #include "../src/mat/impls/sbaij/seq/sbaij.h"
547 #undef __FUNCT__
548 #define __FUNCT__ "MatCholeskyFactorNumeric_SeqBAIJ_N"
549 PetscErrorCode MatCholeskyFactorNumeric_SeqBAIJ_N(Mat C,Mat A,const MatFactorInfo *info)
550 {
551   PetscErrorCode ierr;
552   Mat_SeqBAIJ    *a=(Mat_SeqBAIJ*)A->data;
553   Mat_SeqSBAIJ   *b=(Mat_SeqSBAIJ*)C->data;
554   IS             ip=b->row;
555   const PetscInt *rip;
556   PetscInt       i,j,mbs=a->mbs,bs=A->rmap->bs,*bi=b->i,*bj=b->j,*bcol;
557   PetscInt       *ai=a->i,*aj=a->j;
558   PetscInt       k,jmin,jmax,*jl,*il,col,nexti,ili,nz;
559   MatScalar      *rtmp,*ba=b->a,*bval,*aa=a->a,dk,uikdi;
560   PetscReal      zeropivot,rs,shiftnz;
561   PetscReal      shiftpd;
562   ChShift_Ctx    sctx;
563   PetscInt       newshift;
564 
565   PetscFunctionBegin;
566   if (bs > 1) {
567     if (!a->sbaijMat){
568       ierr = MatConvert(A,MATSEQSBAIJ,MAT_INITIAL_MATRIX,&a->sbaijMat);CHKERRQ(ierr);
569     }
570     ierr = (a->sbaijMat)->ops->choleskyfactornumeric(C,a->sbaijMat,info);CHKERRQ(ierr);
571     ierr = MatDestroy(a->sbaijMat);CHKERRQ(ierr);
572     a->sbaijMat = PETSC_NULL;
573     PetscFunctionReturn(0);
574   }
575 
576   /* initialization */
577   shiftnz   = info->shiftnz;
578   shiftpd   = info->shiftpd;
579   zeropivot = info->zeropivot;
580 
581   ierr  = ISGetIndices(ip,&rip);CHKERRQ(ierr);
582   ierr = PetscMalloc3(mbs,MatScalar,&rtmp,mbs,PetscInt,&il,mbs,PetscInt,&jl);CHKERRQ(ierr);
583 
584   sctx.shift_amount = 0.;
585   sctx.nshift       = 0;
586   do {
587     sctx.chshift = PETSC_FALSE;
588     for (i=0; i<mbs; i++) {
589       rtmp[i] = 0.0; jl[i] = mbs; il[0] = 0.;
590     }
591 
592     for (k = 0; k<mbs; k++){
593       bval = ba + bi[k];
594       /* initialize k-th row by the perm[k]-th row of A */
595       jmin = ai[rip[k]]; jmax = ai[rip[k]+1];
596       for (j = jmin; j < jmax; j++){
597         col = rip[aj[j]];
598         if (col >= k){ /* only take upper triangular entry */
599           rtmp[col] = aa[j];
600           *bval++  = 0.0; /* for in-place factorization */
601         }
602       }
603 
604       /* shift the diagonal of the matrix */
605       if (sctx.nshift) rtmp[k] += sctx.shift_amount;
606 
607       /* modify k-th row by adding in those rows i with U(i,k)!=0 */
608       dk = rtmp[k];
609       i = jl[k]; /* first row to be added to k_th row  */
610 
611       while (i < k){
612         nexti = jl[i]; /* next row to be added to k_th row */
613 
614         /* compute multiplier, update diag(k) and U(i,k) */
615         ili = il[i];  /* index of first nonzero element in U(i,k:bms-1) */
616         uikdi = - ba[ili]*ba[bi[i]];  /* diagonal(k) */
617         dk += uikdi*ba[ili];
618         ba[ili] = uikdi; /* -U(i,k) */
619 
620         /* add multiple of row i to k-th row */
621         jmin = ili + 1; jmax = bi[i+1];
622         if (jmin < jmax){
623           for (j=jmin; j<jmax; j++) rtmp[bj[j]] += uikdi*ba[j];
624           /* update il and jl for row i */
625           il[i] = jmin;
626           j = bj[jmin]; jl[i] = jl[j]; jl[j] = i;
627         }
628         i = nexti;
629       }
630 
631       /* shift the diagonals when zero pivot is detected */
632       /* compute rs=sum of abs(off-diagonal) */
633       rs   = 0.0;
634       jmin = bi[k]+1;
635       nz   = bi[k+1] - jmin;
636       if (nz){
637         bcol = bj + jmin;
638         while (nz--){
639           rs += PetscAbsScalar(rtmp[*bcol]);
640           bcol++;
641         }
642       }
643 
644       sctx.rs = rs;
645       sctx.pv = dk;
646       ierr = MatCholeskyCheckShift_inline(info,sctx,k,newshift);CHKERRQ(ierr);
647       if (newshift == 1) break;
648 
649       /* copy data into U(k,:) */
650       ba[bi[k]] = 1.0/dk; /* U(k,k) */
651       jmin = bi[k]+1; jmax = bi[k+1];
652       if (jmin < jmax) {
653         for (j=jmin; j<jmax; j++){
654           col = bj[j]; ba[j] = rtmp[col]; rtmp[col] = 0.0;
655         }
656         /* add the k-th row into il and jl */
657         il[k] = jmin;
658         i = bj[jmin]; jl[k] = jl[i]; jl[i] = k;
659       }
660     }
661   } while (sctx.chshift);
662   ierr = PetscFree3(rtmp,il,jl);CHKERRQ(ierr);
663 
664   ierr = ISRestoreIndices(ip,&rip);CHKERRQ(ierr);
665   C->assembled    = PETSC_TRUE;
666   C->preallocated = PETSC_TRUE;
667   ierr = PetscLogFlops(C->rmap->N);CHKERRQ(ierr);
668   if (sctx.nshift){
669     if (shiftpd) {
670       ierr = PetscInfo2(A,"number of shiftpd tries %D, shift_amount %G\n",sctx.nshift,sctx.shift_amount);CHKERRQ(ierr);
671     } else if (shiftnz) {
672       ierr = PetscInfo2(A,"number of shiftnz tries %D, shift_amount %G\n",sctx.nshift,sctx.shift_amount);CHKERRQ(ierr);
673     }
674   }
675   PetscFunctionReturn(0);
676 }
677 
678 #undef __FUNCT__
679 #define __FUNCT__ "MatCholeskyFactorNumeric_SeqBAIJ_N_NaturalOrdering"
680 PetscErrorCode MatCholeskyFactorNumeric_SeqBAIJ_N_NaturalOrdering(Mat C,Mat A,const MatFactorInfo *info)
681 {
682   Mat_SeqBAIJ    *a=(Mat_SeqBAIJ*)A->data;
683   Mat_SeqSBAIJ   *b=(Mat_SeqSBAIJ*)C->data;
684   PetscErrorCode ierr;
685   PetscInt       i,j,am=a->mbs;
686   PetscInt       *ai=a->i,*aj=a->j,*bi=b->i,*bj=b->j;
687   PetscInt       k,jmin,*jl,*il,nexti,ili,*acol,*bcol,nz;
688   MatScalar      *rtmp,*ba=b->a,*aa=a->a,dk,uikdi,*aval,*bval;
