Actual source code: mpiaijAssemble.cu

petsc-3.5.4 2015-05-23
Report Typos and Errors
  1: #include <petscconf.h>
  2: PETSC_CUDA_EXTERN_C_BEGIN
  3: #include <../src/mat/impls/aij/seq/aij.h>          /*I "petscmat.h" I*/
  4: #include <../src/mat/impls/aij/mpi/mpiaij.h>
  5: #include <petscbt.h>
  6: #include <../src/vec/vec/impls/dvecimpl.h>
  7: #include <petsc-private/vecimpl.h>
  8: PETSC_CUDA_EXTERN_C_END
  9: #undef VecType
 10: #include <../src/mat/impls/aij/seq/seqcusp/cuspmatimpl.h>

 12: #include <thrust/reduce.h>
 13: #include <thrust/inner_product.h>

 15: #include <cusp/array1d.h>
 16: #include <cusp/print.h>
 17: #include <cusp/coo_matrix.h>

 19: #include <cusp/io/matrix_market.h>

 21: #include <thrust/iterator/counting_iterator.h>
 22: #include <thrust/iterator/transform_iterator.h>
 23: #include <thrust/iterator/permutation_iterator.h>
 24: #include <thrust/functional.h>
 25: #include <thrust/partition.h>
 26: #include <thrust/remove.h>

 28: // this example illustrates how to make repeated access to a range of values
 29: // examples:
 30: //   repeated_range([0, 1, 2, 3], 1) -> [0, 1, 2, 3]
 31: //   repeated_range([0, 1, 2, 3], 2) -> [0, 0, 1, 1, 2, 2, 3, 3]
 32: //   repeated_range([0, 1, 2, 3], 3) -> [0, 0, 0, 1, 1, 1, 2, 2, 2, 3, 3, 3]
 33: //   ...

 35: template <typename Iterator>
 36: class repeated_range
 37: {
 38: public:

 40:   typedef typename thrust::iterator_difference<Iterator>::type difference_type;

 42:   struct repeat_functor : public thrust::unary_function<difference_type,difference_type>
 43:   {
 44:     difference_type repeats;

 46:     repeat_functor(difference_type repeats) : repeats(repeats) {}

 48:     __host__ __device__
 49:     difference_type operator()(const difference_type &i) const
 50:     {
 51:       return i / repeats;
 52:     }
 53:   };

 55:   typedef typename thrust::counting_iterator<difference_type>                   CountingIterator;
 56:   typedef typename thrust::transform_iterator<repeat_functor, CountingIterator> TransformIterator;
 57:   typedef typename thrust::permutation_iterator<Iterator,TransformIterator>     PermutationIterator;

 59:   // type of the repeated_range iterator
 60:   typedef PermutationIterator iterator;

 62:   // construct repeated_range for the range [first,last)
 63:   repeated_range(Iterator first, Iterator last, difference_type repeats)
 64:     : first(first), last(last), repeats(repeats) {}

 66:   iterator begin(void) const
 67:   {
 68:     return PermutationIterator(first, TransformIterator(CountingIterator(0), repeat_functor(repeats)));
 69:   }

 71:   iterator end(void) const
 72:   {
 73:     return begin() + repeats * (last - first);
 74:   }

 76: protected:
 77:   difference_type repeats;
 78:   Iterator        first;
 79:   Iterator        last;

 81: };

 83: // this example illustrates how to repeat blocks in a range multiple times
 84: // examples:
 85: //   tiled_range([0, 1, 2, 3], 2)    -> [0, 1, 2, 3, 0, 1, 2, 3]
 86: //   tiled_range([0, 1, 2, 3], 4, 2) -> [0, 1, 2, 3, 0, 1, 2, 3]
 87: //   tiled_range([0, 1, 2, 3], 2, 2) -> [0, 1, 0, 1, 2, 3, 2, 3]
 88: //   tiled_range([0, 1, 2, 3], 2, 3) -> [0, 1, 0, 1 0, 1, 2, 3, 2, 3, 2, 3]
 89: //   ...

