1 /* Copyright (c) 2013-2021. The SimGrid Team.
2 * All rights reserved. */
4 /* This program is free software; you can redistribute it and/or modify it
5 * under the terms of the license (GNU LGPL) which comes with this package. */
8 * Copyright (c) 2004-2005 The Trustees of Indiana University and Indiana
9 * University Research and Technology
10 * Corporation. All rights reserved.
11 * Copyright (c) 2004-2009 The University of Tennessee and The University
12 * of Tennessee Research Foundation. All rights
14 * Copyright (c) 2004-2005 High Performance Computing Center Stuttgart,
15 * University of Stuttgart. All rights reserved.
16 * Copyright (c) 2004-2005 The Regents of the University of California.
17 * All rights reserved.
18 * Copyright (c) 2009 University of Houston. All rights reserved.
20 * Additional copyrights may follow
22 * Redistribution and use in source and binary forms, with or without
23 * modification, are permitted provided that the following conditions are
26 * - Redistributions of source code must retain the above copyright
27 * notice, this list of conditions and the following disclaimer.
29 * - Redistributions in binary form must reproduce the above copyright
30 * notice, this list of conditions and the following disclaimer listed
31 * in this license in the documentation and/or other materials
32 * provided with the distribution.
34 * - Neither the name of the copyright holders nor the names of its
35 * contributors may be used to endorse or promote products derived from
36 * this software without specific prior written permission.
38 * The copyright holders provide no reassurances that the source code
39 * provided does not infringe any patent, copyright, or any other
40 * intellectual property rights of third parties. The copyright holders
41 * disclaim any liability to any recipient for claims brought against
42 * recipient by any third party for infringement of that parties
43 * intellectual property rights.
45 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
46 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
47 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
48 * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
49 * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
50 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
51 * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
52 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
53 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
54 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
55 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
58 #include "../coll_tuned_topo.hpp"
59 #include "../colls_private.hpp"
60 #define MAXTREEFANOUT 32
64 int bcast__ompi_split_bintree( void* buffer,
66 MPI_Datatype datatype,
70 unsigned int segsize ;
72 int segindex, i, lr, pair;
73 int segcount[2]; /* Number ompi_request_wait_allof elements sent with each segment */
75 int num_segments[2]; /* Number of segments */
76 int sendcount[2]; /* the same like segcount, except for the last segment */
77 size_t realsegsize[2];
80 ptrdiff_t type_extent;
83 MPI_Request base_req, new_req;
84 ompi_coll_tree_t *tree;
85 // mca_coll_tuned_module_t *tuned_module = (mca_coll_tuned_module_t*) module;
86 // mca_coll_tuned_comm_t *data = tuned_module->tuned_data;
92 //compute again segsize
93 const size_t intermediate_message_size = 370728;
94 size_t message_size = datatype->size() * (unsigned long)count;
95 if(message_size < intermediate_message_size)
100 XBT_DEBUG("ompi_coll_tuned_bcast_intra_split_bintree rank %d root %d ss %5u", rank, root, segsize);
106 /* setup the binary tree topology. */
107 tree = ompi_coll_tuned_topo_build_tree(2,comm,root);
109 type_size = datatype->size();
111 /* Determine number of segments and number of elements per segment */
113 if (count % 2 != 0) counts[0]++;
114 counts[1] = count - counts[0];
116 /* Note that ompi_datatype_type_size() will never return a negative
117 value in typelng; it returns an int [vs. an unsigned type]
118 because of the MPI spec. */
119 if (segsize < ((uint32_t)type_size)) {
120 segsize = type_size; /* push segsize up to hold one type */
122 segcount[0] = segcount[1] = segsize / type_size;
123 num_segments[0] = counts[0] / segcount[0];
124 if ((counts[0] % segcount[0]) != 0)
126 num_segments[1] = counts[1] / segcount[1];
127 if ((counts[1] % segcount[1]) != 0)
130 /* if the message is too small to be split into segments */
131 if( (counts[0] == 0 || counts[1] == 0) ||
132 (segsize > counts[0] * type_size) ||
133 (segsize > counts[1] * type_size) ) {
134 /* call linear version here ! */
135 return bcast__SMP_linear( buffer, count, datatype, root, comm);
137 type_extent = datatype->get_extent();
140 /* Determine real segment size */
141 realsegsize[0] = segcount[0] * type_extent;
142 realsegsize[1] = segcount[1] * type_extent;
144 /* set the buffer pointers */
145 tmpbuf[0] = (char *) buffer;
146 tmpbuf[1] = (char *) buffer+counts[0] * type_extent;
149 Root splits the buffer in 2 and sends segmented message down the branches.
150 Left subtree of the tree receives first half of the buffer, while right
151 subtree receives the remaining message.