689   PetscReal      zeropivot,rs,shiftnz;
690   PetscReal      shiftpd;
691   ChShift_Ctx    sctx;
692   PetscInt       newshift;
693 
694   PetscFunctionBegin;
695   /* initialization */
696   shiftnz   = info->shiftnz;
697   shiftpd   = info->shiftpd;
698   zeropivot = info->zeropivot;
699 
700   ierr = PetscMalloc3(am,MatScalar,&rtmp,am,PetscInt,&il,am,PetscInt,&jl);CHKERRQ(ierr);
701 
702   sctx.shift_amount = 0.;
703   sctx.nshift       = 0;
704   do {
705     sctx.chshift = PETSC_FALSE;
706     for (i=0; i<am; i++) {
707       rtmp[i] = 0.0; jl[i] = am; il[0] = 0.;
708     }
709 
710     for (k = 0; k<am; k++){
711     /* initialize k-th row with elements nonzero in row perm(k) of A */
712       nz   = ai[k+1] - ai[k];
713       acol = aj + ai[k];
714       aval = aa + ai[k];
715       bval = ba + bi[k];
716       while (nz -- ){
717         if (*acol < k) { /* skip lower triangular entries */
718           acol++; aval++;
719         } else {
720           rtmp[*acol++] = *aval++;
721           *bval++       = 0.0; /* for in-place factorization */
722         }
723       }
724 
725       /* shift the diagonal of the matrix */
726       if (sctx.nshift) rtmp[k] += sctx.shift_amount;
727 
728       /* modify k-th row by adding in those rows i with U(i,k)!=0 */
729       dk = rtmp[k];
730       i  = jl[k]; /* first row to be added to k_th row  */
731 
732       while (i < k){
733         nexti = jl[i]; /* next row to be added to k_th row */
734         /* compute multiplier, update D(k) and U(i,k) */
735         ili   = il[i];  /* index of first nonzero element in U(i,k:bms-1) */
736         uikdi = - ba[ili]*ba[bi[i]];
737         dk   += uikdi*ba[ili];
738         ba[ili] = uikdi; /* -U(i,k) */
739 
740         /* add multiple of row i to k-th row ... */
741         jmin = ili + 1;
742         nz   = bi[i+1] - jmin;
743         if (nz > 0){
744           bcol = bj + jmin;
745           bval = ba + jmin;
746           while (nz --) rtmp[*bcol++] += uikdi*(*bval++);
747           /* update il and jl for i-th row */
748           il[i] = jmin;
749           j = bj[jmin]; jl[i] = jl[j]; jl[j] = i;
750         }
751         i = nexti;
752       }
753 
754       /* shift the diagonals when zero pivot is detected */
755       /* compute rs=sum of abs(off-diagonal) */
756       rs   = 0.0;
757       jmin = bi[k]+1;
758       nz   = bi[k+1] - jmin;
759       if (nz){
760         bcol = bj + jmin;
761         while (nz--){
762           rs += PetscAbsScalar(rtmp[*bcol]);
763           bcol++;
764         }
765       }
766 
767       sctx.rs = rs;
768       sctx.pv = dk;
769       ierr = MatCholeskyCheckShift_inline(info,sctx,k,newshift);CHKERRQ(ierr);
770       if (newshift == 1) break;    /* sctx.shift_amount is updated */
771 
772       /* copy data into U(k,:) */
773       ba[bi[k]] = 1.0/dk;
774       jmin      = bi[k]+1;
775       nz        = bi[k+1] - jmin;
776       if (nz){
777         bcol = bj + jmin;
778         bval = ba + jmin;
779         while (nz--){
780           *bval++       = rtmp[*bcol];
781           rtmp[*bcol++] = 0.0;
782         }
783         /* add k-th row into il and jl */
784         il[k] = jmin;
785         i = bj[jmin]; jl[k] = jl[i]; jl[i] = k;
786       }
787     }
788   } while (sctx.chshift);
789   ierr = PetscFree3(rtmp,il,jl);CHKERRQ(ierr);
790 
791   C->ops->solve                 = MatSolve_SeqSBAIJ_1_NaturalOrdering;
792   C->ops->solvetranspose        = MatSolve_SeqSBAIJ_1_NaturalOrdering;
793   C->assembled    = PETSC_TRUE;
794   C->preallocated = PETSC_TRUE;
795   ierr = PetscLogFlops(C->rmap->N);CHKERRQ(ierr);
796     if (sctx.nshift){
797     if (shiftnz) {
798       ierr = PetscInfo2(A,"number of shiftnz tries %D, shift_amount %G\n",sctx.nshift,sctx.shift_amount);CHKERRQ(ierr);
799     } else if (shiftpd) {
800       ierr = PetscInfo2(A,"number of shiftpd tries %D, shift_amount %G\n",sctx.nshift,sctx.shift_amount);CHKERRQ(ierr);
801     }
802   }
803   PetscFunctionReturn(0);
804 }
805 
806 #include "petscbt.h"
807 #include "../src/mat/utils/freespace.h"
808 #undef __FUNCT__
809 #define __FUNCT__ "MatICCFactorSymbolic_SeqBAIJ"
810 PetscErrorCode MatICCFactorSymbolic_SeqBAIJ(Mat fact,Mat A,IS perm,const MatFactorInfo *info)
811 {
812   Mat_SeqBAIJ        *a = (Mat_SeqBAIJ*)A->data;
813   Mat_SeqSBAIJ       *b;
814   Mat                B;
815   PetscErrorCode     ierr;
816   PetscTruth         perm_identity;
817   PetscInt           reallocs=0,i,*ai=a->i,*aj=a->j,am=a->mbs,bs=A->rmap->bs,*ui;
818   const PetscInt     *rip;
819   PetscInt           jmin,jmax,nzk,k,j,*jl,prow,*il,nextprow;
820   PetscInt           nlnk,*lnk,*lnk_lvl=PETSC_NULL,ncols,ncols_upper,*cols,*cols_lvl,*uj,**uj_ptr,**uj_lvl_ptr;
821   PetscReal          fill=info->fill,levels=info->levels;
822   PetscFreeSpaceList free_space=PETSC_NULL,current_space=PETSC_NULL;
823   PetscFreeSpaceList free_space_lvl=PETSC_NULL,current_space_lvl=PETSC_NULL;
824   PetscBT            lnkbt;
825 
826   PetscFunctionBegin;
827   if (bs > 1){
828     if (!a->sbaijMat){
829       ierr = MatConvert(A,MATSEQSBAIJ,MAT_INITIAL_MATRIX,&a->sbaijMat);CHKERRQ(ierr);
830     }
831     (fact)->ops->iccfactorsymbolic = MatICCFactorSymbolic_SeqSBAIJ;  /* undue the change made in MatGetFactor_seqbaij_petsc */
832     ierr = MatICCFactorSymbolic(fact,a->sbaijMat,perm,info);CHKERRQ(ierr);
