Mercurial > hg > sv-dependency-builds

comparison src/fftw-3.3.3/doc/html/FFTW-MPI-Wisdom.html @ 10:37bf6b4a2645

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Add FFTW3
author Chris Cannam
date Wed, 20 Mar 2013 15:35:50 +0000
parents
children
comparison
equal deleted inserted replaced
9:c0fb53affa76 10:37bf6b4a2645
1 <html lang="en">
2 <head>
3 <title>FFTW MPI Wisdom - FFTW 3.3.3</title>
4 <meta http-equiv="Content-Type" content="text/html">
5 <meta name="description" content="FFTW 3.3.3">
6 <meta name="generator" content="makeinfo 4.13">
7 <link title="Top" rel="start" href="index.html#Top">
8 <link rel="up" href="Distributed_002dmemory-FFTW-with-MPI.html#Distributed_002dmemory-FFTW-with-MPI" title="Distributed-memory FFTW with MPI">
9 <link rel="prev" href="FFTW-MPI-Transposes.html#FFTW-MPI-Transposes" title="FFTW MPI Transposes">
10 <link rel="next" href="Avoiding-MPI-Deadlocks.html#Avoiding-MPI-Deadlocks" title="Avoiding MPI Deadlocks">
11 <link href="http://www.gnu.org/software/texinfo/" rel="generator-home" title="Texinfo Homepage">
12 <!--
13 This manual is for FFTW
14 (version 3.3.3, 25 November 2012).
15
16 Copyright (C) 2003 Matteo Frigo.
17
18 Copyright (C) 2003 Massachusetts Institute of Technology.
19
20 Permission is granted to make and distribute verbatim copies of
21 this manual provided the copyright notice and this permission
22 notice are preserved on all copies.
23
24 Permission is granted to copy and distribute modified versions of
25 this manual under the conditions for verbatim copying, provided
26 that the entire resulting derived work is distributed under the
27 terms of a permission notice identical to this one.
28
29 Permission is granted to copy and distribute translations of this
30 manual into another language, under the above conditions for
31 modified versions, except that this permission notice may be
32 stated in a translation approved by the Free Software Foundation.
33 -->
34 <meta http-equiv="Content-Style-Type" content="text/css">
35 <style type="text/css"><!--
36 pre.display { font-family:inherit }
37 pre.format { font-family:inherit }
38 pre.smalldisplay { font-family:inherit; font-size:smaller }
39 pre.smallformat { font-family:inherit; font-size:smaller }
40 pre.smallexample { font-size:smaller }
41 pre.smalllisp { font-size:smaller }
42 span.sc { font-variant:small-caps }
43 span.roman { font-family:serif; font-weight:normal; }
44 span.sansserif { font-family:sans-serif; font-weight:normal; }
45 --></style>
46 </head>
47 <body>
48 <div class="node">
49 <a name="FFTW-MPI-Wisdom"></a>
50 <p>
51 Next:&nbsp;<a rel="next" accesskey="n" href="Avoiding-MPI-Deadlocks.html#Avoiding-MPI-Deadlocks">Avoiding MPI Deadlocks</a>,
52 Previous:&nbsp;<a rel="previous" accesskey="p" href="FFTW-MPI-Transposes.html#FFTW-MPI-Transposes">FFTW MPI Transposes</a>,
53 Up:&nbsp;<a rel="up" accesskey="u" href="Distributed_002dmemory-FFTW-with-MPI.html#Distributed_002dmemory-FFTW-with-MPI">Distributed-memory FFTW with MPI</a>
54 <hr>
55 </div>
56
57 <h3 class="section">6.8 FFTW MPI Wisdom</h3>
58
59 <p><a name="index-wisdom-410"></a><a name="index-saving-plans-to-disk-411"></a>
60 FFTW's &ldquo;wisdom&rdquo; facility (see <a href="Words-of-Wisdom_002dSaving-Plans.html#Words-of-Wisdom_002dSaving-Plans">Words of Wisdom-Saving Plans</a>) can
61 be used to save MPI plans as well as to save uniprocessor plans.
