Actual source code: matimpl.h
1: #pragma once
3: #include <petscmat.h>
4: #include <petscmatcoarsen.h>
5: #include <petsc/private/petscimpl.h>
7: PETSC_EXTERN PetscBool MatRegisterAllCalled;
8: PETSC_EXTERN PetscBool MatSeqAIJRegisterAllCalled;
9: PETSC_EXTERN PetscBool MatOrderingRegisterAllCalled;
10: PETSC_EXTERN PetscBool MatColoringRegisterAllCalled;
11: PETSC_EXTERN PetscBool MatPartitioningRegisterAllCalled;
12: PETSC_EXTERN PetscBool MatCoarsenRegisterAllCalled;
13: PETSC_EXTERN PetscErrorCode MatRegisterAll(void);
14: PETSC_EXTERN PetscErrorCode MatOrderingRegisterAll(void);
15: PETSC_EXTERN PetscErrorCode MatColoringRegisterAll(void);
16: PETSC_EXTERN PetscErrorCode MatPartitioningRegisterAll(void);
17: PETSC_EXTERN PetscErrorCode MatCoarsenRegisterAll(void);
18: PETSC_EXTERN PetscErrorCode MatSeqAIJRegisterAll(void);
20: /* Gets the root type of the input matrix's type (e.g., MATAIJ for MATSEQAIJ) */
21: PETSC_EXTERN PetscErrorCode MatGetRootType_Private(Mat, MatType *);
23: /* Gets the MPI type corresponding to the input matrix's type (e.g., MATMPIAIJ for MATSEQAIJ) */
24: PETSC_EXTERN PetscErrorCode MatGetMPIMatType_Private(Mat, MatType *);
26: /*
27: This file defines the parts of the matrix data structure that are
28: shared by all matrix types.
29: */
31: /*
32: If you add entries here also add them to the MATOP enum
33: in include/petscmat.h and src/mat/f90-mod/petscmat.h
34: */
35: typedef struct _MatOps *MatOps;
36: struct _MatOps {
37: /* 0*/
38: PetscErrorCode (*setvalues)(Mat, PetscInt, const PetscInt[], PetscInt, const PetscInt[], const PetscScalar[], InsertMode);
39: PetscErrorCode (*getrow)(Mat, PetscInt, PetscInt *, PetscInt *[], PetscScalar *[]);
40: PetscErrorCode (*restorerow)(Mat, PetscInt, PetscInt *, PetscInt *[], PetscScalar *[]);
41: PetscErrorCode (*mult)(Mat, Vec, Vec);
42: PetscErrorCode (*multadd)(Mat, Vec, Vec, Vec);
43: /* 5*/
44: PetscErrorCode (*multtranspose)(Mat, Vec, Vec);
45: PetscErrorCode (*multtransposeadd)(Mat, Vec, Vec, Vec);
46: PetscErrorCode (*solve)(Mat, Vec, Vec);
47: PetscErrorCode (*solveadd)(Mat, Vec, Vec, Vec);
48: PetscErrorCode (*solvetranspose)(Mat, Vec, Vec);
49: /*10*/
50: PetscErrorCode (*solvetransposeadd)(Mat, Vec, Vec, Vec);
51: PetscErrorCode (*lufactor)(Mat, IS, IS, const MatFactorInfo *);
52: PetscErrorCode (*choleskyfactor)(Mat, IS, const MatFactorInfo *);
53: PetscErrorCode (*sor)(Mat, Vec, PetscReal, MatSORType, PetscReal, PetscInt, PetscInt, Vec);
54: PetscErrorCode (*transpose)(Mat, MatReuse, Mat *);
55: /*15*/
56: PetscErrorCode (*getinfo)(Mat, MatInfoType, MatInfo *);
57: PetscErrorCode (*equal)(Mat, Mat, PetscBool *);
58: PetscErrorCode (*getdiagonal)(Mat, Vec);
59: PetscErrorCode (*diagonalscale)(Mat, Vec, Vec);
60: PetscErrorCode (*norm)(Mat, NormType, PetscReal *);
61: /*20*/
62: PetscErrorCode (*assemblybegin)(Mat, MatAssemblyType);
63: PetscErrorCode (*assemblyend)(Mat, MatAssemblyType);
64: PetscErrorCode (*setoption)(Mat, MatOption, PetscBool);
65: PetscErrorCode (*zeroentries)(Mat);
66: /*24*/
67: PetscErrorCode (*zerorows)(Mat, PetscInt, const PetscInt[], PetscScalar, Vec, Vec);
68: PetscErrorCode (*lufactorsymbolic)(Mat, Mat, IS, IS, const MatFactorInfo *);
69: PetscErrorCode (*lufactornumeric)(Mat, Mat, const MatFactorInfo *);
70: PetscErrorCode (*choleskyfactorsymbolic)(Mat, Mat, IS, const MatFactorInfo *);
71: PetscErrorCode (*choleskyfactornumeric)(Mat, Mat, const MatFactorInfo *);
72: /*29*/
73: PetscErrorCode (*setup)(Mat);
74: PetscErrorCode (*ilufactorsymbolic)(Mat, Mat, IS, IS, const MatFactorInfo *);
75: PetscErrorCode (*iccfactorsymbolic)(Mat, Mat, IS, const MatFactorInfo *);
76: PetscErrorCode (*getdiagonalblock)(Mat, Mat *);
77: PetscErrorCode (*setinf)(Mat);
78: /*34*/
79: PetscErrorCode (*duplicate)(Mat, MatDuplicateOption, Mat *);
80: PetscErrorCode (*forwardsolve)(Mat, Vec, Vec);
81: PetscErrorCode (*backwardsolve)(Mat, Vec, Vec);
82: PetscErrorCode (*ilufactor)(Mat, IS, IS, const MatFactorInfo *);
83: PetscErrorCode (*iccfactor)(Mat, IS, const MatFactorInfo *);
84: /*39*/
85: PetscErrorCode (*axpy)(Mat, PetscScalar, Mat, MatStructure);
86: PetscErrorCode (*createsubmatrices)(Mat, PetscInt, const IS[], const IS[], MatReuse, Mat *[]);
87: PetscErrorCode (*increaseoverlap)(Mat, PetscInt, IS[], PetscInt);
88: PetscErrorCode (*getvalues)(Mat, PetscInt, const PetscInt[], PetscInt, const PetscInt[], PetscScalar[]);
89: PetscErrorCode (*copy)(Mat, Mat, MatStructure);
90: /*44*/
91: PetscErrorCode (*getrowmax)(Mat, Vec, PetscInt[]);
92: PetscErrorCode (*scale)(Mat, PetscScalar);
93: PetscErrorCode (*shift)(Mat, PetscScalar);
94: PetscErrorCode (*diagonalset)(Mat, Vec, InsertMode);
95: PetscErrorCode (*zerorowscolumns)(Mat, PetscInt, const PetscInt[], PetscScalar, Vec, Vec);
96: /*49*/
97: PetscErrorCode (*setrandom)(Mat, PetscRandom);
98: PetscErrorCode (*getrowij)(Mat, PetscInt, PetscBool, PetscBool, PetscInt *, const PetscInt *[], const PetscInt *[], PetscBool *);
99: PetscErrorCode (*restorerowij)(Mat, PetscInt, PetscBool, PetscBool, PetscInt *, const PetscInt *[], const PetscInt *[], PetscBool *);
100: PetscErrorCode (*getcolumnij)(Mat, PetscInt, PetscBool, PetscBool, PetscInt *, const PetscInt *[], const PetscInt *[], PetscBool *);
101: PetscErrorCode (*restorecolumnij)(Mat, PetscInt, PetscBool, PetscBool, PetscInt *, const PetscInt *[], const PetscInt *[], PetscBool *);
102: /*54*/
103: PetscErrorCode (*fdcoloringcreate)(Mat, ISColoring, MatFDColoring);
104: PetscErrorCode (*coloringpatch)(Mat, PetscInt, PetscInt, ISColoringValue[], ISColoring *);
105: PetscErrorCode (*setunfactored)(Mat);
106: PetscErrorCode (*permute)(Mat, IS, IS, Mat *);
107: PetscErrorCode (*setvaluesblocked)(Mat, PetscInt, const PetscInt[], PetscInt, const PetscInt[], const PetscScalar[], InsertMode);
108: /*59*/
109: PetscErrorCode (*createsubmatrix)(Mat, IS, IS, MatReuse, Mat *);
110: PetscErrorCode (*destroy)(Mat);
111: PetscErrorCode (*view)(Mat, PetscViewer);
112: PetscErrorCode (*convertfrom)(Mat, MatType, MatReuse, Mat *);
113: PetscErrorCode (*placeholder_63)(void);
114: /*64*/
115: PetscErrorCode (*matmatmultsymbolic)(Mat, Mat, Mat, PetscReal, Mat);
116: PetscErrorCode (*matmatmultnumeric)(Mat, Mat, Mat, Mat);
117: PetscErrorCode (*setlocaltoglobalmapping)(Mat, ISLocalToGlobalMapping, ISLocalToGlobalMapping);
118: PetscErrorCode (*setvalueslocal)(Mat, PetscInt, const PetscInt[], PetscInt, const PetscInt[], const PetscScalar[], InsertMode);
119: PetscErrorCode (*zerorowslocal)(Mat, PetscInt, const PetscInt[], PetscScalar, Vec, Vec);
120: /*69*/
121: PetscErrorCode (*getrowmaxabs)(Mat, Vec, PetscInt[]);
122: PetscErrorCode (*getrowminabs)(Mat, Vec, PetscInt[]);
123: PetscErrorCode (*convert)(Mat, MatType, MatReuse, Mat *);
124: PetscErrorCode (*hasoperation)(Mat, MatOperation, PetscBool *);
125: PetscErrorCode (*placeholder_73)(void);
126: /*74*/
127: PetscErrorCode (*setvaluesadifor)(Mat, PetscInt, void *);
128: PetscErrorCode (*fdcoloringapply)(Mat, MatFDColoring, Vec, void *);
129: PetscErrorCode (*setfromoptions)(Mat, PetscOptionItems *);
130: PetscErrorCode (*placeholder_77)(void);
131: PetscErrorCode (*placeholder_78)(void);
132: /*79*/
133: PetscErrorCode (*findzerodiagonals)(Mat, IS *);
134: PetscErrorCode (*mults)(Mat, Vecs, Vecs);
135: PetscErrorCode (*solves)(Mat, Vecs, Vecs);
136: PetscErrorCode (*getinertia)(Mat, PetscInt *, PetscInt *, PetscInt *);
137: PetscErrorCode (*load)(Mat, PetscViewer);
138: /*84*/
139: PetscErrorCode (*issymmetric)(Mat, PetscReal, PetscBool *);
140: PetscErrorCode (*ishermitian)(Mat, PetscReal, PetscBool *);
141: PetscErrorCode (*isstructurallysymmetric)(Mat, PetscBool *);
142: PetscErrorCode (*setvaluesblockedlocal)(Mat, PetscInt, const PetscInt[], PetscInt, const PetscInt[], const PetscScalar[], InsertMode);
143: PetscErrorCode (*getvecs)(Mat, Vec *, Vec *);
144: /*89*/
145: PetscErrorCode (*placeholder_89)(void);
146: PetscErrorCode (*matmultsymbolic)(Mat, Mat, PetscReal, Mat);
147: PetscErrorCode (*matmultnumeric)(Mat, Mat, Mat);
148: PetscErrorCode (*placeholder_92)(void);
149: PetscErrorCode (*ptapsymbolic)(Mat, Mat, PetscReal, Mat); /* double dispatch wrapper routine */
150: /*94*/
151: PetscErrorCode (*ptapnumeric)(Mat, Mat, Mat); /* double dispatch wrapper routine */
152: PetscErrorCode (*placeholder_95)(void);
153: PetscErrorCode (*mattransposemultsymbolic)(Mat, Mat, PetscReal, Mat);
154: PetscErrorCode (*mattransposemultnumeric)(Mat, Mat, Mat);
155: PetscErrorCode (*bindtocpu)(Mat, PetscBool);
156: /*99*/
157: PetscErrorCode (*productsetfromoptions)(Mat);
158: PetscErrorCode (*productsymbolic)(Mat);
159: PetscErrorCode (*productnumeric)(Mat);
160: PetscErrorCode (*conjugate)(Mat); /* complex conjugate */
161: PetscErrorCode (*viewnative)(Mat, PetscViewer);
162: /*104*/
163: PetscErrorCode (*setvaluesrow)(Mat, PetscInt, const PetscScalar[]);
164: PetscErrorCode (*realpart)(Mat);
165: PetscErrorCode (*imaginarypart)(Mat);
166: PetscErrorCode (*getrowuppertriangular)(Mat);
167: PetscErrorCode (*restorerowuppertriangular)(Mat);
168: /*109*/
169: PetscErrorCode (*matsolve)(Mat, Mat, Mat);
170: PetscErrorCode (*matsolvetranspose)(Mat, Mat, Mat);
171: PetscErrorCode (*getrowmin)(Mat, Vec, PetscInt[]);
172: PetscErrorCode (*getcolumnvector)(Mat, Vec, PetscInt);
173: PetscErrorCode (*missingdiagonal)(Mat, PetscBool *, PetscInt *);
174: /*114*/
175: PetscErrorCode (*getseqnonzerostructure)(Mat, Mat *);
176: PetscErrorCode (*create)(Mat);
177: PetscErrorCode (*getghosts)(Mat, PetscInt *, const PetscInt *[]);
178: PetscErrorCode (*getlocalsubmatrix)(Mat, IS, IS, Mat *);
179: PetscErrorCode (*restorelocalsubmatrix)(Mat, IS, IS, Mat *);
180: /*119*/
181: PetscErrorCode (*multdiagonalblock)(Mat, Vec, Vec);
182: PetscErrorCode (*hermitiantranspose)(Mat, MatReuse, Mat *);
183: PetscErrorCode (*multhermitiantranspose)(Mat, Vec, Vec);
184: PetscErrorCode (*multhermitiantransposeadd)(Mat, Vec, Vec, Vec);
185: PetscErrorCode (*getmultiprocblock)(Mat, MPI_Comm, MatReuse, Mat *);
186: /*124*/
187: PetscErrorCode (*findnonzerorows)(Mat, IS *);
188: PetscErrorCode (*getcolumnreductions)(Mat, PetscInt, PetscReal *);
189: PetscErrorCode (*invertblockdiagonal)(Mat, const PetscScalar **);
190: PetscErrorCode (*invertvariableblockdiagonal)(Mat, PetscInt, const PetscInt *, PetscScalar *);
191: PetscErrorCode (*createsubmatricesmpi)(Mat, PetscInt, const IS[], const IS[], MatReuse, Mat **);
192: /*129*/
193: PetscErrorCode (*setvaluesbatch)(Mat, PetscInt, PetscInt, PetscInt *, const PetscScalar *);
