Actual source code: matimpl.h
petsc-3.10.3 2018-12-18
2: #ifndef __MATIMPL_H
5: #include <petscmat.h>
6: #include <petscmatcoarsen.h>
7: #include <petsc/private/petscimpl.h>
9: PETSC_EXTERN PetscBool MatRegisterAllCalled;
10: PETSC_EXTERN PetscBool MatSeqAIJRegisterAllCalled;
11: PETSC_EXTERN PetscBool MatOrderingRegisterAllCalled;
12: PETSC_EXTERN PetscBool MatColoringRegisterAllCalled;
13: PETSC_EXTERN PetscBool MatPartitioningRegisterAllCalled;
14: PETSC_EXTERN PetscBool MatCoarsenRegisterAllCalled;
15: PETSC_EXTERN PetscErrorCode MatRegisterAll(void);
16: PETSC_EXTERN PetscErrorCode MatOrderingRegisterAll(void);
17: PETSC_EXTERN PetscErrorCode MatColoringRegisterAll(void);
18: PETSC_EXTERN PetscErrorCode MatPartitioningRegisterAll(void);
19: PETSC_EXTERN PetscErrorCode MatCoarsenRegisterAll(void);
20: PETSC_EXTERN PetscErrorCode MatSeqAIJRegisterAll(void);
22: /*
23: This file defines the parts of the matrix data structure that are
24: shared by all matrix types.
25: */
27: /*
28: If you add entries here also add them to the MATOP enum
29: in include/petscmat.h and include/petsc/finclude/petscmat.h
30: */
31: typedef struct _MatOps *MatOps;
32: struct _MatOps {
33: /* 0*/
34: PetscErrorCode (*setvalues)(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const PetscScalar[],InsertMode);
35: PetscErrorCode (*getrow)(Mat,PetscInt,PetscInt *,PetscInt*[],PetscScalar*[]);
36: PetscErrorCode (*restorerow)(Mat,PetscInt,PetscInt *,PetscInt *[],PetscScalar *[]);
37: PetscErrorCode (*mult)(Mat,Vec,Vec);
38: PetscErrorCode (*multadd)(Mat,Vec,Vec,Vec);
39: /* 5*/
40: PetscErrorCode (*multtranspose)(Mat,Vec,Vec);
41: PetscErrorCode (*multtransposeadd)(Mat,Vec,Vec,Vec);
42: PetscErrorCode (*solve)(Mat,Vec,Vec);
43: PetscErrorCode (*solveadd)(Mat,Vec,Vec,Vec);
44: PetscErrorCode (*solvetranspose)(Mat,Vec,Vec);
45: /*10*/
46: PetscErrorCode (*solvetransposeadd)(Mat,Vec,Vec,Vec);
47: PetscErrorCode (*lufactor)(Mat,IS,IS,const MatFactorInfo*);
48: PetscErrorCode (*choleskyfactor)(Mat,IS,const MatFactorInfo*);
49: PetscErrorCode (*sor)(Mat,Vec,PetscReal,MatSORType,PetscReal,PetscInt,PetscInt,Vec);
50: PetscErrorCode (*transpose)(Mat,MatReuse,Mat *);
51: /*15*/
52: PetscErrorCode (*getinfo)(Mat,MatInfoType,MatInfo*);
53: PetscErrorCode (*equal)(Mat,Mat,PetscBool *);
54: PetscErrorCode (*getdiagonal)(Mat,Vec);
55: PetscErrorCode (*diagonalscale)(Mat,Vec,Vec);
56: PetscErrorCode (*norm)(Mat,NormType,PetscReal*);
57: /*20*/
58: PetscErrorCode (*assemblybegin)(Mat,MatAssemblyType);
59: PetscErrorCode (*assemblyend)(Mat,MatAssemblyType);
60: PetscErrorCode (*setoption)(Mat,MatOption,PetscBool );
61: PetscErrorCode (*zeroentries)(Mat);
62: /*24*/
63: PetscErrorCode (*zerorows)(Mat,PetscInt,const PetscInt[],PetscScalar,Vec,Vec);
64: PetscErrorCode (*lufactorsymbolic)(Mat,Mat,IS,IS,const MatFactorInfo*);
65: PetscErrorCode (*lufactornumeric)(Mat,Mat,const MatFactorInfo*);
66: PetscErrorCode (*choleskyfactorsymbolic)(Mat,Mat,IS,const MatFactorInfo*);
67: PetscErrorCode (*choleskyfactornumeric)(Mat,Mat,const MatFactorInfo*);
68: /*29*/
69: PetscErrorCode (*setup)(Mat);
70: PetscErrorCode (*ilufactorsymbolic)(Mat,Mat,IS,IS,const MatFactorInfo*);
71: PetscErrorCode (*iccfactorsymbolic)(Mat,Mat,IS,const MatFactorInfo*);
72: PetscErrorCode (*getdiagonalblock)(Mat,Mat*);
73: PetscErrorCode (*placeholder_33)(Mat);
74: /*34*/
75: PetscErrorCode (*duplicate)(Mat,MatDuplicateOption,Mat*);
76: PetscErrorCode (*forwardsolve)(Mat,Vec,Vec);
77: PetscErrorCode (*backwardsolve)(Mat,Vec,Vec);
78: PetscErrorCode (*ilufactor)(Mat,IS,IS,const MatFactorInfo*);
79: PetscErrorCode (*iccfactor)(Mat,IS,const MatFactorInfo*);
80: /*39*/
81: PetscErrorCode (*axpy)(Mat,PetscScalar,Mat,MatStructure);
82: PetscErrorCode (*createsubmatrices)(Mat,PetscInt,const IS[],const IS[],MatReuse,Mat *[]);
83: PetscErrorCode (*increaseoverlap)(Mat,PetscInt,IS[],PetscInt);
84: PetscErrorCode (*getvalues)(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],PetscScalar []);
85: PetscErrorCode (*copy)(Mat,Mat,MatStructure);
86: /*44*/
87: PetscErrorCode (*getrowmax)(Mat,Vec,PetscInt[]);
88: PetscErrorCode (*scale)(Mat,PetscScalar);
89: PetscErrorCode (*shift)(Mat,PetscScalar);
90: PetscErrorCode (*diagonalset)(Mat,Vec,InsertMode);
91: PetscErrorCode (*zerorowscolumns)(Mat,PetscInt,const PetscInt[],PetscScalar,Vec,Vec);
92: /*49*/
93: PetscErrorCode (*setrandom)(Mat,PetscRandom);
94: PetscErrorCode (*getrowij)(Mat,PetscInt,PetscBool ,PetscBool ,PetscInt*,const PetscInt *[],const PetscInt *[],PetscBool *);
95: PetscErrorCode (*restorerowij)(Mat,PetscInt,PetscBool ,PetscBool ,PetscInt *,const PetscInt *[],const PetscInt *[],PetscBool *);
96: PetscErrorCode (*getcolumnij)(Mat,PetscInt,PetscBool ,PetscBool ,PetscInt*,const PetscInt *[],const PetscInt *[],PetscBool *);
97: PetscErrorCode (*restorecolumnij)(Mat,PetscInt,PetscBool ,PetscBool ,PetscInt*,const PetscInt *[],const PetscInt *[],PetscBool *);
98: /*54*/
99: PetscErrorCode (*fdcoloringcreate)(Mat,ISColoring,MatFDColoring);
100: PetscErrorCode (*coloringpatch)(Mat,PetscInt,PetscInt,ISColoringValue[],ISColoring*);
101: PetscErrorCode (*setunfactored)(Mat);
102: PetscErrorCode (*permute)(Mat,IS,IS,Mat*);
103: PetscErrorCode (*setvaluesblocked)(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const PetscScalar[],InsertMode);
104: /*59*/
105: PetscErrorCode (*createsubmatrix)(Mat,IS,IS,MatReuse,Mat*);
106: PetscErrorCode (*destroy)(Mat);
107: PetscErrorCode (*view)(Mat,PetscViewer);
108: PetscErrorCode (*convertfrom)(Mat, MatType,MatReuse,Mat*);
109: PetscErrorCode (*matmatmult)(Mat,Mat,Mat,MatReuse,PetscReal,Mat*);
110: /*64*/
111: PetscErrorCode (*matmatmultsymbolic)(Mat,Mat,Mat,PetscReal,Mat*);
112: PetscErrorCode (*matmatmultnumeric)(Mat,Mat,Mat,Mat);
113: PetscErrorCode (*setlocaltoglobalmapping)(Mat,ISLocalToGlobalMapping,ISLocalToGlobalMapping);
114: PetscErrorCode (*setvalueslocal)(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const PetscScalar[],InsertMode);
115: PetscErrorCode (*zerorowslocal)(Mat,PetscInt,const PetscInt[],PetscScalar,Vec,Vec);
116: /*69*/
117: PetscErrorCode (*getrowmaxabs)(Mat,Vec,PetscInt[]);
118: PetscErrorCode (*getrowminabs)(Mat,Vec,PetscInt[]);
119: PetscErrorCode (*convert)(Mat, MatType,MatReuse,Mat*);
120: PetscErrorCode (*hasoperation)(Mat,MatOperation,PetscBool*);
121: PetscErrorCode (*placeholder_73)(Mat,void*);
122: /*74*/
123: PetscErrorCode (*setvaluesadifor)(Mat,PetscInt,void*);
124: PetscErrorCode (*fdcoloringapply)(Mat,MatFDColoring,Vec,void*);
125: PetscErrorCode (*setfromoptions)(PetscOptionItems*,Mat);
126: PetscErrorCode (*multconstrained)(Mat,Vec,Vec);
127: PetscErrorCode (*multtransposeconstrained)(Mat,Vec,Vec);
128: /*79*/
129: PetscErrorCode (*findzerodiagonals)(Mat,IS*);
130: PetscErrorCode (*mults)(Mat, Vecs, Vecs);
131: PetscErrorCode (*solves)(Mat, Vecs, Vecs);
132: PetscErrorCode (*getinertia)(Mat,PetscInt*,PetscInt*,PetscInt*);
133: PetscErrorCode (*load)(Mat, PetscViewer);
134: /*84*/
135: PetscErrorCode (*issymmetric)(Mat,PetscReal,PetscBool *);
136: PetscErrorCode (*ishermitian)(Mat,PetscReal,PetscBool *);
137: PetscErrorCode (*isstructurallysymmetric)(Mat,PetscBool *);
138: PetscErrorCode (*setvaluesblockedlocal)(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const PetscScalar[],InsertMode);
139: PetscErrorCode (*getvecs)(Mat,Vec*,Vec*);
140: /*89*/
141: PetscErrorCode (*matmult)(Mat,Mat,MatReuse,PetscReal,Mat*);
142: PetscErrorCode (*matmultsymbolic)(Mat,Mat,PetscReal,Mat*);
143: PetscErrorCode (*matmultnumeric)(Mat,Mat,Mat);
144: PetscErrorCode (*ptap)(Mat,Mat,MatReuse,PetscReal,Mat*);
145: PetscErrorCode (*ptapsymbolic)(Mat,Mat,PetscReal,Mat*); /* double dispatch wrapper routine */
146: /*94*/
147: PetscErrorCode (*ptapnumeric)(Mat,Mat,Mat); /* double dispatch wrapper routine */
148: PetscErrorCode (*mattransposemult)(Mat,Mat,MatReuse,PetscReal,Mat*);
149: PetscErrorCode (*mattransposemultsymbolic)(Mat,Mat,PetscReal,Mat*);
150: PetscErrorCode (*mattransposemultnumeric)(Mat,Mat,Mat);
151: PetscErrorCode (*placeholder_98)(Mat);
152: /*99*/
153: PetscErrorCode (*placeholder_99)(Mat);
154: PetscErrorCode (*placeholder_100)(Mat);
155: PetscErrorCode (*placeholder_101)(Mat);
156: PetscErrorCode (*conjugate)(Mat); /* complex conjugate */
