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, Mat *);
216: PetscErrorCode (*dummy)(Mat);
217: /*150*/
218: PetscErrorCode (*transposesymbolic)(Mat, Mat *);
219: PetscErrorCode (*eliminatezeros)(Mat, PetscBool);
220: };
221: /*
222: If you add MatOps entries above also add them to the MATOP enum
223: in include/petscmat.h and src/mat/f90-mod/petscmat.h
224: */
226: #include <petscsys.h>
228: typedef struct _p_MatRootName *MatRootName;
229: struct _p_MatRootName {
230: char *rname, *sname, *mname;
231: MatRootName next;
232: };
234: PETSC_EXTERN MatRootName MatRootNameList;
236: /*
237: Utility private matrix routines used outside Mat
238: */
239: PETSC_SINGLE_LIBRARY_INTERN PetscErrorCode MatFindNonzeroRowsOrCols_Basic(Mat, PetscBool, PetscReal, IS *);
241: /*
242: Utility private matrix routines
243: */
244: PETSC_INTERN PetscErrorCode MatConvert_Basic(Mat, MatType, MatReuse, Mat *);
245: PETSC_INTERN PetscErrorCode MatConvert_Shell(Mat, MatType, MatReuse, Mat *);
246: PETSC_INTERN PetscErrorCode MatConvertFrom_Shell(Mat, MatType, MatReuse, Mat *);
247: PETSC_INTERN PetscErrorCode MatCopy_Basic(Mat, Mat, MatStructure);
248: PETSC_INTERN PetscErrorCode MatDiagonalSet_Default(Mat, Vec, InsertMode);
249: #if defined(PETSC_HAVE_SCALAPACK)
250: PETSC_INTERN PetscErrorCode MatConvert_Dense_ScaLAPACK(Mat, MatType, MatReuse, Mat *);
251: #endif
252: PETSC_INTERN PetscErrorCode MatSetPreallocationCOO_Basic(Mat, PetscCount, PetscInt[], PetscInt[]);
253: PETSC_INTERN PetscErrorCode MatSetValuesCOO_Basic(Mat, const PetscScalar[], InsertMode);
255: /* these callbacks rely on the old matrix function pointers for
256: matmat operations. They are unsafe, and should be removed.
257: However, the amount of work needed to clean up all the
258: implementations is not negligible */
259: PETSC_INTERN PetscErrorCode MatProductSymbolic_AB(Mat);
260: PETSC_INTERN PetscErrorCode MatProductNumeric_AB(Mat);
261: PETSC_INTERN PetscErrorCode MatProductSymbolic_AtB(Mat);
262: PETSC_INTERN PetscErrorCode MatProductNumeric_AtB(Mat);
263: PETSC_INTERN PetscErrorCode MatProductSymbolic_ABt(Mat);
264: PETSC_INTERN PetscErrorCode MatProductNumeric_ABt(Mat);
265: PETSC_INTERN PetscErrorCode MatProductNumeric_PtAP(Mat);
266: PETSC_INTERN PetscErrorCode MatProductNumeric_RARt(Mat);
267: PETSC_INTERN PetscErrorCode MatProductSymbolic_ABC(Mat);
268: PETSC_INTERN PetscErrorCode MatProductNumeric_ABC(Mat);
270: PETSC_INTERN PetscErrorCode MatProductCreate_Private(Mat, Mat, Mat, Mat);
271: /* this callback handles all the different triple products and
272: does not rely on the function pointers; used by cuSPARSE/hipSPARSE and KOKKOS-KERNELS */
273: PETSC_INTERN PetscErrorCode MatProductSymbolic_ABC_Basic(Mat);
275: /* CreateGraph is common to AIJ seq and mpi */
276: PETSC_INTERN PetscErrorCode MatCreateGraph_Simple_AIJ(Mat, PetscBool, PetscBool, PetscReal, Mat *);
278: #if defined(PETSC_CLANG_STATIC_ANALYZER)
279: template <typename Tm>
280: extern void MatCheckPreallocated(Tm, int);
281: template <typename Tm>
282: extern void MatCheckProduct(Tm, int);
283: #else /* PETSC_CLANG_STATIC_ANALYZER */
284: #define MatCheckPreallocated(A, arg) \
285: do { \
286: if (!(A)->preallocated) PetscCall(MatSetUp(A)); \
287: } while (0)
289: #if defined(PETSC_USE_DEBUG)
290: #define MatCheckProduct(A, arg) \
291: do { \
292: PetscCheck((A)->product, PETSC_COMM_SELF, PETSC_ERR_ARG_WRONGSTATE, "Argument %d \"%s\" is not a matrix obtained from MatProductCreate()", (arg), #A); \
293: } while (0)
294: #else
295: #define MatCheckProduct(A, arg) \
296: do { \
297: } while (0)
298: #endif
299: #endif /* PETSC_CLANG_STATIC_ANALYZER */
301: /*
302: The stash is used to temporarily store inserted matrix values that
303: belong to another processor. During the assembly phase the stashed
304: values are moved to the correct processor and
305: */
307: typedef struct _MatStashSpace *PetscMatStashSpace;
309: struct _MatStashSpace {
310: PetscMatStashSpace next;
311: PetscScalar *space_head, *val;
312: PetscInt *idx, *idy;
313: PetscInt total_space_size;
314: PetscInt local_used;
315: PetscInt local_remaining;
316: };
318: PETSC_EXTERN PetscErrorCode PetscMatStashSpaceGet(PetscInt, PetscInt, PetscMatStashSpace *);
319: PETSC_EXTERN PetscErrorCode PetscMatStashSpaceContiguous(PetscInt, PetscMatStashSpace *, PetscScalar *, PetscInt *, PetscInt *);
320: PETSC_EXTERN PetscErrorCode PetscMatStashSpaceDestroy(PetscMatStashSpace *);
322: typedef struct {
323: PetscInt count;
324: } MatStashHeader;
326: typedef struct {
327: void *buffer; /* Of type blocktype, dynamically constructed */
328: PetscInt count;
329: char pending;
330: } MatStashFrame;
332: typedef struct _MatStash MatStash;
333: struct _MatStash {
334: PetscInt nmax; /* maximum stash size */
335: PetscInt umax; /* user specified max-size */
336: PetscInt oldnmax; /* the nmax value used previously */
337: PetscInt n; /* stash size */
338: PetscInt bs; /* block size of the stash */
339: PetscInt reallocs; /* preserve the no of mallocs invoked */
340: PetscMatStashSpace space_head, space; /* linked list to hold stashed global row/column numbers and matrix values */
342: PetscErrorCode (*ScatterBegin)(Mat, MatStash *, PetscInt *);
343: PetscErrorCode (*ScatterGetMesg)(MatStash *, PetscMPIInt *, PetscInt **, PetscInt **, PetscScalar **, PetscInt *);
344: PetscErrorCode (*ScatterEnd)(MatStash *);
345: PetscErrorCode (*ScatterDestroy)(MatStash *);
347: /* The following variables are used for communication */
348: MPI_Comm comm;
349: PetscMPIInt size, rank;
350: PetscMPIInt tag1, tag2;
351: MPI_Request *send_waits; /* array of send requests */
352: MPI_Request *recv_waits; /* array of receive requests */
353: MPI_Status *send_status; /* array of send status */
354: PetscInt nsends, nrecvs; /* numbers of sends and receives */
355: PetscScalar *svalues; /* sending data */
356: PetscInt *sindices;
357: PetscScalar **rvalues; /* receiving data (values) */
358: PetscInt **rindices; /* receiving data (indices) */
359: PetscInt nprocessed; /* number of messages already processed */
360: PetscMPIInt *flg_v; /* indicates what messages have arrived so far and from whom */
361: PetscBool reproduce;
362: PetscInt reproduce_count;
364: /* The following variables are used for BTS communication */
365: PetscBool first_assembly_done; /* Is the first time matrix assembly done? */
366: PetscBool use_status; /* Use MPI_Status to determine number of items in each message */
367: PetscMPIInt nsendranks;
368: PetscMPIInt nrecvranks;
369: PetscMPIInt *sendranks;
370: PetscMPIInt *recvranks;
371: MatStashHeader *sendhdr, *recvhdr;
372: MatStashFrame *sendframes; /* pointers to the main messages */
373: MatStashFrame *recvframes;
374: MatStashFrame *recvframe_active;
375: PetscInt recvframe_i; /* index of block within active frame */
376: PetscMPIInt recvframe_count; /* Count actually sent for current frame */
377: PetscInt recvcount; /* Number of receives processed so far */
378: PetscMPIInt *some_indices; /* From last call to MPI_Waitsome */
379: MPI_Status *some_statuses; /* Statuses from last call to MPI_Waitsome */
380: PetscMPIInt some_count; /* Number of requests completed in last call to MPI_Waitsome */
381: PetscMPIInt some_i; /* Index of request currently being processed */
382: MPI_Request *sendreqs;
383: MPI_Request *recvreqs;
384: PetscSegBuffer segsendblocks;
385: PetscSegBuffer segrecvframe;
386: PetscSegBuffer segrecvblocks;
387: MPI_Datatype blocktype;
388: size_t blocktype_size;
389: InsertMode *insertmode; /* Pointer to check mat->insertmode and set upon message arrival in case no local values have been set. */
390: };
392: #if !defined(PETSC_HAVE_MPIUNI)
393: PETSC_INTERN PetscErrorCode MatStashScatterDestroy_BTS(MatStash *);
394: #endif
395: PETSC_INTERN PetscErrorCode MatStashCreate_Private(MPI_Comm, PetscInt, MatStash *);
396: PETSC_INTERN PetscErrorCode MatStashDestroy_Private(MatStash *);
397: PETSC_INTERN PetscErrorCode MatStashScatterEnd_Private(MatStash *);
398: PETSC_INTERN PetscErrorCode MatStashSetInitialSize_Private(MatStash *, PetscInt);
399: PETSC_INTERN PetscErrorCode MatStashGetInfo_Private(MatStash *, PetscInt *, PetscInt *);
400: PETSC_INTERN PetscErrorCode MatStashValuesRow_Private(MatStash *, PetscInt, PetscInt, const PetscInt[], const PetscScalar[], PetscBool);
401: PETSC_INTERN PetscErrorCode MatStashValuesCol_Private(MatStash *, PetscInt, PetscInt, const PetscInt[], const PetscScalar[], PetscInt, PetscBool);
402: PETSC_INTERN PetscErrorCode MatStashValuesRowBlocked_Private(MatStash *, PetscInt, PetscInt, const PetscInt[], const PetscScalar[], PetscInt, PetscInt, PetscInt);
403: PETSC_INTERN PetscErrorCode MatStashValuesColBlocked_Private(MatStash *, PetscInt, PetscInt, const PetscInt[], const PetscScalar[], PetscInt, PetscInt, PetscInt);
404: PETSC_INTERN PetscErrorCode MatStashScatterBegin_Private(Mat, MatStash *, PetscInt *);
