allreduce.c 34.9 KB
Newer Older
1
/* -*- Mode: C; c-basic-offset:4 ; -*- */
2
/*
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
 *  (C) 2001 by Argonne National Laboratory.
 *      See COPYRIGHT in top-level directory.
 */

#include "mpiimpl.h"

/* -- Begin Profiling Symbol Block for routine MPI_Allreduce */
#if defined(HAVE_PRAGMA_WEAK)
#pragma weak MPI_Allreduce = PMPI_Allreduce
#elif defined(HAVE_PRAGMA_HP_SEC_DEF)
#pragma _HP_SECONDARY_DEF PMPI_Allreduce  MPI_Allreduce
#elif defined(HAVE_PRAGMA_CRI_DUP)
#pragma _CRI duplicate MPI_Allreduce as PMPI_Allreduce
#endif
/* -- End Profiling Symbol Block */

/* Define MPICH_MPI_FROM_PMPI if weak symbols are not supported to build
   the MPI routines */
#ifndef MPICH_MPI_FROM_PMPI
#undef MPI_Allreduce
#define MPI_Allreduce PMPI_Allreduce

25
26
/* The order of entries in this table must match the definitions in 
   mpi.h.in */
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
MPI_User_function *MPIR_Op_table[] = { MPIR_MAXF, MPIR_MINF, MPIR_SUM,
                                       MPIR_PROD, MPIR_LAND,
                                       MPIR_BAND, MPIR_LOR, MPIR_BOR,
                                       MPIR_LXOR, MPIR_BXOR,
                                       MPIR_MINLOC, MPIR_MAXLOC, };

MPIR_Op_check_dtype_fn *MPIR_Op_check_dtype_table[] = {
    MPIR_MAXF_check_dtype, MPIR_MINF_check_dtype,
    MPIR_SUM_check_dtype,
    MPIR_PROD_check_dtype, MPIR_LAND_check_dtype,
    MPIR_BAND_check_dtype, MPIR_LOR_check_dtype, MPIR_BOR_check_dtype,
    MPIR_LXOR_check_dtype, MPIR_BXOR_check_dtype,
    MPIR_MINLOC_check_dtype, MPIR_MAXLOC_check_dtype, }; 


/* This is the default implementation of allreduce. The algorithm is:
   
   Algorithm: MPI_Allreduce

   For the heterogeneous case, we call MPI_Reduce followed by MPI_Bcast
   in order to meet the requirement that all processes must have the
   same result. For the homogeneous case, we use the following algorithms.


   For long messages and for builtin ops and if count >= pof2 (where
   pof2 is the nearest power-of-two less than or equal to the number
   of processes), we use Rabenseifner's algorithm (see 
   http://www.hlrs.de/organization/par/services/models/mpi/myreduce.html ).
   This algorithm implements the allreduce in two steps: first a
   reduce-scatter, followed by an allgather. A recursive-halving
   algorithm (beginning with processes that are distance 1 apart) is
   used for the reduce-scatter, and a recursive doubling 
   algorithm is used for the allgather. The non-power-of-two case is
   handled by dropping to the nearest lower power-of-two: the first
   few even-numbered processes send their data to their right neighbors
   (rank+1), and the reduce-scatter and allgather happen among the remaining
   power-of-two processes. At the end, the first few even-numbered
   processes get the result from their right neighbors.

   For the power-of-two case, the cost for the reduce-scatter is 
   lgp.alpha + n.((p-1)/p).beta + n.((p-1)/p).gamma. The cost for the
   allgather lgp.alpha + n.((p-1)/p).beta. Therefore, the
   total cost is:
   Cost = 2.lgp.alpha + 2.n.((p-1)/p).beta + n.((p-1)/p).gamma

   For the non-power-of-two case, 
   Cost = (2.floor(lgp)+2).alpha + (2.((p-1)/p) + 2).n.beta + n.(1+(p-1)/p).gamma

   
76
77
78
79
80
81
82
   For short messages, for user-defined ops, and for count < pof2 
   we use a recursive doubling algorithm (similar to the one in
   MPI_Allgather). We use this algorithm in the case of user-defined ops
   because in this case derived datatypes are allowed, and the user
   could pass basic datatypes on one process and derived on another as
   long as the type maps are the same. Breaking up derived datatypes
   to do the reduce-scatter is tricky. 
83
84
85
86
87
88
89
90

   Cost = lgp.alpha + n.lgp.beta + n.lgp.gamma

   Possible improvements: 