154 /* determine if I am left (0) or right (1), (root is right) */
155 lr = ((rank + size - root)%size + 1)%2;
159 /* determine segment count */
160 sendcount[0] = segcount[0];
161 sendcount[1] = segcount[1];
162 /* for each segment */
163 for (segindex = 0; segindex < num_segments[0]; segindex++) {
165 for( i = 0; i < tree->tree_nextsize && i < 2; i++ ) {
166 if (segindex >= num_segments[i]) { /* no more segments */
169 /* determine how many elements are being sent in this round */
170 if(segindex == (num_segments[i] - 1))
171 sendcount[i] = counts[i] - segindex*segcount[i];
173 Request::send(tmpbuf[i], sendcount[i], datatype,
174 tree->tree_next[i], COLL_TAG_BCAST, comm);
175 /* update tmp buffer */
176 tmpbuf[i] += realsegsize[i];
181 /* intermediate nodes code */
182 else if( tree->tree_nextsize > 0 ) {
183 /* Intermediate nodes:
184 * It will receive segments only from one half of the data.
185 * Which one is determined by whether the node belongs to the "left" or "right"
186 * subtree. Topology building function builds binary tree such that
187 * odd "shifted ranks" ((rank + size - root)%size) are on the left subtree,
188 * and even on the right subtree.
190 * Create the pipeline. We first post the first receive, then in the loop we
191 * post the next receive and after that wait for the previous receive to complete
192 * and we disseminating the data to all children.
194 sendcount[lr] = segcount[lr];
195 base_req = Request::irecv(tmpbuf[lr], sendcount[lr], datatype, tree->tree_prev, COLL_TAG_BCAST, comm);
197 for (segindex = 1; segindex < num_segments[lr]; segindex++) {
198 /* determine how many elements to expect in this round */
199 if (segindex == (num_segments[lr] - 1))
200 sendcount[lr] = counts[lr] - segindex * segcount[lr];
202 new_req = Request::irecv(tmpbuf[lr] + realsegsize[lr], sendcount[lr], datatype, tree->tree_prev, COLL_TAG_BCAST,
205 /* wait for and forward current segment */
206 Request::waitall(1, &base_req, MPI_STATUSES_IGNORE);
207 for (i = 0; i < tree->tree_nextsize; i++) { /* send data to children (segcount[lr]) */
208 Request::send(tmpbuf[lr], segcount[lr], datatype, tree->tree_next[i], COLL_TAG_BCAST, comm);
209 } /* end of for each child */
211 /* update the base request */
213 /* go to the next buffer (ie. the one corresponding to the next recv) */
214 tmpbuf[lr] += realsegsize[lr];
215 } /* end of for segindex */
217 /* wait for the last segment and forward current segment */
218 Request::waitall( 1, &base_req, MPI_STATUSES_IGNORE );
219 for( i = 0; i < tree->tree_nextsize; i++ ) { /* send data to children */
220 Request::send(tmpbuf[lr], sendcount[lr], datatype,
221 tree->tree_next[i], COLL_TAG_BCAST, comm);
222 } /* end of for each child */
227 /* Just consume segments as fast as possible */
228 sendcount[lr] = segcount[lr];
229 for (segindex = 0; segindex < num_segments[lr]; segindex++) {
230 /* determine how many elements to expect in this round */
231 if (segindex == (num_segments[lr] - 1)) sendcount[lr] = counts[lr] - segindex*segcount[lr];
232 /* receive segments */
233 Request::recv(tmpbuf[lr], sendcount[lr], datatype,
234 tree->tree_prev, COLL_TAG_BCAST,
235 comm, MPI_STATUS_IGNORE);
236 /* update the initial pointer to the buffer */
237 tmpbuf[lr] += realsegsize[lr];
241 /* reset the buffer pointers */
242 tmpbuf[0] = (char *) buffer;
243 tmpbuf[1] = (char *) buffer+counts[0] * type_extent;
246 Find your immediate pair (identical node in opposite subtree) and SendRecv
247 data buffer with them.
248 The tree building function ensures that
250 if we are in the left subtree (lr == 0) our pair is (rank+1)%size.
251 if we are in the right subtree (lr == 1) our pair is (rank-1)%size
252 If we have even number of nodes the rank (size-1) will pair up with root.
255 pair = (rank+1)%size;
257 pair = (rank+size-1)%size;
260 if ( (size%2) != 0 && rank != root) {
262 Request::sendrecv( tmpbuf[lr], counts[lr], datatype,
263 pair, COLL_TAG_BCAST,
264 tmpbuf[(lr+1)%2], counts[(lr+1)%2], datatype,
265 pair, COLL_TAG_BCAST,
266 comm, MPI_STATUS_IGNORE);
267 } else if ( (size%2) == 0 ) {
268 /* root sends right buffer to the last node */
270 Request::send(tmpbuf[1], counts[1], datatype,
271 (root+size-1)%size, COLL_TAG_BCAST, comm);
274 /* last node receives right buffer from the root */
275 else if (rank == (root+size-1)%size) {
276 Request::recv(tmpbuf[1], counts[1], datatype,
277 root, COLL_TAG_BCAST,
278 comm, MPI_STATUS_IGNORE);
280 /* everyone else exchanges buffers */
282 Request::sendrecv( tmpbuf[lr], counts[lr], datatype,
283 pair, COLL_TAG_BCAST,
284 tmpbuf[(lr+1)%2], counts[(lr+1)%2], datatype,
285 pair, COLL_TAG_BCAST,
286 comm, MPI_STATUS_IGNORE);
289 ompi_coll_tuned_topo_destroy_tree(&tree);
290 return (MPI_SUCCESS);