405: PETSC_INTERN PetscErrorCode MatStashScatterGetMesg_Private(MatStash *, PetscMPIInt *, PetscInt **, PetscInt **, PetscScalar **, PetscInt *);
406: PETSC_INTERN PetscErrorCode MatGetInfo_External(Mat, MatInfoType, MatInfo *);
408: typedef struct {
409: PetscInt dim;
410: PetscInt dims[4];
411: PetscInt starts[4];
412: PetscBool noc; /* this is a single component problem, hence user will not set MatStencil.c */
413: } MatStencilInfo;
415: /* Info about using compressed row format */
416: typedef struct {
417: PetscBool use; /* indicates compressed rows have been checked and will be used */
418: PetscInt nrows; /* number of non-zero rows */
419: PetscInt *i; /* compressed row pointer */
420: PetscInt *rindex; /* compressed row index */
421: } Mat_CompressedRow;
422: PETSC_EXTERN PetscErrorCode MatCheckCompressedRow(Mat, PetscInt, Mat_CompressedRow *, PetscInt *, PetscInt, PetscReal);
424: typedef struct { /* used by MatCreateRedundantMatrix() for reusing matredundant */
425: PetscInt nzlocal, nsends, nrecvs;
426: PetscMPIInt *send_rank, *recv_rank;
427: PetscInt *sbuf_nz, *rbuf_nz, *sbuf_j, **rbuf_j;
428: PetscScalar *sbuf_a, **rbuf_a;
429: MPI_Comm subcomm; /* when user does not provide a subcomm */
430: IS isrow, iscol;
431: Mat *matseq;
432: } Mat_Redundant;
434: typedef struct { /* used by MatProduct() */
435: MatProductType type;
436: char *alg;
437: Mat A, B, C, Dwork;
438: 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, .. */
439: 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 .. */
440: PetscBool symbolic_used_the_fact_C_is_symmetric; /* MatMatMult(A,B,MAT_REUSE_MATRIX,..&C) is still legitimate), we need to redo symbolic! */
441: PetscReal fill;
442: 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 */
444: /* Some products may display the information on the algorithm used */
445: PetscErrorCode (*view)(Mat, PetscViewer);
447: /* many products have intermediate data structures, each specific to Mat types and product type */
448: PetscBool clear; /* whether or not to clear the data structures after MatProductNumeric has been called */
449: void *data; /* where to stash those structures */
450: PetscErrorCode (*destroy)(void *); /* destroy routine */
451: } Mat_Product;
453: struct _p_Mat {
454: PETSCHEADER(struct _MatOps);
455: PetscLayout rmap, cmap;
456: void *data; /* implementation-specific data */
457: MatFactorType factortype; /* MAT_FACTOR_LU, ILU, CHOLESKY or ICC */
458: PetscBool trivialsymbolic; /* indicates the symbolic factorization doesn't actually do a symbolic factorization, it is delayed to the numeric factorization */
459: PetscBool canuseordering; /* factorization can use ordering provide to routine (most PETSc implementations) */
460: MatOrderingType preferredordering[MAT_FACTOR_NUM_TYPES]; /* what is the preferred (or default) ordering for the matrix solver type */
461: PetscBool assembled; /* is the matrix assembled? */
462: PetscBool was_assembled; /* new values inserted into assembled mat */
463: PetscInt num_ass; /* number of times matrix has been assembled */
464: PetscObjectState nonzerostate; /* each time new nonzeros locations are introduced into the matrix this is updated */
465: PetscObjectState ass_nonzerostate; /* nonzero state at last assembly */
466: MatInfo info; /* matrix information */
467: InsertMode insertmode; /* have values been inserted in matrix or added? */
468: MatStash stash, bstash; /* used for assembling off-proc mat emements */
469: MatNullSpace nullsp; /* null space (operator is singular) */
470: MatNullSpace transnullsp; /* null space of transpose of operator */
471: MatNullSpace nearnullsp; /* near null space to be used by multigrid methods */
472: PetscInt congruentlayouts; /* are the rows and columns layouts congruent? */
473: PetscBool preallocated;
474: MatStencilInfo stencil; /* information for structured grid */
475: PetscBool3 symmetric, hermitian, structurally_symmetric, spd;
476: PetscBool symmetry_eternal, structural_symmetry_eternal, spd_eternal;
477: PetscBool nooffprocentries, nooffproczerorows;
478: PetscBool assembly_subset; /* set by MAT_SUBSET_OFF_PROC_ENTRIES */
479: PetscBool submat_singleis; /* for efficient PCSetUp_ASM() */
480: PetscBool structure_only;
481: PetscBool sortedfull; /* full, sorted rows are inserted */
482: PetscBool force_diagonals; /* set by MAT_FORCE_DIAGONAL_ENTRIES */
483: #if defined(PETSC_HAVE_DEVICE)
484: PetscOffloadMask offloadmask; /* a mask which indicates where the valid matrix data is (GPU, CPU or both) */
485: PetscBool boundtocpu;
486: PetscBool bindingpropagates;
487: #endif
488: char *defaultrandtype;
489: void *spptr; /* pointer for special library like SuperLU */
490: char *solvertype;
491: PetscBool checksymmetryonassembly, checknullspaceonassembly;
492: PetscReal checksymmetrytol;
493: Mat schur; /* Schur complement matrix */
494: MatFactorSchurStatus schur_status; /* status of the Schur complement matrix */
495: Mat_Redundant *redundant; /* used by MatCreateRedundantMatrix() */
496: PetscBool erroriffailure; /* Generate an error if detected (for example a zero pivot) instead of returning */
497: MatFactorError factorerrortype; /* type of error in factorization */
498: PetscReal factorerror_zeropivot_value; /* If numerical zero pivot was detected this is the computed value */
499: PetscInt factorerror_zeropivot_row; /* Row where zero pivot was detected */
500: PetscInt nblocks, *bsizes; /* support for MatSetVariableBlockSizes() */
501: PetscInt p_cstart, p_rank, p_cend, n_rank; /* Information from parallel MatComputeVariableBlockEnvelope() */
502: PetscBool p_parallel;
503: char *defaultvectype;
504: Mat_Product *product;
505: PetscBool form_explicit_transpose; /* hint to generate an explicit mat tranpsose for operations like MatMultTranspose() */
506: PetscBool transupdated; /* whether or not the explicitly generated transpose is up-to-date */
507: char *factorprefix; /* the prefix to use with factored matrix that is created */
508: PetscBool hash_active; /* indicates MatSetValues() is being handled by hashing */
509: };
511: PETSC_INTERN PetscErrorCode MatAXPY_Basic(Mat, PetscScalar, Mat, MatStructure);
512: PETSC_INTERN PetscErrorCode MatAXPY_BasicWithPreallocation(Mat, Mat, PetscScalar, Mat, MatStructure);
513: PETSC_INTERN PetscErrorCode MatAXPY_Basic_Preallocate(Mat, Mat, Mat *);
514: PETSC_INTERN PetscErrorCode MatAXPY_Dense_Nest(Mat, PetscScalar, Mat);
516: PETSC_INTERN PetscErrorCode MatSetUp_Default(Mat);
518: /*
519: Utility for MatZeroRows
520: */
521: PETSC_INTERN PetscErrorCode MatZeroRowsMapLocal_Private(Mat, PetscInt, const PetscInt *, PetscInt *, PetscInt **);
523: /*
524: Utility for MatView/MatLoad
525: */
526: PETSC_INTERN PetscErrorCode MatView_Binary_BlockSizes(Mat, PetscViewer);
527: PETSC_INTERN PetscErrorCode MatLoad_Binary_BlockSizes(Mat, PetscViewer);
529: /*
530: Object for partitioning graphs
531: */
533: typedef struct _MatPartitioningOps *MatPartitioningOps;
534: struct _MatPartitioningOps {
535: PetscErrorCode (*apply)(MatPartitioning, IS *);
536: PetscErrorCode (*applynd)(MatPartitioning, IS *);
537: PetscErrorCode (*setfromoptions)(MatPartitioning, PetscOptionItems *);
538: PetscErrorCode (*destroy)(MatPartitioning);
539: PetscErrorCode (*view)(MatPartitioning, PetscViewer);
540: PetscErrorCode (*improve)(MatPartitioning, IS *);
541: };
543: struct _p_MatPartitioning {
544: PETSCHEADER(struct _MatPartitioningOps);
545: Mat adj;
546: PetscInt *vertex_weights;
547: PetscReal *part_weights;
548: PetscInt n; /* number of partitions */
549: PetscInt ncon; /* number of vertex weights per vertex */
550: void *data;
551: PetscInt setupcalled;
552: PetscBool use_edge_weights; /* A flag indicates whether or not to use edge weights */
553: };
555: /* needed for parallel nested dissection by ParMetis and PTSCOTCH */
556: PETSC_INTERN PetscErrorCode MatPartitioningSizesToSep_Private(PetscInt, PetscInt[], PetscInt[], PetscInt[]);
558: /*
559: Object for coarsen graphs
560: */
561: typedef struct _MatCoarsenOps *MatCoarsenOps;
562: struct _MatCoarsenOps {
563: PetscErrorCode (*apply)(MatCoarsen);
564: PetscErrorCode (*setfromoptions)(MatCoarsen, PetscOptionItems *);
565: PetscErrorCode (*destroy)(MatCoarsen);
566: PetscErrorCode (*view)(MatCoarsen, PetscViewer);
567: };
569: struct _p_MatCoarsen {
570: PETSCHEADER(struct _MatCoarsenOps);
571: Mat graph;
572: void *subctx;
573: /* */
574: PetscBool strict_aggs;
575: IS perm;
576: PetscCoarsenData *agg_lists;
577: };
579: PETSC_EXTERN PetscErrorCode MatCoarsenMISKSetDistance(MatCoarsen, PetscInt);
580: PETSC_EXTERN PetscErrorCode MatCoarsenMISKGetDistance(MatCoarsen, PetscInt *);
582: /*
583: Used in aijdevice.h
584: */
585: typedef struct {
586: PetscInt *i;
587: PetscInt *j;
588: PetscScalar *a;
589: PetscInt n;
590: PetscInt ignorezeroentries;
591: } PetscCSRDataStructure;
593: /*
594: MatFDColoring is used to compute Jacobian matrices efficiently
595: via coloring. The data structure is explained below in an example.