   End Algorithm: MPI_Allreduce
*/

91
92
93
94
95
#undef FUNCNAME
#define FUNCNAME allreduce_intra_or_coll_fn
#undef FCNAME
#define FCNAME MPIU_QUOTE(FUNCNAME)
static inline int allreduce_intra_or_coll_fn(void *sendbuf, void *recvbuf, int count, MPI_Datatype datatype, MPI_Op op,
96
                                             MPID_Comm *comm_ptr, int *errflag)
97
98
99
100
{
    int mpi_errno = MPI_SUCCESS;

    if (comm_ptr->coll_fns != NULL && comm_ptr->coll_fns->Allreduce != NULL) {
101
	mpi_errno = comm_ptr->coll_fns->Allreduce(sendbuf, recvbuf, count, datatype, op, comm_ptr, errflag);
102
103
        if (mpi_errno) MPIU_ERR_POP(mpi_errno);
    } else {
104
        mpi_errno = MPIR_Allreduce_intra(sendbuf, recvbuf, count, datatype, op, comm_ptr, errflag);
105
106
107
108
109
110
111
112
113
        if (mpi_errno) MPIU_ERR_POP(mpi_errno);
    }
        
 fn_exit:
    return mpi_errno;
 fn_fail:
    goto fn_exit;
}

114
115
116

/* not declared static because a machine-specific function may call this one 
   in some cases */
117
118
119
120
121
#undef FUNCNAME
#define FUNCNAME MPIR_Allreduce_intra
#undef FCNAME
#define FCNAME MPIU_QUOTE(FUNCNAME)
int MPIR_Allreduce_intra ( 
122
123
124
125
126
    void *sendbuf, 
    void *recvbuf, 
    int count, 
    MPI_Datatype datatype, 
    MPI_Op op, 
127
128
    MPID_Comm *comm_ptr,
    int *errflag )
129
130
131
132
133
134
{
    int is_homogeneous;
#ifdef MPID_HAS_HETERO
    int rc;
#endif
    int        comm_size, rank, type_size;
135
136
    int mpi_errno = MPI_SUCCESS;
    int mpi_errno_ret = MPI_SUCCESS;
137
138
139
140
141
142
143
144
145
146
147
148
149
    int mask, dst, is_commutative, pof2, newrank, rem, newdst, i,
        send_idx, recv_idx, last_idx, send_cnt, recv_cnt, *cnts, *disps; 
    MPI_Aint true_extent, true_lb, extent;
    void *tmp_buf;
    MPI_User_function *uop;
    MPID_Op *op_ptr;
    MPI_Comm comm;
    MPIU_THREADPRIV_DECL;
#ifdef HAVE_CXX_BINDING
    int is_cxx_uop = 0;
#endif
    MPIU_CHKLMEM_DECL(3);
    
150
151
152
153
    /* check if multiple threads are calling this collective function */
    MPIDU_ERR_CHECK_MULTIPLE_THREADS_ENTER( comm_ptr );

    if (count == 0) goto fn_exit;
154
155
156
    comm = comm_ptr->handle;

    MPIU_THREADPRIV_GET;
157
158
159
160
161
162
163
164
165
166
167
168

#if defined(USE_SMP_COLLECTIVES)
    /* is the op commutative? We do SMP optimizations only if it is. */ 
    if (HANDLE_GET_KIND(op) == HANDLE_KIND_BUILTIN) {
        is_commutative = 1;
    } else {
        MPID_Op_get_ptr(op, op_ptr);
        is_commutative = (op_ptr->kind == MPID_OP_USER_NONCOMMUTE) ? 0 : 1;
    }

    if (MPIR_Comm_is_node_aware(comm_ptr) && is_commutative) {
        /* on each node, do a reduce to the local root */ 
169
        if (comm_ptr->node_comm != NULL) {
170
171
172
173
174
175
176
177
            /* take care of the MPI_IN_PLACE case. For reduce, 
               MPI_IN_PLACE is specified only on the root; 
               for allreduce it is specified on all processes. */

            if ((sendbuf == MPI_IN_PLACE) && (comm_ptr->node_comm->rank != 0)) {
                /* IN_PLACE and not root of reduce. Data supplied to this
                   allreduce is in recvbuf. Pass that as the sendbuf to reduce. */
			
178
                mpi_errno = MPIR_Reduce_impl(recvbuf, NULL, count, datatype, op, 0, comm_ptr->node_comm, errflag);
179
180
                if (mpi_errno) {
                    /* for communication errors, just record the error but continue */
181
                    *errflag = TRUE;
182
183
184
                    MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
                    MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
                }
185
            } else {
186
                mpi_errno = MPIR_Reduce_impl(sendbuf, recvbuf, count, datatype, op, 0, comm_ptr->node_comm, errflag);
187
188
                if (mpi_errno) {
                    /* for communication errors, just record the error but continue */
189
                    *errflag = TRUE;
190
191
192
                    MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
                    MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
                }
193
            }
194
        } else {
195
196
197
            /* only one process on the node. copy sendbuf to recvbuf */
            if (sendbuf != MPI_IN_PLACE) {
                mpi_errno = MPIR_Localcopy(sendbuf, count, datatype, recvbuf, count, datatype);
198
                if (mpi_errno) MPIU_ERR_POP(mpi_errno);
199
200
201
202
203
            }
        }