194: PetscErrorCode (*placeholder_130)(void);
195: PetscErrorCode (*transposematmultsymbolic)(Mat, Mat, PetscReal, Mat);
196: PetscErrorCode (*transposematmultnumeric)(Mat, Mat, Mat);
197: PetscErrorCode (*transposecoloringcreate)(Mat, ISColoring, MatTransposeColoring);
198: /*134*/
199: PetscErrorCode (*transcoloringapplysptoden)(MatTransposeColoring, Mat, Mat);
200: PetscErrorCode (*transcoloringapplydentosp)(MatTransposeColoring, Mat, Mat);
201: PetscErrorCode (*placeholder_136)(void);
202: PetscErrorCode (*rartsymbolic)(Mat, Mat, PetscReal, Mat); /* double dispatch wrapper routine */
203: PetscErrorCode (*rartnumeric)(Mat, Mat, Mat); /* double dispatch wrapper routine */
204: /*139*/
205: PetscErrorCode (*setblocksizes)(Mat, PetscInt, PetscInt);
206: PetscErrorCode (*aypx)(Mat, PetscScalar, Mat, MatStructure);
207: PetscErrorCode (*residual)(Mat, Vec, Vec, Vec);
208: PetscErrorCode (*fdcoloringsetup)(Mat, ISColoring, MatFDColoring);
209: PetscErrorCode (*findoffblockdiagonalentries)(Mat, IS *);
210: PetscErrorCode (*creatempimatconcatenateseqmat)(MPI_Comm, Mat, PetscInt, MatReuse, Mat *);
211: /*145*/
212: PetscErrorCode (*destroysubmatrices)(PetscInt, Mat *[]);
213: PetscErrorCode (*mattransposesolve)(Mat, Mat, Mat);
214: PetscErrorCode (*getvalueslocal)(Mat, PetscInt, const PetscInt[], PetscInt, const PetscInt[], PetscScalar[]);
215: PetscErrorCode (*creategraph)(Mat, PetscBool, PetscBool, PetscReal, PetscInt, PetscInt[], Mat *);
216: PetscErrorCode (*dummy)(Mat);
217: /*150*/
218: PetscErrorCode (*transposesymbolic)(Mat, Mat *);
219: PetscErrorCode (*eliminatezeros)(Mat, PetscBool);
220: PetscErrorCode (*getrowsumabs)(Mat, Vec);
221: PetscErrorCode (*getfactor)(Mat, MatSolverType, MatFactorType, Mat *);
222: PetscErrorCode (*getblockdiagonal)(Mat, Mat *); // NOTE: the caller of get{block, vblock}diagonal owns the returned matrix;
223: PetscErrorCode (*getvblockdiagonal)(Mat, Mat *); // they must destroy it after use
224: };
225: /*
226: If you add MatOps entries above also add them to the MATOP enum
227: in include/petscmat.h and src/mat/f90-mod/petscmat.h
228: */
230: #include <petscsys.h>
232: typedef struct _p_MatRootName *MatRootName;
233: struct _p_MatRootName {
234: char *rname, *sname, *mname;
235: MatRootName next;
236: };
238: PETSC_EXTERN MatRootName MatRootNameList;
240: /*
241: Utility private matrix routines used outside Mat
242: */
243: PETSC_SINGLE_LIBRARY_INTERN PetscErrorCode MatFindNonzeroRowsOrCols_Basic(Mat, PetscBool, PetscReal, IS *);
244: PETSC_EXTERN PetscErrorCode MatShellGetScalingShifts(Mat, PetscScalar *, PetscScalar *, Vec *, Vec *, Vec *, Mat *, IS *, IS *);
246: #define MAT_SHELL_NOT_ALLOWED (void *)-1
248: /*
249: Utility private matrix routines
250: */
251: PETSC_INTERN PetscErrorCode MatConvert_Basic(Mat, MatType, MatReuse, Mat *);
252: PETSC_INTERN PetscErrorCode MatConvert_Shell(Mat, MatType, MatReuse, Mat *);
253: PETSC_INTERN PetscErrorCode MatConvertFrom_Shell(Mat, MatType, MatReuse, Mat *);
254: PETSC_INTERN PetscErrorCode MatShellSetContext_Immutable(Mat X, void *ctx);
255: PETSC_INTERN PetscErrorCode MatShellSetContextDestroy_Immutable(Mat X, PetscErrorCode (*f)(void *));
256: PETSC_INTERN PetscErrorCode MatShellSetManageScalingShifts_Immutable(Mat X);
257: PETSC_INTERN PetscErrorCode MatCopy_Basic(Mat, Mat, MatStructure);
258: PETSC_INTERN PetscErrorCode MatDiagonalSet_Default(Mat, Vec, InsertMode);
259: #if defined(PETSC_HAVE_SCALAPACK)
260: PETSC_INTERN PetscErrorCode MatConvert_Dense_ScaLAPACK(Mat, MatType, MatReuse, Mat *);
261: #endif
262: PETSC_INTERN PetscErrorCode MatSetPreallocationCOO_Basic(Mat, PetscCount, PetscInt[], PetscInt[]);
263: PETSC_INTERN PetscErrorCode MatSetValuesCOO_Basic(Mat, const PetscScalar[], InsertMode);
265: /* This can be moved to the public header after implementing some missing MatProducts */
266: PETSC_INTERN PetscErrorCode MatCreateFromISLocalToGlobalMapping(ISLocalToGlobalMapping, Mat, PetscBool, PetscBool, MatType, Mat *);
268: /* these callbacks rely on the old matrix function pointers for
269: matmat operations. They are unsafe, and should be removed.
270: However, the amount of work needed to clean up all the
271: implementations is not negligible */
272: PETSC_INTERN PetscErrorCode MatProductSymbolic_AB(Mat);
273: PETSC_INTERN PetscErrorCode MatProductNumeric_AB(Mat);
274: PETSC_INTERN PetscErrorCode MatProductSymbolic_AtB(Mat);
275: PETSC_INTERN PetscErrorCode MatProductNumeric_AtB(Mat);
276: PETSC_INTERN PetscErrorCode MatProductSymbolic_ABt(Mat);
277: PETSC_INTERN PetscErrorCode MatProductNumeric_ABt(Mat);
278: PETSC_INTERN PetscErrorCode MatProductNumeric_PtAP(Mat);
279: PETSC_INTERN PetscErrorCode MatProductNumeric_RARt(Mat);
280: PETSC_INTERN PetscErrorCode MatProductSymbolic_ABC(Mat);
281: PETSC_INTERN PetscErrorCode MatProductNumeric_ABC(Mat);
283: PETSC_INTERN PetscErrorCode MatProductCreate_Private(Mat, Mat, Mat, Mat);
284: /* this callback handles all the different triple products and
285: does not rely on the function pointers; used by cuSPARSE/hipSPARSE and KOKKOS-KERNELS */
286: PETSC_INTERN PetscErrorCode MatProductSymbolic_ABC_Basic(Mat);
288: /* CreateGraph is common to AIJ seq and mpi */
289: PETSC_INTERN PetscErrorCode MatCreateGraph_Simple_AIJ(Mat, PetscBool, PetscBool, PetscReal, PetscInt, PetscInt[], Mat *);
291: #if defined(PETSC_CLANG_STATIC_ANALYZER)
292: template <typename Tm>
293: extern void MatCheckPreallocated(Tm, int);
294: template <typename Tm>
295: extern void MatCheckProduct(Tm, int);
296: #else /* PETSC_CLANG_STATIC_ANALYZER */
297: #define MatCheckPreallocated(A, arg) \
298: do { \
299: if (!(A)->preallocated) PetscCall(MatSetUp(A)); \
300: } while (0)
302: #if defined(PETSC_USE_DEBUG)
303: #define MatCheckProduct(A, arg) \
304: do { \
305: PetscCheck((A)->product, PETSC_COMM_SELF, PETSC_ERR_ARG_WRONGSTATE, "Argument %d \"%s\" is not a matrix obtained from MatProductCreate()", (arg), #A); \
306: } while (0)
307: #else
308: #define MatCheckProduct(A, arg) \
309: do { \
310: } while (0)
311: #endif
312: #endif /* PETSC_CLANG_STATIC_ANALYZER */
314: /*
315: The stash is used to temporarily store inserted matrix values that
316: belong to another processor. During the assembly phase the stashed
317: values are moved to the correct processor and
318: */
320: typedef struct _MatStashSpace *PetscMatStashSpace;
322: struct _MatStashSpace {
323: PetscMatStashSpace next;
324: PetscScalar *space_head, *val;
325: PetscInt *idx, *idy;
326: PetscInt total_space_size;
327: PetscInt local_used;
328: PetscInt local_remaining;
329: };
331: PETSC_EXTERN PetscErrorCode PetscMatStashSpaceGet(PetscInt, PetscInt, PetscMatStashSpace *);
332: PETSC_EXTERN PetscErrorCode PetscMatStashSpaceContiguous(PetscInt, PetscMatStashSpace *, PetscScalar *, PetscInt *, PetscInt *);
333: PETSC_EXTERN PetscErrorCode PetscMatStashSpaceDestroy(PetscMatStashSpace *);
335: typedef struct {
336: PetscInt count;
337: } MatStashHeader;
339: typedef struct {
340: void *buffer; /* Of type blocktype, dynamically constructed */
341: PetscInt count;
342: char pending;
343: } MatStashFrame;
345: typedef struct _MatStash MatStash;
346: struct _MatStash {
347: PetscInt nmax; /* maximum stash size */
348: PetscInt umax; /* user specified max-size */
349: PetscInt oldnmax; /* the nmax value used previously */
350: PetscInt n; /* stash size */
351: PetscInt bs; /* block size of the stash */
352: PetscInt reallocs; /* preserve the no of mallocs invoked */
353: PetscMatStashSpace space_head, space; /* linked list to hold stashed global row/column numbers and matrix values */
355: PetscErrorCode (*ScatterBegin)(Mat, MatStash *, PetscInt *);
356: PetscErrorCode (*ScatterGetMesg)(MatStash *, PetscMPIInt *, PetscInt **, PetscInt **, PetscScalar **, PetscInt *);
357: PetscErrorCode (*ScatterEnd)(MatStash *);
358: PetscErrorCode (*ScatterDestroy)(MatStash *);
360: /* The following variables are used for communication */
361: MPI_Comm comm;
362: PetscMPIInt size, rank;
363: PetscMPIInt tag1, tag2;
364: MPI_Request *send_waits; /* array of send requests */
365: MPI_Request *recv_waits; /* array of receive requests */
366: MPI_Status *send_status; /* array of send status */
367: PetscMPIInt nsends, nrecvs; /* numbers of sends and receives */
368: PetscScalar *svalues; /* sending data */
369: PetscInt *sindices;
370: PetscScalar **rvalues; /* receiving data (values) */
371: PetscInt **rindices; /* receiving data (indices) */
372: PetscMPIInt nprocessed; /* number of messages already processed */
373: PetscMPIInt *flg_v; /* indicates what messages have arrived so far and from whom */
374: PetscBool reproduce;
375: PetscMPIInt reproduce_count;
377: /* The following variables are used for BTS communication */
378: PetscBool first_assembly_done; /* Is the first time matrix assembly done? */
379: PetscBool use_status; /* Use MPI_Status to determine number of items in each message */
380: PetscMPIInt nsendranks;
381: PetscMPIInt nrecvranks;
382: PetscMPIInt *sendranks;
383: PetscMPIInt *recvranks;
384: MatStashHeader *sendhdr, *recvhdr;
385: MatStashFrame *sendframes; /* pointers to the main messages */
386: MatStashFrame *recvframes;
387: MatStashFrame *recvframe_active;
388: PetscInt recvframe_i; /* index of block within active frame */
389: PetscInt recvframe_count; /* Count actually sent for current frame */
390: PetscMPIInt recvcount; /* Number of receives processed so far */
391: PetscMPIInt *some_indices; /* From last call to MPI_Waitsome */
392: MPI_Status *some_statuses; /* Statuses from last call to MPI_Waitsome */
393: PetscMPIInt some_count; /* Number of requests completed in last call to MPI_Waitsome */
394: PetscMPIInt some_i; /* Index of request currently being processed */
395: MPI_Request *sendreqs;
396: MPI_Request *recvreqs;
397: PetscSegBuffer segsendblocks;
398: PetscSegBuffer segrecvframe;
399: PetscSegBuffer segrecvblocks;
400: MPI_Datatype blocktype;
401: size_t blocktype_size;
402: InsertMode *insertmode; /* Pointer to check mat->insertmode and set upon message arrival in case no local values have been set. */
403: };
405: #if !defined(PETSC_HAVE_MPIUNI)
406: PETSC_INTERN PetscErrorCode MatStashScatterDestroy_BTS(MatStash *);
407: #endif
408: PETSC_INTERN PetscErrorCode MatStashCreate_Private(MPI_Comm, PetscInt, MatStash *);
409: PETSC_INTERN PetscErrorCode MatStashDestroy_Private(MatStash *);
410: PETSC_INTERN PetscErrorCode MatStashScatterEnd_Private(MatStash *);