597: Color = 0 1 0 2 | 2 3 0
598: ---------------------------------------------------
599: 00 01 | 05
600: 10 11 | 14 15 Processor 0
601: 22 23 | 25
602: 32 33 |
603: ===================================================
604: | 44 45 46
605: 50 | 55 Processor 1
606: | 64 66
607: ---------------------------------------------------
609: ncolors = 4;
611: ncolumns = {2,1,1,0}
612: columns = {{0,2},{1},{3},{}}
613: nrows = {4,2,3,3}
614: rows = {{0,1,2,3},{0,1},{1,2,3},{0,1,2}}
615: vwscale = {dx(0),dx(1),dx(2),dx(3)} MPI Vec
616: vscale = {dx(0),dx(1),dx(2),dx(3),dx(4),dx(5)} Seq Vec
618: ncolumns = {1,0,1,1}
619: columns = {{6},{},{4},{5}}
620: nrows = {3,0,2,2}
621: rows = {{0,1,2},{},{1,2},{1,2}}
622: vwscale = {dx(4),dx(5),dx(6)} MPI Vec
623: vscale = {dx(0),dx(4),dx(5),dx(6)} Seq Vec
625: See the routine MatFDColoringApply() for how this data is used
626: to compute the Jacobian.
628: */
629: typedef struct {
630: PetscInt row;
631: PetscInt col;
632: PetscScalar *valaddr; /* address of value */
633: } MatEntry;
635: typedef struct {
636: PetscInt row;
637: PetscScalar *valaddr; /* address of value */
638: } MatEntry2;
640: struct _p_MatFDColoring {
641: PETSCHEADER(int);
642: PetscInt M, N, m; /* total rows, columns; local rows */
643: PetscInt rstart; /* first row owned by local processor */
644: PetscInt ncolors; /* number of colors */
645: PetscInt *ncolumns; /* number of local columns for a color */
646: PetscInt **columns; /* lists the local columns of each color (using global column numbering) */
647: IS *isa; /* these are the IS that contain the column values given in columns */
648: PetscInt *nrows; /* number of local rows for each color */
649: MatEntry *matentry; /* holds (row, column, address of value) for Jacobian matrix entry */
650: MatEntry2 *matentry2; /* holds (row, address of value) for Jacobian matrix entry */
651: PetscScalar *dy; /* store a block of F(x+dx)-F(x) when J is in BAIJ format */
652: PetscReal error_rel; /* square root of relative error in computing function */
653: PetscReal umin; /* minimum allowable u'dx value */
654: Vec w1, w2, w3; /* work vectors used in computing Jacobian */
655: PetscBool fset; /* indicates that the initial function value F(X) is set */
656: PetscErrorCode (*f)(void); /* function that defines Jacobian */
657: void *fctx; /* optional user-defined context for use by the function f */
658: Vec vscale; /* holds FD scaling, i.e. 1/dx for each perturbed column */
659: PetscInt currentcolor; /* color for which function evaluation is being done now */
660: const char *htype; /* "wp" or "ds" */
661: ISColoringType ctype; /* IS_COLORING_GLOBAL or IS_COLORING_LOCAL */
662: PetscInt brows, bcols; /* number of block rows or columns for speedup inserting the dense matrix into sparse Jacobian */
663: PetscBool setupcalled; /* true if setup has been called */
664: PetscBool viewed; /* true if the -mat_fd_coloring_view has been triggered already */
665: void (*ftn_func_pointer)(void), *ftn_func_cntx; /* serve the same purpose as *fortran_func_pointers in PETSc objects */
666: PetscObjectId matid; /* matrix this object was created with, must always be the same */
667: };
669: typedef struct _MatColoringOps *MatColoringOps;
670: struct _MatColoringOps {
671: PetscErrorCode (*destroy)(MatColoring);
672: PetscErrorCode (*setfromoptions)(MatColoring, PetscOptionItems *);
673: PetscErrorCode (*view)(MatColoring, PetscViewer);
674: PetscErrorCode (*apply)(MatColoring, ISColoring *);
675: PetscErrorCode (*weights)(MatColoring, PetscReal **, PetscInt **);
676: };
678: struct _p_MatColoring {
679: PETSCHEADER(struct _MatColoringOps);
680: Mat mat;
681: PetscInt dist; /* distance of the coloring */
682: PetscInt maxcolors; /* the maximum number of colors returned, maxcolors=1 for MIS */
683: void *data; /* inner context */
684: PetscBool valid; /* check to see if what is produced is a valid coloring */
685: MatColoringWeightType weight_type; /* type of weight computation to be performed */
686: PetscReal *user_weights; /* custom weights and permutation */
687: PetscInt *user_lperm;
688: PetscBool valid_iscoloring; /* check to see if matcoloring is produced a valid iscoloring */
689: };
691: struct _p_MatTransposeColoring {
692: PETSCHEADER(int);
693: PetscInt M, N, m; /* total rows, columns; local rows */
694: PetscInt rstart; /* first row owned by local processor */
695: PetscInt ncolors; /* number of colors */
696: PetscInt *ncolumns; /* number of local columns for a color */
697: PetscInt *nrows; /* number of local rows for each color */
698: PetscInt currentcolor; /* color for which function evaluation is being done now */
699: ISColoringType ctype; /* IS_COLORING_GLOBAL or IS_COLORING_LOCAL */
701: PetscInt *colorforrow, *colorforcol; /* pointer to rows and columns */
702: PetscInt *rows; /* lists the local rows for each color (using the local row numbering) */
703: PetscInt *den2sp; /* maps (row,color) in the dense matrix to index of sparse matrix array a->a */
704: PetscInt *columns; /* lists the local columns of each color (using global column numbering) */
705: PetscInt brows; /* number of rows for efficient implementation of MatTransColoringApplyDenToSp() */
706: PetscInt *lstart; /* array used for loop over row blocks of Csparse */
707: };
709: /*
710: Null space context for preconditioner/operators
711: */
712: struct _p_MatNullSpace {
713: PETSCHEADER(int);
714: PetscBool has_cnst;
715: PetscInt n;
716: Vec *vecs;
717: PetscScalar *alpha; /* for projections */
718: PetscErrorCode (*remove)(MatNullSpace, Vec, void *); /* for user provided removal function */
719: void *rmctx; /* context for remove() function */
720: };
722: /*
723: Checking zero pivot for LU, ILU preconditioners.