        /* now do an IN_PLACE allreduce among the local roots of all nodes */
        if (comm_ptr->node_roots_comm != NULL) {
204
205
            mpi_errno = allreduce_intra_or_coll_fn(MPI_IN_PLACE, recvbuf, count, datatype, op, comm_ptr->node_roots_comm,
                                                   errflag);
206
207
            if (mpi_errno) {
                /* for communication errors, just record the error but continue */
208
                *errflag = TRUE;
209
210
211
                MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
                MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
            }
212
213
214
215
        }

        /* now broadcast the result among local processes */
        if (comm_ptr->node_comm != NULL) {
216
            mpi_errno = MPIR_Bcast_impl(recvbuf, count, datatype, 0, comm_ptr->node_comm, errflag);
217
218
            if (mpi_errno) {
                /* for communication errors, just record the error but continue */
219
                *errflag = TRUE;
220
221
222
                MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
                MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
            }
223
224
225
226
227
        }
        goto fn_exit;
    }
#endif
            
228
229
230
231
232
233
234
235
236
237
238
    
    is_homogeneous = 1;
#ifdef MPID_HAS_HETERO
    if (comm_ptr->is_hetero)
        is_homogeneous = 0;
#endif
    
#ifdef MPID_HAS_HETERO
    if (!is_homogeneous) {
        /* heterogeneous. To get the same result on all processes, we
           do a reduce to 0 and then broadcast. */
239
        mpi_errno = MPIR_Reduce_impl ( sendbuf, recvbuf, count, datatype,
240
                                       op, 0, comm_ptr, errflag );
241
242
        if (mpi_errno) {
            /* for communication errors, just record the error but continue */
243
            *errflag = TRUE;
244
245
246
247
            MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
            MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
        }

248
        mpi_errno = MPIR_Bcast_impl( recvbuf, count, datatype, 0, comm_ptr, errflag );
249
250
        if (mpi_errno) {
            /* for communication errors, just record the error but continue */
251
            *errflag = TRUE;
252
253
            MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
            MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
        }
    }
    else 
#endif /* MPID_HAS_HETERO */
    {
        /* homogeneous */
        
        /* set op_errno to 0. stored in perthread structure */
        MPIU_THREADPRIV_FIELD(op_errno) = 0;

        comm_size = comm_ptr->local_size;
        rank = comm_ptr->rank;
        
        if (HANDLE_GET_KIND(op) == HANDLE_KIND_BUILTIN) {
            is_commutative = 1;
            /* get the function by indexing into the op table */
            uop = MPIR_Op_table[op%16 - 1];
        }
        else {
            MPID_Op_get_ptr(op, op_ptr);
            if (op_ptr->kind == MPID_OP_USER_NONCOMMUTE)
                is_commutative = 0;
            else
                is_commutative = 1;
#ifdef HAVE_CXX_BINDING            
            if (op_ptr->language == MPID_LANG_CXX) {
                uop = (MPI_User_function *) op_ptr->function.c_function;
		is_cxx_uop = 1;
	    }
	    else
#endif
            if ((op_ptr->language == MPID_LANG_C))
                uop = (MPI_User_function *) op_ptr->function.c_function;
            else
                uop = (MPI_User_function *) op_ptr->function.f77_function;
        }
        
        /* need to allocate temporary buffer to store incoming data*/
292
        MPIR_Type_get_true_extent_impl(datatype, &true_lb, &true_extent);
293
294
        MPID_Datatype_get_extent_macro(datatype, extent);

295
        MPID_Ensure_Aint_fits_in_pointer(count * MPIR_MAX(extent, true_extent));
296
297
298
299
300
301
302
303
304
        MPIU_CHKLMEM_MALLOC(tmp_buf, void *, count*(MPIR_MAX(extent,true_extent)), mpi_errno, "temporary buffer");
	
        /* adjust for potential negative lower bound in datatype */
        tmp_buf = (void *)((char*)tmp_buf - true_lb);
        
        /* copy local data into recvbuf */
        if (sendbuf != MPI_IN_PLACE) {
            mpi_errno = MPIR_Localcopy(sendbuf, count, datatype, recvbuf,
                                       count, datatype);
305
            if (mpi_errno) MPIU_ERR_POP(mpi_errno);
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
        }

        MPID_Datatype_get_size_macro(datatype, type_size);

        /* find nearest power-of-two less than or equal to comm_size */
        pof2 = 1;
        while (pof2 <= comm_size) pof2 <<= 1;
        pof2 >>=1;

        rem = comm_size - pof2;