157: PetscErrorCode (*viewnative)(Mat,PetscViewer);
158: /*104*/
159: PetscErrorCode (*setvaluesrow)(Mat,PetscInt,const PetscScalar[]);
160: PetscErrorCode (*realpart)(Mat);
161: PetscErrorCode (*imaginarypart)(Mat);
162: PetscErrorCode (*getrowuppertriangular)(Mat);
163: PetscErrorCode (*restorerowuppertriangular)(Mat);
164: /*109*/
165: PetscErrorCode (*matsolve)(Mat,Mat,Mat);
166: PetscErrorCode (*matsolvetranspose)(Mat,Mat,Mat);
167: PetscErrorCode (*getrowmin)(Mat,Vec,PetscInt[]);
168: PetscErrorCode (*getcolumnvector)(Mat,Vec,PetscInt);
169: PetscErrorCode (*missingdiagonal)(Mat,PetscBool *,PetscInt*);
170: /*114*/
171: PetscErrorCode (*getseqnonzerostructure)(Mat,Mat *);
172: PetscErrorCode (*create)(Mat);
173: PetscErrorCode (*getghosts)(Mat,PetscInt*,const PetscInt *[]);
174: PetscErrorCode (*getlocalsubmatrix)(Mat,IS,IS,Mat*);
175: PetscErrorCode (*restorelocalsubmatrix)(Mat,IS,IS,Mat*);
176: /*119*/
177: PetscErrorCode (*multdiagonalblock)(Mat,Vec,Vec);
178: PetscErrorCode (*hermitiantranspose)(Mat,MatReuse,Mat*);
179: PetscErrorCode (*multhermitiantranspose)(Mat,Vec,Vec);
180: PetscErrorCode (*multhermitiantransposeadd)(Mat,Vec,Vec,Vec);
181: PetscErrorCode (*getmultiprocblock)(Mat,MPI_Comm,MatReuse,Mat*);
182: /*124*/
183: PetscErrorCode (*findnonzerorows)(Mat,IS*);
184: PetscErrorCode (*getcolumnnorms)(Mat,NormType,PetscReal*);
185: PetscErrorCode (*invertblockdiagonal)(Mat,const PetscScalar**);
186: PetscErrorCode (*invertvariableblockdiagonal)(Mat,PetscInt,const PetscInt*,PetscScalar*);
187: PetscErrorCode (*createsubmatricesmpi)(Mat,PetscInt,const IS[], const IS[], MatReuse, Mat**);
188: /*129*/
189: PetscErrorCode (*setvaluesbatch)(Mat,PetscInt,PetscInt,PetscInt*,const PetscScalar*);
190: PetscErrorCode (*transposematmult)(Mat,Mat,MatReuse,PetscReal,Mat*);
191: PetscErrorCode (*transposematmultsymbolic)(Mat,Mat,PetscReal,Mat*);
192: PetscErrorCode (*transposematmultnumeric)(Mat,Mat,Mat);
193: PetscErrorCode (*transposecoloringcreate)(Mat,ISColoring,MatTransposeColoring);
194: /*134*/
195: PetscErrorCode (*transcoloringapplysptoden)(MatTransposeColoring,Mat,Mat);
196: PetscErrorCode (*transcoloringapplydentosp)(MatTransposeColoring,Mat,Mat);
197: PetscErrorCode (*rart)(Mat,Mat,MatReuse,PetscReal,Mat*);
198: PetscErrorCode (*rartsymbolic)(Mat,Mat,PetscReal,Mat*); /* double dispatch wrapper routine */
199: PetscErrorCode (*rartnumeric)(Mat,Mat,Mat); /* double dispatch wrapper routine */
200: /*139*/
201: PetscErrorCode (*setblocksizes)(Mat,PetscInt,PetscInt);
202: PetscErrorCode (*aypx)(Mat,PetscScalar,Mat,MatStructure);
203: PetscErrorCode (*residual)(Mat,Vec,Vec,Vec);
204: PetscErrorCode (*fdcoloringsetup)(Mat,ISColoring,MatFDColoring);
205: PetscErrorCode (*findoffblockdiagonalentries)(Mat,IS*);
206: /*144*/
207: PetscErrorCode (*creatempimatconcatenateseqmat)(MPI_Comm,Mat,PetscInt,MatReuse,Mat*);
208: PetscErrorCode (*destroysubmatrices)(PetscInt,Mat*[]);
209: PetscErrorCode (*mattransposesolve)(Mat,Mat,Mat);
210: };
211: /*
212: If you add MatOps entries above also add them to the MATOP enum
213: in include/petscmat.h and include/petsc/finclude/petscmat.h
214: */
216: #include <petscsys.h>
217: PETSC_EXTERN PetscErrorCode MatRegisterOp(MPI_Comm, const char[], PetscVoidFunction, const char[], PetscInt, ...);
218: PETSC_EXTERN PetscErrorCode MatQueryOp(MPI_Comm, PetscVoidFunction*, const char[], PetscInt, ...);
220: typedef struct _p_MatBaseName* MatBaseName;
221: struct _p_MatBaseName {
222: char *bname,*sname,*mname;
223: MatBaseName next;
224: };
226: PETSC_EXTERN MatBaseName MatBaseNameList;
228: /*
229: Utility private matrix routines
230: */
231: PETSC_INTERN PetscErrorCode MatFindNonzeroRowsOrCols_Basic(Mat,PetscBool,PetscReal,IS*);
232: PETSC_INTERN PetscErrorCode MatConvert_Basic(Mat,MatType,MatReuse,Mat*);
233: PETSC_INTERN PetscErrorCode MatCopy_Basic(Mat,Mat,MatStructure);
234: PETSC_INTERN PetscErrorCode MatDiagonalSet_Default(Mat,Vec,InsertMode);
236: #if defined(PETSC_USE_DEBUG)
237: # define MatCheckPreallocated(A,arg) do { \
238: if (PetscUnlikely(!(A)->preallocated)) SETERRQ3(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Must call MatXXXSetPreallocation() or MatSetUp() on argument %D \"%s\" before %s()",(arg),#A,PETSC_FUNCTION_NAME); \
239: } while (0)
240: #else
241: # define MatCheckPreallocated(A,arg) do {} while (0)
242: #endif
244: /*
245: The stash is used to temporarily store inserted matrix values that
246: belong to another processor. During the assembly phase the stashed
247: values are moved to the correct processor and
248: */
250: typedef struct _MatStashSpace *PetscMatStashSpace;
252: struct _MatStashSpace {
253: PetscMatStashSpace next;
254: PetscScalar *space_head,*val;
255: PetscInt *idx,*idy;
256: PetscInt total_space_size;
257: PetscInt local_used;
258: PetscInt local_remaining;
259: };
261: PETSC_EXTERN PetscErrorCode PetscMatStashSpaceGet(PetscInt,PetscInt,PetscMatStashSpace *);
262: PETSC_EXTERN PetscErrorCode PetscMatStashSpaceContiguous(PetscInt,PetscMatStashSpace *,PetscScalar *,PetscInt *,PetscInt *);
263: PETSC_EXTERN PetscErrorCode PetscMatStashSpaceDestroy(PetscMatStashSpace*);
265: typedef struct {
266: PetscInt count;
267: } MatStashHeader;
269: typedef struct {
270: void *buffer; /* Of type blocktype, dynamically constructed */
271: PetscInt count;
272: char pending;
273: } MatStashFrame;
275: typedef struct _MatStash MatStash;
276: struct _MatStash {
277: PetscInt nmax; /* maximum stash size */
278: PetscInt umax; /* user specified max-size */
279: PetscInt oldnmax; /* the nmax value used previously */
280: PetscInt n; /* stash size */
281: PetscInt bs; /* block size of the stash */
282: PetscInt reallocs; /* preserve the no of mallocs invoked */
283: PetscMatStashSpace space_head,space; /* linked list to hold stashed global row/column numbers and matrix values */
285: PetscErrorCode (*ScatterBegin)(Mat,MatStash*,PetscInt*);
286: PetscErrorCode (*ScatterGetMesg)(MatStash*,PetscMPIInt*,PetscInt**,PetscInt**,PetscScalar**,PetscInt*);
287: PetscErrorCode (*ScatterEnd)(MatStash*);
288: PetscErrorCode (*ScatterDestroy)(MatStash*);
290: /* The following variables are used for communication */
291: MPI_Comm comm;
292: PetscMPIInt size,rank;
293: PetscMPIInt tag1,tag2;
294: MPI_Request *send_waits; /* array of send requests */
295: MPI_Request *recv_waits; /* array of receive requests */
296: MPI_Status *send_status; /* array of send status */
297: PetscInt nsends,nrecvs; /* numbers of sends and receives */
298: PetscScalar *svalues; /* sending data */
299: PetscInt *sindices;
300: PetscScalar **rvalues; /* receiving data (values) */
301: PetscInt **rindices; /* receiving data (indices) */
302: PetscInt nprocessed; /* number of messages already processed */
303: PetscMPIInt *flg_v; /* indicates what messages have arrived so far and from whom */
304: PetscBool reproduce;
305: PetscInt reproduce_count;
307: /* The following variables are used for BTS communication */
308: PetscBool subset_off_proc; /* Subsequent assemblies will set a subset (perhaps equal) of off-process entries set on first assembly */
309: PetscBool use_status; /* Use MPI_Status to determine number of items in each message */
310: PetscMPIInt nsendranks;
311: PetscMPIInt nrecvranks;
312: PetscMPIInt *sendranks;
313: PetscMPIInt *recvranks;
314: MatStashHeader *sendhdr,*recvhdr;
315: MatStashFrame *sendframes; /* pointers to the main messages */
316: MatStashFrame *recvframes;
317: MatStashFrame *recvframe_active;
318: PetscInt recvframe_i; /* index of block within active frame */
319: PetscMPIInt recvframe_count; /* Count actually sent for current frame */
320: PetscInt recvcount; /* Number of receives processed so far */
321: PetscMPIInt *some_indices; /* From last call to MPI_Waitsome */
322: MPI_Status *some_statuses; /* Statuses from last call to MPI_Waitsome */
323: PetscMPIInt some_count; /* Number of requests completed in last call to MPI_Waitsome */
324: PetscMPIInt some_i; /* Index of request currently being processed */
325: MPI_Request *sendreqs;
326: MPI_Request *recvreqs;
327: PetscSegBuffer segsendblocks;
328: PetscSegBuffer segrecvframe;
329: PetscSegBuffer segrecvblocks;
330: MPI_Datatype blocktype;
331: size_t blocktype_size;