724: */
725: typedef struct {
726: PetscInt nshift, nshift_max;
727: PetscReal shift_amount, shift_lo, shift_hi, shift_top, shift_fraction;
728: PetscBool newshift;
729: PetscReal rs; /* active row sum of abs(off-diagonals) */
730: PetscScalar pv; /* pivot of the active row */
731: } FactorShiftCtx;
733: PETSC_EXTERN PetscErrorCode MatTransposeCheckNonzeroState_Private(Mat, Mat);
735: /*
736: Used by MatTranspose() and potentially other functions to track the matrix used in the generation of another matrix
737: */
738: typedef struct {
739: PetscObjectId id;
740: PetscObjectState state;
741: PetscObjectState nonzerostate;
742: } MatParentState;
744: PETSC_EXTERN PetscErrorCode MatFactorDumpMatrix(Mat);
745: PETSC_INTERN PetscErrorCode MatShift_Basic(Mat, PetscScalar);
746: PETSC_INTERN PetscErrorCode MatSetBlockSizes_Default(Mat, PetscInt, PetscInt);
748: static inline PetscErrorCode MatPivotCheck_nz(PETSC_UNUSED Mat mat, const MatFactorInfo *info, FactorShiftCtx *sctx, PETSC_UNUSED PetscInt row)
749: {
750: PetscReal _rs = sctx->rs;
751: PetscReal _zero = info->zeropivot * _rs;
753: PetscFunctionBegin;
754: if (PetscAbsScalar(sctx->pv) <= _zero && !PetscIsNanScalar(sctx->pv)) {
755: /* force |diag| > zeropivot*rs */
756: if (!sctx->nshift) sctx->shift_amount = info->shiftamount;
757: else sctx->shift_amount *= 2.0;
758: sctx->newshift = PETSC_TRUE;
759: (sctx->nshift)++;
760: } else {
761: sctx->newshift = PETSC_FALSE;
762: }
763: PetscFunctionReturn(PETSC_SUCCESS);
764: }
766: static inline PetscErrorCode MatPivotCheck_pd(PETSC_UNUSED Mat mat, const MatFactorInfo *info, FactorShiftCtx *sctx, PETSC_UNUSED PetscInt row)
767: {
768: PetscReal _rs = sctx->rs;
769: PetscReal _zero = info->zeropivot * _rs;
771: PetscFunctionBegin;
772: if (PetscRealPart(sctx->pv) <= _zero && !PetscIsNanScalar(sctx->pv)) {
773: /* force matfactor to be diagonally dominant */
774: if (sctx->nshift == sctx->nshift_max) {
775: sctx->shift_fraction = sctx->shift_hi;
776: } else {
777: sctx->shift_lo = sctx->shift_fraction;
778: sctx->shift_fraction = (sctx->shift_hi + sctx->shift_lo) / (PetscReal)2.;
779: }
780: sctx->shift_amount = sctx->shift_fraction * sctx->shift_top;
781: sctx->nshift++;
782: sctx->newshift = PETSC_TRUE;
783: } else {
784: sctx->newshift = PETSC_FALSE;
785: }
786: PetscFunctionReturn(PETSC_SUCCESS);
787: }
789: static inline PetscErrorCode MatPivotCheck_inblocks(PETSC_UNUSED Mat mat, const MatFactorInfo *info, FactorShiftCtx *sctx, PETSC_UNUSED PetscInt row)
790: {
791: PetscReal _zero = info->zeropivot;
793: PetscFunctionBegin;
794: if (PetscAbsScalar(sctx->pv) <= _zero && !PetscIsNanScalar(sctx->pv)) {
795: sctx->pv += info->shiftamount;
796: sctx->shift_amount = 0.0;
797: sctx->nshift++;
798: }
799: sctx->newshift = PETSC_FALSE;
800: PetscFunctionReturn(PETSC_SUCCESS);
801: }
803: static inline PetscErrorCode MatPivotCheck_none(Mat fact, Mat mat, const MatFactorInfo *info, FactorShiftCtx *sctx, PetscInt row)
804: {
805: PetscReal _zero = info->zeropivot;
807: PetscFunctionBegin;
808: sctx->newshift = PETSC_FALSE;
809: if (PetscAbsScalar(sctx->pv) <= _zero && !PetscIsNanScalar(sctx->pv)) {
810: 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);
811: PetscCall(PetscInfo(mat, "Detected zero pivot in factorization in row %" PetscInt_FMT " value %g tolerance %g\n", row, (double)PetscAbsScalar(sctx->pv), (double)_zero));
812: fact->factorerrortype = MAT_FACTOR_NUMERIC_ZEROPIVOT;
813: fact->factorerror_zeropivot_value = PetscAbsScalar(sctx->pv);
814: fact->factorerror_zeropivot_row = row;
815: }
816: PetscFunctionReturn(PETSC_SUCCESS);
817: }
819: static inline PetscErrorCode MatPivotCheck(Mat fact, Mat mat, const MatFactorInfo *info, FactorShiftCtx *sctx, PetscInt row)
820: {
821: PetscFunctionBegin;
822: if (info->shifttype == (PetscReal)MAT_SHIFT_NONZERO) PetscCall(MatPivotCheck_nz(mat, info, sctx, row));
823: else if (info->shifttype == (PetscReal)MAT_SHIFT_POSITIVE_DEFINITE) PetscCall(MatPivotCheck_pd(mat, info, sctx, row));
824: else if (info->shifttype == (PetscReal)MAT_SHIFT_INBLOCKS) PetscCall(MatPivotCheck_inblocks(mat, info, sctx, row));
825: else PetscCall(MatPivotCheck_none(fact, mat, info, sctx, row));
826: PetscFunctionReturn(PETSC_SUCCESS);
827: }
829: #include <petscbt.h>
830: /*
831: Create and initialize a linked list
832: Input Parameters:
833: idx_start - starting index of the list
834: lnk_max - max value of lnk indicating the end of the list
835: nlnk - max length of the list
836: Output Parameters:
837: lnk - list initialized
838: bt - PetscBT (bitarray) with all bits set to false
839: lnk_empty - flg indicating the list is empty
840: */
841: #define PetscLLCreate(idx_start, lnk_max, nlnk, lnk, bt) ((PetscErrorCode)(PetscMalloc1(nlnk, &lnk) || PetscBTCreate(nlnk, &(bt)) || (lnk[idx_start] = lnk_max, PETSC_SUCCESS)))
843: #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)))
845: 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)
846: {
847: PetscInt location;
849: PetscFunctionBegin;
850: /* start from the beginning if entry < previous entry */
851: if (!assume_sorted && k && entry < *lnkdata) *lnkdata = idx_start;
852: /* search for insertion location */
853: do {
854: location = *lnkdata;
855: *lnkdata = lnk[location];
856: } while (entry > *lnkdata);
857: /* insertion location is found, add entry into lnk */
858: lnk[location] = entry;
859: lnk[entry] = *lnkdata;
860: ++(*nlnk);
861: *lnkdata = entry; /* next search starts from here if next_entry > entry */
862: PetscFunctionReturn(PETSC_SUCCESS);
863: }
865: 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)
866: {
867: PetscFunctionBegin;
868: *nlnk = 0;
869: for (PetscInt k = 0, lnkdata = idx_start; k < nidx; ++k) {
870: const PetscInt entry = indices[k];
872: if (!PetscBTLookupSet(bt, entry)) PetscCall(PetscLLInsertLocation_Private(assume_sorted, k, idx_start, entry, nlnk, &lnkdata, lnk));
873: }
874: PetscFunctionReturn(PETSC_SUCCESS);
875: }
877: /*
878: Add an index set into a sorted linked list
879: Input Parameters:
880: nidx - number of input indices
881: indices - integer array
882: idx_start - starting index of the list
883: lnk - linked list(an integer array) that is created
884: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
885: output Parameters:
886: nlnk - number of newly added indices
887: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from indices
888: bt - updated PetscBT (bitarray)
889: */
890: static inline PetscErrorCode PetscLLAdd(PetscInt nidx, const PetscInt *PETSC_RESTRICT indices, PetscInt idx_start, PetscInt *PETSC_RESTRICT nlnk, PetscInt *PETSC_RESTRICT lnk, PetscBT bt)
891: {
892: PetscFunctionBegin;
893: PetscCall(PetscLLAdd_Private(nidx, indices, idx_start, nlnk, lnk, bt, PETSC_FALSE));
894: PetscFunctionReturn(PETSC_SUCCESS);
895: }
897: /*
898: Add a SORTED ascending index set into a sorted linked list - same as PetscLLAdd() bus skip 'if (_k && _entry < _lnkdata) _lnkdata = idx_start;'
899: Input Parameters:
900: nidx - number of input indices
901: indices - sorted 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 PetscLLAddSorted(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_TRUE));
914: PetscFunctionReturn(PETSC_SUCCESS);
915: }
917: /*
918: Add a permuted index set into a sorted linked list
919: Input Parameters:
920: nidx - number of input indices
921: indices - integer array
922: perm - permutation of indices
923: idx_start - starting index of the list
924: lnk - linked list(an integer array) that is created
925: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
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: */
931: 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)
932: {
933: PetscFunctionBegin;
934: *nlnk = 0;
935: for (PetscInt k = 0, lnkdata = idx_start; k < nidx; ++k) {
936: const PetscInt entry = perm[indices[k]];
938: if (!PetscBTLookupSet(bt, entry)) PetscCall(PetscLLInsertLocation_Private(PETSC_FALSE, k, idx_start, entry, nlnk, &lnkdata, lnk));