        /* In the non-power-of-two case, all even-numbered
           processes of rank < 2*rem send their data to
           (rank+1). These even-numbered processes no longer
           participate in the algorithm until the very end. The
           remaining processes form a nice power-of-two. */
        
        if (rank < 2*rem) {
            if (rank % 2 == 0) { /* even */
325
326
327
                mpi_errno = MPIC_Send_ft(recvbuf, count, 
                                         datatype, rank+1,
                                         MPIR_ALLREDUCE_TAG, comm, errflag);
328
329
                if (mpi_errno) {
                    /* for communication errors, just record the error but continue */
330
                    *errflag = TRUE;
331
332
333
                    MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
                    MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
                }
334
335
336
337
338
339
340
                
                /* temporarily set the rank to -1 so that this
                   process does not pariticipate in recursive
                   doubling */
                newrank = -1; 
            }
            else { /* odd */
341
342
343
344
                mpi_errno = MPIC_Recv_ft(tmp_buf, count, 
                                         datatype, rank-1,
                                         MPIR_ALLREDUCE_TAG, comm,
                                         MPI_STATUS_IGNORE, errflag);
345
346
                if (mpi_errno) {
                    /* for communication errors, just record the error but continue */
347
                    *errflag = TRUE;
348
349
350
                    MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
                    MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
                }
351

352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
                /* do the reduction on received data. since the
                   ordering is right, it doesn't matter whether
                   the operation is commutative or not. */
#ifdef HAVE_CXX_BINDING
                if (is_cxx_uop) {
                    (*MPIR_Process.cxx_call_op_fn)( tmp_buf, recvbuf, 
                                                    count,
                                                    datatype,
                                                    uop ); 
                }
                else 
#endif
                    (*uop)(tmp_buf, recvbuf, &count, &datatype);
                
                /* change the rank */
                newrank = rank / 2;
            }
        }
        else  /* rank >= 2*rem */
            newrank = rank - rem;
        
373
374
375
376
377
378
379
380
381
        /* If op is user-defined or count is less than pof2, use
           recursive doubling algorithm. Otherwise do a reduce-scatter
           followed by allgather. (If op is user-defined,
           derived datatypes are allowed and the user could pass basic
           datatypes on one process and derived on another as long as
           the type maps are the same. Breaking up derived
           datatypes to do the reduce-scatter is tricky, therefore
           using recursive doubling in that case.) */

382
        if (newrank != -1) {
383
            if ((count*type_size <= MPIR_PARAM_ALLREDUCE_SHORT_MSG_SIZE) ||
384
                (HANDLE_GET_KIND(op) != HANDLE_KIND_BUILTIN) ||  
385
386
387
388
389
390
391
392
393
                (count < pof2)) { /* use recursive doubling */
                mask = 0x1;
                while (mask < pof2) {
                    newdst = newrank ^ mask;
                    /* find real rank of dest */
                    dst = (newdst < rem) ? newdst*2 + 1 : newdst + rem;

                    /* Send the most current data, which is in recvbuf. Recv
                       into tmp_buf */ 
394
395
396
397
398
                    mpi_errno = MPIC_Sendrecv_ft(recvbuf, count, datatype, 
                                                 dst, MPIR_ALLREDUCE_TAG, tmp_buf,
                                                 count, datatype, dst,
                                                 MPIR_ALLREDUCE_TAG, comm,
                                                 MPI_STATUS_IGNORE, errflag);
399
400
                    if (mpi_errno) {
                        /* for communication errors, just record the error but continue */
401
                        *errflag = TRUE;
402
403
404
                        MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
                        MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
                    }
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
                    
                    /* tmp_buf contains data received in this step.
                       recvbuf contains data accumulated so far */
                    
                    if (is_commutative  || (dst < rank)) {
                        /* op is commutative OR the order is already right */
#ifdef HAVE_CXX_BINDING
                        if (is_cxx_uop) {
                            (*MPIR_Process.cxx_call_op_fn)( tmp_buf, recvbuf, 
                                                            count,
                                                            datatype,
                                                            uop ); 
                        }
                        else 
#endif
                            (*uop)(tmp_buf, recvbuf, &count, &datatype);
                    }
                    else {
                        /* op is noncommutative and the order is not right */
#ifdef HAVE_CXX_BINDING
                        if (is_cxx_uop) {
                            (*MPIR_Process.cxx_call_op_fn)( recvbuf, tmp_buf, 
                                                            count,
                                                            datatype,
                                                            uop ); 
                        }
                        else 
#endif
                            (*uop)(recvbuf, tmp_buf, &count, &datatype);
                        
                        /* copy result back into recvbuf */
                        mpi_errno = MPIR_Localcopy(tmp_buf, count, datatype,
                                                   recvbuf, count, datatype);
438
                        if (mpi_errno) MPIU_ERR_POP(mpi_errno);
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
                    }
                    mask <<= 1;
                }
            }

            else {

                /* do a reduce-scatter followed by allgather */

                /* for the reduce-scatter, calculate the count that
                   each process receives and the displacement within
                   the buffer */