332: InsertMode *insertmode; /* Pointer to check mat->insertmode and set upon message arrival in case no local values have been set. */
333: };
335: PETSC_INTERN PetscErrorCode MatStashCreate_Private(MPI_Comm,PetscInt,MatStash*);
336: PETSC_INTERN PetscErrorCode MatStashDestroy_Private(MatStash*);
337: PETSC_INTERN PetscErrorCode MatStashScatterEnd_Private(MatStash*);
338: PETSC_INTERN PetscErrorCode MatStashSetInitialSize_Private(MatStash*,PetscInt);
339: PETSC_INTERN PetscErrorCode MatStashGetInfo_Private(MatStash*,PetscInt*,PetscInt*);
340: PETSC_INTERN PetscErrorCode MatStashValuesRow_Private(MatStash*,PetscInt,PetscInt,const PetscInt[],const PetscScalar[],PetscBool );
341: PETSC_INTERN PetscErrorCode MatStashValuesCol_Private(MatStash*,PetscInt,PetscInt,const PetscInt[],const PetscScalar[],PetscInt,PetscBool );
342: PETSC_INTERN PetscErrorCode MatStashValuesRowBlocked_Private(MatStash*,PetscInt,PetscInt,const PetscInt[],const PetscScalar[],PetscInt,PetscInt,PetscInt);
343: PETSC_INTERN PetscErrorCode MatStashValuesColBlocked_Private(MatStash*,PetscInt,PetscInt,const PetscInt[],const PetscScalar[],PetscInt,PetscInt,PetscInt);
344: PETSC_INTERN PetscErrorCode MatStashScatterBegin_Private(Mat,MatStash*,PetscInt*);
345: PETSC_INTERN PetscErrorCode MatStashScatterGetMesg_Private(MatStash*,PetscMPIInt*,PetscInt**,PetscInt**,PetscScalar**,PetscInt*);
346: PETSC_INTERN PetscErrorCode MatGetInfo_External(Mat,MatInfoType,MatInfo*);
348: typedef struct {
349: PetscInt dim;
350: PetscInt dims[4];
351: PetscInt starts[4];
352: PetscBool noc; /* this is a single component problem, hence user will not set MatStencil.c */
353: } MatStencilInfo;
355: /* Info about using compressed row format */
356: typedef struct {
357: PetscBool use; /* indicates compressed rows have been checked and will be used */
358: PetscInt nrows; /* number of non-zero rows */
359: PetscInt *i; /* compressed row pointer */
360: PetscInt *rindex; /* compressed row index */
361: } Mat_CompressedRow;
362: PETSC_EXTERN PetscErrorCode MatCheckCompressedRow(Mat,PetscInt,Mat_CompressedRow*,PetscInt*,PetscInt,PetscReal);
364: typedef struct { /* used by MatCreateRedundantMatrix() for reusing matredundant */
365: PetscInt nzlocal,nsends,nrecvs;
366: PetscMPIInt *send_rank,*recv_rank;
367: PetscInt *sbuf_nz,*rbuf_nz,*sbuf_j,**rbuf_j;
368: PetscScalar *sbuf_a,**rbuf_a;
369: MPI_Comm subcomm; /* when user does not provide a subcomm */
370: IS isrow,iscol;
371: Mat *matseq;
372: } Mat_Redundant;
374: struct _p_Mat {
375: PETSCHEADER(struct _MatOps);
376: PetscLayout rmap,cmap;
377: void *data; /* implementation-specific data */
378: MatFactorType factortype; /* MAT_FACTOR_LU, ILU, CHOLESKY or ICC */
379: PetscBool assembled; /* is the matrix assembled? */
380: PetscBool was_assembled; /* new values inserted into assembled mat */
381: PetscInt num_ass; /* number of times matrix has been assembled */
382: PetscObjectState nonzerostate; /* each time new nonzeros locations are introduced into the matrix this is updated */
383: MatInfo info; /* matrix information */
384: InsertMode insertmode; /* have values been inserted in matrix or added? */
385: MatStash stash,bstash; /* used for assembling off-proc mat emements */
386: MatNullSpace nullsp; /* null space (operator is singular) */
387: MatNullSpace transnullsp; /* null space of transpose of operator */
388: MatNullSpace nearnullsp; /* near null space to be used by multigrid methods */
389: PetscInt congruentlayouts; /* are the rows and columns layouts congruent? */
390: PetscBool preallocated;
391: MatStencilInfo stencil; /* information for structured grid */
392: PetscBool symmetric,hermitian,structurally_symmetric,spd;
393: PetscBool symmetric_set,hermitian_set,structurally_symmetric_set,spd_set; /* if true, then corresponding flag is correct*/
394: PetscBool symmetric_eternal;
395: PetscBool nooffprocentries,nooffproczerorows;
396: PetscBool subsetoffprocentries;
397: PetscBool submat_singleis; /* for efficient PCSetUP_ASM() */
398: PetscBool structure_only;
399: #if defined(PETSC_HAVE_VIENNACL) || defined(PETSC_HAVE_VECCUDA)
400: PetscOffloadFlag valid_GPU_matrix; /* flag pointing to the matrix on the gpu*/
401: #endif
402: void *spptr; /* pointer for special library like SuperLU */
403: char *solvertype;
404: PetscBool checksymmetryonassembly,checknullspaceonassembly;
405: PetscReal checksymmetrytol;
406: Mat schur; /* Schur complement matrix */
407: MatFactorSchurStatus schur_status; /* status of the Schur complement matrix */
408: Mat_Redundant *redundant; /* used by MatCreateRedundantMatrix() */
409: PetscBool erroriffailure; /* Generate an error if detected (for example a zero pivot) instead of returning */
410: MatFactorError factorerrortype; /* type of error in factorization */
411: PetscReal factorerror_zeropivot_value; /* If numerical zero pivot was detected this is the computed value */
412: PetscInt factorerror_zeropivot_row; /* Row where zero pivot was detected */
413: PetscInt nblocks,*bsizes; /* support for MatSetVariableBlockSizes() */
414: char *defaultvectype;
415: };
417: PETSC_INTERN PetscErrorCode MatAXPY_Basic(Mat,PetscScalar,Mat,MatStructure);
418: PETSC_INTERN PetscErrorCode MatAXPY_BasicWithPreallocation(Mat,Mat,PetscScalar,Mat,MatStructure);
420: /*
421: Utility for MatFactor (Schur complement)
422: */
423: PETSC_INTERN PetscErrorCode MatFactorFactorizeSchurComplement_Private(Mat);
424: PETSC_INTERN PetscErrorCode MatFactorInvertSchurComplement_Private(Mat);
425: PETSC_INTERN PetscErrorCode MatFactorUpdateSchurStatus_Private(Mat);
426: PETSC_INTERN PetscErrorCode MatFactorSetUpInPlaceSchur_Private(Mat);
428: /*
429: Utility for MatZeroRows
430: */
431: PETSC_INTERN PetscErrorCode MatZeroRowsMapLocal_Private(Mat,PetscInt,const PetscInt*,PetscInt*,PetscInt**);
433: /*
434: Object for partitioning graphs
435: */
437: typedef struct _MatPartitioningOps *MatPartitioningOps;
438: struct _MatPartitioningOps {
439: PetscErrorCode (*apply)(MatPartitioning,IS*);
440: PetscErrorCode (*applynd)(MatPartitioning,IS*);
441: PetscErrorCode (*setfromoptions)(PetscOptionItems*,MatPartitioning);
442: PetscErrorCode (*destroy)(MatPartitioning);
443: PetscErrorCode (*view)(MatPartitioning,PetscViewer);
444: };
446: struct _p_MatPartitioning {
447: PETSCHEADER(struct _MatPartitioningOps);
448: Mat adj;
449: PetscInt *vertex_weights;
450: PetscReal *part_weights;
451: PetscInt n; /* number of partitions */
452: void *data;
453: PetscInt setupcalled;
454: };
456: /* needed for parallel nested dissection by ParMetis and PTSCOTCH */
457: PETSC_INTERN PetscErrorCode MatPartitioningSizesToSep_Private(PetscInt,PetscInt[],PetscInt[],PetscInt[]);
459: /*
460: Object for coarsen graphs
461: */
462: typedef struct _MatCoarsenOps *MatCoarsenOps;
463: struct _MatCoarsenOps {
464: PetscErrorCode (*apply)(MatCoarsen);
465: PetscErrorCode (*setfromoptions)(PetscOptionItems*,MatCoarsen);
466: PetscErrorCode (*destroy)(MatCoarsen);
467: PetscErrorCode (*view)(MatCoarsen,PetscViewer);
468: };
470: struct _p_MatCoarsen {
471: PETSCHEADER(struct _MatCoarsenOps);
472: Mat graph;
473: PetscInt setupcalled;
474: void *subctx;
475: /* */
476: PetscBool strict_aggs;
477: IS perm;
478: PetscCoarsenData *agg_lists;
479: };
481: /*
482: MatFDColoring is used to compute Jacobian matrices efficiently
483: via coloring. The data structure is explained below in an example.
485: Color = 0 1 0 2 | 2 3 0
486: ---------------------------------------------------
487: 00 01 | 05
488: 10 11 | 14 15 Processor 0
489: 22 23 | 25
490: 32 33 |
491: ===================================================
492: | 44 45 46
493: 50 | 55 Processor 1
494: | 64 66
495: ---------------------------------------------------
497: ncolors = 4;
499: ncolumns = {2,1,1,0}
500: columns = {{0,2},{1},{3},{}}
501: nrows = {4,2,3,3}
502: rows = {{0,1,2,3},{0,1},{1,2,3},{0,1,2}}
503: vwscale = {dx(0),dx(1),dx(2),dx(3)} MPI Vec
504: vscale = {dx(0),dx(1),dx(2),dx(3),dx(4),dx(5)} Seq Vec
506: ncolumns = {1,0,1,1}
507: columns = {{6},{},{4},{5}}
508: nrows = {3,0,2,2}
509: rows = {{0,1,2},{},{1,2},{1,2}}
510: vwscale = {dx(4),dx(5),dx(6)} MPI Vec
511: vscale = {dx(0),dx(4),dx(5),dx(6)} Seq Vec
513: See the routine MatFDColoringApply() for how this data is used
514: to compute the Jacobian.