939: }
940: PetscFunctionReturn(PETSC_SUCCESS);
941: }
943: #if 0
944: /* this appears to be unused? */
945: static inline PetscErrorCode PetscLLAddSorted_new(PetscInt nidx, PetscInt *indices, PetscInt idx_start, PetscBool *lnk_empty, PetscInt *nlnk, PetscInt *lnk, PetscBT bt)
946: {
947: PetscInt lnkdata = idx_start;
949: PetscFunctionBegin;
950: if (*lnk_empty) {
951: for (PetscInt k = 0; k < nidx; ++k) {
952: const PetscInt entry = indices[k], location = lnkdata;
954: PetscCall(PetscBTSet(bt,entry)); /* mark the new entry */
955: lnkdata = lnk[location];
956: /* insertion location is found, add entry into lnk */
957: lnk[location] = entry;
958: lnk[entry] = lnkdata;
959: lnkdata = entry; /* next search starts from here */
960: }
961: /* lnk[indices[nidx-1]] = lnk[idx_start];
962: lnk[idx_start] = indices[0];
963: PetscCall(PetscBTSet(bt,indices[0]));
964: for (_k=1; _k<nidx; _k++) {
965: PetscCall(PetscBTSet(bt,indices[_k]));
966: lnk[indices[_k-1]] = indices[_k];
967: }
968: */
969: *nlnk = nidx;
970: *lnk_empty = PETSC_FALSE;
971: } else {
972: *nlnk = 0;
973: for (PetscInt k = 0; k < nidx; ++k) {
974: const PetscInt entry = indices[k];
976: if (!PetscBTLookupSet(bt,entry)) PetscCall(PetscLLInsertLocation_Private(PETSC_TRUE,k,idx_start,entry,nlnk,&lnkdata,lnk));
977: }
978: }
979: PetscFunctionReturn(PETSC_SUCCESS);
980: }
981: #endif
983: /*
984: Add a SORTED index set into a sorted linked list used for LUFactorSymbolic()
985: Same as PetscLLAddSorted() with an additional operation:
986: count the number of input indices that are no larger than 'diag'
987: Input Parameters:
988: indices - sorted integer array
989: idx_start - starting index of the list, index of pivot row
990: lnk - linked list(an integer array) that is created
991: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
992: diag - index of the active row in LUFactorSymbolic
993: nzbd - number of input indices with indices <= idx_start
994: im - im[idx_start] is initialized as num of nonzero entries in row=idx_start
995: output Parameters:
996: nlnk - number of newly added indices
997: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from indices
998: bt - updated PetscBT (bitarray)
999: im - im[idx_start]: unchanged if diag is not an entry
1000: : num of entries with indices <= diag if diag is an entry
1001: */
1002: 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)
1003: {
1004: const PetscInt nidx = im[idx_start] - nzbd; /* num of entries with idx_start < index <= diag */
1006: PetscFunctionBegin;
1007: *nlnk = 0;
1008: for (PetscInt k = 0, lnkdata = idx_start; k < nidx; ++k) {
1009: const PetscInt entry = indices[k];
1011: ++nzbd;
1012: if (entry == diag) im[idx_start] = nzbd;
1013: if (!PetscBTLookupSet(bt, entry)) PetscCall(PetscLLInsertLocation_Private(PETSC_TRUE, k, idx_start, entry, nlnk, &lnkdata, lnk));
1014: }
1015: PetscFunctionReturn(PETSC_SUCCESS);
1016: }
1018: /*
1019: Copy data on the list into an array, then initialize the list
1020: Input Parameters:
1021: idx_start - starting index of the list
1022: lnk_max - max value of lnk indicating the end of the list
1023: nlnk - number of data on the list to be copied
1024: lnk - linked list
1025: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1026: output Parameters:
1027: indices - array that contains the copied data
1028: lnk - linked list that is cleaned and initialize
1029: bt - PetscBT (bitarray) with all bits set to false
1030: */
1031: static inline PetscErrorCode PetscLLClean(PetscInt idx_start, PetscInt lnk_max, PetscInt nlnk, PetscInt *PETSC_RESTRICT lnk, PetscInt *PETSC_RESTRICT indices, PetscBT bt)
1032: {
1033: PetscFunctionBegin;
1034: for (PetscInt j = 0, idx = idx_start; j < nlnk; ++j) {
1035: idx = lnk[idx];
1036: indices[j] = idx;
1037: PetscCall(PetscBTClear(bt, idx));
1038: }
1039: lnk[idx_start] = lnk_max;
1040: PetscFunctionReturn(PETSC_SUCCESS);
1041: }
1043: /*
1044: Free memories used by the list
1045: */
1046: #define PetscLLDestroy(lnk, bt) ((PetscErrorCode)(PetscFree(lnk) || PetscBTDestroy(&(bt))))
1048: /* Routines below are used for incomplete matrix factorization */
1049: /*
1050: Create and initialize a linked list and its levels
1051: Input Parameters:
1052: idx_start - starting index of the list
1053: lnk_max - max value of lnk indicating the end of the list
1054: nlnk - max length of the list
1055: Output Parameters:
1056: lnk - list initialized
1057: lnk_lvl - array of size nlnk for storing levels of lnk
1058: bt - PetscBT (bitarray) with all bits set to false
1059: */
1060: #define PetscIncompleteLLCreate(idx_start, lnk_max, nlnk, lnk, lnk_lvl, bt) \
1061: ((PetscErrorCode)(PetscIntMultError(2, nlnk, NULL) || PetscMalloc1(2 * nlnk, &lnk) || PetscBTCreate(nlnk, &(bt)) || (lnk[idx_start] = lnk_max, lnk_lvl = lnk + nlnk, PETSC_SUCCESS)))
1063: 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)
1064: {
1065: PetscFunctionBegin;
1066: PetscCall(PetscLLInsertLocation_Private(assume_sorted, k, idx_start, entry, nlnk, lnkdata, lnk));
1067: lnklvl[entry] = newval;
1068: PetscFunctionReturn(PETSC_SUCCESS);
1069: }
1071: /*
1072: Initialize a sorted linked list used for ILU and ICC
1073: Input Parameters:
1074: nidx - number of input idx
1075: idx - integer array used for storing column indices
1076: idx_start - starting index of the list
1077: perm - indices of an IS
1078: lnk - linked list(an integer array) that is created
1079: lnklvl - levels of lnk
1080: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1081: output Parameters:
1082: nlnk - number of newly added idx
1083: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1084: lnklvl - levels of lnk
1085: bt - updated PetscBT (bitarray)
1086: */
1087: 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)
1088: {
1089: PetscFunctionBegin;
1090: *nlnk = 0;
1091: for (PetscInt k = 0, lnkdata = idx_start; k < nidx; ++k) {
1092: const PetscInt entry = perm[idx[k]];
1094: if (!PetscBTLookupSet(bt, entry)) PetscCall(PetscIncompleteLLInsertLocation_Private(PETSC_FALSE, k, idx_start, entry, nlnk, &lnkdata, lnk, lnklvl, 0));
1095: }
1096: PetscFunctionReturn(PETSC_SUCCESS);
1097: }
1099: 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)
1100: {
1101: PetscFunctionBegin;
1102: *nlnk = 0;
1103: for (PetscInt k = 0, lnkdata = idx_start; k < nidx; ++k) {
1104: const PetscInt incrlev = idxlvl[k] + prow_offset + 1;
1106: if (incrlev <= level) {
1107: const PetscInt entry = idx[k];
1109: if (!PetscBTLookupSet(bt, entry)) PetscCall(PetscIncompleteLLInsertLocation_Private(assume_sorted, k, idx_start, entry, nlnk, &lnkdata, lnk, lnklvl, incrlev));
1110: else if (lnklvl[entry] > incrlev) lnklvl[entry] = incrlev; /* existing entry */
1111: }
1112: }
1113: PetscFunctionReturn(PETSC_SUCCESS);
1114: }
1116: /*
1117: Add a SORTED index set into a sorted linked list for ICC
1118: Input Parameters:
1119: nidx - number of input indices
1120: idx - sorted integer array used for storing column indices
1121: level - level of fill, e.g., ICC(level)
1122: idxlvl - level of idx
1123: idx_start - starting index of the list
1124: lnk - linked list(an integer array) that is created
1125: lnklvl - levels of lnk
1126: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1127: idxlvl_prow - idxlvl[prow], where prow is the row number of the idx
1128: output Parameters:
1129: nlnk - number of newly added indices
1130: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1131: lnklvl - levels of lnk
1132: bt - updated PetscBT (bitarray)
1133: Note: the level of U(i,j) is set as lvl(i,j) = min{ lvl(i,j), lvl(prow,i)+lvl(prow,j)+1)
1134: where idx = non-zero columns of U(prow,prow+1:n-1), prow<i
1135: */
1136: 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)
1137: {
1138: PetscFunctionBegin;
1139: PetscCall(PetscIncompleteLLAdd_Private(nidx, idx, level, idxlvl, idx_start, nlnk, lnk, lnklvl, bt, idxlvl_prow, PETSC_TRUE));