		MPIU_CHKLMEM_MALLOC(cnts, int *, pof2*sizeof(int), mpi_errno, "counts");
		MPIU_CHKLMEM_MALLOC(disps, int *, pof2*sizeof(int), mpi_errno, "displacements");

                for (i=0; i<(pof2-1); i++) 
                    cnts[i] = count/pof2;
                cnts[pof2-1] = count - (count/pof2)*(pof2-1);

                disps[0] = 0;
                for (i=1; i<pof2; i++)
                    disps[i] = disps[i-1] + cnts[i-1];

                mask = 0x1;
                send_idx = recv_idx = 0;
                last_idx = pof2;
                while (mask < pof2) {
                    newdst = newrank ^ mask;
                    /* find real rank of dest */
                    dst = (newdst < rem) ? newdst*2 + 1 : newdst + rem;

                    send_cnt = recv_cnt = 0;
                    if (newrank < newdst) {
                        send_idx = recv_idx + pof2/(mask*2);
                        for (i=send_idx; i<last_idx; i++)
                            send_cnt += cnts[i];
                        for (i=recv_idx; i<send_idx; i++)
                            recv_cnt += cnts[i];
                    }
                    else {
                        recv_idx = send_idx + pof2/(mask*2);
                        for (i=send_idx; i<recv_idx; i++)
                            send_cnt += cnts[i];
                        for (i=recv_idx; i<last_idx; i++)
                            recv_cnt += cnts[i];
                    }

/*                    printf("Rank %d, send_idx %d, recv_idx %d, send_cnt %d, recv_cnt %d, last_idx %d\n", newrank, send_idx, recv_idx,
                           send_cnt, recv_cnt, last_idx);
                           */
                    /* Send data from recvbuf. Recv into tmp_buf */ 
491
492
493
494
495
496
497
498
499
                    mpi_errno = MPIC_Sendrecv_ft((char *) recvbuf +
                                                 disps[send_idx]*extent,
                                                 send_cnt, datatype,  
                                                 dst, MPIR_ALLREDUCE_TAG, 
                                                 (char *) tmp_buf +
                                                 disps[recv_idx]*extent,
                                                 recv_cnt, datatype, dst,
                                                 MPIR_ALLREDUCE_TAG, comm,
                                                 MPI_STATUS_IGNORE, errflag);
500
501
                    if (mpi_errno) {
                        /* for communication errors, just record the error but continue */
502
                        *errflag = TRUE;
503
504
505
                        MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
                        MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
                    }
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
                    
                    /* tmp_buf contains data received in this step.
                       recvbuf contains data accumulated so far */
                    
                    /* This algorithm is used only for predefined ops
                       and predefined ops are always commutative. */

		    (*uop)((char *) tmp_buf + disps[recv_idx]*extent,
			   (char *) recvbuf + disps[recv_idx]*extent, 
			   &recv_cnt, &datatype);
                    
                    /* update send_idx for next iteration */
                    send_idx = recv_idx;
                    mask <<= 1;

                    /* update last_idx, but not in last iteration
                       because the value is needed in the allgather
                       step below. */
                    if (mask < pof2)
                        last_idx = recv_idx + pof2/mask;
                }

                /* now do the allgather */

                mask >>= 1;
                while (mask > 0) {
                    newdst = newrank ^ mask;
                    /* find real rank of dest */
                    dst = (newdst < rem) ? newdst*2 + 1 : newdst + rem;

                    send_cnt = recv_cnt = 0;
                    if (newrank < newdst) {
                        /* update last_idx except on first iteration */
                        if (mask != pof2/2)
                            last_idx = last_idx + pof2/(mask*2);

                        recv_idx = send_idx + pof2/(mask*2);
                        for (i=send_idx; i<recv_idx; i++)
                            send_cnt += cnts[i];
                        for (i=recv_idx; i<last_idx; i++)
                            recv_cnt += cnts[i];
                    }
                    else {
                        recv_idx = send_idx - pof2/(mask*2);
                        for (i=send_idx; i<last_idx; i++)
                            send_cnt += cnts[i];
                        for (i=recv_idx; i<send_idx; i++)
                            recv_cnt += cnts[i];
                    }

556
557
558
559
560
561
562
563
564
                    mpi_errno = MPIC_Sendrecv_ft((char *) recvbuf +
                                                 disps[send_idx]*extent,
                                                 send_cnt, datatype,  
                                                 dst, MPIR_ALLREDUCE_TAG, 
                                                 (char *) recvbuf +
                                                 disps[recv_idx]*extent,
                                                 recv_cnt, datatype, dst,
                                                 MPIR_ALLREDUCE_TAG, comm,
                                                 MPI_STATUS_IGNORE, errflag);
565
566
                    if (mpi_errno) {
                        /* for communication errors, just record the error but continue */
567
                        *errflag = TRUE;
568
569
570
                        MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
                        MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
                    }
571
572
573
574
575
576
577
578
579
580
581
582
583

                    if (newrank > newdst) send_idx = recv_idx;