411: PETSC_INTERN PetscErrorCode MatStashSetInitialSize_Private(MatStash *, PetscInt);
412: PETSC_INTERN PetscErrorCode MatStashGetInfo_Private(MatStash *, PetscInt *, PetscInt *);
413: PETSC_INTERN PetscErrorCode MatStashValuesRow_Private(MatStash *, PetscInt, PetscInt, const PetscInt[], const PetscScalar[], PetscBool);
414: PETSC_INTERN PetscErrorCode MatStashValuesCol_Private(MatStash *, PetscInt, PetscInt, const PetscInt[], const PetscScalar[], PetscInt, PetscBool);
415: PETSC_INTERN PetscErrorCode MatStashValuesRowBlocked_Private(MatStash *, PetscInt, PetscInt, const PetscInt[], const PetscScalar[], PetscInt, PetscInt, PetscInt);
416: PETSC_INTERN PetscErrorCode MatStashValuesColBlocked_Private(MatStash *, PetscInt, PetscInt, const PetscInt[], const PetscScalar[], PetscInt, PetscInt, PetscInt);
417: PETSC_INTERN PetscErrorCode MatStashScatterBegin_Private(Mat, MatStash *, PetscInt *);
418: PETSC_INTERN PetscErrorCode MatStashScatterGetMesg_Private(MatStash *, PetscMPIInt *, PetscInt **, PetscInt **, PetscScalar **, PetscInt *);
419: PETSC_INTERN PetscErrorCode MatGetInfo_External(Mat, MatInfoType, MatInfo *);
421: typedef struct {
422: PetscInt dim;
423: PetscInt dims[4];
424: PetscInt starts[4];
425: PetscBool noc; /* this is a single component problem, hence user will not set MatStencil.c */
426: } MatStencilInfo;
428: /* Info about using compressed row format */
429: typedef struct {
430: PetscBool use; /* indicates compressed rows have been checked and will be used */
431: PetscInt nrows; /* number of non-zero rows */
432: PetscInt *i; /* compressed row pointer */
433: PetscInt *rindex; /* compressed row index */
434: } Mat_CompressedRow;
435: PETSC_EXTERN PetscErrorCode MatCheckCompressedRow(Mat, PetscInt, Mat_CompressedRow *, PetscInt *, PetscInt, PetscReal);
437: typedef struct { /* used by MatCreateRedundantMatrix() for reusing matredundant */
438: PetscInt nzlocal, nsends, nrecvs;
439: PetscMPIInt *send_rank, *recv_rank;
440: PetscInt *sbuf_nz, *rbuf_nz, *sbuf_j, **rbuf_j;
441: PetscScalar *sbuf_a, **rbuf_a;
442: MPI_Comm subcomm; /* when user does not provide a subcomm */
443: IS isrow, iscol;
444: Mat *matseq;
445: } Mat_Redundant;
447: typedef struct { /* used by MatProduct() */
448: MatProductType type;
449: char *alg;
450: Mat A, B, C, Dwork;
451: PetscBool symbolic_used_the_fact_A_is_symmetric; /* Symbolic phase took advantage of the fact that A is symmetric, and optimized e.g. AtB as AB. Then, .. */
452: PetscBool symbolic_used_the_fact_B_is_symmetric; /* .. in the numeric phase, if a new A is not symmetric (but has the same sparsity as the old A therefore .. */
453: PetscBool symbolic_used_the_fact_C_is_symmetric; /* MatMatMult(A,B,MAT_REUSE_MATRIX,..&C) is still legitimate), we need to redo symbolic! */
454: PetscObjectParameterDeclare(PetscReal, fill);
455: PetscBool api_user; /* used to distinguish command line options and to indicate the matrix values are ready to be consumed at symbolic phase if needed */
456: PetscBool setfromoptionscalled;
458: /* Some products may display the information on the algorithm used */
459: PetscErrorCode (*view)(Mat, PetscViewer);
461: /* many products have intermediate data structures, each specific to Mat types and product type */
462: PetscBool clear; /* whether or not to clear the data structures after MatProductNumeric has been called */
463: void *data; /* where to stash those structures */
464: PetscErrorCode (*destroy)(void *); /* destroy routine */
465: } Mat_Product;
467: struct _p_Mat {
468: PETSCHEADER(struct _MatOps);
469: PetscLayout rmap, cmap;
470: void *data; /* implementation-specific data */
471: MatFactorType factortype; /* MAT_FACTOR_LU, ILU, CHOLESKY or ICC */
472: PetscBool trivialsymbolic; /* indicates the symbolic factorization doesn't actually do a symbolic factorization, it is delayed to the numeric factorization */
473: PetscBool canuseordering; /* factorization can use ordering provide to routine (most PETSc implementations) */
474: MatOrderingType preferredordering[MAT_FACTOR_NUM_TYPES]; /* what is the preferred (or default) ordering for the matrix solver type */
475: PetscBool assembled; /* is the matrix assembled? */
476: PetscBool was_assembled; /* new values inserted into assembled mat */
477: PetscInt num_ass; /* number of times matrix has been assembled */
478: PetscObjectState nonzerostate; /* each time new nonzeros locations are introduced into the matrix this is updated */
479: PetscObjectState ass_nonzerostate; /* nonzero state at last assembly */
480: MatInfo info; /* matrix information */
481: InsertMode insertmode; /* have values been inserted in matrix or added? */
482: MatStash stash, bstash; /* used for assembling off-proc mat emements */
483: MatNullSpace nullsp; /* null space (operator is singular) */
484: MatNullSpace transnullsp; /* null space of transpose of operator */
485: MatNullSpace nearnullsp; /* near null space to be used by multigrid methods */
486: PetscInt congruentlayouts; /* are the rows and columns layouts congruent? */
487: PetscBool preallocated;
488: MatStencilInfo stencil; /* information for structured grid */
489: PetscBool3 symmetric, hermitian, structurally_symmetric, spd;
490: PetscBool symmetry_eternal, structural_symmetry_eternal, spd_eternal;
491: PetscBool nooffprocentries, nooffproczerorows;
492: PetscBool assembly_subset; /* set by MAT_SUBSET_OFF_PROC_ENTRIES */
493: PetscBool submat_singleis; /* for efficient PCSetUp_ASM() */
494: PetscBool structure_only;
495: PetscBool sortedfull; /* full, sorted rows are inserted */
496: PetscBool force_diagonals; /* set by MAT_FORCE_DIAGONAL_ENTRIES */
497: #if defined(PETSC_HAVE_DEVICE)
498: PetscOffloadMask offloadmask; /* a mask which indicates where the valid matrix data is (GPU, CPU or both) */
499: PetscBool boundtocpu;
500: PetscBool bindingpropagates;
501: #endif
502: char *defaultrandtype;
503: void *spptr; /* pointer for special library like SuperLU */
504: char *solvertype;
505: PetscBool checksymmetryonassembly, checknullspaceonassembly;
506: PetscReal checksymmetrytol;
507: Mat schur; /* Schur complement matrix */
508: MatFactorSchurStatus schur_status; /* status of the Schur complement matrix */
509: Mat_Redundant *redundant; /* used by MatCreateRedundantMatrix() */
510: PetscBool erroriffailure; /* Generate an error if detected (for example a zero pivot) instead of returning */
511: MatFactorError factorerrortype; /* type of error in factorization */
512: PetscReal factorerror_zeropivot_value; /* If numerical zero pivot was detected this is the computed value */
513: PetscInt factorerror_zeropivot_row; /* Row where zero pivot was detected */
514: PetscInt nblocks, *bsizes; /* support for MatSetVariableBlockSizes() */
515: PetscInt p_cstart, p_rank, p_cend, n_rank; /* Information from parallel MatComputeVariableBlockEnvelope() */
516: PetscBool p_parallel;
517: char *defaultvectype;
518: Mat_Product *product;
519: PetscBool form_explicit_transpose; /* hint to generate an explicit mat tranpsose for operations like MatMultTranspose() */
520: PetscBool transupdated; /* whether or not the explicitly generated transpose is up-to-date */
521: char *factorprefix; /* the prefix to use with factored matrix that is created */
522: PetscBool hash_active; /* indicates MatSetValues() is being handled by hashing */
523: };
525: PETSC_INTERN PetscErrorCode MatAXPY_Basic(Mat, PetscScalar, Mat, MatStructure);
526: PETSC_INTERN PetscErrorCode MatAXPY_BasicWithPreallocation(Mat, Mat, PetscScalar, Mat, MatStructure);
527: PETSC_INTERN PetscErrorCode MatAXPY_Basic_Preallocate(Mat, Mat, Mat *);
528: PETSC_INTERN PetscErrorCode MatAXPY_Dense_Nest(Mat, PetscScalar, Mat);
530: PETSC_INTERN PetscErrorCode MatSetUp_Default(Mat);
532: /*
533: Utility for MatZeroRows
534: */
535: PETSC_INTERN PetscErrorCode MatZeroRowsMapLocal_Private(Mat, PetscInt, const PetscInt *, PetscInt *, PetscInt **);
537: /*
538: Utility for MatView/MatLoad
539: */
540: PETSC_INTERN PetscErrorCode MatView_Binary_BlockSizes(Mat, PetscViewer);
541: PETSC_INTERN PetscErrorCode MatLoad_Binary_BlockSizes(Mat, PetscViewer);
543: /*
544: Object for partitioning graphs
545: */
547: typedef struct _MatPartitioningOps *MatPartitioningOps;
548: struct _MatPartitioningOps {
549: PetscErrorCode (*apply)(MatPartitioning, IS *);
550: PetscErrorCode (*applynd)(MatPartitioning, IS *);
551: PetscErrorCode (*setfromoptions)(MatPartitioning, PetscOptionItems *);
552: PetscErrorCode (*destroy)(MatPartitioning);
553: PetscErrorCode (*view)(MatPartitioning, PetscViewer);
554: PetscErrorCode (*improve)(MatPartitioning, IS *);
555: };
557: struct _p_MatPartitioning {
558: PETSCHEADER(struct _MatPartitioningOps);
559: Mat adj;
560: PetscInt *vertex_weights;
561: PetscReal *part_weights;
562: PetscInt n; /* number of partitions */
563: PetscInt ncon; /* number of vertex weights per vertex */
564: void *data;
565: PetscInt setupcalled;
566: PetscBool use_edge_weights; /* A flag indicates whether or not to use edge weights */
567: };
569: /* needed for parallel nested dissection by ParMetis and PTSCOTCH */
570: PETSC_INTERN PetscErrorCode MatPartitioningSizesToSep_Private(PetscInt, PetscInt[], PetscInt[], PetscInt[]);
572: /*
573: Object for coarsen graphs
574: */
575: typedef struct _MatCoarsenOps *MatCoarsenOps;
576: struct _MatCoarsenOps {
577: PetscErrorCode (*apply)(MatCoarsen);
578: PetscErrorCode (*setfromoptions)(MatCoarsen, PetscOptionItems *);
579: PetscErrorCode (*destroy)(MatCoarsen);
580: PetscErrorCode (*view)(MatCoarsen, PetscViewer);
581: };
583: #define MAT_COARSEN_STRENGTH_INDEX_SIZE 3
584: struct _p_MatCoarsen {
585: PETSCHEADER(struct _MatCoarsenOps);
586: Mat graph;
587: void *subctx;
588: /* */
589: PetscBool strict_aggs;
590: IS perm;
591: PetscCoarsenData *agg_lists;
592: PetscInt max_it; /* number of iterations in HEM */
593: PetscReal threshold; /* HEM can filter interim graphs */
594: PetscInt strength_index_size;
595: PetscInt strength_index[MAT_COARSEN_STRENGTH_INDEX_SIZE];
596: };
598: PETSC_EXTERN PetscErrorCode MatCoarsenMISKSetDistance(MatCoarsen, PetscInt);
599: PETSC_EXTERN PetscErrorCode MatCoarsenMISKGetDistance(MatCoarsen, PetscInt *);
601: /*
602: Used in aijdevice.h
603: */
604: typedef struct {
605: PetscInt *i;
606: PetscInt *j;
607: PetscScalar *a;
608: PetscInt n;
609: PetscInt ignorezeroentries;
610: } PetscCSRDataStructure;
612: /*
613: MatFDColoring is used to compute Jacobian matrices efficiently
614: via coloring. The data structure is explained below in an example.