 91: template <typename Iterator>
 92: class tiled_range
 93: {
 94: public:

 96:   typedef typename thrust::iterator_difference<Iterator>::type difference_type;

 98:   struct tile_functor : public thrust::unary_function<difference_type,difference_type>
 99:   {
100:     difference_type repeats;
101:     difference_type tile_size;

103:     tile_functor(difference_type repeats, difference_type tile_size)
104:       : tile_size(tile_size), repeats(repeats) {}

106:     __host__ __device__
107:     difference_type operator() (const difference_type &i) const
108:     {
109:       return tile_size * (i / (tile_size * repeats)) + i % tile_size;
110:     }
111:   };

113:   typedef typename thrust::counting_iterator<difference_type>                   CountingIterator;
114:   typedef typename thrust::transform_iterator<tile_functor, CountingIterator>   TransformIterator;
115:   typedef typename thrust::permutation_iterator<Iterator,TransformIterator>     PermutationIterator;

117:   // type of the tiled_range iterator
118:   typedef PermutationIterator iterator;

120:   // construct repeated_range for the range [first,last)
121:   tiled_range(Iterator first, Iterator last, difference_type repeats)
122:     : first(first), last(last), repeats(repeats), tile_size(last - first) {}

124:   tiled_range(Iterator first, Iterator last, difference_type repeats, difference_type tile_size)
125:     : first(first), last(last), repeats(repeats), tile_size(tile_size)
126:   {
127:     // ASSERT((last - first) % tile_size == 0)
128:   }

130:   iterator begin(void) const
131:   {
132:     return PermutationIterator(first, TransformIterator(CountingIterator(0), tile_functor(repeats, tile_size)));
133:   }

135:   iterator end(void) const
136:   {
137:     return begin() + repeats * (last - first);
138:   }

140: protected:
141:   difference_type repeats;
142:   difference_type tile_size;
143:   Iterator        first;
144:   Iterator        last;
145: };

147: typedef cusp::device_memory memSpace;
148: typedef int IndexType;
149: typedef PetscScalar ValueType;
150: typedef cusp::array1d<IndexType, memSpace> IndexArray;
151: typedef cusp::array1d<ValueType, memSpace> ValueArray;
152: typedef cusp::array1d<IndexType, cusp::host_memory> IndexHostArray;
153: typedef IndexArray::iterator IndexArrayIterator;
154: typedef ValueArray::iterator ValueArrayIterator;

156: struct is_diag
157: {
158:   IndexType first, last;

160:   is_diag(IndexType first, IndexType last) : first(first), last(last) {}

162:   template <typename Tuple>
163:   __host__ __device__
164:   bool operator()(Tuple t)
165:   {
166:     // Check column
167:     IndexType row = thrust::get<0>(t);
168:     IndexType col = thrust::get<1>(t);
169:     return (row >= first) && (row < last) && (col >= first) && (col < last);
170:   }
171: };

173: struct is_nonlocal
174: {
175:   IndexType first, last;

177:   is_nonlocal(IndexType first, IndexType last) : first(first), last(last) {}

179:   template <typename Tuple>
180:   __host__ __device__
181:   bool operator() (Tuple t)
182:   {
183:     // Check column
184:     IndexType row = thrust::get<0>(t);
185:     return (row < first) || (row >= last);
186:   }
187: };

189: /*@C
190:   MatMPIAIJSetValuesBatch - Set multiple blocks of values into a matrix

192:   Not collective

194:   Input Parameters:
195: + J  - the assembled Mat object
196: . Ne -  the number of blocks (elements)
197: . Nl -  the block size (number of dof per element)
198: . elemRows - List of block row indices, in bunches of length Nl
199: - elemMats - List of block values, in bunches of Nl*Nl