                    mask >>= 1;
                }
            }
        }

        /* In the non-power-of-two case, all odd-numbered
           processes of rank < 2*rem send the result to
           (rank-1), the ranks who didn't participate above. */
        if (rank < 2*rem) {
            if (rank % 2)  /* odd */
584
585
586
                mpi_errno = MPIC_Send_ft(recvbuf, count, 
                                         datatype, rank-1,
                                         MPIR_ALLREDUCE_TAG, comm, errflag);
587
            else  /* even */
588
589
590
591
                mpi_errno = MPIC_Recv_ft(recvbuf, count,
                                         datatype, rank+1,
                                         MPIR_ALLREDUCE_TAG, comm,
                                         MPI_STATUS_IGNORE, errflag);
592
593
            if (mpi_errno) {
                /* for communication errors, just record the error but continue */
594
                *errflag = TRUE;
595
596
597
                MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
                MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
            }
598
599
600
601
602
603
604
        }

        if (MPIU_THREADPRIV_FIELD(op_errno)) 
	    mpi_errno = MPIU_THREADPRIV_FIELD(op_errno);
    }

  fn_exit:
605
606
607
    /* check if multiple threads are calling this collective function */
    MPIDU_ERR_CHECK_MULTIPLE_THREADS_EXIT( comm_ptr );

608
    MPIU_CHKLMEM_FREEALL();
609
610
    if (mpi_errno_ret)
        mpi_errno = mpi_errno_ret;
611
612
613
614
615
616
617
618
619
    return (mpi_errno);

  fn_fail:
    goto fn_exit;
}


/* not declared static because a machine-specific function may call this one 
   in some cases */
620
621
#undef FUNCNAME
#define FUNCNAME MPIR_Allreduce_inter
622
#undef FCNAME
623
#define FCNAME MPIU_QUOTE(FUNCNAME)
624
625
626
627
628
629
int MPIR_Allreduce_inter ( 
    void *sendbuf, 
    void *recvbuf, 
    int count, 
    MPI_Datatype datatype, 
    MPI_Op op, 
630
631
    MPID_Comm *comm_ptr,
    int *errflag )
632
633
634
635
636
637
638
639
640
641
{
/* Intercommunicator Allreduce.
   We first do an intercommunicator reduce to rank 0 on left group,
   then an intercommunicator reduce to rank 0 on right group, followed
   by local intracommunicator broadcasts in each group.

   We don't do local reduces first and then intercommunicator
   broadcasts because it would require allocation of a temporary buffer. 
*/
    int rank, mpi_errno, root;
642
    int mpi_errno_ret = MPI_SUCCESS;
643
644
645
646
647
648
649
650
651
652
    MPID_Comm *newcomm_ptr = NULL;
    
    rank = comm_ptr->rank;

    /* first do a reduce from right group to rank 0 in left group,
       then from left group to rank 0 in right group*/
    if (comm_ptr->is_low_group) {
        /* reduce from right group to rank 0*/
        root = (rank == 0) ? MPI_ROOT : MPI_PROC_NULL;
        mpi_errno = MPIR_Reduce_inter(sendbuf, recvbuf, count, datatype, op,
653
				      root, comm_ptr, errflag);
654
655
        if (mpi_errno) {
            /* for communication errors, just record the error but continue */
656
            *errflag = TRUE;
657
658
659
            MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
            MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
        }
660
661
662
663

        /* reduce to rank 0 of right group */
        root = 0;
        mpi_errno = MPIR_Reduce_inter(sendbuf, recvbuf, count, datatype, op,
664
				      root, comm_ptr, errflag);
665
666
        if (mpi_errno) {
            /* for communication errors, just record the error but continue */
667
            *errflag = TRUE;
668
669
670
            MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
            MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
        }
671
672
673
674
675
    }
    else {
        /* reduce to rank 0 of left group */
        root = 0;
        mpi_errno = MPIR_Reduce_inter(sendbuf, recvbuf, count, datatype, op,
676
				      root, comm_ptr, errflag);
677
678
        if (mpi_errno) {
            /* for communication errors, just record the error but continue */
679
            *errflag = TRUE;
680
681
682
            MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
            MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
        }
683
684
685
686

        /* reduce from right group to rank 0 */
        root = (rank == 0) ? MPI_ROOT : MPI_PROC_NULL;
        mpi_errno = MPIR_Reduce_inter(sendbuf, recvbuf, count, datatype, op,
687
				      root, comm_ptr, errflag);
688
689
        if (mpi_errno) {
            /* for communication errors, just record the error but continue */
690
            *errflag = TRUE;
691
692
693
            MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
            MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
        }
694
695
696
697
698
699
700
701
    }

    /* Get the local intracommunicator */
    if (!comm_ptr->local_comm)
	MPIR_Setup_intercomm_localcomm( comm_ptr );

    newcomm_ptr = comm_ptr->local_comm;