616: Color = 0 1 0 2 | 2 3 0
617: ---------------------------------------------------
618: 00 01 | 05
619: 10 11 | 14 15 Processor 0
620: 22 23 | 25
621: 32 33 |
622: ===================================================
623: | 44 45 46
624: 50 | 55 Processor 1
625: | 64 66
626: ---------------------------------------------------
628: ncolors = 4;
630: ncolumns = {2,1,1,0}
631: columns = {{0,2},{1},{3},{}}
632: nrows = {4,2,3,3}
633: rows = {{0,1,2,3},{0,1},{1,2,3},{0,1,2}}
634: vwscale = {dx(0),dx(1),dx(2),dx(3)} MPI Vec
635: vscale = {dx(0),dx(1),dx(2),dx(3),dx(4),dx(5)} Seq Vec
637: ncolumns = {1,0,1,1}
638: columns = {{6},{},{4},{5}}
639: nrows = {3,0,2,2}
640: rows = {{0,1,2},{},{1,2},{1,2}}
641: vwscale = {dx(4),dx(5),dx(6)} MPI Vec
642: vscale = {dx(0),dx(4),dx(5),dx(6)} Seq Vec
644: See the routine MatFDColoringApply() for how this data is used
645: to compute the Jacobian.
647: */
648: typedef struct {
649: PetscInt row;
650: PetscInt col;
651: PetscScalar *valaddr; /* address of value */
652: } MatEntry;
654: typedef struct {
655: PetscInt row;
656: PetscScalar *valaddr; /* address of value */
657: } MatEntry2;
659: struct _p_MatFDColoring {
660: PETSCHEADER(int);
661: PetscInt M, N, m; /* total rows, columns; local rows */
662: PetscInt rstart; /* first row owned by local processor */
663: PetscInt ncolors; /* number of colors */
664: PetscInt *ncolumns; /* number of local columns for a color */
665: PetscInt **columns; /* lists the local columns of each color (using global column numbering) */
666: IS *isa; /* these are the IS that contain the column values given in columns */
667: PetscInt *nrows; /* number of local rows for each color */
668: MatEntry *matentry; /* holds (row, column, address of value) for Jacobian matrix entry */
669: MatEntry2 *matentry2; /* holds (row, address of value) for Jacobian matrix entry */
670: PetscScalar *dy; /* store a block of F(x+dx)-F(x) when J is in BAIJ format */
671: PetscReal error_rel; /* square root of relative error in computing function */
672: PetscReal umin; /* minimum allowable u'dx value */
673: Vec w1, w2, w3; /* work vectors used in computing Jacobian */
674: PetscBool fset; /* indicates that the initial function value F(X) is set */
675: PetscErrorCode (*f)(void); /* function that defines Jacobian */
676: void *fctx; /* optional user-defined context for use by the function f */
677: Vec vscale; /* holds FD scaling, i.e. 1/dx for each perturbed column */
678: PetscInt currentcolor; /* color for which function evaluation is being done now */
679: const char *htype; /* "wp" or "ds" */
680: ISColoringType ctype; /* IS_COLORING_GLOBAL or IS_COLORING_LOCAL */
681: PetscInt brows, bcols; /* number of block rows or columns for speedup inserting the dense matrix into sparse Jacobian */
682: PetscBool setupcalled; /* true if setup has been called */
683: PetscBool viewed; /* true if the -mat_fd_coloring_view has been triggered already */
684: void (*ftn_func_pointer)(void), *ftn_func_cntx; /* serve the same purpose as *fortran_func_pointers in PETSc objects */
685: PetscObjectId matid; /* matrix this object was created with, must always be the same */
686: };
688: typedef struct _MatColoringOps *MatColoringOps;
689: struct _MatColoringOps {
690: PetscErrorCode (*destroy)(MatColoring);
691: PetscErrorCode (*setfromoptions)(MatColoring, PetscOptionItems *);
692: PetscErrorCode (*view)(MatColoring, PetscViewer);
693: PetscErrorCode (*apply)(MatColoring, ISColoring *);
694: PetscErrorCode (*weights)(MatColoring, PetscReal **, PetscInt **);
695: };
697: struct _p_MatColoring {
698: PETSCHEADER(struct _MatColoringOps);
699: Mat mat;
700: PetscInt dist; /* distance of the coloring */
701: PetscInt maxcolors; /* the maximum number of colors returned, maxcolors=1 for MIS */
702: void *data; /* inner context */
703: PetscBool valid; /* check to see if what is produced is a valid coloring */
704: MatColoringWeightType weight_type; /* type of weight computation to be performed */
705: PetscReal *user_weights; /* custom weights and permutation */
706: PetscInt *user_lperm;
707: PetscBool valid_iscoloring; /* check to see if matcoloring is produced a valid iscoloring */
708: };
710: struct _p_MatTransposeColoring {
711: PETSCHEADER(int);
712: PetscInt M, N, m; /* total rows, columns; local rows */
713: PetscInt rstart; /* first row owned by local processor */
714: PetscInt ncolors; /* number of colors */
715: PetscInt *ncolumns; /* number of local columns for a color */
716: PetscInt *nrows; /* number of local rows for each color */
717: PetscInt currentcolor; /* color for which function evaluation is being done now */
718: ISColoringType ctype; /* IS_COLORING_GLOBAL or IS_COLORING_LOCAL */
720: PetscInt *colorforrow, *colorforcol; /* pointer to rows and columns */
721: PetscInt *rows; /* lists the local rows for each color (using the local row numbering) */
722: PetscInt *den2sp; /* maps (row,color) in the dense matrix to index of sparse matrix array a->a */
723: PetscInt *columns; /* lists the local columns of each color (using global column numbering) */
724: PetscInt brows; /* number of rows for efficient implementation of MatTransColoringApplyDenToSp() */
725: PetscInt *lstart; /* array used for loop over row blocks of Csparse */
726: };
728: /*
729: Null space context for preconditioner/operators
730: */
731: struct _p_MatNullSpace {
732: PETSCHEADER(int);
733: PetscBool has_cnst;
734: PetscInt n;
735: Vec *vecs;
736: PetscScalar *alpha; /* for projections */
737: PetscErrorCode (*remove)(MatNullSpace, Vec, void *); /* for user provided removal function */
738: void *rmctx; /* context for remove() function */
739: };
741: /*
742: Checking zero pivot for LU, ILU preconditioners.
743: */
744: typedef struct {
745: PetscInt nshift, nshift_max;
746: PetscReal shift_amount, shift_lo, shift_hi, shift_top, shift_fraction;
747: PetscBool newshift;
748: PetscReal rs; /* active row sum of abs(off-diagonals) */
749: PetscScalar pv; /* pivot of the active row */
750: } FactorShiftCtx;
752: PETSC_EXTERN PetscErrorCode MatTransposeCheckNonzeroState_Private(Mat, Mat);
754: /*
755: Used by MatTranspose() and potentially other functions to track the matrix used in the generation of another matrix
756: */
757: typedef struct {
758: PetscObjectId id;
759: PetscObjectState state;
760: PetscObjectState nonzerostate;
761: } MatParentState;
763: PETSC_EXTERN PetscErrorCode MatFactorDumpMatrix(Mat);
764: PETSC_INTERN PetscErrorCode MatSetBlockSizes_Default(Mat, PetscInt, PetscInt);
766: PETSC_SINGLE_LIBRARY_INTERN PetscErrorCode MatShift_Basic(Mat, PetscScalar);
768: static inline PetscErrorCode MatPivotCheck_nz(PETSC_UNUSED Mat mat, const MatFactorInfo *info, FactorShiftCtx *sctx, PETSC_UNUSED PetscInt row)
769: {
770: PetscReal _rs = sctx->rs;
771: PetscReal _zero = info->zeropivot * _rs;
773: PetscFunctionBegin;
774: if (PetscAbsScalar(sctx->pv) <= _zero && !PetscIsNanScalar(sctx->pv)) {
775: /* force |diag| > zeropivot*rs */
776: if (!sctx->nshift) sctx->shift_amount = info->shiftamount;
777: else sctx->shift_amount *= 2.0;
778: sctx->newshift = PETSC_TRUE;
779: (sctx->nshift)++;
780: } else {
781: sctx->newshift = PETSC_FALSE;
782: }
783: PetscFunctionReturn(PETSC_SUCCESS);
784: }
786: static inline PetscErrorCode MatPivotCheck_pd(PETSC_UNUSED Mat mat, const MatFactorInfo *info, FactorShiftCtx *sctx, PETSC_UNUSED PetscInt row)
787: {
788: PetscReal _rs = sctx->rs;
789: PetscReal _zero = info->zeropivot * _rs;
791: PetscFunctionBegin;
792: if (PetscRealPart(sctx->pv) <= _zero && !PetscIsNanScalar(sctx->pv)) {
793: /* force matfactor to be diagonally dominant */
794: if (sctx->nshift == sctx->nshift_max) {
795: sctx->shift_fraction = sctx->shift_hi;
796: } else {
797: sctx->shift_lo = sctx->shift_fraction;
798: sctx->shift_fraction = (sctx->shift_hi + sctx->shift_lo) / (PetscReal)2.;
799: }
800: sctx->shift_amount = sctx->shift_fraction * sctx->shift_top;
801: sctx->nshift++;
802: sctx->newshift = PETSC_TRUE;
803: } else {
804: sctx->newshift = PETSC_FALSE;
805: }
806: PetscFunctionReturn(PETSC_SUCCESS);
807: }
809: static inline PetscErrorCode MatPivotCheck_inblocks(PETSC_UNUSED Mat mat, const MatFactorInfo *info, FactorShiftCtx *sctx, PETSC_UNUSED PetscInt row)
810: {
811: PetscReal _zero = info->zeropivot;
813: PetscFunctionBegin;
814: if (PetscAbsScalar(sctx->pv) <= _zero && !PetscIsNanScalar(sctx->pv)) {
815: sctx->pv += info->shiftamount;
816: sctx->shift_amount = 0.0;
817: sctx->nshift++;
818: }
819: sctx->newshift = PETSC_FALSE;
820: PetscFunctionReturn(PETSC_SUCCESS);
821: }
823: static inline PetscErrorCode MatPivotCheck_none(Mat fact, Mat mat, const MatFactorInfo *info, FactorShiftCtx *sctx, PetscInt row)
824: {
825: PetscReal _zero = info->zeropivot;
827: PetscFunctionBegin;
828: sctx->newshift = PETSC_FALSE;
829: if (PetscAbsScalar(sctx->pv) <= _zero && !PetscIsNanScalar(sctx->pv)) {
830: PetscCheck(!mat->erroriffailure, PETSC_COMM_SELF, PETSC_ERR_MAT_LU_ZRPVT, "Zero pivot row %" PetscInt_FMT " value %g tolerance %g", row, (double)PetscAbsScalar(sctx->pv), (double)_zero);
831: PetscCall(PetscInfo(mat, "Detected zero pivot in factorization in row %" PetscInt_FMT " value %g tolerance %g\n", row, (double)PetscAbsScalar(sctx->pv), (double)_zero));
832: fact->factorerrortype = MAT_FACTOR_NUMERIC_ZEROPIVOT;
833: fact->factorerror_zeropivot_value = PetscAbsScalar(sctx->pv);
834: fact->factorerror_zeropivot_row = row;
835: }
836: PetscFunctionReturn(PETSC_SUCCESS);
837: }
839: static inline PetscErrorCode MatPivotCheck(Mat fact, Mat mat, const MatFactorInfo *info, FactorShiftCtx *sctx, PetscInt row)
840: {
841: PetscFunctionBegin;
842: if (info->shifttype == (PetscReal)MAT_SHIFT_NONZERO) PetscCall(MatPivotCheck_nz(mat, info, sctx, row));
843: else if (info->shifttype == (PetscReal)MAT_SHIFT_POSITIVE_DEFINITE) PetscCall(MatPivotCheck_pd(mat, info, sctx, row));
844: else if (info->shifttype == (PetscReal)MAT_SHIFT_INBLOCKS) PetscCall(MatPivotCheck_inblocks(mat, info, sctx, row));
845: else PetscCall(MatPivotCheck_none(fact, mat, info, sctx, row));
846: PetscFunctionReturn(PETSC_SUCCESS);
847: }
849: #include <petscbt.h>
850: /*
851: Create and initialize a linked list
852: Input Parameters:
853: idx_start - starting index of the list
854: lnk_max - max value of lnk indicating the end of the list
855: nlnk - max length of the list
856: Output Parameters:
857: lnk - list initialized
858: bt - PetscBT (bitarray) with all bits set to false
859: lnk_empty - flg indicating the list is empty
860: */
861: #define PetscLLCreate(idx_start, lnk_max, nlnk, lnk, bt) ((PetscErrorCode)(PetscMalloc1(nlnk, &lnk) || PetscBTCreate(nlnk, &(bt)) || (lnk[idx_start] = lnk_max, PETSC_SUCCESS)))
863: #define PetscLLCreate_new(idx_start, lnk_max, nlnk, lnk, bt, lnk_empty) ((PetscErrorCode)(PetscMalloc1(nlnk, &lnk) || PetscBTCreate(nlnk, &(bt)) || (lnk_empty = PETSC_TRUE, 0) || (lnk[idx_start] = lnk_max, PETSC_SUCCESS)))