516: */
517: typedef struct {
518: PetscInt row;
519: PetscInt col;
520: PetscScalar *valaddr; /* address of value */
521: } MatEntry;
523: typedef struct {
524: PetscInt row;
525: PetscScalar *valaddr; /* address of value */
526: } MatEntry2;
528: struct _p_MatFDColoring{
529: PETSCHEADER(int);
530: PetscInt M,N,m; /* total rows, columns; local rows */
531: PetscInt rstart; /* first row owned by local processor */
532: PetscInt ncolors; /* number of colors */
533: PetscInt *ncolumns; /* number of local columns for a color */
534: PetscInt **columns; /* lists the local columns of each color (using global column numbering) */
535: PetscInt *nrows; /* number of local rows for each color */
536: MatEntry *matentry; /* holds (row, column, address of value) for Jacobian matrix entry */
537: MatEntry2 *matentry2; /* holds (row, address of value) for Jacobian matrix entry */
538: PetscScalar *dy; /* store a block of F(x+dx)-F(x) when J is in BAIJ format */
539: PetscReal error_rel; /* square root of relative error in computing function */
540: PetscReal umin; /* minimum allowable u'dx value */
541: Vec w1,w2,w3; /* work vectors used in computing Jacobian */
542: PetscBool fset; /* indicates that the initial function value F(X) is set */
543: PetscErrorCode (*f)(void); /* function that defines Jacobian */
544: void *fctx; /* optional user-defined context for use by the function f */
545: Vec vscale; /* holds FD scaling, i.e. 1/dx for each perturbed column */
546: PetscInt currentcolor; /* color for which function evaluation is being done now */
547: const char *htype; /* "wp" or "ds" */
548: ISColoringType ctype; /* IS_COLORING_GLOBAL or IS_COLORING_LOCAL */
549: PetscInt brows,bcols; /* number of block rows or columns for speedup inserting the dense matrix into sparse Jacobian */
550: PetscBool setupcalled; /* true if setup has been called */
551: PetscBool viewed; /* true if the -mat_fd_coloring_view has been triggered already */
552: void (*ftn_func_pointer)(void),*ftn_func_cntx; /* serve the same purpose as *fortran_func_pointers in PETSc objects */
553: };
555: typedef struct _MatColoringOps *MatColoringOps;
556: struct _MatColoringOps {
557: PetscErrorCode (*destroy)(MatColoring);
558: PetscErrorCode (*setfromoptions)(PetscOptionItems*,MatColoring);
559: PetscErrorCode (*view)(MatColoring,PetscViewer);
560: PetscErrorCode (*apply)(MatColoring,ISColoring*);
561: PetscErrorCode (*weights)(MatColoring,PetscReal**,PetscInt**);
562: };
564: struct _p_MatColoring {
565: PETSCHEADER(struct _MatColoringOps);
566: Mat mat;
567: PetscInt dist; /* distance of the coloring */
568: PetscInt maxcolors; /* the maximum number of colors returned, maxcolors=1 for MIS */
569: void *data; /* inner context */
570: PetscBool valid; /* check to see if what is produced is a valid coloring */
571: MatColoringWeightType weight_type; /* type of weight computation to be performed */
572: PetscReal *user_weights; /* custom weights and permutation */
573: PetscInt *user_lperm;
574: PetscBool valid_iscoloring; /* check to see if matcoloring is produced a valid iscoloring */
575: };
577: struct _p_MatTransposeColoring{
578: PETSCHEADER(int);
579: PetscInt M,N,m; /* total rows, columns; local rows */
580: PetscInt rstart; /* first row owned by local processor */
581: PetscInt ncolors; /* number of colors */
582: PetscInt *ncolumns; /* number of local columns for a color */
583: PetscInt *nrows; /* number of local rows for each color */
584: PetscInt currentcolor; /* color for which function evaluation is being done now */
585: ISColoringType ctype; /* IS_COLORING_GLOBAL or IS_COLORING_LOCAL */
587: PetscInt *colorforrow,*colorforcol; /* pointer to rows and columns */
588: PetscInt *rows; /* lists the local rows for each color (using the local row numbering) */
589: PetscInt *den2sp; /* maps (row,color) in the dense matrix to index of sparse matrix array a->a */
590: PetscInt *columns; /* lists the local columns of each color (using global column numbering) */
591: PetscInt brows; /* number of rows for efficient implementation of MatTransColoringApplyDenToSp() */
592: PetscInt *lstart; /* array used for loop over row blocks of Csparse */
593: };
595: /*
596: Null space context for preconditioner/operators
597: */
598: struct _p_MatNullSpace {
599: PETSCHEADER(int);
600: PetscBool has_cnst;
601: PetscInt n;
602: Vec* vecs;
603: PetscScalar* alpha; /* for projections */
604: PetscErrorCode (*remove)(MatNullSpace,Vec,void*); /* for user provided removal function */
605: void* rmctx; /* context for remove() function */
606: };
608: /*
609: Checking zero pivot for LU, ILU preconditioners.
610: */
611: typedef struct {
612: PetscInt nshift,nshift_max;
613: PetscReal shift_amount,shift_lo,shift_hi,shift_top,shift_fraction;
614: PetscBool newshift;
615: PetscReal rs; /* active row sum of abs(offdiagonals) */
616: PetscScalar pv; /* pivot of the active row */
617: } FactorShiftCtx;
619: /*
620: Used by MatCreateSubMatrices_MPIXAIJ_Local()
621: */
622: #include <petscctable.h>
623: typedef struct { /* used by MatCreateSubMatrices_MPIAIJ_SingleIS_Local() and MatCreateSubMatrices_MPIAIJ_Local */
624: PetscInt id; /* index of submats, only submats[0] is responsible for deleting some arrays below */
625: PetscInt nrqs,nrqr;
626: PetscInt **rbuf1,**rbuf2,**rbuf3,**sbuf1,**sbuf2;
627: PetscInt **ptr;
628: PetscInt *tmp;
629: PetscInt *ctr;
630: PetscInt *pa; /* proc array */
631: PetscInt *req_size,*req_source1,*req_source2;
632: PetscBool allcolumns,allrows;
633: PetscBool singleis;
634: PetscInt *row2proc; /* row to proc map */
635: PetscInt nstages;
636: #if defined(PETSC_USE_CTABLE)
637: PetscTable cmap,rmap;
638: PetscInt *cmap_loc,*rmap_loc;
639: #else
640: PetscInt *cmap,*rmap;
641: #endif
643: PetscErrorCode (*destroy)(Mat);
644: } Mat_SubSppt;
646: PETSC_EXTERN PetscErrorCode MatFactorDumpMatrix(Mat);
647: PETSC_INTERN PetscErrorCode MatShift_Basic(Mat,PetscScalar);
648: PETSC_INTERN PetscErrorCode MatSetBlockSizes_Default(Mat,PetscInt,PetscInt);
650: PETSC_STATIC_INLINE PetscErrorCode MatPivotCheck_nz(Mat mat,const MatFactorInfo *info,FactorShiftCtx *sctx,PetscInt row)
651: {
652: PetscReal _rs = sctx->rs;
653: PetscReal _zero = info->zeropivot*_rs;
656: if (PetscAbsScalar(sctx->pv) <= _zero && !PetscIsNanScalar(sctx->pv)){
657: /* force |diag| > zeropivot*rs */
658: if (!sctx->nshift) sctx->shift_amount = info->shiftamount;
659: else sctx->shift_amount *= 2.0;
660: sctx->newshift = PETSC_TRUE;
661: (sctx->nshift)++;
662: } else {
663: sctx->newshift = PETSC_FALSE;
664: }
665: return(0);
666: }
668: PETSC_STATIC_INLINE PetscErrorCode MatPivotCheck_pd(Mat mat,const MatFactorInfo *info,FactorShiftCtx *sctx,PetscInt row)
669: {
670: PetscReal _rs = sctx->rs;
671: PetscReal _zero = info->zeropivot*_rs;
674: if (PetscRealPart(sctx->pv) <= _zero && !PetscIsNanScalar(sctx->pv)){
675: /* force matfactor to be diagonally dominant */
676: if (sctx->nshift == sctx->nshift_max) {
677: sctx->shift_fraction = sctx->shift_hi;
678: } else {
679: sctx->shift_lo = sctx->shift_fraction;
680: sctx->shift_fraction = (sctx->shift_hi+sctx->shift_lo)/2.;
681: }
682: sctx->shift_amount = sctx->shift_fraction * sctx->shift_top;
683: sctx->nshift++;
684: sctx->newshift = PETSC_TRUE;
685: } else {
686: sctx->newshift = PETSC_FALSE;
687: }
688: return(0);
689: }
691: PETSC_STATIC_INLINE PetscErrorCode MatPivotCheck_inblocks(Mat mat,const MatFactorInfo *info,FactorShiftCtx *sctx,PetscInt row)
692: {
693: PetscReal _zero = info->zeropivot;
696: if (PetscAbsScalar(sctx->pv) <= _zero && !PetscIsNanScalar(sctx->pv)){
697: sctx->pv += info->shiftamount;
698: sctx->shift_amount = 0.0;
699: sctx->nshift++;
700: }
701: sctx->newshift = PETSC_FALSE;
702: return(0);
703: }
705: PETSC_STATIC_INLINE PetscErrorCode MatPivotCheck_none(Mat fact,Mat mat,const MatFactorInfo *info,FactorShiftCtx *sctx,PetscInt row)
706: {
707: PetscReal _zero = info->zeropivot;
711: sctx->newshift = PETSC_FALSE;
712: if (PetscAbsScalar(sctx->pv) <= _zero && !PetscIsNanScalar(sctx->pv)) {
713: if (!mat->erroriffailure) {
714: PetscInfo3(mat,"Detected zero pivot in factorization in row %D value %g tolerance %g\n",row,(double)PetscAbsScalar(sctx->pv),(double)_zero);
715: fact->factorerrortype = MAT_FACTOR_NUMERIC_ZEROPIVOT;
716: fact->factorerror_zeropivot_value = PetscAbsScalar(sctx->pv);
717: fact->factorerror_zeropivot_row = row;
718: } else SETERRQ3(PETSC_COMM_SELF,PETSC_ERR_MAT_LU_ZRPVT,"Zero pivot row %D value %g tolerance %g\n",row,(double)PetscAbsScalar(sctx->pv),(double)_zero);
719: }
720: return(0);
721: }
723: PETSC_STATIC_INLINE PetscErrorCode MatPivotCheck(Mat fact,Mat mat,const MatFactorInfo *info,FactorShiftCtx *sctx,PetscInt row)
724: {
728: if (info->shifttype == (PetscReal) MAT_SHIFT_NONZERO){
729: MatPivotCheck_nz(mat,info,sctx,row);
730: } else if (info->shifttype == (PetscReal) MAT_SHIFT_POSITIVE_DEFINITE){
731: MatPivotCheck_pd(mat,info,sctx,row);
732: } else if (info->shifttype == (PetscReal) MAT_SHIFT_INBLOCKS){
733: MatPivotCheck_inblocks(mat,info,sctx,row);
734: } else {
735: MatPivotCheck_none(fact,mat,info,sctx,row);
736: }
737: return(0);
738: }
740: /*
741: Create and initialize a linked list
742: Input Parameters:
743: idx_start - starting index of the list
744: lnk_max - max value of lnk indicating the end of the list
745: nlnk - max length of the list
746: Output Parameters:
747: lnk - list initialized
748: bt - PetscBT (bitarray) with all bits set to false
749: lnk_empty - flg indicating the list is empty
750: */
751: #define PetscLLCreate(idx_start,lnk_max,nlnk,lnk,bt) \
752: (PetscMalloc1(nlnk,&lnk) || PetscBTCreate(nlnk,&(bt)) || (lnk[idx_start] = lnk_max,0))
754: #define PetscLLCreate_new(idx_start,lnk_max,nlnk,lnk,bt,lnk_empty)\
755: (PetscMalloc1(nlnk,&lnk) || PetscBTCreate(nlnk,&(bt)) || (lnk_empty = PETSC_TRUE,0) ||(lnk[idx_start] = lnk_max,0))
757: /*
758: Add an index set into a sorted linked list
759: Input Parameters:
760: nidx - number of input indices
761: indices - integer array
762: idx_start - starting index of the list
763: lnk - linked list(an integer array) that is created
764: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
765: output Parameters:
766: nlnk - number of newly added indices
767: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from indices
768: bt - updated PetscBT (bitarray)
769: */
770: #define PetscLLAdd(nidx,indices,idx_start,nlnk,lnk,bt) 0;\
771: {\
772: PetscInt _k,_entry,_location,_lnkdata;\
773: nlnk = 0;\
774: _lnkdata = idx_start;\
775: for (_k=0; _k<nidx; _k++){\
776: _entry = indices[_k];\
777: if (!PetscBTLookupSet(bt,_entry)){ /* new entry */\
778: /* search for insertion location */\
779: /* start from the beginning if _entry < previous _entry */\
780: if (_k && _entry < _lnkdata) _lnkdata = idx_start;\
781: do {\
782: _location = _lnkdata;\
783: _lnkdata = lnk[_location];\
784: } while (_entry > _lnkdata);\
785: /* insertion location is found, add entry into lnk */\
786: lnk[_location] = _entry;\
787: lnk[_entry] = _lnkdata;\
788: nlnk++;\
789: _lnkdata = _entry; /* next search starts from here if next_entry > _entry */\
790: }\
791: }\
792: }
794: /*
795: Add a permuted index set into a sorted linked list
796: Input Parameters:
797: nidx - number of input indices
798: indices - integer array
799: perm - permutation of indices
800: idx_start - starting index of the list
801: lnk - linked list(an integer array) that is created
802: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
803: output Parameters:
804: nlnk - number of newly added indices
805: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from indices
806: bt - updated PetscBT (bitarray)
807: */
808: #define PetscLLAddPerm(nidx,indices,perm,idx_start,nlnk,lnk,bt) 0;\
809: {\
810: PetscInt _k,_entry,_location,_lnkdata;\
811: nlnk = 0;\
812: _lnkdata = idx_start;\
813: for (_k=0; _k<nidx; _k++){\
814: _entry = perm[indices[_k]];\
815: if (!PetscBTLookupSet(bt,_entry)){ /* new entry */\
816: /* search for insertion location */\
817: /* start from the beginning if _entry < previous _entry */\
818: if (_k && _entry < _lnkdata) _lnkdata = idx_start;\
819: do {\
820: _location = _lnkdata;\
821: _lnkdata = lnk[_location];\
822: } while (_entry > _lnkdata);\
823: /* insertion location is found, add entry into lnk */\
824: lnk[_location] = _entry;\
825: lnk[_entry] = _lnkdata;\
826: nlnk++;\
827: _lnkdata = _entry; /* next search starts from here if next_entry > _entry */\
828: }\
829: }\
830: }
832: /*
833: Add a SORTED ascending index set into a sorted linked list - same as PetscLLAdd() bus skip 'if (_k && _entry < _lnkdata) _lnkdata = idx_start;'
834: Input Parameters:
835: nidx - number of input indices
836: indices - sorted integer array
837: idx_start - starting index of the list
838: lnk - linked list(an integer array) that is created
839: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
840: output Parameters:
841: nlnk - number of newly added indices
842: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from indices
843: bt - updated PetscBT (bitarray)
844: */
845: #define PetscLLAddSorted(nidx,indices,idx_start,nlnk,lnk,bt) 0;\
846: {\
847: PetscInt _k,_entry,_location,_lnkdata;\
848: nlnk = 0;\
849: _lnkdata = idx_start;\
850: for (_k=0; _k<nidx; _k++){\
851: _entry = indices[_k];\
852: if (!PetscBTLookupSet(bt,_entry)){ /* new entry */\
853: /* search for insertion location */\
854: do {\
855: _location = _lnkdata;\
856: _lnkdata = lnk[_location];\
857: } while (_entry > _lnkdata);\
858: /* insertion location is found, add entry into lnk */\
859: lnk[_location] = _entry;\
860: lnk[_entry] = _lnkdata;\
861: nlnk++;\
862: _lnkdata = _entry; /* next search starts from here */\
863: }\
864: }\
865: }
867: #define PetscLLAddSorted_new(nidx,indices,idx_start,lnk_empty,nlnk,lnk,bt) 0; \
868: {\
869: PetscInt _k,_entry,_location,_lnkdata;\
870: if (lnk_empty){\
871: _lnkdata = idx_start; \
872: for (_k=0; _k<nidx; _k++){ \
873: _entry = indices[_k]; \
874: PetscBTSet(bt,_entry); /* mark the new entry */ \
875: _location = _lnkdata; \
876: _lnkdata = lnk[_location]; \
877: /* insertion location is found, add entry into lnk */ \
878: lnk[_location] = _entry; \
879: lnk[_entry] = _lnkdata; \
880: _lnkdata = _entry; /* next search starts from here */ \
881: } \
882: /*\
883: lnk[indices[nidx-1]] = lnk[idx_start];\
884: lnk[idx_start] = indices[0];\
885: PetscBTSet(bt,indices[0]); \
886: for (_k=1; _k<nidx; _k++){ \
887: PetscBTSet(bt,indices[_k]); \
888: lnk[indices[_k-1]] = indices[_k]; \
889: } \
890: */\
891: nlnk = nidx;\
892: lnk_empty = PETSC_FALSE;\
893: } else {\
894: nlnk = 0; \
895: _lnkdata = idx_start; \
896: for (_k=0; _k<nidx; _k++){ \
897: _entry = indices[_k]; \
898: if (!PetscBTLookupSet(bt,_entry)){ /* new entry */ \
899: /* search for insertion location */ \
900: do { \
901: _location = _lnkdata; \
902: _lnkdata = lnk[_location]; \
903: } while (_entry > _lnkdata); \
904: /* insertion location is found, add entry into lnk */ \
905: lnk[_location] = _entry; \
906: lnk[_entry] = _lnkdata; \
907: nlnk++; \
908: _lnkdata = _entry; /* next search starts from here */ \
909: } \
910: } \
911: } \
912: }
914: /*
915: Add a SORTED index set into a sorted linked list used for LUFactorSymbolic()
916: Same as PetscLLAddSorted() with an additional operation:
917: count the number of input indices that are no larger than 'diag'
918: Input Parameters:
919: indices - sorted integer array
920: idx_start - starting index of the list, index of pivot row
921: lnk - linked list(an integer array) that is created
922: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
923: diag - index of the active row in LUFactorSymbolic
924: nzbd - number of input indices with indices <= idx_start
925: im - im[idx_start] is initialized as num of nonzero entries in row=idx_start
926: output Parameters:
927: nlnk - number of newly added indices
928: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from indices
929: bt - updated PetscBT (bitarray)
930: im - im[idx_start]: unchanged if diag is not an entry
931: : num of entries with indices <= diag if diag is an entry
932: */
933: #define PetscLLAddSortedLU(indices,idx_start,nlnk,lnk,bt,diag,nzbd,im) 0;\
934: {\
935: PetscInt _k,_entry,_location,_lnkdata,_nidx;\
936: nlnk = 0;\
937: _lnkdata = idx_start;\
938: _nidx = im[idx_start] - nzbd; /* num of entries with idx_start < index <= diag */\
939: for (_k=0; _k<_nidx; _k++){\
940: _entry = indices[_k];\
941: nzbd++;\
942: if ( _entry== diag) im[idx_start] = nzbd;\
943: if (!PetscBTLookupSet(bt,_entry)){ /* new entry */\
944: /* search for insertion location */\
945: do {\
946: _location = _lnkdata;\
947: _lnkdata = lnk[_location];\
948: } while (_entry > _lnkdata);\
949: /* insertion location is found, add entry into lnk */\
950: lnk[_location] = _entry;\
951: lnk[_entry] = _lnkdata;\
952: nlnk++;\
953: _lnkdata = _entry; /* next search starts from here */\
954: }\
955: }\
956: }
958: /*
959: Copy data on the list into an array, then initialize the list
960: Input Parameters:
961: idx_start - starting index of the list
962: lnk_max - max value of lnk indicating the end of the list
963: nlnk - number of data on the list to be copied
964: lnk - linked list
965: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
966: output Parameters:
967: indices - array that contains the copied data
968: lnk - linked list that is cleaned and initialize
969: bt - PetscBT (bitarray) with all bits set to false
970: */
971: #define PetscLLClean(idx_start,lnk_max,nlnk,lnk,indices,bt) 0;\
972: {\
973: PetscInt _j,_idx=idx_start;\
974: for (_j=0; _j<nlnk; _j++){\
975: _idx = lnk[_idx];\
976: indices[_j] = _idx;\
977: PetscBTClear(bt,_idx);\
978: }\
979: lnk[idx_start] = lnk_max;\
980: }
981: /*
982: Free memories used by the list
983: */
984: #define PetscLLDestroy(lnk,bt) (PetscFree(lnk) || PetscBTDestroy(&(bt)))
986: /* Routines below are used for incomplete matrix factorization */
987: /*
988: Create and initialize a linked list and its levels
989: Input Parameters:
990: idx_start - starting index of the list
991: lnk_max - max value of lnk indicating the end of the list
992: nlnk - max length of the list
993: Output Parameters:
994: lnk - list initialized
995: lnk_lvl - array of size nlnk for storing levels of lnk
996: bt - PetscBT (bitarray) with all bits set to false
997: */
998: #define PetscIncompleteLLCreate(idx_start,lnk_max,nlnk,lnk,lnk_lvl,bt)\
999: (PetscIntMultError(2,nlnk,NULL) || PetscMalloc1(2*nlnk,&lnk) || PetscBTCreate(nlnk,&(bt)) || (lnk[idx_start] = lnk_max,lnk_lvl = lnk + nlnk,0))
1001: /*
1002: Initialize a sorted linked list used for ILU and ICC
1003: Input Parameters:
1004: nidx - number of input idx
1005: idx - integer array used for storing column indices
1006: idx_start - starting index of the list
1007: perm - indices of an IS
1008: lnk - linked list(an integer array) that is created
1009: lnklvl - levels of lnk
1010: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1011: output Parameters:
1012: nlnk - number of newly added idx
1013: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1014: lnklvl - levels of lnk
1015: bt - updated PetscBT (bitarray)
1016: */
1017: #define PetscIncompleteLLInit(nidx,idx,idx_start,perm,nlnk,lnk,lnklvl,bt) 0;\
1018: {\
1019: PetscInt _k,_entry,_location,_lnkdata;\
1020: nlnk = 0;\
1021: _lnkdata = idx_start;\
1022: for (_k=0; _k<nidx; _k++){\
1023: _entry = perm[idx[_k]];\
1024: if (!PetscBTLookupSet(bt,_entry)){ /* new entry */\
1025: /* search for insertion location */\
1026: if (_k && _entry < _lnkdata) _lnkdata = idx_start;\
1027: do {\
1028: _location = _lnkdata;\
1029: _lnkdata = lnk[_location];\
1030: } while (_entry > _lnkdata);\
1031: /* insertion location is found, add entry into lnk */\
1032: lnk[_location] = _entry;\
1033: lnk[_entry] = _lnkdata;\
1034: lnklvl[_entry] = 0;\
1035: nlnk++;\