702
    mpi_errno = MPIR_Bcast_impl(recvbuf, count, datatype, 0, newcomm_ptr, errflag);
703
704
    if (mpi_errno) {
        /* for communication errors, just record the error but continue */
705
        *errflag = TRUE;
706
707
708
        MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**fail");
        MPIU_ERR_ADD(mpi_errno_ret, mpi_errno);
    }
709
710

  fn_exit:
711
712
    if (mpi_errno_ret)
        mpi_errno = mpi_errno_ret;
713
714
715
    else if (*errflag)
        MPIU_ERR_SET(mpi_errno, MPI_ERR_OTHER, "**coll_fail");

716
717
718
719
720
721
    return mpi_errno;

  fn_fail:
    goto fn_exit;
}

722
723
724
725
726
727
728
729
/* MPIR_Allreduce performs an allreduce using point-to-point messages.
   This is intended to be used by device-specific implementations of
   allreduce.  In all other cases MPIR_Allreduce_impl should be
   used. */
#undef FUNCNAME
#define FUNCNAME MPIR_Allreduce
#undef FCNAME
#define FCNAME MPIU_QUOTE(FUNCNAME)
730
731
int MPIR_Allreduce(void *sendbuf, void *recvbuf, int count, MPI_Datatype datatype, MPI_Op op, MPID_Comm *comm_ptr,
                   int *errflag)
732
733
734
735
736
{
    int mpi_errno = MPI_SUCCESS;

    if (comm_ptr->comm_kind == MPID_INTRACOMM) {
        /* intracommunicator */
737
        mpi_errno = MPIR_Allreduce_intra(sendbuf, recvbuf, count, datatype, op, comm_ptr, errflag);
738
739
740
741
        if (mpi_errno) MPIU_ERR_POP(mpi_errno);
    }
    else {
        /* intercommunicator */
742
        mpi_errno = MPIR_Allreduce_inter(sendbuf, recvbuf, count, datatype, op, comm_ptr, errflag);
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
        if (mpi_errno) MPIU_ERR_POP(mpi_errno);
    }

fn_exit:
    return mpi_errno;
fn_fail:

    goto fn_exit;
}

/* MPIR_Allreduce_impl should be called by any internal component that
   would otherwise call MPI_Allreduce.  This differs from
   MPIR_Allreduce in that this will call the coll_fns version if it
   exists.  */
#undef FUNCNAME
#define FUNCNAME MPIR_Allreduce_impl
#undef FCNAME
#define FCNAME MPIU_QUOTE(FUNCNAME)
761
762
int MPIR_Allreduce_impl(void *sendbuf, void *recvbuf, int count, MPI_Datatype datatype, MPI_Op op, MPID_Comm *comm_ptr,
                        int *errflag)
763
764
765
766
767
{
    int mpi_errno = MPI_SUCCESS;

    if (comm_ptr->coll_fns != NULL && comm_ptr->coll_fns->Allreduce != NULL)
    {
768
	mpi_errno = comm_ptr->coll_fns->Allreduce(sendbuf, recvbuf, count, datatype, op, comm_ptr, errflag);
769
        if (mpi_errno) MPIU_ERR_POP(mpi_errno);
770
771
772
773
774
    }
    else
    {
        if (comm_ptr->comm_kind == MPID_INTRACOMM) {
            /* intracommunicator */
775
            mpi_errno = MPIR_Allreduce_intra(sendbuf, recvbuf, count, datatype, op, comm_ptr, errflag);
776
            if (mpi_errno) MPIU_ERR_POP(mpi_errno);
777
778
779
	}
        else {
            /* intercommunicator */
780
            mpi_errno = MPIR_Allreduce_inter(sendbuf, recvbuf, count, datatype, op, comm_ptr, errflag);
781
            if (mpi_errno) MPIU_ERR_POP(mpi_errno);
782
783
784
785
786
787
788
789
790
791
        }
    }

fn_exit:
    return mpi_errno;
fn_fail:
    goto fn_exit;
}


792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
#endif

#undef FUNCNAME
#define FUNCNAME MPI_Allreduce
#undef FCNAME

/*@
MPI_Allreduce - Combines values from all processes and distributes the result
                back to all processes

Input Parameters:
+ sendbuf - starting address of send buffer (choice) 
. count - number of elements in send buffer (integer) 
. datatype - data type of elements of send buffer (handle) 
. op - operation (handle) 
- comm - communicator (handle) 

Output Parameter:
. recvbuf - starting address of receive buffer (choice) 

.N ThreadSafe

.N Fortran

.N collops

.N Errors
.N MPI_ERR_BUFFER
.N MPI_ERR_COUNT
.N MPI_ERR_TYPE
.N MPI_ERR_OP
.N MPI_ERR_COMM
@*/
int MPI_Allreduce ( void *sendbuf, void *recvbuf, int count, 
		    MPI_Datatype datatype, MPI_Op op, MPI_Comm comm )
{
    static const char FCNAME[] = "MPI_Allreduce";
    int mpi_errno = MPI_SUCCESS;
    MPID_Comm *comm_ptr = NULL;
831
    int errflag = FALSE;
832
833
834
835
    MPID_MPI_STATE_DECL(MPID_STATE_MPI_ALLREDUCE);