865: static inline PetscErrorCode PetscLLInsertLocation_Private(PetscBool assume_sorted, PetscInt k, PetscInt idx_start, PetscInt entry, PetscInt *PETSC_RESTRICT nlnk, PetscInt *PETSC_RESTRICT lnkdata, PetscInt *PETSC_RESTRICT lnk)
866: {
867: PetscInt location;
869: PetscFunctionBegin;
870: /* start from the beginning if entry < previous entry */
871: if (!assume_sorted && k && entry < *lnkdata) *lnkdata = idx_start;
872: /* search for insertion location */
873: do {
874: location = *lnkdata;
875: *lnkdata = lnk[location];
876: } while (entry > *lnkdata);
877: /* insertion location is found, add entry into lnk */
878: lnk[location] = entry;
879: lnk[entry] = *lnkdata;
880: ++(*nlnk);
881: *lnkdata = entry; /* next search starts from here if next_entry > entry */
882: PetscFunctionReturn(PETSC_SUCCESS);
883: }
885: static inline PetscErrorCode PetscLLAdd_Private(PetscInt nidx, const PetscInt *PETSC_RESTRICT indices, PetscInt idx_start, PetscInt *PETSC_RESTRICT nlnk, PetscInt *PETSC_RESTRICT lnk, PetscBT bt, PetscBool assume_sorted)
886: {
887: PetscFunctionBegin;
888: *nlnk = 0;
889: for (PetscInt k = 0, lnkdata = idx_start; k < nidx; ++k) {
890: const PetscInt entry = indices[k];
892: if (!PetscBTLookupSet(bt, entry)) PetscCall(PetscLLInsertLocation_Private(assume_sorted, k, idx_start, entry, nlnk, &lnkdata, lnk));
893: }
894: PetscFunctionReturn(PETSC_SUCCESS);
895: }
897: /*
898: Add an index set into a sorted linked list
899: Input Parameters:
900: nidx - number of input indices
901: indices - integer array
902: idx_start - starting index of the list
903: lnk - linked list(an integer array) that is created
904: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
905: output Parameters:
906: nlnk - number of newly added indices
907: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from indices
908: bt - updated PetscBT (bitarray)
909: */
910: static inline PetscErrorCode PetscLLAdd(PetscInt nidx, const PetscInt *PETSC_RESTRICT indices, PetscInt idx_start, PetscInt *PETSC_RESTRICT nlnk, PetscInt *PETSC_RESTRICT lnk, PetscBT bt)
911: {
912: PetscFunctionBegin;
913: PetscCall(PetscLLAdd_Private(nidx, indices, idx_start, nlnk, lnk, bt, PETSC_FALSE));
914: PetscFunctionReturn(PETSC_SUCCESS);
915: }
917: /*
918: Add a SORTED ascending index set into a sorted linked list - same as PetscLLAdd() bus skip 'if (_k && _entry < _lnkdata) _lnkdata = idx_start;'
919: Input Parameters:
920: nidx - number of input indices
921: indices - sorted integer array
922: idx_start - starting index of the list
923: lnk - linked list(an integer array) that is created
924: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
925: output Parameters:
926: nlnk - number of newly added indices
927: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from indices
928: bt - updated PetscBT (bitarray)
929: */
930: static inline PetscErrorCode PetscLLAddSorted(PetscInt nidx, const PetscInt *PETSC_RESTRICT indices, PetscInt idx_start, PetscInt *PETSC_RESTRICT nlnk, PetscInt *PETSC_RESTRICT lnk, PetscBT bt)
931: {
932: PetscFunctionBegin;
933: PetscCall(PetscLLAdd_Private(nidx, indices, idx_start, nlnk, lnk, bt, PETSC_TRUE));
934: PetscFunctionReturn(PETSC_SUCCESS);
935: }
937: /*
938: Add a permuted index set into a sorted linked list
939: Input Parameters:
940: nidx - number of input indices
941: indices - integer array
942: perm - permutation of indices
943: idx_start - starting index of the list
944: lnk - linked list(an integer array) that is created
945: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
946: output Parameters:
947: nlnk - number of newly added indices
948: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from indices
949: bt - updated PetscBT (bitarray)
950: */
951: static inline PetscErrorCode PetscLLAddPerm(PetscInt nidx, const PetscInt *PETSC_RESTRICT indices, const PetscInt *PETSC_RESTRICT perm, PetscInt idx_start, PetscInt *PETSC_RESTRICT nlnk, PetscInt *PETSC_RESTRICT lnk, PetscBT bt)
952: {
953: PetscFunctionBegin;
954: *nlnk = 0;
955: for (PetscInt k = 0, lnkdata = idx_start; k < nidx; ++k) {
956: const PetscInt entry = perm[indices[k]];
958: if (!PetscBTLookupSet(bt, entry)) PetscCall(PetscLLInsertLocation_Private(PETSC_FALSE, k, idx_start, entry, nlnk, &lnkdata, lnk));
959: }
960: PetscFunctionReturn(PETSC_SUCCESS);
961: }
963: #if 0
964: /* this appears to be unused? */
965: static inline PetscErrorCode PetscLLAddSorted_new(PetscInt nidx, PetscInt *indices, PetscInt idx_start, PetscBool *lnk_empty, PetscInt *nlnk, PetscInt *lnk, PetscBT bt)
966: {
967: PetscInt lnkdata = idx_start;
969: PetscFunctionBegin;
970: if (*lnk_empty) {
971: for (PetscInt k = 0; k < nidx; ++k) {
972: const PetscInt entry = indices[k], location = lnkdata;
974: PetscCall(PetscBTSet(bt,entry)); /* mark the new entry */
975: lnkdata = lnk[location];
976: /* insertion location is found, add entry into lnk */
977: lnk[location] = entry;
978: lnk[entry] = lnkdata;
979: lnkdata = entry; /* next search starts from here */
980: }
981: /* lnk[indices[nidx-1]] = lnk[idx_start];
982: lnk[idx_start] = indices[0];
983: PetscCall(PetscBTSet(bt,indices[0]));
984: for (_k=1; _k<nidx; _k++) {
985: PetscCall(PetscBTSet(bt,indices[_k]));
986: lnk[indices[_k-1]] = indices[_k];
987: }
988: */
989: *nlnk = nidx;
990: *lnk_empty = PETSC_FALSE;
991: } else {
992: *nlnk = 0;
993: for (PetscInt k = 0; k < nidx; ++k) {
994: const PetscInt entry = indices[k];
996: if (!PetscBTLookupSet(bt,entry)) PetscCall(PetscLLInsertLocation_Private(PETSC_TRUE,k,idx_start,entry,nlnk,&lnkdata,lnk));
997: }
998: }
999: PetscFunctionReturn(PETSC_SUCCESS);
1000: }
1001: #endif
1003: /*
1004: Add a SORTED index set into a sorted linked list used for LUFactorSymbolic()
1005: Same as PetscLLAddSorted() with an additional operation:
1006: count the number of input indices that are no larger than 'diag'
1007: Input Parameters:
1008: indices - sorted integer array
1009: idx_start - starting index of the list, index of pivot row
1010: lnk - linked list(an integer array) that is created
1011: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1012: diag - index of the active row in LUFactorSymbolic
1013: nzbd - number of input indices with indices <= idx_start
1014: im - im[idx_start] is initialized as num of nonzero entries in row=idx_start
1015: output Parameters:
1016: nlnk - number of newly added indices
1017: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from indices
1018: bt - updated PetscBT (bitarray)
1019: im - im[idx_start]: unchanged if diag is not an entry
1020: : num of entries with indices <= diag if diag is an entry
1021: */
1022: static inline PetscErrorCode PetscLLAddSortedLU(const PetscInt *PETSC_RESTRICT indices, PetscInt idx_start, PetscInt *PETSC_RESTRICT nlnk, PetscInt *PETSC_RESTRICT lnk, PetscBT bt, PetscInt diag, PetscInt nzbd, PetscInt *PETSC_RESTRICT im)
1023: {
1024: const PetscInt nidx = im[idx_start] - nzbd; /* num of entries with idx_start < index <= diag */
1026: PetscFunctionBegin;
1027: *nlnk = 0;
1028: for (PetscInt k = 0, lnkdata = idx_start; k < nidx; ++k) {
1029: const PetscInt entry = indices[k];
1031: ++nzbd;
1032: if (entry == diag) im[idx_start] = nzbd;
1033: if (!PetscBTLookupSet(bt, entry)) PetscCall(PetscLLInsertLocation_Private(PETSC_TRUE, k, idx_start, entry, nlnk, &lnkdata, lnk));
1034: }
1035: PetscFunctionReturn(PETSC_SUCCESS);
1036: }
1038: /*
1039: Copy data on the list into an array, then initialize the list
1040: Input Parameters:
1041: idx_start - starting index of the list
1042: lnk_max - max value of lnk indicating the end of the list
1043: nlnk - number of data on the list to be copied
1044: lnk - linked list
1045: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1046: output Parameters:
1047: indices - array that contains the copied data
1048: lnk - linked list that is cleaned and initialize
1049: bt - PetscBT (bitarray) with all bits set to false
1050: */
1051: static inline PetscErrorCode PetscLLClean(PetscInt idx_start, PetscInt lnk_max, PetscInt nlnk, PetscInt *PETSC_RESTRICT lnk, PetscInt *PETSC_RESTRICT indices, PetscBT bt)
1052: {
1053: PetscFunctionBegin;
1054: for (PetscInt j = 0, idx = idx_start; j < nlnk; ++j) {
1055: idx = lnk[idx];
1056: indices[j] = idx;
1057: PetscCall(PetscBTClear(bt, idx));
1058: }
1059: lnk[idx_start] = lnk_max;
1060: PetscFunctionReturn(PETSC_SUCCESS);
1061: }
1063: /*
1064: Free memories used by the list
1065: */
1066: #define PetscLLDestroy(lnk, bt) ((PetscErrorCode)(PetscFree(lnk) || PetscBTDestroy(&(bt))))
1068: /* Routines below are used for incomplete matrix factorization */
1069: /*
1070: Create and initialize a linked list and its levels
1071: Input Parameters:
1072: idx_start - starting index of the list
1073: lnk_max - max value of lnk indicating the end of the list
1074: nlnk - max length of the list
1075: Output Parameters:
1076: lnk - list initialized
1077: lnk_lvl - array of size nlnk for storing levels of lnk
1078: bt - PetscBT (bitarray) with all bits set to false
1079: */
1080: #define PetscIncompleteLLCreate(idx_start, lnk_max, nlnk, lnk, lnk_lvl, bt) \
1081: ((PetscErrorCode)(PetscIntMultError(2, nlnk, NULL) || PetscMalloc1(2 * nlnk, &lnk) || PetscBTCreate(nlnk, &(bt)) || (lnk[idx_start] = lnk_max, lnk_lvl = lnk + nlnk, PETSC_SUCCESS)))
1083: static inline PetscErrorCode PetscIncompleteLLInsertLocation_Private(PetscBool assume_sorted, PetscInt k, PetscInt idx_start, PetscInt entry, PetscInt *PETSC_RESTRICT nlnk, PetscInt *PETSC_RESTRICT lnkdata, PetscInt *PETSC_RESTRICT lnk, PetscInt *PETSC_RESTRICT lnklvl, PetscInt newval)
1084: {
1085: PetscFunctionBegin;
1086: PetscCall(PetscLLInsertLocation_Private(assume_sorted, k, idx_start, entry, nlnk, lnkdata, lnk));
1087: lnklvl[entry] = newval;
1088: PetscFunctionReturn(PETSC_SUCCESS);
1089: }
1091: /*
1092: Initialize a sorted linked list used for ILU and ICC
1093: Input Parameters:
1094: nidx - number of input idx
1095: idx - integer array used for storing column indices
1096: idx_start - starting index of the list
1097: perm - indices of an IS
1098: lnk - linked list(an integer array) that is created
1099: lnklvl - levels of lnk
1100: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1101: output Parameters:
1102: nlnk - number of newly added idx
1103: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1104: lnklvl - levels of lnk
1105: bt - updated PetscBT (bitarray)
1106: */
1107: static inline PetscErrorCode PetscIncompleteLLInit(PetscInt nidx, const PetscInt *PETSC_RESTRICT idx, PetscInt idx_start, const PetscInt *PETSC_RESTRICT perm, PetscInt *PETSC_RESTRICT nlnk, PetscInt *PETSC_RESTRICT lnk, PetscInt *PETSC_RESTRICT lnklvl, PetscBT bt)
1108: {
1109: PetscFunctionBegin;
1110: *nlnk = 0;
1111: for (PetscInt k = 0, lnkdata = idx_start; k < nidx; ++k) {
1112: const PetscInt entry = perm[idx[k]];