62 However, for MPI there are several unavoidable complications.
63
64 <p><a name="index-MPI-I_002fO-412"></a>First, the MPI standard does not guarantee that every process can
65 perform file I/O (at least, not using C stdio routines)&mdash;in general,
66 we may only assume that process 0 is capable of I/O.<a rel="footnote" href="#fn-1" name="fnd-1"><sup>1</sup></a> So, if we
67 want to export the wisdom from a single process to a file, we must
68 first export the wisdom to a string, then send it to process 0, then
69 write it to a file.
70
71 <p>Second, in principle we may want to have separate wisdom for every
72 process, since in general the processes may run on different hardware
73 even for a single MPI program. However, in practice FFTW's MPI code
74 is designed for the case of homogeneous hardware (see <a href="Load-balancing.html#Load-balancing">Load balancing</a>), and in this case it is convenient to use the same wisdom
75 for every process. Thus, we need a mechanism to synchronize the wisdom.
76
77 <p>To address both of these problems, FFTW provides the following two
78 functions:
79
80 <pre class="example"> void fftw_mpi_broadcast_wisdom(MPI_Comm comm);
81 void fftw_mpi_gather_wisdom(MPI_Comm comm);
82 </pre>
83 <p><a name="index-fftw_005fmpi_005fgather_005fwisdom-413"></a><a name="index-fftw_005fmpi_005fbroadcast_005fwisdom-414"></a>
84 Given a communicator <code>comm</code>, <code>fftw_mpi_broadcast_wisdom</code>
85 will broadcast the wisdom from process 0 to all other processes.
86 Conversely, <code>fftw_mpi_gather_wisdom</code> will collect wisdom from all
87 processes onto process 0. (If the plans created for the same problem
88 by different processes are not the same, <code>fftw_mpi_gather_wisdom</code>
89 will arbitrarily choose one of the plans.) Both of these functions
90 may result in suboptimal plans for different processes if the
91 processes are running on non-identical hardware. Both of these
92 functions are <em>collective</em> calls, which means that they must be
93 executed by all processes in the communicator.
94 <a name="index-collective-function-415"></a>
95
96 <p>So, for example, a typical code snippet to import wisdom from a file
97 and use it on all processes would be:
98
99 <pre class="example"> {
100 int rank;
101
102 fftw_mpi_init();
103 MPI_Comm_rank(MPI_COMM_WORLD, &amp;rank);
104 if (rank == 0) fftw_import_wisdom_from_filename("mywisdom");
105 fftw_mpi_broadcast_wisdom(MPI_COMM_WORLD);
106 }
107 </pre>
108 <p>(Note that we must call <code>fftw_mpi_init</code> before importing any
109 wisdom that might contain MPI plans.) Similarly, a typical code
110 snippet to export wisdom from all processes to a file is:
111 <a name="index-fftw_005fmpi_005finit-416"></a>
112 <pre class="example"> {
113 int rank;
114
115 fftw_mpi_gather_wisdom(MPI_COMM_WORLD);
116 MPI_Comm_rank(MPI_COMM_WORLD, &amp;rank);
117 if (rank == 0) fftw_export_wisdom_to_filename("mywisdom");
118 }
119 </pre>
120 <!-- -->
121 <div class="footnote">
122 <hr>
123 <h4>Footnotes</h4><p class="footnote"><small>[<a name="fn-1" href="#fnd-1">1</a>]</small> In fact,
124 even this assumption is not technically guaranteed by the standard,
125 although it seems to be universal in actual MPI implementations and is
126 widely assumed by MPI-using software. Technically, you need to query
127 the <code>MPI_IO</code> attribute of <code>MPI_COMM_WORLD</code> with
128 <code>MPI_Attr_get</code>. If this attribute is <code>MPI_PROC_NULL</code>, no
129 I/O is possible. If it is <code>MPI_ANY_SOURCE</code>, any process can
130 perform I/O. Otherwise, it is the rank of a process that can perform
131 I/O ... but since it is not guaranteed to yield the <em>same</em> rank
132 on all processes, you have to do an <code>MPI_Allreduce</code> of some kind
133 if you want all processes to agree about which is going to do I/O.
134 And even then, the standard only guarantees that this process can
135 perform output, but not input. See e.g. <cite>Parallel Programming
136 with MPI</cite> by P. S. Pacheco, section 8.1.3. Needless to say, in our
137 experience virtually no MPI programmers worry about this.</p>
138
139 <hr></div>
140
141 </body></html>
142