833     PetscFunctionReturn(0);
834   }
835 
836   ierr = ISIdentity(perm,&perm_identity);CHKERRQ(ierr);
837   ierr = ISGetIndices(perm,&rip);CHKERRQ(ierr);
838 
839   /* special case that simply copies fill pattern */
840   if (!levels && perm_identity) {
841     ierr = MatMarkDiagonal_SeqBAIJ(A);CHKERRQ(ierr);
842     ierr = PetscMalloc((am+1)*sizeof(PetscInt),&ui);CHKERRQ(ierr);
843     for (i=0; i<am; i++) {
844       ui[i] = ai[i+1] - a->diag[i]; /* ui: rowlengths - changes when !perm_identity */
845     }
846     B = fact;
847     ierr = MatSeqSBAIJSetPreallocation(B,1,0,ui);CHKERRQ(ierr);
848 
849 
850     b  = (Mat_SeqSBAIJ*)B->data;
851     uj = b->j;
852     for (i=0; i<am; i++) {
853       aj = a->j + a->diag[i];
854       for (j=0; j<ui[i]; j++){
855         *uj++ = *aj++;
856       }
857       b->ilen[i] = ui[i];
858     }
859     ierr = PetscFree(ui);CHKERRQ(ierr);
860     B->factor = MAT_FACTOR_NONE;
861     ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr);
862     ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr);
863     B->factor = MAT_FACTOR_ICC;
864 
865     B->ops->choleskyfactornumeric = MatCholeskyFactorNumeric_SeqBAIJ_N_NaturalOrdering;
866     PetscFunctionReturn(0);
867   }
868 
869   /* initialization */
870   ierr  = PetscMalloc((am+1)*sizeof(PetscInt),&ui);CHKERRQ(ierr);
871   ui[0] = 0.;
872   ierr  = PetscMalloc((2*am+1)*sizeof(PetscInt),&cols_lvl);CHKERRQ(ierr);
873 
874   /* jl: linked list for storing indices of the pivot rows
875      il: il[i] points to the 1st nonzero entry of U(i,k:am-1) */
876   ierr = PetscMalloc4(am,PetscInt*,&uj_ptr,am,PetscInt*,&uj_lvl_ptr,am,PetscInt,&il,am,PetscInt,&jl);CHKERRQ(ierr);
877   for (i=0; i<am; i++){
878     jl[i] = am; il[i] = 0.;
879   }
880 
881   /* create and initialize a linked list for storing column indices of the active row k */
882   nlnk = am + 1;
883   ierr = PetscIncompleteLLCreate(am,am,nlnk,lnk,lnk_lvl,lnkbt);CHKERRQ(ierr);
884 
885   /* initial FreeSpace size is fill*(ai[am]+1) */
886   ierr = PetscFreeSpaceGet((PetscInt)(fill*(ai[am]+1)),&free_space);CHKERRQ(ierr);
887   current_space = free_space;
888   ierr = PetscFreeSpaceGet((PetscInt)(fill*(ai[am]+1)),&free_space_lvl);CHKERRQ(ierr);
889   current_space_lvl = free_space_lvl;
890 
891   for (k=0; k<am; k++){  /* for each active row k */
892     /* initialize lnk by the column indices of row rip[k] of A */
893     nzk   = 0;
894     ncols = ai[rip[k]+1] - ai[rip[k]];
895     ncols_upper = 0;
896     cols        = cols_lvl + am;
897     for (j=0; j<ncols; j++){
898       i = rip[*(aj + ai[rip[k]] + j)];
899       if (i >= k){ /* only take upper triangular entry */
900         cols[ncols_upper] = i;
901         cols_lvl[ncols_upper] = -1;  /* initialize level for nonzero entries */
902         ncols_upper++;
903       }
904     }
905     ierr = PetscIncompleteLLAdd(ncols_upper,cols,levels,cols_lvl,am,nlnk,lnk,lnk_lvl,lnkbt);CHKERRQ(ierr);
906     nzk += nlnk;
907 
908     /* update lnk by computing fill-in for each pivot row to be merged in */
909     prow = jl[k]; /* 1st pivot row */
910 
911     while (prow < k){
912       nextprow = jl[prow];
913 
914       /* merge prow into k-th row */
915       jmin = il[prow] + 1;  /* index of the 2nd nzero entry in U(prow,k:am-1) */
916       jmax = ui[prow+1];
917       ncols = jmax-jmin;
918       i     = jmin - ui[prow];
919       cols = uj_ptr[prow] + i; /* points to the 2nd nzero entry in U(prow,k:am-1) */
920       for (j=0; j<ncols; j++) cols_lvl[j] = *(uj_lvl_ptr[prow] + i + j);
921       ierr = PetscIncompleteLLAddSorted(ncols,cols,levels,cols_lvl,am,nlnk,lnk,lnk_lvl,lnkbt);CHKERRQ(ierr);
922       nzk += nlnk;
923 
924       /* update il and jl for prow */
925       if (jmin < jmax){
926         il[prow] = jmin;
927         j = *cols; jl[prow] = jl[j]; jl[j] = prow;
928       }
929       prow = nextprow;
930     }
931 
932     /* if free space is not available, make more free space */
933     if (current_space->local_remaining<nzk) {
934       i = am - k + 1; /* num of unfactored rows */
935       i = PetscMin(i*nzk, i*(i-1)); /* i*nzk, i*(i-1): estimated and max additional space needed */
936       ierr = PetscFreeSpaceGet(i,&current_space);CHKERRQ(ierr);
937       ierr = PetscFreeSpaceGet(i,&current_space_lvl);CHKERRQ(ierr);
938       reallocs++;
939     }
940 
941     /* copy data into free_space and free_space_lvl, then initialize lnk */
942     ierr = PetscIncompleteLLClean(am,am,nzk,lnk,lnk_lvl,current_space->array,current_space_lvl->array,lnkbt);CHKERRQ(ierr);
943 
944     /* add the k-th row into il and jl */
945     if (nzk-1 > 0){
946       i = current_space->array[1]; /* col value of the first nonzero element in U(k, k+1:am-1) */
947       jl[k] = jl[i]; jl[i] = k;
948       il[k] = ui[k] + 1;
949     }
950     uj_ptr[k]     = current_space->array;
951     uj_lvl_ptr[k] = current_space_lvl->array;
952 
953     current_space->array           += nzk;
954     current_space->local_used      += nzk;
955     current_space->local_remaining -= nzk;
956 
957     current_space_lvl->array           += nzk;
958     current_space_lvl->local_used      += nzk;
959     current_space_lvl->local_remaining -= nzk;
960 
961     ui[k+1] = ui[k] + nzk;
962   }
963 
964 #if defined(PETSC_USE_INFO)