1036: _lnkdata = _entry; /* next search starts from here if next_entry > _entry */\
1037: }\
1038: }\
1039: }
1041: /*
1042: Add a SORTED index set into a sorted linked list for ILU
1043: Input Parameters:
1044: nidx - number of input indices
1045: idx - sorted integer array used for storing column indices
1046: level - level of fill, e.g., ICC(level)
1047: idxlvl - level of idx
1048: idx_start - starting index of the list
1049: lnk - linked list(an integer array) that is created
1050: lnklvl - levels of lnk
1051: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1052: prow - the row number of idx
1053: output Parameters:
1054: nlnk - number of newly added idx
1055: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1056: lnklvl - levels of lnk
1057: bt - updated PetscBT (bitarray)
1059: Note: the level of factor(i,j) is set as lvl(i,j) = min{ lvl(i,j), lvl(i,prow)+lvl(prow,j)+1)
1060: where idx = non-zero columns of U(prow,prow+1:n-1), prow<i
1061: */
1062: #define PetscILULLAddSorted(nidx,idx,level,idxlvl,idx_start,nlnk,lnk,lnklvl,bt,lnklvl_prow) 0;\
1063: {\
1064: PetscInt _k,_entry,_location,_lnkdata,_incrlev,_lnklvl_prow=lnklvl[prow];\
1065: nlnk = 0;\
1066: _lnkdata = idx_start;\
1067: for (_k=0; _k<nidx; _k++){\
1068: _incrlev = idxlvl[_k] + _lnklvl_prow + 1;\
1069: if (_incrlev > level) continue;\
1070: _entry = idx[_k];\
1071: if (!PetscBTLookupSet(bt,_entry)){ /* new entry */\
1072: /* search for insertion location */\
1073: do {\
1074: _location = _lnkdata;\
1075: _lnkdata = lnk[_location];\
1076: } while (_entry > _lnkdata);\
1077: /* insertion location is found, add entry into lnk */\
1078: lnk[_location] = _entry;\
1079: lnk[_entry] = _lnkdata;\
1080: lnklvl[_entry] = _incrlev;\
1081: nlnk++;\
1082: _lnkdata = _entry; /* next search starts from here if next_entry > _entry */\
1083: } else { /* existing entry: update lnklvl */\
1084: if (lnklvl[_entry] > _incrlev) lnklvl[_entry] = _incrlev;\
1085: }\
1086: }\
1087: }
1089: /*
1090: Add a index set into a sorted linked list
1091: Input Parameters:
1092: nidx - number of input idx
1093: idx - integer array used for storing column indices
1094: level - level of fill, e.g., ICC(level)
1095: idxlvl - level of idx
1096: idx_start - starting index of the list
1097: lnk - linked list(an integer array) that is created
1098: lnklvl - levels of lnk
1099: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1100: output Parameters:
1101: nlnk - number of newly added idx
1102: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1103: lnklvl - levels of lnk
1104: bt - updated PetscBT (bitarray)
1105: */
1106: #define PetscIncompleteLLAdd(nidx,idx,level,idxlvl,idx_start,nlnk,lnk,lnklvl,bt) 0;\
1107: {\
1108: PetscInt _k,_entry,_location,_lnkdata,_incrlev;\
1109: nlnk = 0;\
1110: _lnkdata = idx_start;\
1111: for (_k=0; _k<nidx; _k++){\
1112: _incrlev = idxlvl[_k] + 1;\
1113: if (_incrlev > level) continue;\
1114: _entry = idx[_k];\
1115: if (!PetscBTLookupSet(bt,_entry)){ /* new entry */\
1116: /* search for insertion location */\
1117: if (_k && _entry < _lnkdata) _lnkdata = idx_start;\
1118: do {\
1119: _location = _lnkdata;\
1120: _lnkdata = lnk[_location];\
1121: } while (_entry > _lnkdata);\
1122: /* insertion location is found, add entry into lnk */\
1123: lnk[_location] = _entry;\
1124: lnk[_entry] = _lnkdata;\
1125: lnklvl[_entry] = _incrlev;\
1126: nlnk++;\
1127: _lnkdata = _entry; /* next search starts from here if next_entry > _entry */\
1128: } else { /* existing entry: update lnklvl */\
1129: if (lnklvl[_entry] > _incrlev) lnklvl[_entry] = _incrlev;\
1130: }\
1131: }\
1132: }
1134: /*
1135: Add a SORTED index set into a sorted linked list
1136: Input Parameters:
1137: nidx - number of input indices
1138: idx - sorted integer array used for storing column indices
1139: level - level of fill, e.g., ICC(level)
1140: idxlvl - level of idx
1141: idx_start - starting index of the list
1142: lnk - linked list(an integer array) that is created
1143: lnklvl - levels of lnk
1144: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1145: output Parameters:
1146: nlnk - number of newly added idx
1147: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1148: lnklvl - levels of lnk
1149: bt - updated PetscBT (bitarray)
1150: */
1151: #define PetscIncompleteLLAddSorted(nidx,idx,level,idxlvl,idx_start,nlnk,lnk,lnklvl,bt) 0;\
1152: {\
1153: PetscInt _k,_entry,_location,_lnkdata,_incrlev;\
1154: nlnk = 0;\
1155: _lnkdata = idx_start;\
1156: for (_k=0; _k<nidx; _k++){\
1157: _incrlev = idxlvl[_k] + 1;\
1158: if (_incrlev > level) continue;\
1159: _entry = idx[_k];\
1160: if (!PetscBTLookupSet(bt,_entry)){ /* new entry */\
1161: /* search for insertion location */\
1162: do {\
1163: _location = _lnkdata;\
1164: _lnkdata = lnk[_location];\
1165: } while (_entry > _lnkdata);\
1166: /* insertion location is found, add entry into lnk */\
1167: lnk[_location] = _entry;\
1168: lnk[_entry] = _lnkdata;\
1169: lnklvl[_entry] = _incrlev;\
1170: nlnk++;\
1171: _lnkdata = _entry; /* next search starts from here */\
1172: } else { /* existing entry: update lnklvl */\
1173: if (lnklvl[_entry] > _incrlev) lnklvl[_entry] = _incrlev;\
1174: }\
1175: }\
1176: }
1178: /*
1179: Add a SORTED index set into a sorted linked list for ICC
1180: Input Parameters:
1181: nidx - number of input indices
1182: idx - sorted integer array used for storing column indices
1183: level - level of fill, e.g., ICC(level)
1184: idxlvl - level of idx
1185: idx_start - starting index of the list
1186: lnk - linked list(an integer array) that is created
1187: lnklvl - levels of lnk
1188: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1189: idxlvl_prow - idxlvl[prow], where prow is the row number of the idx
1190: output Parameters:
1191: nlnk - number of newly added indices
1192: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1193: lnklvl - levels of lnk
1194: bt - updated PetscBT (bitarray)
1195: Note: the level of U(i,j) is set as lvl(i,j) = min{ lvl(i,j), lvl(prow,i)+lvl(prow,j)+1)
1196: where idx = non-zero columns of U(prow,prow+1:n-1), prow<i
1197: */
1198: #define PetscICCLLAddSorted(nidx,idx,level,idxlvl,idx_start,nlnk,lnk,lnklvl,bt,idxlvl_prow) 0;\
1199: {\
1200: PetscInt _k,_entry,_location,_lnkdata,_incrlev;\
1201: nlnk = 0;\
1202: _lnkdata = idx_start;\
1203: for (_k=0; _k<nidx; _k++){\
1204: _incrlev = idxlvl[_k] + idxlvl_prow + 1;\
1205: if (_incrlev > level) continue;\
1206: _entry = idx[_k];\
1207: if (!PetscBTLookupSet(bt,_entry)){ /* new entry */\
1208: /* search for insertion location */\
1209: do {\
1210: _location = _lnkdata;\
1211: _lnkdata = lnk[_location];\
1212: } while (_entry > _lnkdata);\
1213: /* insertion location is found, add entry into lnk */\
1214: lnk[_location] = _entry;\
1215: lnk[_entry] = _lnkdata;\
1216: lnklvl[_entry] = _incrlev;\
1217: nlnk++;\
1218: _lnkdata = _entry; /* next search starts from here */\
1219: } else { /* existing entry: update lnklvl */\
1220: if (lnklvl[_entry] > _incrlev) lnklvl[_entry] = _incrlev;\
1221: }\
1222: }\
1223: }
1225: /*
1226: Copy data on the list into an array, then initialize the list
1227: Input Parameters:
1228: idx_start - starting index of the list
1229: lnk_max - max value of lnk indicating the end of the list
1230: nlnk - number of data on the list to be copied
1231: lnk - linked list
1232: lnklvl - level of lnk
1233: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1234: output Parameters:
1235: indices - array that contains the copied data
1236: lnk - linked list that is cleaned and initialize
1237: lnklvl - level of lnk that is reinitialized
1238: bt - PetscBT (bitarray) with all bits set to false
1239: */
1240: #define PetscIncompleteLLClean(idx_start,lnk_max,nlnk,lnk,lnklvl,indices,indiceslvl,bt) 0;\
1241: {\
1242: PetscInt _j,_idx=idx_start;\
1243: for (_j=0; _j<nlnk; _j++){\
1244: _idx = lnk[_idx];\
1245: *(indices+_j) = _idx;\
1246: *(indiceslvl+_j) = lnklvl[_idx];\
1247: lnklvl[_idx] = -1;\
1248: PetscBTClear(bt,_idx);\
1249: }\
1250: lnk[idx_start] = lnk_max;\
1251: }
1252: /*
1253: Free memories used by the list
1254: */
1255: #define PetscIncompleteLLDestroy(lnk,bt) (PetscFree(lnk) || PetscBTDestroy(&(bt)))
1257: #define MatCheckSameLocalSize(A,ar1,B,ar2) \
1259: if ((A->rmap->n != B->rmap->n) || (A->cmap->n != B->cmap->n)) SETERRQ6(PETSC_COMM_SELF,PETSC_ERR_ARG_INCOMP,"Incompatible matrix local sizes: parameter # %d (%D x %D) != parameter # %d (%D x %D)",ar1,A->rmap->n,A->cmap->n,ar2,B->rmap->n,B->cmap->n);
1260:
1261: #define MatCheckSameSize(A,ar1,B,ar2) \
1262: if ((A->rmap->N != B->rmap->N) || (A->cmap->N != B->cmap->N)) SETERRQ6(PetscObjectComm((PetscObject)A),PETSC_ERR_ARG_INCOMP,"Incompatible matrix global sizes: parameter # %d (%D x %D) != parameter # %d (%D x %D)",ar1,A->rmap->N,A->cmap->N,ar2,B->rmap->N,B->cmap->N);\
1263: MatCheckSameLocalSize(A,ar1,B,ar2);
1264:
1265: #define VecCheckMatCompatible(M,x,ar1,b,ar2) \
1266: if (M->cmap->N != x->map->N) SETERRQ3(PetscObjectComm((PetscObject)M),PETSC_ERR_ARG_SIZ,"Vector global length incompatible with matrix: parameter # %d global size %D != matrix column global size %D",ar1,x->map->N,M->cmap->N);\
1267: if (M->rmap->N != b->map->N) SETERRQ3(PetscObjectComm((PetscObject)M),PETSC_ERR_ARG_SIZ,"Vector global length incompatible with matrix: parameter # %d global size %D != matrix row global size %D",ar2,b->map->N,M->rmap->N);
1269: /* -------------------------------------------------------------------------------------------------------*/
1270: #include <petscbt.h>
1271: /*
1272: Create and initialize a condensed linked list -
1273: same as PetscLLCreate(), but uses a scalable array 'lnk' with size of max number of entries, not O(N).
1274: Barry suggested this approach (Dec. 6, 2011):
1275: I've thought of an alternative way of representing a linked list that is efficient but doesn't have the O(N) scaling issue
1276: (it may be faster than the O(N) even sequentially due to less crazy memory access).