1140: PetscFunctionReturn(PETSC_SUCCESS);
1141: }
1143: /*
1144: Add a SORTED index set into a sorted linked list for ILU
1145: Input Parameters:
1146: nidx - number of input indices
1147: idx - sorted integer array used for storing column indices
1148: level - level of fill, e.g., ICC(level)
1149: idxlvl - level of idx
1150: idx_start - starting index of the list
1151: lnk - linked list(an integer array) that is created
1152: lnklvl - levels of lnk
1153: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1154: prow - the row number of idx
1155: output Parameters:
1156: nlnk - number of newly added idx
1157: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1158: lnklvl - levels of lnk
1159: bt - updated PetscBT (bitarray)
1161: Note: the level of factor(i,j) is set as lvl(i,j) = min{ lvl(i,j), lvl(i,prow)+lvl(prow,j)+1)
1162: where idx = non-zero columns of U(prow,prow+1:n-1), prow<i
1163: */
1164: 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)
1165: {
1166: PetscFunctionBegin;
1167: PetscCall(PetscIncompleteLLAdd_Private(nidx, idx, level, idxlvl, idx_start, nlnk, lnk, lnklvl, bt, lnklvl[prow], PETSC_TRUE));
1168: PetscFunctionReturn(PETSC_SUCCESS);
1169: }
1171: /*
1172: Add a index set into a sorted linked list
1173: Input Parameters:
1174: nidx - number of input idx
1175: idx - integer array used for storing column indices
1176: level - level of fill, e.g., ICC(level)
1177: idxlvl - level of idx
1178: idx_start - starting index of the list
1179: lnk - linked list(an integer array) that is created
1180: lnklvl - levels of lnk
1181: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1182: output Parameters:
1183: nlnk - number of newly added idx
1184: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1185: lnklvl - levels of lnk
1186: bt - updated PetscBT (bitarray)
1187: */
1188: 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)
1189: {
1190: PetscFunctionBegin;
1191: PetscCall(PetscIncompleteLLAdd_Private(nidx, idx, level, idxlvl, idx_start, nlnk, lnk, lnklvl, bt, 0, PETSC_FALSE));
1192: PetscFunctionReturn(PETSC_SUCCESS);
1193: }
1195: /*
1196: Add a SORTED index set into a sorted linked list
1197: Input Parameters:
1198: nidx - number of input indices
1199: idx - sorted integer array used for storing column indices
1200: level - level of fill, e.g., ICC(level)
1201: idxlvl - level of idx
1202: idx_start - starting index of the list
1203: lnk - linked list(an integer array) that is created
1204: lnklvl - levels of lnk
1205: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1206: output Parameters:
1207: nlnk - number of newly added idx
1208: lnk - the sorted(increasing order) linked list containing new and non-redundate entries from idx
1209: lnklvl - levels of lnk
1210: bt - updated PetscBT (bitarray)
1211: */
1212: 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)
1213: {
1214: PetscFunctionBegin;
1215: PetscCall(PetscIncompleteLLAdd_Private(nidx, idx, level, idxlvl, idx_start, nlnk, lnk, lnklvl, bt, 0, PETSC_TRUE));
1216: PetscFunctionReturn(PETSC_SUCCESS);
1217: }
1219: /*
1220: Copy data on the list into an array, then initialize the list
1221: Input Parameters:
1222: idx_start - starting index of the list
1223: lnk_max - max value of lnk indicating the end of the list
1224: nlnk - number of data on the list to be copied
1225: lnk - linked list
1226: lnklvl - level of lnk
1227: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1228: output Parameters:
1229: indices - array that contains the copied data
1230: lnk - linked list that is cleaned and initialize
1231: lnklvl - level of lnk that is reinitialized
1232: bt - PetscBT (bitarray) with all bits set to false
1233: */
1234: 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)
1235: {
1236: PetscFunctionBegin;
1237: for (PetscInt j = 0, idx = idx_start; j < nlnk; ++j) {
1238: idx = lnk[idx];
1239: indices[j] = idx;
1240: indiceslvl[j] = lnklvl[idx];
1241: lnklvl[idx] = -1;
1242: PetscCall(PetscBTClear(bt, idx));
1243: }
1244: lnk[idx_start] = lnk_max;
1245: PetscFunctionReturn(PETSC_SUCCESS);
1246: }
1248: /*
1249: Free memories used by the list
1250: */
1251: #define PetscIncompleteLLDestroy(lnk, bt) ((PetscErrorCode)(PetscFree(lnk) || PetscBTDestroy(&(bt))))
1253: #if !defined(PETSC_CLANG_STATIC_ANALYZER)
1254: #define MatCheckSameLocalSize(A, ar1, B, ar2) \
1255: do { \
1256: PetscCheckSameComm(A, ar1, B, ar2); \
1257: 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, \
1258: (A)->rmap->n, (A)->cmap->n, ar2, (B)->rmap->n, (B)->cmap->n); \
1259: } while (0)
1260: #define MatCheckSameSize(A, ar1, B, ar2) \
1261: do { \
1262: 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, \
1263: (A)->rmap->N, (A)->cmap->N, ar2, (B)->rmap->N, (B)->cmap->N); \
1264: MatCheckSameLocalSize(A, ar1, B, ar2); \
1265: } while (0)
1266: #else
1267: template <typename Tm>
1268: extern void MatCheckSameLocalSize(Tm, int, Tm, int);
1269: template <typename Tm>
1270: extern void MatCheckSameSize(Tm, int, Tm, int);
1271: #endif
1273: #define VecCheckMatCompatible(M, x, ar1, b, ar2) \
1274: do { \
1275: 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, \
1276: (M)->cmap->N); \
1277: 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, \
1278: (M)->rmap->N); \
1279: } while (0)
1281: /* -------------------------------------------------------------------------------------------------------*/
1282: /*
1283: Create and initialize a condensed linked list -
1284: same as PetscLLCreate(), but uses a scalable array 'lnk' with size of max number of entries, not O(N).
1285: Barry suggested this approach (Dec. 6, 2011):
1286: I've thought of an alternative way of representing a linked list that is efficient but doesn't have the O(N) scaling issue
1287: (it may be faster than the O(N) even sequentially due to less crazy memory access).
1289: 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
1290: 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
1291: 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
1292: 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.
1293: 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
1294: to each other so memory access is much better than using the big array.
1296: Example:
1297: nlnk_max=5, lnk_max=36:
1298: Initial list: [0, 0 | 36, 2 | 0, 0 | 0, 0 | 0, 0 | 0, 0 | 0, 0]
1299: here, head_node has index 2 with value lnk[2]=lnk_max=36,
1300: 0-th entry is used to store the number of entries in the list,
1301: The initial lnk represents head -> tail(marked by 36) with number of entries = lnk[0]=0.
1303: Now adding a sorted set {2,4}, the list becomes
1304: [2, 0 | 36, 4 |2, 6 | 4, 2 | 0, 0 | 0, 0 | 0, 0 ]
1305: represents head -> 2 -> 4 -> tail with number of entries = lnk[0]=2.
1307: Then adding a sorted set {0,3,35}, the list
1308: [5, 0 | 36, 8 | 2, 10 | 4, 12 | 0, 4 | 3, 6 | 35, 2 ]
1309: represents head -> 0 -> 2 -> 3 -> 4 -> 35 -> tail with number of entries = lnk[0]=5.
1311: Input Parameters:
1312: nlnk_max - max length of the list
1313: lnk_max - max value of the entries
1314: Output Parameters:
1315: lnk - list created and initialized
1316: bt - PetscBT (bitarray) with all bits set to false. Note: bt has size lnk_max, not nln_max!
1317: */
1318: static inline PetscErrorCode PetscLLCondensedCreate(PetscInt nlnk_max, PetscInt lnk_max, PetscInt **lnk, PetscBT *bt)
1319: {
1320: PetscInt *llnk, lsize = 0;
1322: PetscFunctionBegin;
1323: PetscCall(PetscIntMultError(2, nlnk_max + 2, &lsize));
1324: PetscCall(PetscMalloc1(lsize, lnk));
1325: PetscCall(PetscBTCreate(lnk_max, bt));
1326: llnk = *lnk;
1327: llnk[0] = 0; /* number of entries on the list */
1328: llnk[2] = lnk_max; /* value in the head node */
1329: llnk[3] = 2; /* next for the head node */
1330: PetscFunctionReturn(PETSC_SUCCESS);
1331: }
1333: /*
1334: Add a SORTED ascending index set into a sorted linked list. See PetscLLCondensedCreate() for detailed description.
