1 /* Copyright (c) 2013-2017. 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. */
7 #include "../colls_private.hpp"
9 /*****************************************************************************
11 Copyright (c) 2006, Ahmad Faraj & Xin Yuan,
14 Redistribution and use in source and binary forms, with or without
15 modification, are permitted provided that the following conditions are met:
17 * Redistributions of source code must retain the above copyright notice,
18 this list of conditions and the following disclaimer.
20 * Redistributions in binary form must reproduce the above copyright notice,
21 this list of conditions and the following disclaimer in the documentation
22 and/or other materials provided with the distribution.
24 * Neither the name of the Florida State University nor the names of its
25 contributors may be used to endorse or promote products derived from this
26 software without specific prior written permission.
28 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
29 ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
30 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
31 DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
32 ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
33 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
34 LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
35 ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
36 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
37 SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
39 *************************************************************************
40 * Any results obtained from executing this software require the *
41 * acknowledgment and citation of the software and its owners. *
42 * The full citation is given below: *
44 * A. Faraj and X. Yuan. "Automatic Generation and Tuning of MPI *
45 * Collective Communication Routines." The 19th ACM International *
46 * Conference on Supercomputing (ICS), Cambridge, Massachusetts, *
48 *************************************************************************
50 *****************************************************************************/
52 /*****************************************************************************
55 * num: the number of processors in a communicator
59 * descp: takes a number and tries to find a factoring of x*y*z mesh out of it
60 ****************************************************************************/
63 static int is_3dmesh(int num, int *i, int *j, int *k)
69 if ((num % (x * x)) == 0) {
79 /*****************************************************************************
80 * Function: allgather_3dmesh_shoot
82 * send_buff: send input buffer
83 * send_count: number of elements to send
84 * send_type: data type of elements being sent
85 * recv_buff: receive output buffer
86 * recv_count: number of elements to received
87 * recv_type: data type of elements being received
89 * Descrp: Function realizes the allgather operation using the 2dmesh
90 * algorithm. Allgather ommunication occurs first in the x dimension, y
91 * dimension, and then in the z dimension. Communication in each dimension
94 ****************************************************************************/
99 int Coll_allgather_3dmesh::allgather(void *send_buff, int send_count,
100 MPI_Datatype send_type, void *recv_buff,
101 int recv_count, MPI_Datatype recv_type,
104 MPI_Request *req, *req_ptr;
107 int i, src, dst, rank, num_procs, block_size, my_z_base;
108 int my_z, X, Y, Z, send_offset, recv_offset;
109 int two_dsize, my_row_base, my_col_base, src_row_base, src_z_base, num_reqs;
110 int tag = COLL_TAG_ALLGATHER;
113 num_procs = comm->size();
114 extent = send_type->get_extent();
116 if (not is_3dmesh(num_procs, &X, &Y, &Z))
117 THROWF(arg_error,0, "allgather_3dmesh algorithm can't be used with this number of processes! ");
128 my_z = rank / two_dsize;
130 my_row_base = (rank / X) * X;
131 my_col_base = (rank % Y) + (my_z * two_dsize);
132 my_z_base = my_z * two_dsize;
134 block_size = extent * send_count;
136 req = (MPI_Request *) xbt_malloc(num_reqs * sizeof(MPI_Request));
140 // do local allgather/local copy
141 recv_offset = rank * block_size;
142 Datatype::copy(send_buff, send_count, send_type, (char *)recv_buff + recv_offset,
143 recv_count, recv_type);
146 for (i = 0; i < Y; i++) {
147 src = i + my_row_base;
150 recv_offset = src * block_size;
151 *(req_ptr++) = Request::irecv((char *)recv_buff + recv_offset, send_count, recv_type, src, tag,
155 for (i = 0; i < Y; i++) {
156 dst = i + my_row_base;
159 Request::send(send_buff, send_count, send_type, dst, tag, comm);
162 Request::waitall(Y - 1, req, MPI_STATUSES_IGNORE);
165 // do colwise comm, it does not matter here if i*X or i *Y since X == Y
167 for (i = 0; i < X; i++) {
168 src = (i * Y + my_col_base);
172 src_row_base = (src / X) * X;
173 recv_offset = src_row_base * block_size;
174 *(req_ptr++) = Request::irecv((char *)recv_buff + recv_offset, recv_count * Y, recv_type, src, tag,
178 send_offset = my_row_base * block_size;
180 for (i = 0; i < X; i++) {
181 dst = (i * Y + my_col_base);
184 Request::send((char *)recv_buff + send_offset, send_count * Y, send_type, dst, tag,
188 Request::waitall(X - 1, req, MPI_STATUSES_IGNORE);
191 for (i = 1; i < Z; i++) {
192 src = (rank + i * two_dsize) % num_procs;
193 src_z_base = (src / two_dsize) * two_dsize;
195 recv_offset = (src_z_base * block_size);
197 *(req_ptr++) = Request::irecv((char *)recv_buff + recv_offset, recv_count * two_dsize, recv_type,
201 for (i = 1; i < Z; i++) {
202 dst = (rank + i * two_dsize) % num_procs;
203 send_offset = my_z_base * block_size;
204 Request::send((char *)recv_buff + send_offset, send_count * two_dsize, send_type,
207 Request::waitall(Z - 1, req, MPI_STATUSES_IGNORE);