1114: if (!PetscBTLookupSet(bt, entry)) PetscCall(PetscIncompleteLLInsertLocation_Private(PETSC_FALSE, k, idx_start, entry, nlnk, &lnkdata, lnk, lnklvl, 0));
1115: }
1116: PetscFunctionReturn(PETSC_SUCCESS);
1117: }
1119: static inline PetscErrorCode PetscIncompleteLLAdd_Private(PetscInt nidx, const PetscInt *PETSC_RESTRICT idx, PetscReal level, const PetscInt *PETSC_RESTRICT idxlvl, PetscInt idx_start, PetscInt *PETSC_RESTRICT nlnk, PetscInt *PETSC_RESTRICT lnk, PetscInt *PETSC_RESTRICT lnklvl, PetscBT bt, PetscInt prow_offset, PetscBool assume_sorted)
1120: {
1121: PetscFunctionBegin;
1122: *nlnk = 0;
1123: for (PetscInt k = 0, lnkdata = idx_start; k < nidx; ++k) {
1124: const PetscInt incrlev = idxlvl[k] + prow_offset + 1;
1126: if (incrlev <= level) {
1127: const PetscInt entry = idx[k];
1129: if (!PetscBTLookupSet(bt, entry)) PetscCall(PetscIncompleteLLInsertLocation_Private(assume_sorted, k, idx_start, entry, nlnk, &lnkdata, lnk, lnklvl, incrlev));
1130: else if (lnklvl[entry] > incrlev) lnklvl[entry] = incrlev; /* existing entry */
1131: }
1132: }
1133: PetscFunctionReturn(PETSC_SUCCESS);
1134: }
1136: /*
1137: Add a SORTED index set into a sorted linked list for ICC
1138: Input Parameters:
1139: nidx - number of input indices
1140: idx - sorted integer array used for storing column indices
1141: level - level of fill, e.g., ICC(level)
1142: idxlvl - level of idx
1143: idx_start - starting index of the list
1144: lnk - linked list(an integer array) that is created
1145: lnklvl - levels of lnk
1146: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1147: idxlvl_prow - idxlvl[prow], where prow is the row number of the idx
1148: output Parameters:
1149: nlnk - number of newly added indices
1150: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1151: lnklvl - levels of lnk
1152: bt - updated PetscBT (bitarray)
1153: Note: the level of U(i,j) is set as lvl(i,j) = min{ lvl(i,j), lvl(prow,i)+lvl(prow,j)+1)
1154: where idx = non-zero columns of U(prow,prow+1:n-1), prow<i
1155: */
1156: static inline PetscErrorCode PetscICCLLAddSorted(PetscInt nidx, const PetscInt *PETSC_RESTRICT idx, PetscReal level, const PetscInt *PETSC_RESTRICT idxlvl, PetscInt idx_start, PetscInt *PETSC_RESTRICT nlnk, PetscInt *PETSC_RESTRICT lnk, PetscInt *PETSC_RESTRICT lnklvl, PetscBT bt, PetscInt idxlvl_prow)
1157: {
1158: PetscFunctionBegin;
1159: PetscCall(PetscIncompleteLLAdd_Private(nidx, idx, level, idxlvl, idx_start, nlnk, lnk, lnklvl, bt, idxlvl_prow, PETSC_TRUE));
1160: PetscFunctionReturn(PETSC_SUCCESS);
1161: }
1163: /*
1164: Add a SORTED index set into a sorted linked list for ILU
1165: Input Parameters:
1166: nidx - number of input indices
1167: idx - sorted integer array used for storing column indices
1168: level - level of fill, e.g., ICC(level)
1169: idxlvl - level of idx
1170: idx_start - starting index of the list
1171: lnk - linked list(an integer array) that is created
1172: lnklvl - levels of lnk
1173: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1174: prow - the row number of idx
1175: output Parameters:
1176: nlnk - number of newly added idx
1177: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1178: lnklvl - levels of lnk
1179: bt - updated PetscBT (bitarray)
1181: Note: the level of factor(i,j) is set as lvl(i,j) = min{ lvl(i,j), lvl(i,prow)+lvl(prow,j)+1)
1182: where idx = non-zero columns of U(prow,prow+1:n-1), prow<i
1183: */
1184: static inline PetscErrorCode PetscILULLAddSorted(PetscInt nidx, const PetscInt *PETSC_RESTRICT idx, PetscInt level, const PetscInt *PETSC_RESTRICT idxlvl, PetscInt idx_start, PetscInt *PETSC_RESTRICT nlnk, PetscInt *PETSC_RESTRICT lnk, PetscInt *PETSC_RESTRICT lnklvl, PetscBT bt, PetscInt prow)
1185: {
1186: PetscFunctionBegin;
1187: PetscCall(PetscIncompleteLLAdd_Private(nidx, idx, level, idxlvl, idx_start, nlnk, lnk, lnklvl, bt, lnklvl[prow], PETSC_TRUE));
1188: PetscFunctionReturn(PETSC_SUCCESS);
1189: }
1191: /*
1192: Add a index set into a sorted linked list
1193: Input Parameters:
1194: nidx - number of input idx
1195: idx - integer array used for storing column indices
1196: level - level of fill, e.g., ICC(level)
1197: idxlvl - level of idx
1198: idx_start - starting index of the list
1199: lnk - linked list(an integer array) that is created
1200: lnklvl - levels of lnk
1201: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1202: output Parameters:
1203: nlnk - number of newly added idx
1204: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1205: lnklvl - levels of lnk
1206: bt - updated PetscBT (bitarray)
1207: */
1208: static inline PetscErrorCode PetscIncompleteLLAdd(PetscInt nidx, const PetscInt *PETSC_RESTRICT idx, PetscReal level, const PetscInt *PETSC_RESTRICT idxlvl, PetscInt idx_start, PetscInt *PETSC_RESTRICT nlnk, PetscInt *PETSC_RESTRICT lnk, PetscInt *PETSC_RESTRICT lnklvl, PetscBT bt)
1209: {
1210: PetscFunctionBegin;
1211: PetscCall(PetscIncompleteLLAdd_Private(nidx, idx, level, idxlvl, idx_start, nlnk, lnk, lnklvl, bt, 0, PETSC_FALSE));
1212: PetscFunctionReturn(PETSC_SUCCESS);
1213: }
1215: /*
1216: Add a SORTED index set into a sorted linked list
1217: Input Parameters:
1218: nidx - number of input indices
1219: idx - sorted integer array used for storing column indices
1220: level - level of fill, e.g., ICC(level)
1221: idxlvl - level of idx
1222: idx_start - starting index of the list
1223: lnk - linked list(an integer array) that is created
1224: lnklvl - levels of lnk
1225: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1226: output Parameters:
1227: nlnk - number of newly added idx
1228: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1229: lnklvl - levels of lnk
1230: bt - updated PetscBT (bitarray)
1231: */
1232: static inline PetscErrorCode PetscIncompleteLLAddSorted(PetscInt nidx, const PetscInt *PETSC_RESTRICT idx, PetscReal level, const PetscInt *PETSC_RESTRICT idxlvl, PetscInt idx_start, PetscInt *PETSC_RESTRICT nlnk, PetscInt *PETSC_RESTRICT lnk, PetscInt *PETSC_RESTRICT lnklvl, PetscBT bt)
1233: {
1234: PetscFunctionBegin;
1235: PetscCall(PetscIncompleteLLAdd_Private(nidx, idx, level, idxlvl, idx_start, nlnk, lnk, lnklvl, bt, 0, PETSC_TRUE));
1236: PetscFunctionReturn(PETSC_SUCCESS);
1237: }
1239: /*
1240: Copy data on the list into an array, then initialize the list
1241: Input Parameters:
1242: idx_start - starting index of the list
1243: lnk_max - max value of lnk indicating the end of the list
1244: nlnk - number of data on the list to be copied
1245: lnk - linked list
1246: lnklvl - level of lnk
1247: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1248: output Parameters:
1249: indices - array that contains the copied data
1250: lnk - linked list that is cleaned and initialize
1251: lnklvl - level of lnk that is reinitialized
1252: bt - PetscBT (bitarray) with all bits set to false
1253: */
1254: static inline PetscErrorCode PetscIncompleteLLClean(PetscInt idx_start, PetscInt lnk_max, PetscInt nlnk, PetscInt *PETSC_RESTRICT lnk, PetscInt *PETSC_RESTRICT lnklvl, PetscInt *PETSC_RESTRICT indices, PetscInt *PETSC_RESTRICT indiceslvl, PetscBT bt)
1255: {
1256: PetscFunctionBegin;
1257: for (PetscInt j = 0, idx = idx_start; j < nlnk; ++j) {
1258: idx = lnk[idx];
1259: indices[j] = idx;
1260: indiceslvl[j] = lnklvl[idx];
1261: lnklvl[idx] = -1;
1262: PetscCall(PetscBTClear(bt, idx));
1263: }
1264: lnk[idx_start] = lnk_max;
1265: PetscFunctionReturn(PETSC_SUCCESS);
1266: }
1268: /*
1269: Free memories used by the list
1270: */
1271: #define PetscIncompleteLLDestroy(lnk, bt) ((PetscErrorCode)(PetscFree(lnk) || PetscBTDestroy(&(bt))))
1273: #if !defined(PETSC_CLANG_STATIC_ANALYZER)
1274: #define MatCheckSameLocalSize(A, ar1, B, ar2) \
1275: do { \
1276: PetscCheckSameComm(A, ar1, B, ar2); \
1277: PetscCheck(((A)->rmap->n == (B)->rmap->n) && ((A)->cmap->n == (B)->cmap->n), PETSC_COMM_SELF, PETSC_ERR_ARG_INCOMP, "Incompatible matrix local sizes: parameter # %d (%" PetscInt_FMT " x %" PetscInt_FMT ") != parameter # %d (%" PetscInt_FMT " x %" PetscInt_FMT ")", ar1, \
1278: (A)->rmap->n, (A)->cmap->n, ar2, (B)->rmap->n, (B)->cmap->n); \
1279: } while (0)
1280: #define MatCheckSameSize(A, ar1, B, ar2) \
1281: do { \
1282: PetscCheck(((A)->rmap->N == (B)->rmap->N) && ((A)->cmap->N == (B)->cmap->N), PetscObjectComm((PetscObject)(A)), PETSC_ERR_ARG_INCOMP, "Incompatible matrix global sizes: parameter # %d (%" PetscInt_FMT " x %" PetscInt_FMT ") != parameter # %d (%" PetscInt_FMT " x %" PetscInt_FMT ")", ar1, \
1283: (A)->rmap->N, (A)->cmap->N, ar2, (B)->rmap->N, (B)->cmap->N); \
1284: MatCheckSameLocalSize(A, ar1, B, ar2); \
1285: } while (0)
1286: #else
1287: template <typename Tm>
1288: extern void MatCheckSameLocalSize(Tm, int, Tm, int);
1289: template <typename Tm>
1290: extern void MatCheckSameSize(Tm, int, Tm, int);
1291: #endif
1293: #define VecCheckMatCompatible(M, x, ar1, b, ar2) \
1294: do { \
1295: PetscCheck((M)->cmap->N == (x)->map->N, PetscObjectComm((PetscObject)(M)), PETSC_ERR_ARG_SIZ, "Vector global length incompatible with matrix: parameter # %d global size %" PetscInt_FMT " != matrix column global size %" PetscInt_FMT, ar1, (x)->map->N, \
1296: (M)->cmap->N); \
1297: PetscCheck((M)->rmap->N == (b)->map->N, PetscObjectComm((PetscObject)(M)), PETSC_ERR_ARG_SIZ, "Vector global length incompatible with matrix: parameter # %d global size %" PetscInt_FMT " != matrix row global size %" PetscInt_FMT, ar2, (b)->map->N, \
1298: (M)->rmap->N); \
1299: } while (0)
1301: /* -------------------------------------------------------------------------------------------------------*/
1302: /*
1303: Create and initialize a condensed linked list -
1304: same as PetscLLCreate(), but uses a scalable array 'lnk' with size of max number of entries, not O(N).
1305: Barry suggested this approach (Dec. 6, 2011):
1306: I've thought of an alternative way of representing a linked list that is efficient but doesn't have the O(N) scaling issue
1307: (it may be faster than the O(N) even sequentially due to less crazy memory access).
1309: 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
1310: 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
1311: 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
1312: 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.
1313: 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
1314: to each other so memory access is much better than using the big array.
1316: Example:
1317: nlnk_max=5, lnk_max=36:
1318: Initial list: [0, 0 | 36, 2 | 0, 0 | 0, 0 | 0, 0 | 0, 0 | 0, 0]
1319: here, head_node has index 2 with value lnk[2]=lnk_max=36,
1320: 0-th entry is used to store the number of entries in the list,
1321: The initial lnk represents head -> tail(marked by 36) with number of entries = lnk[0]=0.
1323: Now adding a sorted set {2,4}, the list becomes
1324: [2, 0 | 36, 4 |2, 6 | 4, 2 | 0, 0 | 0, 0 | 0, 0 ]
1325: represents head -> 2 -> 4 -> tail with number of entries = lnk[0]=2.