201:   Level: advanced

203: .seealso MatSetValues()
204: @*/
207: PetscErrorCode MatSetValuesBatch_MPIAIJCUSP(Mat J, PetscInt Ne, PetscInt Nl, PetscInt *elemRows, const PetscScalar *elemMats)
208: {
209:   // Assumptions:
210:   //   1) Each elemMat is square, of size Nl x Nl
211:   //
212:   //      This means that any nonlocal entry (i,j) where i is owned by another process is matched to
213:   //        a) an offdiagonal entry (j,i) if j is diagonal, or
214:   //        b) another nonlocal entry (j,i) if j is offdiagonal
215:   //
216:   //      No - numSendEntries: The number of on-process  diagonal+offdiagonal entries
217:   //      numRecvEntries:      The number of off-process diagonal+offdiagonal entries
218:   //
219:   //  Glossary:
220:   //     diagonal: (i,j) such that i,j in [firstRow, lastRow)
221:   //  offdiagonal: (i,j) such that i in [firstRow, lastRow), and j not in [firstRow, lastRow)
222:   //     nonlocal: (i,j) such that i not in [firstRow, lastRow)
223:   //  nondiagonal: (i,j) such that i not in [firstRow, lastRow), or j not in [firstRow, lastRow)
224:   //   on-process: entries provided by elemMats
225:   //  off-process: entries received from other processes
226:   MPI_Comm       comm;
227:   Mat_MPIAIJ     *j   = (Mat_MPIAIJ*) J->data;
228:   size_t         N    = Ne * Nl;     // Length of elemRows (dimension of unassembled space)
229:   size_t         No   = Ne * Nl*Nl;  // Length of elemMats (total number of values)
230:   PetscInt       Nr;                 // Size of J          (dimension of assembled space)
231:   PetscInt       firstRow, lastRow, firstCol;
232:   const PetscInt *rowRanges;
233:   PetscInt       numNonlocalRows;    // Number of rows in elemRows not owned by this process
234:   PetscInt       numSendEntries;     // Number of (i,j,v) entries sent to other processes
235:   PetscInt       numRecvEntries;     // Number of (i,j,v) entries received from other processes
236:   PetscInt       Nc;
237:   PetscMPIInt    numProcs, rank;

240:   // copy elemRows and elemMat to device
241:   IndexArray d_elemRows(elemRows, elemRows + N);
242:   ValueArray d_elemMats(elemMats, elemMats + No);

245:   PetscObjectGetComm((PetscObject)J,&comm);
246:   MPI_Comm_size(comm, &numProcs);
247:   MPI_Comm_rank(comm, &rank);
248:   // get matrix information
249:   MatGetLocalSize(J, &Nr, NULL);
250:   MatGetOwnershipRange(J, &firstRow, &lastRow);
251:   MatGetOwnershipRanges(J, &rowRanges);
252:   MatGetOwnershipRangeColumn(J, &firstCol, NULL);
253:   PetscInfo3(J, "Assembling matrix of size %d (rows %d -- %d)\n", Nr, firstRow, lastRow);

255:   // repeat elemRows entries Nl times
256:   PetscInfo(J, "Making row indices\n");
257:   repeated_range<IndexArrayIterator> rowInd(d_elemRows.begin(), d_elemRows.end(), Nl);

259:   // tile rows of elemRows Nl times
260:   PetscInfo(J, "Making column indices\n");
261:   tiled_range<IndexArrayIterator> colInd(d_elemRows.begin(), d_elemRows.end(), Nl, Nl);

263:   // Find number of nonlocal rows, convert nonlocal rows to procs, and send sizes of off-proc entries (could send diag and offdiag sizes)
264:   // TODO: Ask Nathan how to do this on GPU
265:   PetscLogEventBegin(MAT_SetValuesBatchI,0,0,0,0);
266:   PetscMPIInt *procSendSizes, *procRecvSizes;