1278: Instead of having some like a 2 -> 4 -> 11 -> 22 list that uses slot 2 4 11 and 22 in a big array use a small array with two slots
1279: for each entry for example [ 2 1 | 4 3 | 22 -1 | 11 2] so the first number (of the pair) is the value while the second tells you where
1280: in the list the next entry is. Inserting a new link means just append another pair at the end. For example say we want to insert 13 into the
1281: list it would then become [2 1 | 4 3 | 22 -1 | 11 4 | 13 2 ] you just add a pair at the end and fix the point for the one that points to it.
1282: That is 11 use to point to the 2 slot, after the change 11 points to the 4th slot which has the value 13. Note that values are always next
1283: to each other so memory access is much better than using the big array.
1285: Example:
1286: nlnk_max=5, lnk_max=36:
1287: Initial list: [0, 0 | 36, 2 | 0, 0 | 0, 0 | 0, 0 | 0, 0 | 0, 0]
1288: here, head_node has index 2 with value lnk[2]=lnk_max=36,
1289: 0-th entry is used to store the number of entries in the list,
1290: The initial lnk represents head -> tail(marked by 36) with number of entries = lnk[0]=0.
1292: Now adding a sorted set {2,4}, the list becomes
1293: [2, 0 | 36, 4 |2, 6 | 4, 2 | 0, 0 | 0, 0 | 0, 0 ]
1294: represents head -> 2 -> 4 -> tail with number of entries = lnk[0]=2.
1296: Then adding a sorted set {0,3,35}, the list
1297: [5, 0 | 36, 8 | 2, 10 | 4, 12 | 0, 4 | 3, 6 | 35, 2 ]
1298: represents head -> 0 -> 2 -> 3 -> 4 -> 35 -> tail with number of entries = lnk[0]=5.
1300: Input Parameters:
1301: nlnk_max - max length of the list
1302: lnk_max - max value of the entries
1303: Output Parameters:
1304: lnk - list created and initialized
1305: bt - PetscBT (bitarray) with all bits set to false. Note: bt has size lnk_max, not nln_max!
1306: */
1307: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedCreate(PetscInt nlnk_max,PetscInt lnk_max,PetscInt **lnk,PetscBT *bt)
1308: {
1310: PetscInt *llnk,lsize = 0;
1313: PetscIntMultError(2,nlnk_max+2,&lsize);
1314: PetscMalloc1(lsize,lnk);
1315: PetscBTCreate(lnk_max,bt);
1316: llnk = *lnk;
1317: llnk[0] = 0; /* number of entries on the list */
1318: llnk[2] = lnk_max; /* value in the head node */
1319: llnk[3] = 2; /* next for the head node */
1320: return(0);
1321: }
1323: /*
1324: Add a SORTED ascending index set into a sorted linked list. See PetscLLCondensedCreate() for detailed description.
1325: Input Parameters:
1326: nidx - number of input indices
1327: indices - sorted integer array
1328: lnk - condensed linked list(an integer array) that is created
1329: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1330: output Parameters:
1331: lnk - the sorted(increasing order) linked list containing previous and newly added non-redundate indices
1332: bt - updated PetscBT (bitarray)
1333: */
1334: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedAddSorted(PetscInt nidx,const PetscInt indices[],PetscInt lnk[],PetscBT bt)
1335: {
1336: PetscInt _k,_entry,_location,_next,_lnkdata,_nlnk,_newnode;
1339: _nlnk = lnk[0]; /* num of entries on the input lnk */
1340: _location = 2; /* head */
1341: for (_k=0; _k<nidx; _k++){
1342: _entry = indices[_k];
1343: if (!PetscBTLookupSet(bt,_entry)){ /* new entry */
1344: /* search for insertion location */
1345: do {
1346: _next = _location + 1; /* link from previous node to next node */
1347: _location = lnk[_next]; /* idx of next node */
1348: _lnkdata = lnk[_location];/* value of next node */
1349: } while (_entry > _lnkdata);
1350: /* insertion location is found, add entry into lnk */
1351: _newnode = 2*(_nlnk+2); /* index for this new node */
1352: lnk[_next] = _newnode; /* connect previous node to the new node */
1353: lnk[_newnode] = _entry; /* set value of the new node */
1354: lnk[_newnode+1] = _location; /* connect new node to next node */
1355: _location = _newnode; /* next search starts from the new node */
1356: _nlnk++;
1357: } \
1358: }\
1359: lnk[0] = _nlnk; /* number of entries in the list */
1360: return(0);
1361: }
1363: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedClean(PetscInt lnk_max,PetscInt nidx,PetscInt *indices,PetscInt lnk[],PetscBT bt)
1364: {
1366: PetscInt _k,_next,_nlnk;
1369: _next = lnk[3]; /* head node */
1370: _nlnk = lnk[0]; /* num of entries on the list */
1371: for (_k=0; _k<_nlnk; _k++){
1372: indices[_k] = lnk[_next];
1373: _next = lnk[_next + 1];
1374: PetscBTClear(bt,indices[_k]);
1375: }
1376: lnk[0] = 0; /* num of entries on the list */
1377: lnk[2] = lnk_max; /* initialize head node */
1378: lnk[3] = 2; /* head node */
1379: return(0);
1380: }
1382: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedView(PetscInt *lnk)
1383: {
1385: PetscInt k;
1388: PetscPrintf(PETSC_COMM_SELF,"LLCondensed of size %D, (val, next)\n",lnk[0]);
1389: for (k=2; k< lnk[0]+2; k++){
1390: PetscPrintf(PETSC_COMM_SELF," %D: (%D, %D)\n",2*k,lnk[2*k],lnk[2*k+1]);
1391: }
1392: return(0);
1393: }
1395: /*
1396: Free memories used by the list
1397: */
1398: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedDestroy(PetscInt *lnk,PetscBT bt)
1399: {
1403: PetscFree(lnk);
1404: PetscBTDestroy(&bt);
1405: return(0);
1406: }
1408: /* -------------------------------------------------------------------------------------------------------*/
1409: /*
1410: Same as PetscLLCondensedCreate(), but does not use non-scalable O(lnk_max) bitarray
1411: Input Parameters:
1412: nlnk_max - max length of the list
1413: Output Parameters:
1414: lnk - list created and initialized
1415: */
1416: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedCreate_Scalable(PetscInt nlnk_max,PetscInt **lnk)
1417: {
1419: PetscInt *llnk,lsize = 0;
1422: PetscIntMultError(2,nlnk_max+2,&lsize);
1423: PetscMalloc1(lsize,lnk);
1424: llnk = *lnk;
1425: llnk[0] = 0; /* number of entries on the list */
1426: llnk[2] = PETSC_MAX_INT; /* value in the head node */
1427: llnk[3] = 2; /* next for the head node */
1428: return(0);
1429: }
1431: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedExpand_Scalable(PetscInt nlnk_max,PetscInt **lnk)
1432: {
1434: PetscInt lsize = 0;
1437: PetscIntMultError(2,nlnk_max+2,&lsize);
1438: PetscRealloc(lsize*sizeof(PetscInt),lnk);
1439: return(0);
1440: }
1442: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedAddSorted_Scalable(PetscInt nidx,const PetscInt indices[],PetscInt lnk[])
1443: {
1444: PetscInt _k,_entry,_location,_next,_lnkdata,_nlnk,_newnode;
1445: _nlnk = lnk[0]; /* num of entries on the input lnk */
1446: _location = 2; /* head */ \
1447: for (_k=0; _k<nidx; _k++){
1448: _entry = indices[_k];
1449: /* search for insertion location */
1450: do {
1451: _next = _location + 1; /* link from previous node to next node */
1452: _location = lnk[_next]; /* idx of next node */
1453: _lnkdata = lnk[_location];/* value of next node */
1454: } while (_entry > _lnkdata);
1455: if (_entry < _lnkdata) {
1456: /* insertion location is found, add entry into lnk */
1457: _newnode = 2*(_nlnk+2); /* index for this new node */
1458: lnk[_next] = _newnode; /* connect previous node to the new node */
1459: lnk[_newnode] = _entry; /* set value of the new node */
1460: lnk[_newnode+1] = _location; /* connect new node to next node */
1461: _location = _newnode; /* next search starts from the new node */
1462: _nlnk++;
1463: }
1464: }
1465: lnk[0] = _nlnk; /* number of entries in the list */
1466: return 0;
1467: }
1469: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedClean_Scalable(PetscInt nidx,PetscInt *indices,PetscInt *lnk)
1470: {
1471: PetscInt _k,_next,_nlnk;
1472: _next = lnk[3]; /* head node */
1473: _nlnk = lnk[0];
1474: for (_k=0; _k<_nlnk; _k++){
1475: indices[_k] = lnk[_next];
1476: _next = lnk[_next + 1];
1477: }
1478: lnk[0] = 0; /* num of entries on the list */
1479: lnk[3] = 2; /* head node */
1480: return 0;
1481: }
1483: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedDestroy_Scalable(PetscInt *lnk)
1484: {
1485: return PetscFree(lnk);
1486: }
1488: /* -------------------------------------------------------------------------------------------------------*/
1489: /*
1490: lnk[0] number of links
1491: lnk[1] number of entries
1492: lnk[3n] value
1493: lnk[3n+1] len
1494: lnk[3n+2] link to next value
1496: The next three are always the first link
1498: lnk[3] PETSC_MIN_INT+1
1499: lnk[4] 1
1500: lnk[5] link to first real entry
1502: The next three are always the last link
1504: lnk[6] PETSC_MAX_INT - 1
1505: lnk[7] 1
1506: lnk[8] next valid link (this is the same as lnk[0] but without the decreases)
1507: */
1509: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedCreate_fast(PetscInt nlnk_max,PetscInt **lnk)
1510: {
1512: PetscInt *llnk,lsize = 0;
1515: PetscIntMultError(3,nlnk_max+3,&lsize);
1516: PetscMalloc1(lsize,lnk);