1327: Then adding a sorted set {0,3,35}, the list
1328: [5, 0 | 36, 8 | 2, 10 | 4, 12 | 0, 4 | 3, 6 | 35, 2 ]
1329: represents head -> 0 -> 2 -> 3 -> 4 -> 35 -> tail with number of entries = lnk[0]=5.
1331: Input Parameters:
1332: nlnk_max - max length of the list
1333: lnk_max - max value of the entries
1334: Output Parameters:
1335: lnk - list created and initialized
1336: bt - PetscBT (bitarray) with all bits set to false. Note: bt has size lnk_max, not nln_max!
1337: */
1338: static inline PetscErrorCode PetscLLCondensedCreate(PetscInt nlnk_max, PetscInt lnk_max, PetscInt **lnk, PetscBT *bt)
1339: {
1340: PetscInt *llnk, lsize = 0;
1342: PetscFunctionBegin;
1343: PetscCall(PetscIntMultError(2, nlnk_max + 2, &lsize));
1344: PetscCall(PetscMalloc1(lsize, lnk));
1345: PetscCall(PetscBTCreate(lnk_max, bt));
1346: llnk = *lnk;
1347: llnk[0] = 0; /* number of entries on the list */
1348: llnk[2] = lnk_max; /* value in the head node */
1349: llnk[3] = 2; /* next for the head node */
1350: PetscFunctionReturn(PETSC_SUCCESS);
1351: }
1353: /*
1354: Add a SORTED ascending index set into a sorted linked list. See PetscLLCondensedCreate() for detailed description.
1355: Input Parameters:
1356: nidx - number of input indices
1357: indices - sorted integer array
1358: lnk - condensed linked list(an integer array) that is created
1359: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1360: output Parameters:
1361: lnk - the sorted(increasing order) linked list containing previous and newly added non-redundate indices
1362: bt - updated PetscBT (bitarray)
1363: */
1364: static inline PetscErrorCode PetscLLCondensedAddSorted(PetscInt nidx, const PetscInt indices[], PetscInt lnk[], PetscBT bt)
1365: {
1366: PetscInt location = 2; /* head */
1367: PetscInt nlnk = lnk[0]; /* num of entries on the input lnk */
1369: PetscFunctionBegin;
1370: for (PetscInt k = 0; k < nidx; k++) {
1371: const PetscInt entry = indices[k];
1372: if (!PetscBTLookupSet(bt, entry)) { /* new entry */
1373: PetscInt next, lnkdata;
1375: /* search for insertion location */
1376: do {
1377: next = location + 1; /* link from previous node to next node */
1378: location = lnk[next]; /* idx of next node */
1379: lnkdata = lnk[location]; /* value of next node */
1380: } while (entry > lnkdata);
1381: /* insertion location is found, add entry into lnk */
1382: const PetscInt newnode = 2 * (nlnk + 2); /* index for this new node */
1383: lnk[next] = newnode; /* connect previous node to the new node */
1384: lnk[newnode] = entry; /* set value of the new node */
1385: lnk[newnode + 1] = location; /* connect new node to next node */
1386: location = newnode; /* next search starts from the new node */
1387: nlnk++;
1388: }
1389: }
1390: lnk[0] = nlnk; /* number of entries in the list */
1391: PetscFunctionReturn(PETSC_SUCCESS);
1392: }
1394: static inline PetscErrorCode PetscLLCondensedClean(PetscInt lnk_max, PETSC_UNUSED PetscInt nidx, PetscInt *indices, PetscInt lnk[], PetscBT bt)
1395: {
1396: const PetscInt nlnk = lnk[0]; /* num of entries on the list */
1397: PetscInt next = lnk[3]; /* head node */
1399: PetscFunctionBegin;
1400: for (PetscInt k = 0; k < nlnk; k++) {
1401: indices[k] = lnk[next];
1402: next = lnk[next + 1];
1403: PetscCall(PetscBTClear(bt, indices[k]));
1404: }
1405: lnk[0] = 0; /* num of entries on the list */
1406: lnk[2] = lnk_max; /* initialize head node */
1407: lnk[3] = 2; /* head node */
1408: PetscFunctionReturn(PETSC_SUCCESS);
1409: }
1411: static inline PetscErrorCode PetscLLCondensedView(PetscInt *lnk)
1412: {
1413: PetscFunctionBegin;
1414: PetscCall(PetscPrintf(PETSC_COMM_SELF, "LLCondensed of size %" PetscInt_FMT ", (val, next)\n", lnk[0]));
1415: for (PetscInt k = 2; k < lnk[0] + 2; ++k) PetscCall(PetscPrintf(PETSC_COMM_SELF, " %" PetscInt_FMT ": (%" PetscInt_FMT ", %" PetscInt_FMT ")\n", 2 * k, lnk[2 * k], lnk[2 * k + 1]));
1416: PetscFunctionReturn(PETSC_SUCCESS);
1417: }
1419: /*
1420: Free memories used by the list
1421: */
1422: static inline PetscErrorCode PetscLLCondensedDestroy(PetscInt *lnk, PetscBT bt)
1423: {
1424: PetscFunctionBegin;
1425: PetscCall(PetscFree(lnk));
1426: PetscCall(PetscBTDestroy(&bt));
1427: PetscFunctionReturn(PETSC_SUCCESS);
1428: }
1430: /* -------------------------------------------------------------------------------------------------------*/
1431: /*
1432: Same as PetscLLCondensedCreate(), but does not use non-scalable O(lnk_max) bitarray
1433: Input Parameters:
1434: nlnk_max - max length of the list
1435: Output Parameters:
1436: lnk - list created and initialized
1437: */
1438: static inline PetscErrorCode PetscLLCondensedCreate_Scalable(PetscInt nlnk_max, PetscInt **lnk)
1439: {
1440: PetscInt *llnk, lsize = 0;
1442: PetscFunctionBegin;
1443: PetscCall(PetscIntMultError(2, nlnk_max + 2, &lsize));
1444: PetscCall(PetscMalloc1(lsize, lnk));
1445: llnk = *lnk;
1446: llnk[0] = 0; /* number of entries on the list */
1447: llnk[2] = PETSC_INT_MAX; /* value in the head node */
1448: llnk[3] = 2; /* next for the head node */
1449: PetscFunctionReturn(PETSC_SUCCESS);
1450: }
1452: static inline PetscErrorCode PetscLLCondensedExpand_Scalable(PetscInt nlnk_max, PetscInt **lnk)
1453: {
1454: PetscInt lsize = 0;
1456: PetscFunctionBegin;
1457: PetscCall(PetscIntMultError(2, nlnk_max + 2, &lsize));
1458: PetscCall(PetscRealloc(lsize * sizeof(PetscInt), lnk));
1459: PetscFunctionReturn(PETSC_SUCCESS);
1460: }
1462: static inline PetscErrorCode PetscLLCondensedAddSorted_Scalable(PetscInt nidx, const PetscInt indices[], PetscInt lnk[])
1463: {
1464: PetscInt location = 2; /* head */
1465: PetscInt nlnk = lnk[0]; /* num of entries on the input lnk */
1467: for (PetscInt k = 0; k < nidx; k++) {
1468: const PetscInt entry = indices[k];
1469: PetscInt next, lnkdata;
1471: /* search for insertion location */
1472: do {
1473: next = location + 1; /* link from previous node to next node */
1474: location = lnk[next]; /* idx of next node */
1475: lnkdata = lnk[location]; /* value of next node */
1476: } while (entry > lnkdata);
1477: if (entry < lnkdata) {
1478: /* insertion location is found, add entry into lnk */
1479: const PetscInt newnode = 2 * (nlnk + 2); /* index for this new node */
1480: lnk[next] = newnode; /* connect previous node to the new node */
1481: lnk[newnode] = entry; /* set value of the new node */
1482: lnk[newnode + 1] = location; /* connect new node to next node */
1483: location = newnode; /* next search starts from the new node */
1484: nlnk++;
1485: }
1486: }
1487: lnk[0] = nlnk; /* number of entries in the list */
1488: return PETSC_SUCCESS;
1489: }
1491: static inline PetscErrorCode PetscLLCondensedClean_Scalable(PETSC_UNUSED PetscInt nidx, PetscInt *indices, PetscInt *lnk)
1492: {
1493: const PetscInt nlnk = lnk[0];
1494: PetscInt next = lnk[3]; /* head node */
1496: for (PetscInt k = 0; k < nlnk; k++) {
1497: indices[k] = lnk[next];
1498: next = lnk[next + 1];
1499: }
1500: lnk[0] = 0; /* num of entries on the list */
1501: lnk[3] = 2; /* head node */
1502: return PETSC_SUCCESS;
1503: }
1505: static inline PetscErrorCode PetscLLCondensedDestroy_Scalable(PetscInt *lnk)
1506: {
1507: return PetscFree(lnk);
1508: }
1510: /* -------------------------------------------------------------------------------------------------------*/
1511: /*
1512: lnk[0] number of links
1513: lnk[1] number of entries
1514: lnk[3n] value
1515: lnk[3n+1] len
1516: lnk[3n+2] link to next value
1518: The next three are always the first link
1520: lnk[3] PETSC_INT_MIN+1
1521: lnk[4] 1
1522: lnk[5] link to first real entry
1524: The next three are always the last link
1526: lnk[6] PETSC_INT_MAX - 1
1527: lnk[7] 1
1528: lnk[8] next valid link (this is the same as lnk[0] but without the decreases)
1529: */
1531: static inline PetscErrorCode PetscLLCondensedCreate_fast(PetscInt nlnk_max, PetscInt **lnk)
1532: {
1533: PetscInt *llnk;
1534: PetscInt lsize = 0;
1536: PetscFunctionBegin;
1537: PetscCall(PetscIntMultError(3, nlnk_max + 3, &lsize));
1538: PetscCall(PetscMalloc1(lsize, lnk));
1539: llnk = *lnk;
1540: llnk[0] = 0; /* nlnk: number of entries on the list */
1541: llnk[1] = 0; /* number of integer entries represented in list */
1542: llnk[3] = PETSC_INT_MIN + 1; /* value in the first node */
1543: llnk[4] = 1; /* count for the first node */
1544: llnk[5] = 6; /* next for the first node */
1545: llnk[6] = PETSC_INT_MAX - 1; /* value in the last node */
1546: llnk[7] = 1; /* count for the last node */
1547: llnk[8] = 0; /* next valid node to be used */
1548: PetscFunctionReturn(PETSC_SUCCESS);
1549: }
1551: static inline PetscErrorCode PetscLLCondensedAddSorted_fast(PetscInt nidx, const PetscInt indices[], PetscInt lnk[])
1552: {
1553: for (PetscInt k = 0, prev = 3 /* first value */; k < nidx; k++) {
1554: const PetscInt entry = indices[k];
1555: PetscInt next = lnk[prev + 2];
1557: /* search for insertion location */
1558: while (entry >= lnk[next]) {
1559: prev = next;
1560: next = lnk[next + 2];
1561: }
1562: /* entry is in range of previous list */
1563: if (entry < lnk[prev] + lnk[prev + 1]) continue;
1564: lnk[1]++;
1565: /* entry is right after previous list */
1566: if (entry == lnk[prev] + lnk[prev + 1]) {
1567: lnk[prev + 1]++;
1568: if (lnk[next] == entry + 1) { /* combine two contiguous strings */
1569: lnk[prev + 1] += lnk[next + 1];
1570: lnk[prev + 2] = lnk[next + 2];
1571: next = lnk[next + 2];
1572: lnk[0]--;
1573: }
1574: continue;
1575: }
1576: /* entry is right before next list */
1577: if (entry == lnk[next] - 1) {
1578: lnk[next]--;
1579: lnk[next + 1]++;
1580: prev = next;
1581: next = lnk[prev + 2];
1582: continue;
1583: }