268:   PetscCalloc2(numProcs, &procSendSizes, numProcs, &procRecvSizes);

270:   numNonlocalRows = 0;
271:   for (size_t i = 0; i < N; ++i) {
272:     const PetscInt row = elemRows[i];

274:     if ((row < firstRow) || (row >= lastRow)) {
275:       numNonlocalRows++;
276:       for (IndexType p = 0; p < numProcs; ++p) {
277:         if ((row >= rowRanges[p]) && (row < rowRanges[p+1])) {
278:           procSendSizes[p] += Nl;
279:           break;
280:         }
281:       }
282:     }
283:   }
284:   numSendEntries = numNonlocalRows*Nl;

286:   PetscInfo2(J, "Nonlocal rows %d total entries %d\n", numNonlocalRows, No);
287:   MPI_Alltoall(procSendSizes, 1, MPIU_INT, procRecvSizes, 1, MPIU_INT, comm);

289:   numRecvEntries = 0;
290:   for (PetscInt p = 0; p < numProcs; ++p) numRecvEntries += procRecvSizes[p];
291:   PetscInfo2(j->A, "Send entries %d Recv Entries %d\n", numSendEntries, numRecvEntries);
292:   PetscLogEventEnd(MAT_SetValuesBatchI,0,0,0,0);
293:   // Allocate storage for "fat" COO representation of matrix
294:   PetscLogEventBegin(MAT_SetValuesBatchII,0,0,0,0);
295:   PetscInfo2(j->A, "Making COO matrices, diag entries %d, nondiag entries %d\n", No-numSendEntries+numRecvEntries, numSendEntries*2);
296:   cusp::coo_matrix<IndexType,ValueType, memSpace> diagCOO(Nr, Nr, No-numSendEntries+numRecvEntries); // ALLOC: This is oversized because I also count offdiagonal entries
297:   IndexArray nondiagonalRows(numSendEntries+numSendEntries); // ALLOC: This is oversized because numSendEntries > on-process offdiagonal entries
298:   IndexArray nondiagonalCols(numSendEntries+numSendEntries); // ALLOC: This is oversized because numSendEntries > on-process offdiagonal entries
299:   ValueArray nondiagonalVals(numSendEntries+numSendEntries); // ALLOC: This is oversized because numSendEntries > on-process offdiagonal entries
300:   IndexArray nonlocalRows(numSendEntries);
301:   IndexArray nonlocalCols(numSendEntries);
302:   ValueArray nonlocalVals(numSendEntries);
303:   // partition on-process entries into diagonal and off-diagonal+nonlocal
304:   PetscInfo(J, "Splitting on-process entries into diagonal and off-diagonal+nonlocal\n");
305:   thrust::fill(diagCOO.row_indices.begin(), diagCOO.row_indices.end(), -1);
306:   thrust::fill(nondiagonalRows.begin(),     nondiagonalRows.end(),     -1);
307:   thrust::partition_copy(thrust::make_zip_iterator(thrust::make_tuple(rowInd.begin(), colInd.begin(), d_elemMats.begin())),
308:                          thrust::make_zip_iterator(thrust::make_tuple(rowInd.end(),   colInd.end(),   d_elemMats.end())),
309:                          thrust::make_zip_iterator(thrust::make_tuple(diagCOO.row_indices.begin(),    diagCOO.column_indices.begin(), diagCOO.values.begin())),
310:                          thrust::make_zip_iterator(thrust::make_tuple(nondiagonalRows.begin(),        nondiagonalCols.begin(),        nondiagonalVals.begin())),
311:                          is_diag(firstRow, lastRow));
312:   // Current size without off-proc entries
313:   PetscInt diagonalSize    = (diagCOO.row_indices.end() - diagCOO.row_indices.begin()) - thrust::count(diagCOO.row_indices.begin(), diagCOO.row_indices.end(), -1);
314:   PetscInt nondiagonalSize = No - diagonalSize;
315:   PetscInfo2(j->A, "Diagonal size %d Nondiagonal size %d\n", diagonalSize, nondiagonalSize);
316:   ///cusp::print(diagCOO);
317:   ///cusp::print(nondiagonalRows);
318:   // partition on-process entries again into off-diagonal and nonlocal
319:   PetscInfo(J, "Splitting on-process entries into off-diagonal and nonlocal\n");
320:   thrust::stable_partition(thrust::make_zip_iterator(thrust::make_tuple(nondiagonalRows.begin(), nondiagonalCols.begin(), nondiagonalVals.begin())),
321:                            thrust::make_zip_iterator(thrust::make_tuple(nondiagonalRows.end(),   nondiagonalCols.end(),   nondiagonalVals.end())),
322:                            is_nonlocal(firstRow, lastRow));
323:   PetscInt nonlocalSize    = numSendEntries;
324:   PetscInt offdiagonalSize = nondiagonalSize - nonlocalSize;
325:   PetscInfo2(j->A, "Nonlocal size %d Offdiagonal size %d\n", nonlocalSize, offdiagonalSize);
326:   PetscLogEventEnd(MAT_SetValuesBatchII,0,0,0,0);
327:   ///cusp::print(nondiagonalRows);
328:   // send off-proc entries (pack this up later)
329:   PetscLogEventBegin(MAT_SetValuesBatchIII,0,0,0,0);
330:   PetscMPIInt *procSendDispls, *procRecvDispls;
331:   PetscInt    *sendRows, *recvRows;
332:   PetscInt    *sendCols, *recvCols;
333:   PetscScalar *sendVals, *recvVals;