965   if (ai[am] != 0) {
966     PetscReal af = ((PetscReal)(2*ui[am]-am))/((PetscReal)ai[am]);
967     ierr = PetscInfo3(A,"Reallocs %D Fill ratio:given %G needed %G\n",reallocs,fill,af);CHKERRQ(ierr);
968     ierr = PetscInfo1(A,"Run with -pc_factor_fill %G or use \n",af);CHKERRQ(ierr);
969     ierr = PetscInfo1(A,"PCFactorSetFill(pc,%G) for best performance.\n",af);CHKERRQ(ierr);
970   } else {
971     ierr = PetscInfo(A,"Empty matrix.\n");CHKERRQ(ierr);
972   }
973 #endif
974 
975   ierr = ISRestoreIndices(perm,&rip);CHKERRQ(ierr);
976   ierr = PetscFree4(uj_ptr,uj_lvl_ptr,il,jl);CHKERRQ(ierr);
977   ierr = PetscFree(cols_lvl);CHKERRQ(ierr);
978 
979   /* destroy list of free space and other temporary array(s) */
980   ierr = PetscMalloc((ui[am]+1)*sizeof(PetscInt),&uj);CHKERRQ(ierr);
981   ierr = PetscFreeSpaceContiguous(&free_space,uj);CHKERRQ(ierr);
982   ierr = PetscIncompleteLLDestroy(lnk,lnkbt);CHKERRQ(ierr);
983   ierr = PetscFreeSpaceDestroy(free_space_lvl);CHKERRQ(ierr);
984 
985   /* put together the new matrix in MATSEQSBAIJ format */
986   B = fact;
987   ierr = MatSeqSBAIJSetPreallocation(B,1,MAT_SKIP_ALLOCATION,PETSC_NULL);CHKERRQ(ierr);
988 
989   b = (Mat_SeqSBAIJ*)B->data;
990   b->singlemalloc = PETSC_FALSE;
991   b->free_a       = PETSC_TRUE;
992   b->free_ij       = PETSC_TRUE;
993   ierr = PetscMalloc((ui[am]+1)*sizeof(MatScalar),&b->a);CHKERRQ(ierr);
994   b->j    = uj;
995   b->i    = ui;
996   b->diag = 0;
997   b->ilen = 0;
998   b->imax = 0;
999   b->row  = perm;
1000   b->pivotinblocks = PETSC_FALSE; /* need to get from MatFactorInfo */
1001   ierr    = PetscObjectReference((PetscObject)perm);CHKERRQ(ierr);
1002   b->icol = perm;
1003   ierr    = PetscObjectReference((PetscObject)perm);CHKERRQ(ierr);
1004   ierr    = PetscMalloc((am+1)*sizeof(PetscScalar),&b->solve_work);CHKERRQ(ierr);
1005   ierr    = PetscLogObjectMemory(B,(ui[am]-am)*(sizeof(PetscInt)+sizeof(MatScalar)));CHKERRQ(ierr);
1006   b->maxnz = b->nz = ui[am];
1007 
1008   B->info.factor_mallocs    = reallocs;
1009   B->info.fill_ratio_given  = fill;
1010   if (ai[am] != 0.) {
1011     B->info.fill_ratio_needed = ((PetscReal)ui[am])/((PetscReal)ai[am]);
1012   } else {
1013     B->info.fill_ratio_needed = 0.0;
1014   }
1015   if (perm_identity){
1016     B->ops->solve           = MatSolve_SeqSBAIJ_1_NaturalOrdering;
1017     B->ops->solvetranspose  = MatSolve_SeqSBAIJ_1_NaturalOrdering;
1018     B->ops->choleskyfactornumeric = MatCholeskyFactorNumeric_SeqBAIJ_N_NaturalOrdering;
1019   } else {
1020     (fact)->ops->choleskyfactornumeric = MatCholeskyFactorNumeric_SeqBAIJ_N;
1021   }
1022   PetscFunctionReturn(0);
1023 }
1024 
1025 #undef __FUNCT__
1026 #define __FUNCT__ "MatCholeskyFactorSymbolic_SeqBAIJ"
1027 PetscErrorCode MatCholeskyFactorSymbolic_SeqBAIJ(Mat fact,Mat A,IS perm,const MatFactorInfo *info)
1028 {
1029   Mat_SeqBAIJ        *a = (Mat_SeqBAIJ*)A->data;
1030   Mat_SeqSBAIJ       *b;
1031   Mat                B;
1032   PetscErrorCode     ierr;
1033   PetscTruth         perm_identity;
1034   PetscReal          fill = info->fill;
1035   const PetscInt     *rip;
1036   PetscInt           i,mbs=a->mbs,bs=A->rmap->bs,*ai=a->i,*aj=a->j,reallocs=0,prow;
1037   PetscInt           *jl,jmin,jmax,nzk,*ui,k,j,*il,nextprow;
1038   PetscInt           nlnk,*lnk,ncols,ncols_upper,*cols,*uj,**ui_ptr,*uj_ptr;
1039   PetscFreeSpaceList free_space=PETSC_NULL,current_space=PETSC_NULL;
1040   PetscBT            lnkbt;
1041 
1042   PetscFunctionBegin;
1043   if (bs > 1) { /* convert to seqsbaij */
1044     if (!a->sbaijMat){
1045       ierr = MatConvert(A,MATSEQSBAIJ,MAT_INITIAL_MATRIX,&a->sbaijMat);CHKERRQ(ierr);
1046     }
1047     (fact)->ops->choleskyfactorsymbolic = MatCholeskyFactorSymbolic_SeqSBAIJ; /* undue the change made in MatGetFactor_seqbaij_petsc */
1048     ierr = MatCholeskyFactorSymbolic(fact,a->sbaijMat,perm,info);CHKERRQ(ierr);
1049     PetscFunctionReturn(0);
1050   }
1051 
1052   /* check whether perm is the identity mapping */
1053   ierr = ISIdentity(perm,&perm_identity);CHKERRQ(ierr);
1054   if (!perm_identity) SETERRQ(PETSC_ERR_SUP,"Matrix reordering is not supported");
1055   ierr = ISGetIndices(perm,&rip);CHKERRQ(ierr);
1056 
1057   /* initialization */
1058   ierr  = PetscMalloc((mbs+1)*sizeof(PetscInt),&ui);CHKERRQ(ierr);
1059   ui[0] = 0.;
1060 
1061   /* jl: linked list for storing indices of the pivot rows
1062      il: il[i] points to the 1st nonzero entry of U(i,k:mbs-1) */
1063   ierr = PetscMalloc4(mbs,PetscInt*,&ui_ptr,mbs,PetscInt,&il,mbs,PetscInt,&jl,mbs,PetscInt,&cols);CHKERRQ(ierr);
1064   for (i=0; i<mbs; i++){
1065     jl[i] = mbs; il[i] = 0.;
1066   }
1067 
1068   /* create and initialize a linked list for storing column indices of the active row k */
1069   nlnk = mbs + 1;
1070   ierr = PetscLLCreate(mbs,mbs,nlnk,lnk,lnkbt);CHKERRQ(ierr);
1071 
1072   /* initial FreeSpace size is fill*(ai[mbs]+1) */
1073   ierr = PetscFreeSpaceGet((PetscInt)(fill*(ai[mbs]+1)),&free_space);CHKERRQ(ierr);
1074   current_space = free_space;
1075 
1076   for (k=0; k<mbs; k++){  /* for each active row k */
1077     /* initialize lnk by the column indices of row rip[k] of A */
1078     nzk   = 0;
1079     ncols = ai[rip[k]+1] - ai[rip[k]];
1080     ncols_upper = 0.;
1081     for (j=0; j<ncols; j++){