1517: llnk = *lnk;
1518: llnk[0] = 0; /* nlnk: number of entries on the list */
1519: llnk[1] = 0; /* number of integer entries represented in list */
1520: llnk[3] = PETSC_MIN_INT+1; /* value in the first node */
1521: llnk[4] = 1; /* count for the first node */
1522: llnk[5] = 6; /* next for the first node */
1523: llnk[6] = PETSC_MAX_INT-1; /* value in the last node */
1524: llnk[7] = 1; /* count for the last node */
1525: llnk[8] = 0; /* next valid node to be used */
1526: return(0);
1527: }
1529: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedAddSorted_fast(PetscInt nidx,const PetscInt indices[],PetscInt lnk[])
1530: {
1531: PetscInt k,entry,prev,next;
1532: prev = 3; /* first value */
1533: next = lnk[prev+2];
1534: for (k=0; k<nidx; k++){
1535: entry = indices[k];
1536: /* search for insertion location */
1537: while (entry >= lnk[next]) {
1538: prev = next;
1539: next = lnk[next+2];
1540: }
1541: /* entry is in range of previous list */
1542: if (entry < lnk[prev]+lnk[prev+1]) continue;
1543: lnk[1]++;
1544: /* entry is right after previous list */
1545: if (entry == lnk[prev]+lnk[prev+1]) {
1546: lnk[prev+1]++;
1547: if (lnk[next] == entry+1) { /* combine two contiguous strings */
1548: lnk[prev+1] += lnk[next+1];
1549: lnk[prev+2] = lnk[next+2];
1550: next = lnk[next+2];
1551: lnk[0]--;
1552: }
1553: continue;
1554: }
1555: /* entry is right before next list */
1556: if (entry == lnk[next]-1) {
1557: lnk[next]--;
1558: lnk[next+1]++;
1559: prev = next;
1560: next = lnk[prev+2];
1561: continue;
1562: }
1563: /* add entry into lnk */
1564: lnk[prev+2] = 3*((lnk[8]++)+3); /* connect previous node to the new node */
1565: prev = lnk[prev+2];
1566: lnk[prev] = entry; /* set value of the new node */
1567: lnk[prev+1] = 1; /* number of values in contiguous string is one to start */
1568: lnk[prev+2] = next; /* connect new node to next node */
1569: lnk[0]++;
1570: }
1571: return 0;
1572: }
1574: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedClean_fast(PetscInt nidx,PetscInt *indices,PetscInt *lnk)
1575: {
1576: PetscInt _k,_next,_nlnk,cnt,j;
1577: _next = lnk[5]; /* first node */
1578: _nlnk = lnk[0];
1579: cnt = 0;
1580: for (_k=0; _k<_nlnk; _k++){
1581: for (j=0; j<lnk[_next+1]; j++) {
1582: indices[cnt++] = lnk[_next] + j;
1583: }
1584: _next = lnk[_next + 2];
1585: }
1586: lnk[0] = 0; /* nlnk: number of links */
1587: lnk[1] = 0; /* number of integer entries represented in list */
1588: lnk[3] = PETSC_MIN_INT+1; /* value in the first node */
1589: lnk[4] = 1; /* count for the first node */
1590: lnk[5] = 6; /* next for the first node */
1591: lnk[6] = PETSC_MAX_INT-1; /* value in the last node */
1592: lnk[7] = 1; /* count for the last node */
1593: lnk[8] = 0; /* next valid location to make link */
1594: return 0;
1595: }
1597: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedView_fast(PetscInt *lnk)
1598: {
1599: PetscInt k,next,nlnk;
1600: next = lnk[5]; /* first node */
1601: nlnk = lnk[0];
1602: for (k=0; k<nlnk; k++){
1603: #if 0 /* Debugging code */
1604: printf("%d value %d len %d next %d\n",next,lnk[next],lnk[next+1],lnk[next+2]);
1605: #endif
1606: next = lnk[next + 2];
1607: }
1608: return 0;
1609: }
1611: PETSC_STATIC_INLINE PetscErrorCode PetscLLCondensedDestroy_fast(PetscInt *lnk)
1612: {
1613: return PetscFree(lnk);
1614: }
1616: /* this is extern because it is used in MatFDColoringUseDM() which is in the DM library */
1617: PETSC_EXTERN PetscErrorCode MatFDColoringApply_AIJ(Mat,MatFDColoring,Vec,void*);
1619: PETSC_EXTERN PetscLogEvent MAT_Mult;
1620: PETSC_EXTERN PetscLogEvent MAT_MultMatrixFree;
1621: PETSC_EXTERN PetscLogEvent MAT_Mults;
1622: PETSC_EXTERN PetscLogEvent MAT_MultConstrained;
1623: PETSC_EXTERN PetscLogEvent MAT_MultAdd;
1624: PETSC_EXTERN PetscLogEvent MAT_MultTranspose;
1625: PETSC_EXTERN PetscLogEvent MAT_MultTransposeConstrained;
1626: PETSC_EXTERN PetscLogEvent MAT_MultTransposeAdd;
1627: PETSC_EXTERN PetscLogEvent MAT_Solve;
1628: PETSC_EXTERN PetscLogEvent MAT_Solves;
1629: PETSC_EXTERN PetscLogEvent MAT_SolveAdd;
1630: PETSC_EXTERN PetscLogEvent MAT_SolveTranspose;
1631: PETSC_EXTERN PetscLogEvent MAT_SolveTransposeAdd;
1632: PETSC_EXTERN PetscLogEvent MAT_SOR;
1633: PETSC_EXTERN PetscLogEvent MAT_ForwardSolve;
1634: PETSC_EXTERN PetscLogEvent MAT_BackwardSolve;
1635: PETSC_EXTERN PetscLogEvent MAT_LUFactor;
1636: PETSC_EXTERN PetscLogEvent MAT_LUFactorSymbolic;
1637: PETSC_EXTERN PetscLogEvent MAT_LUFactorNumeric;
1638: PETSC_EXTERN PetscLogEvent MAT_CholeskyFactor;
1639: PETSC_EXTERN PetscLogEvent MAT_CholeskyFactorSymbolic;
1640: PETSC_EXTERN PetscLogEvent MAT_CholeskyFactorNumeric;
1641: PETSC_EXTERN PetscLogEvent MAT_ILUFactor;
1642: PETSC_EXTERN PetscLogEvent MAT_ILUFactorSymbolic;
1643: PETSC_EXTERN PetscLogEvent MAT_ICCFactorSymbolic;
1644: PETSC_EXTERN PetscLogEvent MAT_Copy;
1645: PETSC_EXTERN PetscLogEvent MAT_Convert;
1646: PETSC_EXTERN PetscLogEvent MAT_Scale;
1647: PETSC_EXTERN PetscLogEvent MAT_AssemblyBegin;
1648: PETSC_EXTERN PetscLogEvent MAT_AssemblyEnd;
1649: PETSC_EXTERN PetscLogEvent MAT_SetValues;
1650: PETSC_EXTERN PetscLogEvent MAT_GetValues;
1651: PETSC_EXTERN PetscLogEvent MAT_GetRow;
1652: PETSC_EXTERN PetscLogEvent MAT_GetRowIJ;
1653: PETSC_EXTERN PetscLogEvent MAT_CreateSubMats;
1654: PETSC_EXTERN PetscLogEvent MAT_GetColoring;
1655: PETSC_EXTERN PetscLogEvent MAT_GetOrdering;
1656: PETSC_EXTERN PetscLogEvent MAT_RedundantMat;
1657: PETSC_EXTERN PetscLogEvent MAT_IncreaseOverlap;
1658: PETSC_EXTERN PetscLogEvent MAT_Partitioning;
1659: PETSC_EXTERN PetscLogEvent MAT_PartitioningND;
1660: PETSC_EXTERN PetscLogEvent MAT_Coarsen;
1661: PETSC_EXTERN PetscLogEvent MAT_ZeroEntries;
1662: PETSC_EXTERN PetscLogEvent MAT_Load;
1663: PETSC_EXTERN PetscLogEvent MAT_View;
1664: PETSC_EXTERN PetscLogEvent MAT_AXPY;
1665: PETSC_EXTERN PetscLogEvent MAT_FDColoringCreate;
1666: PETSC_EXTERN PetscLogEvent MAT_TransposeColoringCreate;
1667: PETSC_EXTERN PetscLogEvent MAT_FDColoringSetUp;
1668: PETSC_EXTERN PetscLogEvent MAT_FDColoringApply;
1669: PETSC_EXTERN PetscLogEvent MAT_Transpose;
1670: PETSC_EXTERN PetscLogEvent MAT_FDColoringFunction;
1671: PETSC_EXTERN PetscLogEvent MAT_CreateSubMat;
1672: PETSC_EXTERN PetscLogEvent MAT_MatMult;
1673: PETSC_EXTERN PetscLogEvent MAT_MatSolve;
1674: PETSC_EXTERN PetscLogEvent MAT_MatTrSolve;
1675: PETSC_EXTERN PetscLogEvent MAT_MatMultSymbolic;
1676: PETSC_EXTERN PetscLogEvent MAT_MatMultNumeric;
1677: PETSC_EXTERN PetscLogEvent MAT_Getlocalmatcondensed;
1678: PETSC_EXTERN PetscLogEvent MAT_GetBrowsOfAcols;
1679: PETSC_EXTERN PetscLogEvent MAT_GetBrowsOfAocols;
1680: PETSC_EXTERN PetscLogEvent MAT_PtAP;
1681: PETSC_EXTERN PetscLogEvent MAT_PtAPSymbolic;
1682: PETSC_EXTERN PetscLogEvent MAT_PtAPNumeric;
1683: PETSC_EXTERN PetscLogEvent MAT_Seqstompinum;
1684: PETSC_EXTERN PetscLogEvent MAT_Seqstompisym;
1685: PETSC_EXTERN PetscLogEvent MAT_Seqstompi;
1686: PETSC_EXTERN PetscLogEvent MAT_Getlocalmat;
1687: PETSC_EXTERN PetscLogEvent MAT_RARt;
1688: PETSC_EXTERN PetscLogEvent MAT_RARtSymbolic;
1689: PETSC_EXTERN PetscLogEvent MAT_RARtNumeric;
1690: PETSC_EXTERN PetscLogEvent MAT_MatTransposeMult;
1691: PETSC_EXTERN PetscLogEvent MAT_MatTransposeMultSymbolic;
1692: PETSC_EXTERN PetscLogEvent MAT_MatTransposeMultNumeric;
1693: PETSC_EXTERN PetscLogEvent MAT_TransposeMatMult;
1694: PETSC_EXTERN PetscLogEvent MAT_TransposeMatMultSymbolic;
1695: PETSC_EXTERN PetscLogEvent MAT_TransposeMatMultNumeric;
1696: PETSC_EXTERN PetscLogEvent MAT_MatMatMult;
1697: PETSC_EXTERN PetscLogEvent MAT_MatMatMultSymbolic;
1698: PETSC_EXTERN PetscLogEvent MAT_MatMatMultNumeric;
1699: PETSC_EXTERN PetscLogEvent MAT_Applypapt;
1700: PETSC_EXTERN PetscLogEvent MAT_Applypapt_symbolic;
1701: PETSC_EXTERN PetscLogEvent MAT_Applypapt_numeric;
1702: PETSC_EXTERN PetscLogEvent MAT_Getsymtranspose;
1703: PETSC_EXTERN PetscLogEvent MAT_Transpose_SeqAIJ;
1704: PETSC_EXTERN PetscLogEvent MAT_Getsymtransreduced;
1705: PETSC_EXTERN PetscLogEvent MAT_GetSequentialNonzeroStructure;
1706: PETSC_EXTERN PetscLogEvent MATMFFD_Mult;
1707: PETSC_EXTERN PetscLogEvent MAT_GetMultiProcBlock;
1708: PETSC_EXTERN PetscLogEvent MAT_CUSPARSECopyToGPU;
1709: PETSC_EXTERN PetscLogEvent MAT_SetValuesBatch;
1710: PETSC_EXTERN PetscLogEvent MAT_ViennaCLCopyToGPU;
1711: PETSC_EXTERN PetscLogEvent MAT_Merge;
1712: PETSC_EXTERN PetscLogEvent MAT_Residual;
1713: PETSC_EXTERN PetscLogEvent MAT_SetRandom;
1714: PETSC_EXTERN PetscLogEvent MATCOLORING_Apply;
1715: PETSC_EXTERN PetscLogEvent MATCOLORING_Comm;
1716: PETSC_EXTERN PetscLogEvent MATCOLORING_Local;
1717: PETSC_EXTERN PetscLogEvent MATCOLORING_ISCreate;
1718: PETSC_EXTERN PetscLogEvent MATCOLORING_SetUp;
1719: PETSC_EXTERN PetscLogEvent MATCOLORING_Weights;
1721: #endif