335:   PetscMalloc2(numProcs, &procSendDispls, numProcs, &procRecvDispls);
336:   PetscMalloc3(numSendEntries, &sendRows, numSendEntries, &sendCols, numSendEntries, &sendVals);
337:   PetscMalloc3(numRecvEntries, &recvRows, numRecvEntries, &recvCols, numRecvEntries, &recvVals);

339:   procSendDispls[0] = procRecvDispls[0] = 0;
340:   for (PetscInt p = 1; p < numProcs; ++p) {
341:     procSendDispls[p] = procSendDispls[p-1] + procSendSizes[p-1];
342:     procRecvDispls[p] = procRecvDispls[p-1] + procRecvSizes[p-1];
343:   }
344: #if 0
345:   thrust::copy(nondiagonalRows.begin(), nondiagonalRows.begin()+nonlocalSize, sendRows);
346:   thrust::copy(nondiagonalCols.begin(), nondiagonalCols.begin()+nonlocalSize, sendCols);
347:   thrust::copy(nondiagonalVals.begin(), nondiagonalVals.begin()+nonlocalSize, sendVals);
348: #else
349:   thrust::copy(thrust::make_zip_iterator(thrust::make_tuple(nondiagonalRows.begin(), nondiagonalCols.begin(), nondiagonalVals.begin())),
350:                thrust::make_zip_iterator(thrust::make_tuple(nondiagonalRows.begin(), nondiagonalCols.begin(), nondiagonalVals.begin()))+nonlocalSize,
351:                thrust::make_zip_iterator(thrust::make_tuple(sendRows,                sendCols,                sendVals)));
352: #endif
353:   //   could pack into a struct and unpack
354:   MPI_Alltoallv(sendRows, procSendSizes, procSendDispls, MPIU_INT,    recvRows, procRecvSizes, procRecvDispls, MPIU_INT,    comm);
355:   MPI_Alltoallv(sendCols, procSendSizes, procSendDispls, MPIU_INT,    recvCols, procRecvSizes, procRecvDispls, MPIU_INT,    comm);
356:   MPI_Alltoallv(sendVals, procSendSizes, procSendDispls, MPIU_SCALAR, recvVals, procRecvSizes, procRecvDispls, MPIU_SCALAR, comm);
357:   PetscFree2(procSendSizes, procRecvSizes);
358:   PetscFree2(procSendDispls, procRecvDispls);
359:   PetscFree3(sendRows, sendCols, sendVals);
360:   PetscLogEventEnd(MAT_SetValuesBatchIII,0,0,0,0);