1082       i = rip[*(aj + ai[rip[k]] + j)];
1083       if (i >= k){ /* only take upper triangular entry */
1084         cols[ncols_upper] = i;
1085         ncols_upper++;
1086       }
1087     }
1088     ierr = PetscLLAdd(ncols_upper,cols,mbs,nlnk,lnk,lnkbt);CHKERRQ(ierr);
1089     nzk += nlnk;
1090 
1091     /* update lnk by computing fill-in for each pivot row to be merged in */
1092     prow = jl[k]; /* 1st pivot row */
1093 
1094     while (prow < k){
1095       nextprow = jl[prow];
1096       /* merge prow into k-th row */
1097       jmin = il[prow] + 1;  /* index of the 2nd nzero entry in U(prow,k:mbs-1) */
1098       jmax = ui[prow+1];
1099       ncols = jmax-jmin;
1100       uj_ptr = ui_ptr[prow] + jmin - ui[prow]; /* points to the 2nd nzero entry in U(prow,k:mbs-1) */
1101       ierr = PetscLLAddSorted(ncols,uj_ptr,mbs,nlnk,lnk,lnkbt);CHKERRQ(ierr);
1102       nzk += nlnk;
1103 
1104       /* update il and jl for prow */
1105       if (jmin < jmax){
1106         il[prow] = jmin;
1107         j = *uj_ptr; jl[prow] = jl[j]; jl[j] = prow;
1108       }
1109       prow = nextprow;
1110     }
1111 
1112     /* if free space is not available, make more free space */
1113     if (current_space->local_remaining<nzk) {
1114       i = mbs - k + 1; /* num of unfactored rows */
1115       i = PetscMin(i*nzk, i*(i-1)); /* i*nzk, i*(i-1): estimated and max additional space needed */
1116       ierr = PetscFreeSpaceGet(i,&current_space);CHKERRQ(ierr);
1117       reallocs++;
1118     }
1119 
1120     /* copy data into free space, then initialize lnk */
1121     ierr = PetscLLClean(mbs,mbs,nzk,lnk,current_space->array,lnkbt);CHKERRQ(ierr);
1122 
1123     /* add the k-th row into il and jl */
1124     if (nzk-1 > 0){
1125       i = current_space->array[1]; /* col value of the first nonzero element in U(k, k+1:mbs-1) */
1126       jl[k] = jl[i]; jl[i] = k;
1127       il[k] = ui[k] + 1;
1128     }
1129     ui_ptr[k] = current_space->array;
1130     current_space->array           += nzk;
1131     current_space->local_used      += nzk;
1132     current_space->local_remaining -= nzk;
1133 
1134     ui[k+1] = ui[k] + nzk;
1135   }
1136 
1137 #if defined(PETSC_USE_INFO)
1138   if (ai[mbs] != 0.) {
1139     PetscReal af = ((PetscReal)ui[mbs])/((PetscReal)ai[mbs]);
1140     ierr = PetscInfo3(A,"Reallocs %D Fill ratio:given %G needed %G\n",reallocs,fill,af);CHKERRQ(ierr);
1141     ierr = PetscInfo1(A,"Run with -pc_factor_fill %G or use \n",af);CHKERRQ(ierr);
1142     ierr = PetscInfo1(A,"PCFactorSetFill(pc,%G) for best performance.\n",af);CHKERRQ(ierr);
1143   } else {
1144     ierr = PetscInfo(A,"Empty matrix.\n");CHKERRQ(ierr);
1145   }
1146 #endif
1147 
1148   ierr = ISRestoreIndices(perm,&rip);CHKERRQ(ierr);
1149   ierr = PetscFree4(ui_ptr,il,jl,cols);CHKERRQ(ierr);
1150 
1151   /* destroy list of free space and other temporary array(s) */
1152   ierr = PetscMalloc((ui[mbs]+1)*sizeof(PetscInt),&uj);CHKERRQ(ierr);
1153   ierr = PetscFreeSpaceContiguous(&free_space,uj);CHKERRQ(ierr);
1154   ierr = PetscLLDestroy(lnk,lnkbt);CHKERRQ(ierr);
1155 
1156   /* put together the new matrix in MATSEQSBAIJ format */
1157   B    = fact;
1158   ierr = MatSeqSBAIJSetPreallocation(B,bs,MAT_SKIP_ALLOCATION,PETSC_NULL);CHKERRQ(ierr);
1159 
1160   b = (Mat_SeqSBAIJ*)B->data;
1161   b->singlemalloc = PETSC_FALSE;
1162   b->free_a       = PETSC_TRUE;
1163   b->free_ij      = PETSC_TRUE;
1164   ierr = PetscMalloc((ui[mbs]+1)*sizeof(MatScalar),&b->a);CHKERRQ(ierr);
1165   b->j    = uj;
1166   b->i    = ui;
1167   b->diag = 0;
1168   b->ilen = 0;
1169   b->imax = 0;
1170   b->row  = perm;
1171   b->pivotinblocks = PETSC_FALSE; /* need to get from MatFactorInfo */
1172   ierr    = PetscObjectReference((PetscObject)perm);CHKERRQ(ierr);
1173   b->icol = perm;
1174   ierr    = PetscObjectReference((PetscObject)perm);CHKERRQ(ierr);
1175   ierr    = PetscMalloc((mbs+1)*sizeof(PetscScalar),&b->solve_work);CHKERRQ(ierr);
1176   ierr    = PetscLogObjectMemory(B,(ui[mbs]-mbs)*(sizeof(PetscInt)+sizeof(MatScalar)));CHKERRQ(ierr);
1177   b->maxnz = b->nz = ui[mbs];
1178 
1179   B->info.factor_mallocs    = reallocs;
1180   B->info.fill_ratio_given  = fill;
1181   if (ai[mbs] != 0.) {
1182     B->info.fill_ratio_needed = ((PetscReal)ui[mbs])/((PetscReal)ai[mbs]);
1183   } else {
1184     B->info.fill_ratio_needed = 0.0;
1185   }
1186   if (perm_identity){
1187     B->ops->choleskyfactornumeric = MatCholeskyFactorNumeric_SeqBAIJ_N_NaturalOrdering;
1188   } else {
1189     B->ops->choleskyfactornumeric = MatCholeskyFactorNumeric_SeqBAIJ_N;
1190   }
1191   PetscFunctionReturn(0);
1192 }
1193 
1194 #undef __FUNCT__
1195 #define __FUNCT__ "MatSolve_SeqBAIJ_N_NaturalOrdering_newdatastruct"
1196 PetscErrorCode MatSolve_SeqBAIJ_N_NaturalOrdering_newdatastruct(Mat A,Vec bb,Vec xx)
1197 {
1198   Mat_SeqBAIJ    *a=(Mat_SeqBAIJ *)A->data;
1199   PetscErrorCode ierr;
1200   const PetscInt *ai=a->i,*aj=a->j,*adiag=a->diag,*vi;
1201   PetscInt       i,k,n=a->mbs;
1202   PetscInt       nz,bs=A->rmap->bs,bs2=a->bs2;
1203   MatScalar      *aa=a->a,*v;
1204   PetscScalar    *x,*b,*s,*t,*ls;
1205 
1206   PetscFunctionBegin;
1207   ierr = VecGetArray(bb,&b);CHKERRQ(ierr);
1208   ierr = VecGetArray(xx,&x);CHKERRQ(ierr);
1209   t  = a->solve_work;
1210 
1211   /* forward solve the lower triangular */