1584: /* add entry into lnk */
1585: lnk[prev + 2] = 3 * ((lnk[8]++) + 3); /* connect previous node to the new node */
1586: prev = lnk[prev + 2];
1587: lnk[prev] = entry; /* set value of the new node */
1588: lnk[prev + 1] = 1; /* number of values in contiguous string is one to start */
1589: lnk[prev + 2] = next; /* connect new node to next node */
1590: lnk[0]++;
1591: }
1592: return PETSC_SUCCESS;
1593: }
1595: static inline PetscErrorCode PetscLLCondensedClean_fast(PETSC_UNUSED PetscInt nidx, PetscInt *indices, PetscInt *lnk)
1596: {
1597: const PetscInt nlnk = lnk[0];
1598: PetscInt next = lnk[5]; /* first node */
1600: for (PetscInt k = 0, cnt = 0; k < nlnk; k++) {
1601: for (PetscInt j = 0; j < lnk[next + 1]; j++) indices[cnt++] = lnk[next] + j;
1602: next = lnk[next + 2];
1603: }
1604: lnk[0] = 0; /* nlnk: number of links */
1605: lnk[1] = 0; /* number of integer entries represented in list */
1606: lnk[3] = PETSC_INT_MIN + 1; /* value in the first node */
1607: lnk[4] = 1; /* count for the first node */
1608: lnk[5] = 6; /* next for the first node */
1609: lnk[6] = PETSC_INT_MAX - 1; /* value in the last node */
1610: lnk[7] = 1; /* count for the last node */
1611: lnk[8] = 0; /* next valid location to make link */
1612: return PETSC_SUCCESS;
1613: }
1615: static inline PetscErrorCode PetscLLCondensedView_fast(const PetscInt *lnk)
1616: {
1617: const PetscInt nlnk = lnk[0];
1618: PetscInt next = lnk[5]; /* first node */
1620: for (PetscInt k = 0; k < nlnk; k++) {
1621: #if 0 /* Debugging code */
1622: printf("%d value %d len %d next %d\n", next, lnk[next], lnk[next + 1], lnk[next + 2]);
1623: #endif
1624: next = lnk[next + 2];
1625: }
1626: return PETSC_SUCCESS;
1627: }
1629: static inline PetscErrorCode PetscLLCondensedDestroy_fast(PetscInt *lnk)
1630: {
1631: return PetscFree(lnk);
1632: }
1634: PETSC_EXTERN PetscErrorCode PetscCDCreate(PetscInt, PetscCoarsenData **);
1635: PETSC_EXTERN PetscErrorCode PetscCDDestroy(PetscCoarsenData *);
1636: PETSC_EXTERN PetscErrorCode PetscCDIntNdSetID(PetscCDIntNd *, PetscInt);
1637: PETSC_EXTERN PetscErrorCode PetscCDIntNdGetID(const PetscCDIntNd *, PetscInt *);
1638: PETSC_EXTERN PetscErrorCode PetscCDAppendID(PetscCoarsenData *, PetscInt, PetscInt);
1639: PETSC_EXTERN PetscErrorCode PetscCDMoveAppend(PetscCoarsenData *, PetscInt, PetscInt);
1640: PETSC_EXTERN PetscErrorCode PetscCDAppendNode(PetscCoarsenData *, PetscInt, PetscCDIntNd *);
1641: PETSC_EXTERN PetscErrorCode PetscCDRemoveNextNode(PetscCoarsenData *, PetscInt, PetscCDIntNd *);
1642: PETSC_EXTERN PetscErrorCode PetscCDCountAt(const PetscCoarsenData *, PetscInt, PetscInt *);
1643: PETSC_EXTERN PetscErrorCode PetscCDIsEmptyAt(const PetscCoarsenData *, PetscInt, PetscBool *);
1644: PETSC_EXTERN PetscErrorCode PetscCDSetChunkSize(PetscCoarsenData *, PetscInt);
1645: PETSC_EXTERN PetscErrorCode PetscCDPrint(const PetscCoarsenData *, PetscInt, MPI_Comm);
1646: PETSC_EXTERN PetscErrorCode PetscCDGetNonemptyIS(PetscCoarsenData *, IS *);
1647: PETSC_EXTERN PetscErrorCode PetscCDGetMat(PetscCoarsenData *, Mat *);
1648: PETSC_EXTERN PetscErrorCode PetscCDSetMat(PetscCoarsenData *, Mat);
1649: PETSC_EXTERN PetscErrorCode PetscCDClearMat(PetscCoarsenData *);
1650: PETSC_EXTERN PetscErrorCode PetscCDRemoveAllAt(PetscCoarsenData *, PetscInt);
1651: PETSC_EXTERN PetscErrorCode PetscCDCount(const PetscCoarsenData *, PetscInt *_sz);
1653: PETSC_EXTERN PetscErrorCode PetscCDGetHeadPos(const PetscCoarsenData *, PetscInt, PetscCDIntNd **);
1654: PETSC_EXTERN PetscErrorCode PetscCDGetNextPos(const PetscCoarsenData *, PetscInt, PetscCDIntNd **);
1655: PETSC_EXTERN PetscErrorCode PetscCDGetASMBlocks(const PetscCoarsenData *, const PetscInt, PetscInt *, IS **);
1657: /* this is extern because it is used in MatFDColoringUseDM() which is in the DM library */
1658: PETSC_EXTERN PetscErrorCode MatFDColoringApply_AIJ(Mat, MatFDColoring, Vec, void *);
1660: PETSC_EXTERN PetscLogEvent MAT_Mult;
1661: PETSC_EXTERN PetscLogEvent MAT_MultAdd;
1662: PETSC_EXTERN PetscLogEvent MAT_MultTranspose;
1663: PETSC_EXTERN PetscLogEvent MAT_MultHermitianTranspose;
1664: PETSC_EXTERN PetscLogEvent MAT_MultTransposeAdd;
1665: PETSC_EXTERN PetscLogEvent MAT_MultHermitianTransposeAdd;
1666: PETSC_EXTERN PetscLogEvent MAT_Solve;
1667: PETSC_EXTERN PetscLogEvent MAT_Solves;
1668: PETSC_EXTERN PetscLogEvent MAT_SolveAdd;
1669: PETSC_EXTERN PetscLogEvent MAT_SolveTranspose;
1670: PETSC_EXTERN PetscLogEvent MAT_SolveTransposeAdd;
1671: PETSC_EXTERN PetscLogEvent MAT_SOR;
1672: PETSC_EXTERN PetscLogEvent MAT_ForwardSolve;
1673: PETSC_EXTERN PetscLogEvent MAT_BackwardSolve;
1674: PETSC_EXTERN PetscLogEvent MAT_LUFactor;
1675: PETSC_EXTERN PetscLogEvent MAT_LUFactorSymbolic;
1676: PETSC_EXTERN PetscLogEvent MAT_LUFactorNumeric;
1677: PETSC_EXTERN PetscLogEvent MAT_QRFactor;
1678: PETSC_EXTERN PetscLogEvent MAT_QRFactorSymbolic;
1679: PETSC_EXTERN PetscLogEvent MAT_QRFactorNumeric;
1680: PETSC_EXTERN PetscLogEvent MAT_CholeskyFactor;
1681: PETSC_EXTERN PetscLogEvent MAT_CholeskyFactorSymbolic;
1682: PETSC_EXTERN PetscLogEvent MAT_CholeskyFactorNumeric;
1683: PETSC_EXTERN PetscLogEvent MAT_ILUFactor;
1684: PETSC_EXTERN PetscLogEvent MAT_ILUFactorSymbolic;
1685: PETSC_EXTERN PetscLogEvent MAT_ICCFactorSymbolic;
1686: PETSC_EXTERN PetscLogEvent MAT_Copy;
1687: PETSC_EXTERN PetscLogEvent MAT_Convert;
1688: PETSC_EXTERN PetscLogEvent MAT_Scale;
1689: PETSC_EXTERN PetscLogEvent MAT_AssemblyBegin;
1690: PETSC_EXTERN PetscLogEvent MAT_AssemblyEnd;
1691: PETSC_EXTERN PetscLogEvent MAT_SetValues;
1692: PETSC_EXTERN PetscLogEvent MAT_GetValues;
1693: PETSC_EXTERN PetscLogEvent MAT_GetRow;
1694: PETSC_EXTERN PetscLogEvent MAT_GetRowIJ;
1695: PETSC_EXTERN PetscLogEvent MAT_CreateSubMats;
1696: PETSC_EXTERN PetscLogEvent MAT_GetOrdering;
1697: PETSC_EXTERN PetscLogEvent MAT_RedundantMat;
1698: PETSC_EXTERN PetscLogEvent MAT_IncreaseOverlap;
1699: PETSC_EXTERN PetscLogEvent MAT_Partitioning;
1700: PETSC_EXTERN PetscLogEvent MAT_PartitioningND;
1701: PETSC_EXTERN PetscLogEvent MAT_Coarsen;
1702: PETSC_EXTERN PetscLogEvent MAT_ZeroEntries;
1703: PETSC_EXTERN PetscLogEvent MAT_Load;
1704: PETSC_EXTERN PetscLogEvent MAT_View;
1705: PETSC_EXTERN PetscLogEvent MAT_AXPY;
1706: PETSC_EXTERN PetscLogEvent MAT_FDColoringCreate;
1707: PETSC_EXTERN PetscLogEvent MAT_TransposeColoringCreate;
1708: PETSC_EXTERN PetscLogEvent MAT_FDColoringSetUp;
1709: PETSC_EXTERN PetscLogEvent MAT_FDColoringApply;
1710: PETSC_EXTERN PetscLogEvent MAT_Transpose;
1711: PETSC_EXTERN PetscLogEvent MAT_FDColoringFunction;
1712: PETSC_EXTERN PetscLogEvent MAT_CreateSubMat;
1713: PETSC_EXTERN PetscLogEvent MAT_MatSolve;
1714: PETSC_EXTERN PetscLogEvent MAT_MatTrSolve;
1715: PETSC_EXTERN PetscLogEvent MAT_MatMultSymbolic;
1716: PETSC_EXTERN PetscLogEvent MAT_MatMultNumeric;
1717: PETSC_EXTERN PetscLogEvent MAT_Getlocalmatcondensed;
1718: PETSC_EXTERN PetscLogEvent MAT_GetBrowsOfAcols;
1719: PETSC_EXTERN PetscLogEvent MAT_GetBrowsOfAocols;
1720: PETSC_EXTERN PetscLogEvent MAT_PtAPSymbolic;
1721: PETSC_EXTERN PetscLogEvent MAT_PtAPNumeric;
1722: PETSC_EXTERN PetscLogEvent MAT_Seqstompinum;
1723: PETSC_EXTERN PetscLogEvent MAT_Seqstompisym;
1724: PETSC_EXTERN PetscLogEvent MAT_Seqstompi;
1725: PETSC_EXTERN PetscLogEvent MAT_Getlocalmat;
1726: PETSC_EXTERN PetscLogEvent MAT_RARtSymbolic;
1727: PETSC_EXTERN PetscLogEvent MAT_RARtNumeric;
1728: PETSC_EXTERN PetscLogEvent MAT_MatTransposeMultSymbolic;
1729: PETSC_EXTERN PetscLogEvent MAT_MatTransposeMultNumeric;
1730: PETSC_EXTERN PetscLogEvent MAT_TransposeMatMultSymbolic;
1731: PETSC_EXTERN PetscLogEvent MAT_TransposeMatMultNumeric;
1732: PETSC_EXTERN PetscLogEvent MAT_MatMatMultSymbolic;
1733: PETSC_EXTERN PetscLogEvent MAT_MatMatMultNumeric;
1734: PETSC_EXTERN PetscLogEvent MAT_Getsymtransreduced;
1735: PETSC_EXTERN PetscLogEvent MAT_GetSeqNonzeroStructure;
1736: PETSC_EXTERN PetscLogEvent MATMFFD_Mult;
1737: PETSC_EXTERN PetscLogEvent MAT_GetMultiProcBlock;
1738: PETSC_EXTERN PetscLogEvent MAT_CUSPARSECopyToGPU;
1739: PETSC_EXTERN PetscLogEvent MAT_CUSPARSECopyFromGPU;
1740: PETSC_EXTERN PetscLogEvent MAT_CUSPARSEGenerateTranspose;
1741: PETSC_EXTERN PetscLogEvent MAT_CUSPARSESolveAnalysis;
1742: PETSC_EXTERN PetscLogEvent MAT_HIPSPARSECopyToGPU;
1743: PETSC_EXTERN PetscLogEvent MAT_HIPSPARSECopyFromGPU;
1744: PETSC_EXTERN PetscLogEvent MAT_HIPSPARSEGenerateTranspose;
1745: PETSC_EXTERN PetscLogEvent MAT_HIPSPARSESolveAnalysis;
1746: PETSC_EXTERN PetscLogEvent MAT_SetValuesBatch;
1747: PETSC_EXTERN PetscLogEvent MAT_CreateGraph;
1748: PETSC_EXTERN PetscLogEvent MAT_ViennaCLCopyToGPU;
1749: PETSC_EXTERN PetscLogEvent MAT_DenseCopyToGPU;
1750: PETSC_EXTERN PetscLogEvent MAT_DenseCopyFromGPU;
1751: PETSC_EXTERN PetscLogEvent MAT_Merge;
1752: PETSC_EXTERN PetscLogEvent MAT_Residual;
1753: PETSC_EXTERN PetscLogEvent MAT_SetRandom;
1754: PETSC_EXTERN PetscLogEvent MAT_FactorFactS;
1755: PETSC_EXTERN PetscLogEvent MAT_FactorInvS;
1756: PETSC_EXTERN PetscLogEvent MAT_PreallCOO;
1757: PETSC_EXTERN PetscLogEvent MAT_SetVCOO;
1758: PETSC_EXTERN PetscLogEvent MATCOLORING_Apply;
1759: PETSC_EXTERN PetscLogEvent MATCOLORING_Comm;
1760: PETSC_EXTERN PetscLogEvent MATCOLORING_Local;
1761: PETSC_EXTERN PetscLogEvent MATCOLORING_ISCreate;
1762: PETSC_EXTERN PetscLogEvent MATCOLORING_SetUp;
1763: PETSC_EXTERN PetscLogEvent MATCOLORING_Weights;
1764: PETSC_EXTERN PetscLogEvent MAT_H2Opus_Build;
1765: PETSC_EXTERN PetscLogEvent MAT_H2Opus_Compress;
1766: PETSC_EXTERN PetscLogEvent MAT_H2Opus_Orthog;
1767: PETSC_EXTERN PetscLogEvent MAT_H2Opus_LR;
1768: PETSC_EXTERN PetscLogEvent MAT_CUDACopyToGPU;
1769: PETSC_EXTERN PetscLogEvent MAT_HIPCopyToGPU;