1335: Input Parameters:
1336: nidx - number of input indices
1337: indices - sorted integer array
1338: lnk - condensed linked list(an integer array) that is created
1339: bt - PetscBT (bitarray), bt[idx]=true marks idx is in lnk
1340: output Parameters:
1341: lnk - the sorted(increasing order) linked list containing previous and newly added non-redundate indices
1342: bt - updated PetscBT (bitarray)
1343: */
1344: static inline PetscErrorCode PetscLLCondensedAddSorted(PetscInt nidx, const PetscInt indices[], PetscInt lnk[], PetscBT bt)
1345: {
1346: PetscInt location = 2; /* head */
1347: PetscInt nlnk = lnk[0]; /* num of entries on the input lnk */
1349: PetscFunctionBegin;
1350: for (PetscInt k = 0; k < nidx; k++) {
1351: const PetscInt entry = indices[k];
1352: if (!PetscBTLookupSet(bt, entry)) { /* new entry */
1353: PetscInt next, lnkdata;
1355: /* search for insertion location */
1356: do {
1357: next = location + 1; /* link from previous node to next node */
1358: location = lnk[next]; /* idx of next node */
1359: lnkdata = lnk[location]; /* value of next node */
1360: } while (entry > lnkdata);
1361: /* insertion location is found, add entry into lnk */
1362: const PetscInt newnode = 2 * (nlnk + 2); /* index for this new node */
1363: lnk[next] = newnode; /* connect previous node to the new node */
1364: lnk[newnode] = entry; /* set value of the new node */
1365: lnk[newnode + 1] = location; /* connect new node to next node */
1366: location = newnode; /* next search starts from the new node */
1367: nlnk++;
1368: }
1369: }
1370: lnk[0] = nlnk; /* number of entries in the list */
1371: PetscFunctionReturn(PETSC_SUCCESS);
1372: }
1374: static inline PetscErrorCode PetscLLCondensedClean(PetscInt lnk_max, PETSC_UNUSED PetscInt nidx, PetscInt *indices, PetscInt lnk[], PetscBT bt)
1375: {
1376: const PetscInt nlnk = lnk[0]; /* num of entries on the list */
1377: PetscInt next = lnk[3]; /* head node */
1379: PetscFunctionBegin;
1380: for (PetscInt k = 0; k < nlnk; k++) {
1381: indices[k] = lnk[next];
1382: next = lnk[next + 1];
1383: PetscCall(PetscBTClear(bt, indices[k]));
1384: }
1385: lnk[0] = 0; /* num of entries on the list */
1386: lnk[2] = lnk_max; /* initialize head node */
1387: lnk[3] = 2; /* head node */
1388: PetscFunctionReturn(PETSC_SUCCESS);
1389: }
1391: static inline PetscErrorCode PetscLLCondensedView(PetscInt *lnk)
1392: {
1393: PetscFunctionBegin;
1394: PetscCall(PetscPrintf(PETSC_COMM_SELF, "LLCondensed of size %" PetscInt_FMT ", (val, next)\n", lnk[0]));
1395: 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]));
1396: PetscFunctionReturn(PETSC_SUCCESS);
1397: }
1399: /*
1400: Free memories used by the list
1401: */
1402: static inline PetscErrorCode PetscLLCondensedDestroy(PetscInt *lnk, PetscBT bt)
1403: {
1404: PetscFunctionBegin;
1405: PetscCall(PetscFree(lnk));
1406: PetscCall(PetscBTDestroy(&bt));
1407: PetscFunctionReturn(PETSC_SUCCESS);
1408: }
1410: /* -------------------------------------------------------------------------------------------------------*/
1411: /*
1412: Same as PetscLLCondensedCreate(), but does not use non-scalable O(lnk_max) bitarray
1413: Input Parameters:
1414: nlnk_max - max length of the list
1415: Output Parameters:
1416: lnk - list created and initialized
1417: */
1418: static inline PetscErrorCode PetscLLCondensedCreate_Scalable(PetscInt nlnk_max, PetscInt **lnk)
1419: {
1420: PetscInt *llnk, lsize = 0;
1422: PetscFunctionBegin;
1423: PetscCall(PetscIntMultError(2, nlnk_max + 2, &lsize));
1424: PetscCall(PetscMalloc1(lsize, lnk));
1425: llnk = *lnk;
1426: llnk[0] = 0; /* number of entries on the list */
1427: llnk[2] = PETSC_MAX_INT; /* value in the head node */
1428: llnk[3] = 2; /* next for the head node */
1429: PetscFunctionReturn(PETSC_SUCCESS);
1430: }
1432: static inline PetscErrorCode PetscLLCondensedExpand_Scalable(PetscInt nlnk_max, PetscInt **lnk)
1433: {
1434: PetscInt lsize = 0;
1436: PetscFunctionBegin;
1437: PetscCall(PetscIntMultError(2, nlnk_max + 2, &lsize));
1438: PetscCall(PetscRealloc(lsize * sizeof(PetscInt), lnk));
1439: PetscFunctionReturn(PETSC_SUCCESS);
1440: }
1442: static inline PetscErrorCode PetscLLCondensedAddSorted_Scalable(PetscInt nidx, const PetscInt indices[], PetscInt lnk[])
1443: {
1444: PetscInt location = 2; /* head */
1445: PetscInt nlnk = lnk[0]; /* num of entries on the input lnk */
1447: for (PetscInt k = 0; k < nidx; k++) {
1448: const PetscInt entry = indices[k];
1449: PetscInt next, lnkdata;
1451: /* search for insertion location */
1452: do {
1453: next = location + 1; /* link from previous node to next node */
1454: location = lnk[next]; /* idx of next node */
1455: lnkdata = lnk[location]; /* value of next node */
1456: } while (entry > lnkdata);
1457: if (entry < lnkdata) {
1458: /* insertion location is found, add entry into lnk */
1459: const PetscInt newnode = 2 * (nlnk + 2); /* index for this new node */
1460: lnk[next] = newnode; /* connect previous node to the new node */
1461: lnk[newnode] = entry; /* set value of the new node */
1462: lnk[newnode + 1] = location; /* connect new node to next node */
1463: location = newnode; /* next search starts from the new node */
1464: nlnk++;
1465: }
1466: }
1467: lnk[0] = nlnk; /* number of entries in the list */
1468: return PETSC_SUCCESS;
1469: }
1471: static inline PetscErrorCode PetscLLCondensedClean_Scalable(PETSC_UNUSED PetscInt nidx, PetscInt *indices, PetscInt *lnk)
1472: {
1473: const PetscInt nlnk = lnk[0];
1474: PetscInt next = lnk[3]; /* head node */
1476: for (PetscInt k = 0; k < nlnk; k++) {
1477: indices[k] = lnk[next];
1478: next = lnk[next + 1];
1479: }
1480: lnk[0] = 0; /* num of entries on the list */
1481: lnk[3] = 2; /* head node */
1482: return PETSC_SUCCESS;
1483: }
1485: static inline PetscErrorCode PetscLLCondensedDestroy_Scalable(PetscInt *lnk)
1486: {
1487: return PetscFree(lnk);
1488: }
1490: /* -------------------------------------------------------------------------------------------------------*/
1491: /*
1492: lnk[0] number of links
1493: lnk[1] number of entries
1494: lnk[3n] value
1495: lnk[3n+1] len
1496: lnk[3n+2] link to next value
1498: The next three are always the first link
1500: lnk[3] PETSC_MIN_INT+1
1501: lnk[4] 1
1502: lnk[5] link to first real entry
1504: The next three are always the last link
1506: lnk[6] PETSC_MAX_INT - 1
1507: lnk[7] 1
1508: lnk[8] next valid link (this is the same as lnk[0] but without the decreases)
1509: */
1511: static inline PetscErrorCode PetscLLCondensedCreate_fast(PetscInt nlnk_max, PetscInt **lnk)
1512: {
1513: PetscInt *llnk;
1514: PetscInt lsize = 0;
1516: PetscFunctionBegin;
1517: PetscCall(PetscIntMultError(3, nlnk_max + 3, &lsize));
1518: PetscCall(PetscMalloc1(lsize, lnk));
1519: llnk = *lnk;
1520: llnk[0] = 0; /* nlnk: number of entries on the list */
1521: llnk[1] = 0; /* number of integer entries represented in list */
1522: llnk[3] = PETSC_MIN_INT + 1; /* value in the first node */
1523: llnk[4] = 1; /* count for the first node */
1524: llnk[5] = 6; /* next for the first node */
1525: llnk[6] = PETSC_MAX_INT - 1; /* value in the last node */
1526: llnk[7] = 1; /* count for the last node */
1527: llnk[8] = 0; /* next valid node to be used */
1528: PetscFunctionReturn(PETSC_SUCCESS);
1529: }
1531: static inline PetscErrorCode PetscLLCondensedAddSorted_fast(PetscInt nidx, const PetscInt indices[], PetscInt lnk[])
1532: {
1533: for (PetscInt k = 0, prev = 3 /* first value */; k < nidx; k++) {
1534: const PetscInt entry = indices[k];
1535: PetscInt next = lnk[prev + 2];
1537: /* search for insertion location */
1538: while (entry >= lnk[next]) {
1539: prev = next;
1540: next = lnk[next + 2];
1541: }
1542: /* entry is in range of previous list */
1543: if (entry < lnk[prev] + lnk[prev + 1]) continue;
1544: lnk[1]++;
1545: /* entry is right after previous list */
1546: if (entry == lnk[prev] + lnk[prev + 1]) {