1212   ierr = PetscMemcpy(t,b,bs*sizeof(PetscScalar));CHKERRQ(ierr); /* copy 1st block of b to t */
1213 
1214   for (i=1; i<n; i++) {
1215     v   = aa + bs2*ai[i];
1216     vi  = aj + ai[i];
1217     nz = ai[i+1] - ai[i];
1218     s = t + bs*i;
1219     ierr = PetscMemcpy(s,b+bs*i,bs*sizeof(PetscScalar));CHKERRQ(ierr); /* copy i_th block of b to t */
1220     for(k=0;k<nz;k++){
1221       Kernel_v_gets_v_minus_A_times_w(bs,s,v,t+bs*vi[k]);
1222       v += bs2;
1223     }
1224   }
1225 
1226   /* backward solve the upper triangular */
1227   ls = a->solve_work + A->cmap->n;
1228   for (i=n-1; i>=0; i--){
1229     v  = aa + bs2*(adiag[i+1]+1);
1230     vi = aj + adiag[i+1]+1;
1231     nz = adiag[i] - adiag[i+1]-1;
1232     ierr = PetscMemcpy(ls,t+i*bs,bs*sizeof(PetscScalar));CHKERRQ(ierr);
1233     for(k=0;k<nz;k++){
1234       Kernel_v_gets_v_minus_A_times_w(bs,ls,v,t+bs*vi[k]);
1235       v += bs2;
1236     }
1237     Kernel_w_gets_A_times_v(bs,ls,aa+bs2*adiag[i],t+i*bs); /* *inv(diagonal[i]) */
1238     ierr = PetscMemcpy(x+i*bs,t+i*bs,bs*sizeof(PetscScalar));CHKERRQ(ierr);
1239   }
1240 
1241   ierr = VecRestoreArray(bb,&b);CHKERRQ(ierr);
1242   ierr = VecRestoreArray(xx,&x);CHKERRQ(ierr);
1243   ierr = PetscLogFlops(2.0*(a->bs2)*(a->nz) - A->rmap->bs*A->cmap->n);CHKERRQ(ierr);
1244   PetscFunctionReturn(0);
1245 }
1246 
1247 #undef __FUNCT__
1248 #define __FUNCT__ "MatSolve_SeqBAIJ_N_newdatastruct"
1249 PetscErrorCode MatSolve_SeqBAIJ_N_newdatastruct(Mat A,Vec bb,Vec xx)
1250 {
1251   Mat_SeqBAIJ    *a=(Mat_SeqBAIJ *)A->data;
1252   IS             iscol=a->col,isrow=a->row;
1253   PetscErrorCode ierr;
1254   const PetscInt *r,*c,*rout,*cout,*ai=a->i,*aj=a->j,*adiag=a->diag,*vi;
1255   PetscInt       i,m,n=a->mbs;
1256   PetscInt       nz,bs=A->rmap->bs,bs2=a->bs2;
1257   MatScalar      *aa=a->a,*v;
1258   PetscScalar    *x,*b,*s,*t,*ls;
1259 
1260   PetscFunctionBegin;
1261   ierr = VecGetArray(bb,&b);CHKERRQ(ierr);
1262   ierr = VecGetArray(xx,&x);CHKERRQ(ierr);
1263   t  = a->solve_work;
1264 
1265   ierr = ISGetIndices(isrow,&rout);CHKERRQ(ierr); r = rout;
1266   ierr = ISGetIndices(iscol,&cout);CHKERRQ(ierr); c = cout;
1267 
1268   /* forward solve the lower triangular */
1269   ierr = PetscMemcpy(t,b+bs*r[0],bs*sizeof(PetscScalar));CHKERRQ(ierr);
1270   for (i=1; i<n; i++) {
1271     v   = aa + bs2*ai[i];
1272     vi  = aj + ai[i];
1273     nz = ai[i+1] - ai[i];
1274     s = t + bs*i;
1275     ierr = PetscMemcpy(s,b+bs*r[i],bs*sizeof(PetscScalar));CHKERRQ(ierr);
1276     for(m=0;m<nz;m++){
1277       Kernel_v_gets_v_minus_A_times_w(bs,s,v,t+bs*vi[m]);
1278       v += bs2;
1279     }
1280   }
1281 
1282   /* backward solve the upper triangular */
1283   ls = a->solve_work + A->cmap->n;
1284   for (i=n-1; i>=0; i--){
1285     v  = aa + bs2*(adiag[i+1]+1);
1286     vi = aj + adiag[i+1]+1;
1287     nz = adiag[i] - adiag[i+1] - 1;
1288     ierr = PetscMemcpy(ls,t+i*bs,bs*sizeof(PetscScalar));CHKERRQ(ierr);
1289     for(m=0;m<nz;m++){
1290       Kernel_v_gets_v_minus_A_times_w(bs,ls,v,t+bs*vi[m]);
1291       v += bs2;
1292     }
1293     Kernel_w_gets_A_times_v(bs,ls,v,t+i*bs); /* *inv(diagonal[i]) */
1294     ierr = PetscMemcpy(x + bs*c[i],t+i*bs,bs*sizeof(PetscScalar));CHKERRQ(ierr);
1295   }
1296   ierr = ISRestoreIndices(isrow,&rout);CHKERRQ(ierr);
1297   ierr = ISRestoreIndices(iscol,&cout);CHKERRQ(ierr);
1298   ierr = VecRestoreArray(bb,&b);CHKERRQ(ierr);
1299   ierr = VecRestoreArray(xx,&x);CHKERRQ(ierr);
1300   ierr = PetscLogFlops(2.0*(a->bs2)*(a->nz) - A->rmap->bs*A->cmap->n);CHKERRQ(ierr);
1301   PetscFunctionReturn(0);
1302 }
1303 
1304 #undef __FUNCT__
1305 #define __FUNCT__ "BlockAbs_privat"
1306 PetscErrorCode BlockAbs_private(PetscInt nbs,PetscInt bs2,PetscScalar *blockarray,PetscReal *absarray)
1307 {
1308   PetscErrorCode     ierr;
1309   PetscInt           i,j;
1310   PetscFunctionBegin;
1311   ierr = PetscMemzero(absarray,(nbs+1)*sizeof(PetscReal));CHKERRQ(ierr);
1312   for (i=0; i<nbs; i++){
1313     for (j=0; j<bs2; j++){
1314       if (absarray[i] < PetscAbsScalar(blockarray[i*nbs+j])) absarray[i] = PetscAbsScalar(blockarray[i*nbs+j]);
1315     }
1316   }
1317   PetscFunctionReturn(0);
1318 }
1319 
1320 #undef __FUNCT__
1321 #define __FUNCT__ "MatILUDTFactor_SeqBAIJ"
1322 PetscErrorCode MatILUDTFactor_SeqBAIJ(Mat A,IS isrow,IS iscol,const MatFactorInfo *info,Mat *fact)
1323 {
1324   Mat                B = *fact;
1325   Mat_SeqBAIJ        *a=(Mat_SeqBAIJ*)A->data,*b;
1326   IS                 isicol;
1327   PetscErrorCode     ierr;
1328   const PetscInt     *r,*ic;
1329   PetscInt           i,mbs=a->mbs,bs=A->rmap->bs,bs2=a->bs2,*ai=a->i,*aj=a->j,*ajtmp,*adiag;
1330   PetscInt           *bi,*bj,*bdiag;
1331 
1332   PetscInt           row,nzi,nzi_bl,nzi_bu,*im,dtcount,nzi_al,nzi_au;
1333   PetscInt           nlnk,*lnk;
1334   PetscBT            lnkbt;
1335   PetscTruth         row_identity,icol_identity,both_identity;
1336   MatScalar          *aatmp,*pv,*batmp,*ba,*rtmp,*pc,*multiplier,*vtmp;
1337   const PetscInt     *ics;
1338   PetscInt           j,nz,*pj,*bjtmp,k,ncut,*jtmp;
1339 
1340   PetscReal          dt=info->dt; /* shift=info->shiftinblocks; */
1341   PetscInt           nnz_max;
1342   PetscTruth         missing;
1343   PetscReal          *vtmp_abs;
1344   MatScalar          *v_work;