    MPIR_ERRTEST_INITIALIZED_ORDIE();
    
836
    MPIU_THREAD_CS_ENTER(ALLFUNC,);
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
    MPID_MPI_COLL_FUNC_ENTER(MPID_STATE_MPI_ALLREDUCE);

    /* Validate parameters, especially handles needing to be converted */
#   ifdef HAVE_ERROR_CHECKING
    {
        MPID_BEGIN_ERROR_CHECKS;
        {
	    MPIR_ERRTEST_COMM(comm, mpi_errno);
            if (mpi_errno != MPI_SUCCESS) goto fn_fail;
	}
        MPID_END_ERROR_CHECKS;
    }
#   endif /* HAVE_ERROR_CHECKING */

    /* Convert MPI object handles to object pointers */
    MPID_Comm_get_ptr( comm, comm_ptr );

    /* Validate parameters and objects (post conversion) */
#   ifdef HAVE_ERROR_CHECKING
    {
        MPID_BEGIN_ERROR_CHECKS;
        {
            MPID_Datatype *datatype_ptr = NULL;
            MPID_Op *op_ptr = NULL;

            MPID_Comm_valid_ptr( comm_ptr, mpi_errno );
            if (mpi_errno != MPI_SUCCESS) goto fn_fail;
	    MPIR_ERRTEST_COUNT(count, mpi_errno);
	    MPIR_ERRTEST_DATATYPE(datatype, "datatype", mpi_errno);
	    MPIR_ERRTEST_OP(op, mpi_errno);
	    
            if (HANDLE_GET_KIND(datatype) != HANDLE_KIND_BUILTIN) {
                MPID_Datatype_get_ptr(datatype, datatype_ptr);
                MPID_Datatype_valid_ptr( datatype_ptr, mpi_errno );
                MPID_Datatype_committed_ptr( datatype_ptr, mpi_errno );
            }

	    if (comm_ptr->comm_kind == MPID_INTERCOMM)
                MPIR_ERRTEST_SENDBUF_INPLACE(sendbuf, count, mpi_errno);
            
            if (sendbuf != MPI_IN_PLACE) 
                MPIR_ERRTEST_USERBUFFER(sendbuf,count,datatype,mpi_errno);

            MPIR_ERRTEST_RECVBUF_INPLACE(recvbuf, count, mpi_errno);
	    MPIR_ERRTEST_USERBUFFER(recvbuf,count,datatype,mpi_errno);

	    if (mpi_errno != MPI_SUCCESS) goto fn_fail;

            if (HANDLE_GET_KIND(op) != HANDLE_KIND_BUILTIN) {
                MPID_Op_get_ptr(op, op_ptr);
                MPID_Op_valid_ptr( op_ptr, mpi_errno );
            }
            if (HANDLE_GET_KIND(op) == HANDLE_KIND_BUILTIN) {
                mpi_errno = 
                    ( * MPIR_Op_check_dtype_table[op%16 - 1] )(datatype); 
            }
893
894
895
	    if (count != 0) {
		MPIR_ERRTEST_ALIAS_COLL(sendbuf, recvbuf, mpi_errno);
	    }
896
	    if (mpi_errno != MPI_SUCCESS) goto fn_fail;
897
	}
898
899
900
901
902
903
        MPID_END_ERROR_CHECKS;
    }
#   endif /* HAVE_ERROR_CHECKING */

    /* ... body of routine ...  */

904
    mpi_errno = MPIR_Allreduce_impl(sendbuf, recvbuf, count, datatype, op, comm_ptr, &errflag);
905
    if (mpi_errno) goto fn_fail;
906
907
908
909
910

    /* ... end of body of routine ... */
    
  fn_exit:
    MPID_MPI_COLL_FUNC_EXIT(MPID_STATE_MPI_ALLREDUCE);
911
    MPIU_THREAD_CS_EXIT(ALLFUNC,);
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
    return mpi_errno;

  fn_fail:
    /* --BEGIN ERROR HANDLING-- */
#   ifdef HAVE_ERROR_CHECKING
    {
	mpi_errno = MPIR_Err_create_code(
	    mpi_errno, MPIR_ERR_RECOVERABLE, FCNAME, __LINE__, MPI_ERR_OTHER, "**mpi_allreduce",
	    "**mpi_allreduce %p %p %d %D %O %C", sendbuf, recvbuf, count, datatype, op, comm);
    }
#   endif
    mpi_errno = MPIR_Err_return_comm( comm_ptr, FCNAME, mpi_errno );
    goto fn_exit;
    /* --END ERROR HANDLING-- */
}