362:   PetscLogEventBegin(MAT_SetValuesBatchIV,0,0,0,0);
363:   // Create off-diagonal matrix
364:   cusp::coo_matrix<IndexType,ValueType, memSpace> offdiagCOO(Nr, Nr, offdiagonalSize+numRecvEntries); // ALLOC: This is oversizes because we count diagonal entries in numRecvEntries
365:   // partition again into diagonal and off-diagonal
366:   IndexArray d_recvRows(recvRows, recvRows+numRecvEntries);
367:   IndexArray d_recvCols(recvCols, recvCols+numRecvEntries);
368:   ValueArray d_recvVals(recvVals, recvVals+numRecvEntries);
369: #if 0
370:   thrust::copy(nondiagonalRows.end()-offdiagonalSize, nondiagonalRows.end(), offdiagCOO.row_indices.begin());
371:   thrust::copy(nondiagonalCols.end()-offdiagonalSize, nondiagonalCols.end(), offdiagCOO.column_indices.begin());
372:   thrust::copy(nondiagonalVals.end()-offdiagonalSize, nondiagonalVals.end(), offdiagCOO.values.begin());
373: #else
374:   thrust::copy(thrust::make_zip_iterator(thrust::make_tuple(nondiagonalRows.end(),          nondiagonalCols.end(),             nondiagonalVals.end()))-offdiagonalSize,
375:                thrust::make_zip_iterator(thrust::make_tuple(nondiagonalRows.end(),          nondiagonalCols.end(),             nondiagonalVals.end())),
376:                thrust::make_zip_iterator(thrust::make_tuple(offdiagCOO.row_indices.begin(), offdiagCOO.column_indices.begin(), offdiagCOO.values.begin())));
377: #endif
378:   thrust::fill(diagCOO.row_indices.begin()+diagonalSize,       diagCOO.row_indices.end(),    -1);
379:   thrust::fill(offdiagCOO.row_indices.begin()+offdiagonalSize, offdiagCOO.row_indices.end(), -1);
380:   thrust::partition_copy(thrust::make_zip_iterator(thrust::make_tuple(d_recvRows.begin(), d_recvCols.begin(), d_recvVals.begin())),
381:                          thrust::make_zip_iterator(thrust::make_tuple(d_recvRows.end(),   d_recvCols.end(),   d_recvVals.end())),
382:                          thrust::make_zip_iterator(thrust::make_tuple(diagCOO.row_indices.begin()+diagonalSize, diagCOO.column_indices.begin()+diagonalSize, diagCOO.values.begin()+diagonalSize)),
383:                          thrust::make_zip_iterator(thrust::make_tuple(offdiagCOO.row_indices.begin()+offdiagonalSize, offdiagCOO.column_indices.begin()+offdiagonalSize, offdiagCOO.values.begin()+offdiagonalSize)),
384:                          is_diag(firstRow, lastRow));

386:   PetscFree3(recvRows, recvCols, recvVals);

388:   diagonalSize    = (diagCOO.row_indices.end()    - diagCOO.row_indices.begin())    - thrust::count(diagCOO.row_indices.begin(),    diagCOO.row_indices.end(),    -1);
389:   offdiagonalSize = (offdiagCOO.row_indices.end() - offdiagCOO.row_indices.begin()) - thrust::count(offdiagCOO.row_indices.begin(), offdiagCOO.row_indices.end(), -1);