1345   PetscInt           *v_pivots;
1346 
1347   PetscFunctionBegin;
1348   /* ------- symbolic factorization, can be reused ---------*/
1349   ierr = MatMissingDiagonal(A,&missing,&i);CHKERRQ(ierr);
1350   if (missing) SETERRQ1(PETSC_ERR_ARG_WRONGSTATE,"Matrix is missing diagonal entry %D",i);
1351   adiag=a->diag;
1352 
1353   ierr = ISInvertPermutation(iscol,PETSC_DECIDE,&isicol);CHKERRQ(ierr);
1354 
1355   /* bdiag is location of diagonal in factor */
1356   ierr = PetscMalloc((mbs+1)*sizeof(PetscInt),&bdiag);CHKERRQ(ierr);
1357 
1358   /* allocate row pointers bi */
1359   ierr = PetscMalloc((2*mbs+2)*sizeof(PetscInt),&bi);CHKERRQ(ierr);
1360 
1361   /* allocate bj and ba; max num of nonzero entries is (ai[n]+2*n*dtcount+2) */
1362   dtcount = (PetscInt)info->dtcount;
1363   if (dtcount > mbs-1) dtcount = mbs-1;
1364   nnz_max  = ai[mbs]+2*mbs*dtcount +2;
1365   /* printf("MatILUDTFactor_SeqBAIJ, bs %d, ai[mbs] %d, nnz_max  %d, dtcount %d\n",bs,ai[mbs],nnz_max,dtcount); */
1366   ierr = PetscMalloc(nnz_max*sizeof(PetscInt),&bj);CHKERRQ(ierr);
1367   nnz_max = nnz_max*bs2;
1368   ierr = PetscMalloc(nnz_max*sizeof(MatScalar),&ba);CHKERRQ(ierr);
1369 
1370   /* put together the new matrix */
1371   ierr = MatSeqBAIJSetPreallocation_SeqBAIJ(B,bs,MAT_SKIP_ALLOCATION,PETSC_NULL);CHKERRQ(ierr);
1372   ierr = PetscLogObjectParent(B,isicol);CHKERRQ(ierr);
1373   b    = (Mat_SeqBAIJ*)(B)->data;
1374   b->free_a       = PETSC_TRUE;
1375   b->free_ij      = PETSC_TRUE;
1376   b->singlemalloc = PETSC_FALSE;
1377   b->a          = ba;
1378   b->j          = bj;
1379   b->i          = bi;
1380   b->diag       = bdiag;
1381   b->ilen       = 0;
1382   b->imax       = 0;
1383   b->row        = isrow;
1384   b->col        = iscol;
1385   ierr          = PetscObjectReference((PetscObject)isrow);CHKERRQ(ierr);
1386   ierr          = PetscObjectReference((PetscObject)iscol);CHKERRQ(ierr);
1387   b->icol       = isicol;
1388   ierr = PetscMalloc((bs*(mbs+1))*sizeof(PetscScalar),&b->solve_work);CHKERRQ(ierr);
1389 
1390   ierr = PetscLogObjectMemory(B,nnz_max*(sizeof(PetscInt)+sizeof(MatScalar)));CHKERRQ(ierr);
1391   b->maxnz = nnz_max/bs2;
1392 
1393   (B)->factor                = MAT_FACTOR_ILUDT;
1394   (B)->info.factor_mallocs   = 0;
1395   (B)->info.fill_ratio_given = ((PetscReal)nnz_max)/((PetscReal)(ai[mbs]*bs2));
1396   CHKMEMQ;
1397   /* ------- end of symbolic factorization ---------*/
1398   ierr = ISGetIndices(isrow,&r);CHKERRQ(ierr);
1399   ierr = ISGetIndices(isicol,&ic);CHKERRQ(ierr);
1400   ics  = ic;
1401 
1402   /* linked list for storing column indices of the active row */
1403   nlnk = mbs + 1;
1404   ierr = PetscLLCreate(mbs,mbs,nlnk,lnk,lnkbt);CHKERRQ(ierr);
1405 
1406   /* im: used by PetscLLAddSortedLU(); jtmp: working array for column indices of active row */
1407   ierr = PetscMalloc2(mbs,PetscInt,&im,mbs,PetscInt,&jtmp);CHKERRQ(ierr);
1408   /* rtmp, vtmp: working arrays for sparse and contiguous row entries of active row */
1409   ierr = PetscMalloc2(mbs*bs2,MatScalar,&rtmp,mbs*bs2,MatScalar,&vtmp);CHKERRQ(ierr);
1410   ierr = PetscMalloc((mbs+1)*sizeof(PetscReal),&vtmp_abs);CHKERRQ(ierr);
1411   ierr = PetscMalloc3(bs,MatScalar,&v_work,bs2,MatScalar,&multiplier,bs,PetscInt,&v_pivots);CHKERRQ(ierr);
1412 
1413   bi[0]    = 0.;
1414   bdiag[0] = (nnz_max/bs2)-1; /* location of diagonal in factor B */
1415   bi[2*mbs+1] = bdiag[0]+1; /* endof bj and ba array */
1416   for (i=0; i<mbs; i++) {
1417     /* copy initial fill into linked list */
1418     nzi = 0; /* nonzeros for active row i */
1419     nzi = ai[r[i]+1] - ai[r[i]];
1420     if (!nzi) SETERRQ2(PETSC_ERR_MAT_LU_ZRPVT,"Empty row in matrix: row in original ordering %D in permuted ordering %D",r[i],i);
1421     nzi_al = adiag[r[i]] - ai[r[i]];
1422     nzi_au = ai[r[i]+1] - adiag[r[i]] -1;
1423     /* printf("row %d, nzi_al/au %d %d\n",i,nzi_al,nzi_au); */
1424 
1425     /* load in initial unfactored row */
1426     ajtmp = aj + ai[r[i]];
1427     ierr = PetscLLAddPerm(nzi,ajtmp,ic,mbs,nlnk,lnk,lnkbt);CHKERRQ(ierr);
1428     ierr = PetscMemzero(rtmp,mbs*bs2*sizeof(PetscScalar));CHKERRQ(ierr);
1429     aatmp = a->a + bs2*ai[r[i]];
1430     for (j=0; j<nzi; j++) {
1431       ierr = PetscMemcpy(rtmp+bs2*ic[ajtmp[j]],aatmp+bs2*j,bs2*sizeof(MatScalar));CHKERRQ(ierr);
1432     }
1433 
1434     /* add pivot rows into linked list */
1435     row = lnk[mbs];
1436     while (row < i) {
1437       nzi_bl = bi[row+1] - bi[row] + 1;
1438       bjtmp = bj + bdiag[row+1]+1; /* points to 1st column next to the diagonal in U */
1439       ierr  = PetscLLAddSortedLU(bjtmp,row,nlnk,lnk,lnkbt,i,nzi_bl,im);CHKERRQ(ierr);
1440       nzi  += nlnk;
1441       row   = lnk[row];
1442     }
1443 
1444     /* copy data from lnk into jtmp, then initialize lnk */
1445     ierr = PetscLLClean(mbs,mbs,nzi,lnk,jtmp,lnkbt);CHKERRQ(ierr);
1446 
1447     /* numerical factorization */
1448     bjtmp = jtmp;
1449     row   = *bjtmp++; /* 1st pivot row */
1450 
1451     while  (row < i) {
1452       pc = rtmp + bs2*row;
1453       pv = ba + bs2*bdiag[row]; /* inv(diag) of the pivot row */
1454       Kernel_A_gets_A_times_B(bs,pc,pv,multiplier); /* pc= multiplier = pc*inv(diag[row]) */
1455       ierr = BlockAbs_private(1,bs2,pc,vtmp_abs);CHKERRQ(ierr);
1456       if (vtmp_abs[0] > dt){ /* apply tolerance dropping rule */