1547: lnk[prev + 1]++;
1548: if (lnk[next] == entry + 1) { /* combine two contiguous strings */
1549: lnk[prev + 1] += lnk[next + 1];
1550: lnk[prev + 2] = lnk[next + 2];
1551: next = lnk[next + 2];
1552: lnk[0]--;
1553: }
1554: continue;
1555: }
1556: /* entry is right before next list */
1557: if (entry == lnk[next] - 1) {
1558: lnk[next]--;
1559: lnk[next + 1]++;
1560: prev = next;
1561: next = lnk[prev + 2];
1562: continue;
1563: }
1564: /* add entry into lnk */
1565: lnk[prev + 2] = 3 * ((lnk[8]++) + 3); /* connect previous node to the new node */
1566: prev = lnk[prev + 2];
1567: lnk[prev] = entry; /* set value of the new node */
1568: lnk[prev + 1] = 1; /* number of values in contiguous string is one to start */
1569: lnk[prev + 2] = next; /* connect new node to next node */
1570: lnk[0]++;
1571: }
1572: return PETSC_SUCCESS;
1573: }
1575: static inline PetscErrorCode PetscLLCondensedClean_fast(PETSC_UNUSED PetscInt nidx, PetscInt *indices, PetscInt *lnk)
1576: {
1577: const PetscInt nlnk = lnk[0];
1578: PetscInt next = lnk[5]; /* first node */
1580: for (PetscInt k = 0, cnt = 0; k < nlnk; k++) {
1581: for (PetscInt j = 0; j < lnk[next + 1]; j++) indices[cnt++] = lnk[next] + j;
1582: next = lnk[next + 2];
1583: }
1584: lnk[0] = 0; /* nlnk: number of links */
1585: lnk[1] = 0; /* number of integer entries represented in list */
1586: lnk[3] = PETSC_MIN_INT + 1; /* value in the first node */
1587: lnk[4] = 1; /* count for the first node */
1588: lnk[5] = 6; /* next for the first node */
1589: lnk[6] = PETSC_MAX_INT - 1; /* value in the last node */
1590: lnk[7] = 1; /* count for the last node */
1591: lnk[8] = 0; /* next valid location to make link */
1592: return PETSC_SUCCESS;
1593: }
1595: static inline PetscErrorCode PetscLLCondensedView_fast(const PetscInt *lnk)
1596: {
1597: const PetscInt nlnk = lnk[0];
1598: PetscInt next = lnk[5]; /* first node */
1600: for (PetscInt k = 0; k < nlnk; k++) {
1601: #if 0 /* Debugging code */
1602: printf("%d value %d len %d next %d\n", next, lnk[next], lnk[next + 1], lnk[next + 2]);
1603: #endif
1604: next = lnk[next + 2];
1605: }
1606: return PETSC_SUCCESS;
1607: }
1609: static inline PetscErrorCode PetscLLCondensedDestroy_fast(PetscInt *lnk)
1610: {
1611: return PetscFree(lnk);
1612: }
1614: /* this is extern because it is used in MatFDColoringUseDM() which is in the DM library */
1615: PETSC_EXTERN PetscErrorCode MatFDColoringApply_AIJ(Mat, MatFDColoring, Vec, void *);
1617: PETSC_EXTERN PetscLogEvent MAT_Mult;
1618: PETSC_EXTERN PetscLogEvent MAT_MultAdd;
1619: PETSC_EXTERN PetscLogEvent MAT_MultTranspose;
1620: PETSC_EXTERN PetscLogEvent MAT_MultHermitianTranspose;
1621: PETSC_EXTERN PetscLogEvent MAT_MultTransposeAdd;
1622: PETSC_EXTERN PetscLogEvent MAT_MultHermitianTransposeAdd;
1623: PETSC_EXTERN PetscLogEvent MAT_Solve;
1624: PETSC_EXTERN PetscLogEvent MAT_Solves;
1625: PETSC_EXTERN PetscLogEvent MAT_SolveAdd;
1626: PETSC_EXTERN PetscLogEvent MAT_SolveTranspose;
1627: PETSC_EXTERN PetscLogEvent MAT_SolveTransposeAdd;
1628: PETSC_EXTERN PetscLogEvent MAT_SOR;
1629: PETSC_EXTERN PetscLogEvent MAT_ForwardSolve;
1630: PETSC_EXTERN PetscLogEvent MAT_BackwardSolve;
1631: PETSC_EXTERN PetscLogEvent MAT_LUFactor;
1632: PETSC_EXTERN PetscLogEvent MAT_LUFactorSymbolic;
1633: PETSC_EXTERN PetscLogEvent MAT_LUFactorNumeric;
1634: PETSC_EXTERN PetscLogEvent MAT_QRFactor;
1635: PETSC_EXTERN PetscLogEvent MAT_QRFactorSymbolic;
1636: PETSC_EXTERN PetscLogEvent MAT_QRFactorNumeric;
1637: PETSC_EXTERN PetscLogEvent MAT_CholeskyFactor;
1638: PETSC_EXTERN PetscLogEvent MAT_CholeskyFactorSymbolic;
1639: PETSC_EXTERN PetscLogEvent MAT_CholeskyFactorNumeric;
1640: PETSC_EXTERN PetscLogEvent MAT_ILUFactor;
1641: PETSC_EXTERN PetscLogEvent MAT_ILUFactorSymbolic;
1642: PETSC_EXTERN PetscLogEvent MAT_ICCFactorSymbolic;
1643: PETSC_EXTERN PetscLogEvent MAT_Copy;
1644: PETSC_EXTERN PetscLogEvent MAT_Convert;
1645: PETSC_EXTERN PetscLogEvent MAT_Scale;
1646: PETSC_EXTERN PetscLogEvent MAT_AssemblyBegin;
1647: PETSC_EXTERN PetscLogEvent MAT_AssemblyEnd;
1648: PETSC_EXTERN PetscLogEvent MAT_SetValues;
1649: PETSC_EXTERN PetscLogEvent MAT_GetValues;
1650: PETSC_EXTERN PetscLogEvent MAT_GetRow;
1651: PETSC_EXTERN PetscLogEvent MAT_GetRowIJ;
1652: PETSC_EXTERN PetscLogEvent MAT_CreateSubMats;
1653: PETSC_EXTERN PetscLogEvent MAT_GetOrdering;
1654: PETSC_EXTERN PetscLogEvent MAT_RedundantMat;
1655: PETSC_EXTERN PetscLogEvent MAT_IncreaseOverlap;
1656: PETSC_EXTERN PetscLogEvent MAT_Partitioning;
1657: PETSC_EXTERN PetscLogEvent MAT_PartitioningND;
1658: PETSC_EXTERN PetscLogEvent MAT_Coarsen;
1659: PETSC_EXTERN PetscLogEvent MAT_ZeroEntries;
1660: PETSC_EXTERN PetscLogEvent MAT_Load;
1661: PETSC_EXTERN PetscLogEvent MAT_View;
1662: PETSC_EXTERN PetscLogEvent MAT_AXPY;
1663: PETSC_EXTERN PetscLogEvent MAT_FDColoringCreate;
1664: PETSC_EXTERN PetscLogEvent MAT_TransposeColoringCreate;
1665: PETSC_EXTERN PetscLogEvent MAT_FDColoringSetUp;
1666: PETSC_EXTERN PetscLogEvent MAT_FDColoringApply;
1667: PETSC_EXTERN PetscLogEvent MAT_Transpose;
1668: PETSC_EXTERN PetscLogEvent MAT_FDColoringFunction;
1669: PETSC_EXTERN PetscLogEvent MAT_CreateSubMat;
1670: PETSC_EXTERN PetscLogEvent MAT_MatSolve;
1671: PETSC_EXTERN PetscLogEvent MAT_MatTrSolve;
1672: PETSC_EXTERN PetscLogEvent MAT_MatMultSymbolic;
1673: PETSC_EXTERN PetscLogEvent MAT_MatMultNumeric;
1674: PETSC_EXTERN PetscLogEvent MAT_Getlocalmatcondensed;
1675: PETSC_EXTERN PetscLogEvent MAT_GetBrowsOfAcols;
1676: PETSC_EXTERN PetscLogEvent MAT_GetBrowsOfAocols;
1677: PETSC_EXTERN PetscLogEvent MAT_PtAPSymbolic;
1678: PETSC_EXTERN PetscLogEvent MAT_PtAPNumeric;
1679: PETSC_EXTERN PetscLogEvent MAT_Seqstompinum;
1680: PETSC_EXTERN PetscLogEvent MAT_Seqstompisym;
1681: PETSC_EXTERN PetscLogEvent MAT_Seqstompi;
1682: PETSC_EXTERN PetscLogEvent MAT_Getlocalmat;
1683: PETSC_EXTERN PetscLogEvent MAT_RARtSymbolic;
1684: PETSC_EXTERN PetscLogEvent MAT_RARtNumeric;
1685: PETSC_EXTERN PetscLogEvent MAT_MatTransposeMultSymbolic;
1686: PETSC_EXTERN PetscLogEvent MAT_MatTransposeMultNumeric;
1687: PETSC_EXTERN PetscLogEvent MAT_TransposeMatMultSymbolic;
1688: PETSC_EXTERN PetscLogEvent MAT_TransposeMatMultNumeric;
1689: PETSC_EXTERN PetscLogEvent MAT_MatMatMultSymbolic;
1690: PETSC_EXTERN PetscLogEvent MAT_MatMatMultNumeric;
1691: PETSC_EXTERN PetscLogEvent MAT_Getsymtransreduced;
1692: PETSC_EXTERN PetscLogEvent MAT_GetSeqNonzeroStructure;
1693: PETSC_EXTERN PetscLogEvent MATMFFD_Mult;
1694: PETSC_EXTERN PetscLogEvent MAT_GetMultiProcBlock;
1695: PETSC_EXTERN PetscLogEvent MAT_CUSPARSECopyToGPU;
1696: PETSC_EXTERN PetscLogEvent MAT_CUSPARSECopyFromGPU;
1697: PETSC_EXTERN PetscLogEvent MAT_CUSPARSEGenerateTranspose;
1698: PETSC_EXTERN PetscLogEvent MAT_CUSPARSESolveAnalysis;
1699: PETSC_EXTERN PetscLogEvent MAT_HIPSPARSECopyToGPU;
1700: PETSC_EXTERN PetscLogEvent MAT_HIPSPARSECopyFromGPU;
1701: PETSC_EXTERN PetscLogEvent MAT_HIPSPARSEGenerateTranspose;
1702: PETSC_EXTERN PetscLogEvent MAT_HIPSPARSESolveAnalysis;
1703: PETSC_EXTERN PetscLogEvent MAT_SetValuesBatch;
1704: PETSC_EXTERN PetscLogEvent MAT_ViennaCLCopyToGPU;
1705: PETSC_EXTERN PetscLogEvent MAT_DenseCopyToGPU;
1706: PETSC_EXTERN PetscLogEvent MAT_DenseCopyFromGPU;
1707: PETSC_EXTERN PetscLogEvent MAT_Merge;
1708: PETSC_EXTERN PetscLogEvent MAT_Residual;
1709: PETSC_EXTERN PetscLogEvent MAT_SetRandom;
1710: PETSC_EXTERN PetscLogEvent MAT_FactorFactS;
1711: PETSC_EXTERN PetscLogEvent MAT_FactorInvS;
1712: PETSC_EXTERN PetscLogEvent MAT_PreallCOO;
1713: PETSC_EXTERN PetscLogEvent MAT_SetVCOO;
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;
1720: PETSC_EXTERN PetscLogEvent MAT_H2Opus_Build;
1721: PETSC_EXTERN PetscLogEvent MAT_H2Opus_Compress;
1722: PETSC_EXTERN PetscLogEvent MAT_H2Opus_Orthog;
1723: PETSC_EXTERN PetscLogEvent MAT_H2Opus_LR;
1724: PETSC_EXTERN PetscLogEvent MAT_CUDACopyToGPU;