391:   // sort COO format by (i,j), this is the most costly step
392:   PetscInfo(J, "Sorting rows and columns\n");
393:   diagCOO.sort_by_row_and_column();
394:   offdiagCOO.sort_by_row_and_column();
395:   PetscInt diagonalOffset    = (diagCOO.row_indices.end()    - diagCOO.row_indices.begin())    - diagonalSize;
396:   PetscInt offdiagonalOffset = (offdiagCOO.row_indices.end() - offdiagCOO.row_indices.begin()) - offdiagonalSize;

398:   // print the "fat" COO representation
399:   if (PetscLogPrintInfo) {
400: #if !defined(PETSC_USE_COMPLEX)
401:     cusp::print(diagCOO);
402:     cusp::print(offdiagCOO);
403: #endif
404:   }

406:   // Figure out the number of unique off-diagonal columns
407:   Nc = thrust::inner_product(offdiagCOO.column_indices.begin()+offdiagonalOffset,
408:                              offdiagCOO.column_indices.end()   - 1,
409:                              offdiagCOO.column_indices.begin()+offdiagonalOffset + 1,
410:                              size_t(1), thrust::plus<size_t>(), thrust::not_equal_to<IndexType>());

412:   // compute number of unique (i,j) entries
413:   //   this counts the number of changes as we move along the (i,j) list
414:   PetscInfo(J, "Computing number of unique entries\n");
415:   size_t num_diag_entries = thrust::inner_product
416:                               (thrust::make_zip_iterator(thrust::make_tuple(diagCOO.row_indices.begin(), diagCOO.column_indices.begin())) + diagonalOffset,
417:                               thrust::make_zip_iterator(thrust::make_tuple(diagCOO.row_indices.end(),   diagCOO.column_indices.end())) - 1,
418:                               thrust::make_zip_iterator(thrust::make_tuple(diagCOO.row_indices.begin(), diagCOO.column_indices.begin())) + diagonalOffset + 1,
419:                               size_t(1),
420:                               thrust::plus<size_t>(),
421:                               thrust::not_equal_to< thrust::tuple<IndexType,IndexType> >());
422:   size_t num_offdiag_entries = thrust::inner_product
423:                                  (thrust::make_zip_iterator(thrust::make_tuple(offdiagCOO.row_indices.begin(), offdiagCOO.column_indices.begin())) + offdiagonalOffset,
424:                                  thrust::make_zip_iterator(thrust::make_tuple(offdiagCOO.row_indices.end(),   offdiagCOO.column_indices.end())) - 1,
425:                                  thrust::make_zip_iterator(thrust::make_tuple(offdiagCOO.row_indices.begin(), offdiagCOO.column_indices.begin())) + offdiagonalOffset + 1,
426:                                  size_t(1),
427:                                  thrust::plus<size_t>(),
428:                                  thrust::not_equal_to< thrust::tuple<IndexType,IndexType> >());

430:   // allocate COO storage for final matrix
431:   PetscInfo(J, "Allocating compressed matrices\n");
432:   cusp::coo_matrix<IndexType, ValueType, memSpace> A(Nr, Nr, num_diag_entries);
433:   cusp::coo_matrix<IndexType, ValueType, memSpace> B(Nr, Nc, num_offdiag_entries);

435:   // sum values with the same (i,j) index
436:   // XXX thrust::reduce_by_key is unoptimized right now, so we provide a SpMV-based one in cusp::detail
437:   //     the Cusp one is 2x faster, but still not optimal
438:   // This could possibly be done in-place
439:   PetscInfo(J, "Compressing matrices\n");
440:   thrust::reduce_by_key(thrust::make_zip_iterator(thrust::make_tuple(diagCOO.row_indices.begin(), diagCOO.column_indices.begin())) + diagonalOffset,
441:                         thrust::make_zip_iterator(thrust::make_tuple(diagCOO.row_indices.end(),   diagCOO.column_indices.end())),
442:                         diagCOO.values.begin() + diagonalOffset,
443:                         thrust::make_zip_iterator(thrust::make_tuple(A.row_indices.begin(), A.column_indices.begin())),
444:                         A.values.begin(),
445:                         thrust::equal_to< thrust::tuple<IndexType,IndexType> >(),
446:                         thrust::plus<ValueType>());