1457         pj         = bj + bdiag[row+1] + 1; /* point to 1st entry of U(row,:) */
1458         pv         = ba + bs2*(bdiag[row+1] + 1);
1459         nz         = bdiag[row] - bdiag[row+1] - 1; /* num of entries in U(row,:), excluding diagonal */
1460         for (j=0; j<nz; j++){
1461           Kernel_A_gets_A_minus_B_times_C(bs,rtmp+bs2*pj[j],pc,pv+bs2*j);
1462         }
1463         /* ierr = PetscLogFlops(bslog*(nz+1.0)-bs);CHKERRQ(ierr); */
1464       }
1465       row = *bjtmp++;
1466     }
1467 
1468     /* copy sparse rtmp into contiguous vtmp; separate L and U part */
1469     nzi_bl = 0; j = 0;
1470     while (jtmp[j] < i){ /* L-part. Note: jtmp is sorted */
1471       ierr = PetscMemcpy(vtmp+bs2*j,rtmp+bs2*jtmp[j],bs2*sizeof(MatScalar));CHKERRQ(ierr);
1472       nzi_bl++; j++;
1473     }
1474     nzi_bu = nzi - nzi_bl -1;
1475     /* printf("nzi %d, nzi_bl %d, nzi_bu %d\n",nzi,nzi_bl,nzi_bu); */
1476 
1477     while (j < nzi){ /* U-part */
1478       ierr = PetscMemcpy(vtmp+bs2*j,rtmp+bs2*jtmp[j],bs2*sizeof(MatScalar));CHKERRQ(ierr);
1479       /*
1480       printf(" col %d: ",jtmp[j]);
1481       for (j1=0; j1<bs2; j1++) printf(" %g",*(vtmp+bs2*j+j1));
1482       printf(" \n");
1483       */
1484       j++;
1485     }
1486 
1487     ierr = BlockAbs_private(nzi,bs2,vtmp,vtmp_abs);CHKERRQ(ierr);
1488     /*
1489     printf(" row %d, nzi %d, vtmp_abs\n",i,nzi);
1490     for (j1=0; j1<nzi; j1++) printf(" (%d %g),",jtmp[j1],vtmp_abs[j1]);
1491     printf(" \n");
1492     */
1493     bjtmp = bj + bi[i];
1494     batmp = ba + bs2*bi[i];
1495     /* apply level dropping rule to L part */
1496     ncut = nzi_al + dtcount;
1497     if (ncut < nzi_bl){
1498       ierr = PetscSortSplitReal(ncut,nzi_bl,vtmp_abs,jtmp);CHKERRQ(ierr);
1499       ierr = PetscSortIntWithScalarArray(ncut,jtmp,vtmp);CHKERRQ(ierr);
1500     } else {
1501       ncut = nzi_bl;
1502     }
1503     for (j=0; j<ncut; j++){
1504       bjtmp[j] = jtmp[j];
1505       ierr = PetscMemcpy(batmp+bs2*j,rtmp+bs2*bjtmp[j],bs2*sizeof(MatScalar));CHKERRQ(ierr);
1506       /*
1507       printf(" col %d: ",bjtmp[j]);
1508       for (j1=0; j1<bs2; j1++) printf(" %g,",*(batmp+bs2*j+j1));
1509       printf("\n");
1510       */
1511     }
1512     bi[i+1] = bi[i] + ncut;
1513     nzi = ncut + 1;
1514 
1515     /* apply level dropping rule to U part */
1516     ncut = nzi_au + dtcount;
1517     if (ncut < nzi_bu){
1518       ierr = PetscSortSplitReal(ncut,nzi_bu,vtmp_abs+nzi_bl+1,jtmp+nzi_bl+1);CHKERRQ(ierr);
1519       ierr = PetscSortIntWithScalarArray(ncut,jtmp+nzi_bl+1,vtmp+nzi_bl+1);CHKERRQ(ierr);
1520     } else {
1521       ncut = nzi_bu;
1522     }
1523     nzi += ncut;
1524 
1525     /* mark bdiagonal */
1526     bdiag[i+1]    = bdiag[i] - (ncut + 1);
1527     bi[2*mbs - i] = bi[2*mbs - i +1] - (ncut + 1);
1528 
1529     bjtmp = bj + bdiag[i];
1530     batmp = ba + bs2*bdiag[i];
1531     ierr = PetscMemcpy(batmp,rtmp+bs2*i,bs2*sizeof(MatScalar));CHKERRQ(ierr);
1532     *bjtmp = i;
1533     /*
1534     printf(" diag %d: ",*bjtmp);
1535     for (j=0; j<bs2; j++){
1536       printf(" %g,",batmp[j]);
1537     }
1538     printf("\n");
1539     */
1540     bjtmp = bj + bdiag[i+1]+1;
1541     batmp = ba + (bdiag[i+1]+1)*bs2;
1542 
1543     for (k=0; k<ncut; k++){
1544       bjtmp[k] = jtmp[nzi_bl+1+k];
1545       ierr = PetscMemcpy(batmp+bs2*k,rtmp+bs2*bjtmp[k],bs2*sizeof(MatScalar));CHKERRQ(ierr);
1546       /*
1547       printf(" col %d:",bjtmp[k]);
1548       for (j1=0; j1<bs2; j1++) printf(" %g,",*(batmp+bs2*k+j1));
1549       printf("\n");
1550       */
1551     }
1552 
1553     im[i] = nzi; /* used by PetscLLAddSortedLU() */
1554 
1555     /* invert diagonal block for simplier triangular solves - add shift??? */
1556     batmp = ba + bs2*bdiag[i];
1557     ierr = Kernel_A_gets_inverse_A(bs,batmp,v_pivots,v_work);CHKERRQ(ierr);
1558   } /* for (i=0; i<mbs; i++) */
1559   ierr = PetscFree3(v_work,multiplier,v_pivots);CHKERRQ(ierr);
1560 
1561   /* printf("end of L %d, beginning of U %d\n",bi[mbs],bdiag[mbs]); */
1562   if (bi[mbs] >= bdiag[mbs]) SETERRQ2(PETSC_ERR_ARG_SIZ,"end of L array %d cannot >= the beginning of U array %d",bi[mbs],bdiag[mbs]);
1563 
1564   ierr = ISRestoreIndices(isrow,&r);CHKERRQ(ierr);
1565   ierr = ISRestoreIndices(isicol,&ic);CHKERRQ(ierr);
1566 
1567   ierr = PetscLLDestroy(lnk,lnkbt);CHKERRQ(ierr);
1568 
1569   ierr = PetscFree2(im,jtmp);CHKERRQ(ierr);
1570   ierr = PetscFree2(rtmp,vtmp);CHKERRQ(ierr);
1571 
1572   ierr = PetscLogFlops(bs2*B->cmap->n);CHKERRQ(ierr);
1573   b->maxnz = b->nz = bi[mbs] + bdiag[0] - bdiag[mbs];
1574 
1575   ierr = ISIdentity(isrow,&row_identity);CHKERRQ(ierr);
1576   ierr = ISIdentity(isicol,&icol_identity);CHKERRQ(ierr);
1577   both_identity = (PetscTruth) (row_identity && icol_identity);
1578   if (row_identity && icol_identity) {
1579     B->ops->solve = MatSolve_SeqBAIJ_N_NaturalOrdering_newdatastruct;
1580   } else {
1581     B->ops->solve = MatSolve_SeqBAIJ_N_newdatastruct;
1582   }
1583 
1584   B->ops->solveadd          = 0;
1585   B->ops->solvetranspose    = 0;
1586   B->ops->solvetransposeadd = 0;
1587   B->ops->matsolve          = 0;
1588   B->assembled              = PETSC_TRUE;
1589   B->preallocated           = PETSC_TRUE;
1590   PetscFunctionReturn(0);
1591 }
1592