448:   thrust::reduce_by_key(thrust::make_zip_iterator(thrust::make_tuple(offdiagCOO.row_indices.begin(), offdiagCOO.column_indices.begin())) + offdiagonalOffset,
449:                         thrust::make_zip_iterator(thrust::make_tuple(offdiagCOO.row_indices.end(),   offdiagCOO.column_indices.end())),
450:                         offdiagCOO.values.begin() + offdiagonalOffset,
451:                         thrust::make_zip_iterator(thrust::make_tuple(B.row_indices.begin(), B.column_indices.begin())),
452:                         B.values.begin(),
453:                         thrust::equal_to< thrust::tuple<IndexType,IndexType> >(),
454:                         thrust::plus<ValueType>());

456:   // Convert row and column numbers
457:   if (firstRow) {
458:     thrust::transform(A.row_indices.begin(), A.row_indices.end(), thrust::make_constant_iterator(firstRow), A.row_indices.begin(), thrust::minus<IndexType>());
459:     thrust::transform(B.row_indices.begin(), B.row_indices.end(), thrust::make_constant_iterator(firstRow), B.row_indices.begin(), thrust::minus<IndexType>());
460:   }
461:   if (firstCol) {
462:     thrust::transform(A.column_indices.begin(), A.column_indices.end(), thrust::make_constant_iterator(firstCol), A.column_indices.begin(), thrust::minus<IndexType>());
463:   }
464: #if 0 // This is done by MatSetUpMultiply_MPIAIJ()
465:   //   TODO: Get better code from Nathan
466:   IndexArray d_colmap(Nc);
467:   thrust::unique_copy(B.column_indices.begin(), B.column_indices.end(), d_colmap.begin());
468:   IndexHostArray colmap(d_colmap.begin(), d_colmap.end());
469:   IndexType newCol = 0;
470:   for (IndexHostArray::iterator c_iter = colmap.begin(); c_iter != colmap.end(); ++c_iter, ++newCol) {
471:     thrust::replace(B.column_indices.begin(), B.column_indices.end(), *c_iter, newCol);
472:   }
473: #endif

475:   // print the final matrix
476:   if (PetscLogPrintInfo) {
477: #if !defined(PETSC_USE_COMPLEX)
478:     cusp::print(A);
479:     cusp::print(B);
480: #endif
481:   }

483:   PetscInfo(J, "Converting to PETSc matrix\n");
484:   MatSetType(J, MATMPIAIJCUSP);
485:   CUSPMATRIX *Agpu = new CUSPMATRIX;
486:   CUSPMATRIX *Bgpu = new CUSPMATRIX;
487:   cusp::convert(A, *Agpu);
488:   cusp::convert(B, *Bgpu);
489:   if (PetscLogPrintInfo) {
490: #if !defined(PETSC_USE_COMPLEX)
491:     cusp::print(*Agpu);
492:     cusp::print(*Bgpu);
493: #endif
494:   }
495:   {
496:     PetscInfo(J, "Copying to CPU matrix");
497:     MatCUSPCopyFromGPU(j->A, Agpu);
498:     MatCUSPCopyFromGPU(j->B, Bgpu);
499: #if 0 // This is done by MatSetUpMultiply_MPIAIJ()

501:     // Create the column map
502:     PetscFree(j->garray);
503:     PetscMalloc1(Nc, &j->garray);
504:     PetscInt c = 0;
505:     for (IndexHostArray::iterator c_iter = colmap.begin(); c_iter != colmap.end(); ++c_iter, ++c) {
506:       j->garray[c] = *c_iter;
507:     }
508: #endif
509:   }
510:   PetscLogEventEnd(MAT_SetValuesBatchIV,0,0,0,0);
511:   return(0);
512: }