Chris@42: <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd"> Chris@42: <html> Chris@42: <!-- This manual is for FFTW Chris@42: (version 3.3.5, 30 July 2016). Chris@42: Chris@42: Copyright (C) 2003 Matteo Frigo. Chris@42: Chris@42: Copyright (C) 2003 Massachusetts Institute of Technology. Chris@42: Chris@42: Permission is granted to make and distribute verbatim copies of this Chris@42: manual provided the copyright notice and this permission notice are Chris@42: preserved on all copies. Chris@42: Chris@42: Permission is granted to copy and distribute modified versions of this Chris@42: manual under the conditions for verbatim copying, provided that the Chris@42: entire resulting derived work is distributed under the terms of a Chris@42: permission notice identical to this one. Chris@42: Chris@42: Permission is granted to copy and distribute translations of this manual Chris@42: into another language, under the above conditions for modified versions, Chris@42: except that this permission notice may be stated in a translation Chris@42: approved by the Free Software Foundation. --> Chris@42: <!-- Created by GNU Texinfo 5.2, http://www.gnu.org/software/texinfo/ --> Chris@42: <head> Chris@42: <title>FFTW 3.3.5: Advanced distributed-transpose interface</title> Chris@42: Chris@42: <meta name="description" content="FFTW 3.3.5: Advanced distributed-transpose interface"> Chris@42: <meta name="keywords" content="FFTW 3.3.5: Advanced distributed-transpose interface"> Chris@42: <meta name="resource-type" content="document"> Chris@42: <meta name="distribution" content="global"> Chris@42: <meta name="Generator" content="makeinfo"> Chris@42: <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> Chris@42: <link href="index.html#Top" rel="start" title="Top"> Chris@42: <link href="Concept-Index.html#Concept-Index" rel="index" title="Concept Index"> Chris@42: <link href="index.html#SEC_Contents" rel="contents" title="Table of Contents"> Chris@42: <link href="FFTW-MPI-Transposes.html#FFTW-MPI-Transposes" rel="up" title="FFTW MPI Transposes"> Chris@42: <link href="An-improved-replacement-for-MPI_005fAlltoall.html#An-improved-replacement-for-MPI_005fAlltoall" rel="next" title="An improved replacement for MPI_Alltoall"> Chris@42: <link href="Basic-distributed_002dtranspose-interface.html#Basic-distributed_002dtranspose-interface" rel="prev" title="Basic distributed-transpose interface"> Chris@42: <style type="text/css"> Chris@42: <!-- Chris@42: a.summary-letter {text-decoration: none} Chris@42: blockquote.smallquotation {font-size: smaller} Chris@42: div.display {margin-left: 3.2em} Chris@42: div.example {margin-left: 3.2em} Chris@42: div.indentedblock {margin-left: 3.2em} Chris@42: div.lisp {margin-left: 3.2em} Chris@42: div.smalldisplay {margin-left: 3.2em} Chris@42: div.smallexample {margin-left: 3.2em} Chris@42: div.smallindentedblock {margin-left: 3.2em; font-size: smaller} Chris@42: div.smalllisp {margin-left: 3.2em} Chris@42: kbd {font-style:oblique} Chris@42: pre.display {font-family: inherit} Chris@42: pre.format {font-family: inherit} Chris@42: pre.menu-comment {font-family: serif} Chris@42: pre.menu-preformatted {font-family: serif} Chris@42: pre.smalldisplay {font-family: inherit; font-size: smaller} Chris@42: pre.smallexample {font-size: smaller} Chris@42: pre.smallformat {font-family: inherit; font-size: smaller} Chris@42: pre.smalllisp {font-size: smaller} Chris@42: span.nocodebreak {white-space:nowrap} Chris@42: span.nolinebreak {white-space:nowrap} Chris@42: span.roman {font-family:serif; font-weight:normal} Chris@42: span.sansserif {font-family:sans-serif; font-weight:normal} Chris@42: ul.no-bullet {list-style: none} Chris@42: --> Chris@42: </style> Chris@42: Chris@42: Chris@42: </head> Chris@42: Chris@42: <body lang="en" bgcolor="#FFFFFF" text="#000000" link="#0000FF" vlink="#800080" alink="#FF0000"> Chris@42: <a name="Advanced-distributed_002dtranspose-interface"></a> Chris@42: <div class="header"> Chris@42: <p> Chris@42: Next: <a href="An-improved-replacement-for-MPI_005fAlltoall.html#An-improved-replacement-for-MPI_005fAlltoall" accesskey="n" rel="next">An improved replacement for MPI_Alltoall</a>, Previous: <a href="Basic-distributed_002dtranspose-interface.html#Basic-distributed_002dtranspose-interface" accesskey="p" rel="prev">Basic distributed-transpose interface</a>, Up: <a href="FFTW-MPI-Transposes.html#FFTW-MPI-Transposes" accesskey="u" rel="up">FFTW MPI Transposes</a> [<a href="index.html#SEC_Contents" title="Table of contents" rel="contents">Contents</a>][<a href="Concept-Index.html#Concept-Index" title="Index" rel="index">Index</a>]</p> Chris@42: </div> Chris@42: <hr> Chris@42: <a name="Advanced-distributed_002dtranspose-interface-1"></a> Chris@42: <h4 class="subsection">6.7.2 Advanced distributed-transpose interface</h4> Chris@42: Chris@42: <p>The above routines are for a transpose of a matrix of numbers (of type Chris@42: <code>double</code>), using FFTW’s default block sizes. More generally, one Chris@42: can perform transposes of <em>tuples</em> of numbers, with Chris@42: user-specified block sizes for the input and output: Chris@42: </p> Chris@42: <div class="example"> Chris@42: <pre class="example">fftw_plan fftw_mpi_plan_many_transpose Chris@42: (ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t howmany, Chris@42: ptrdiff_t block0, ptrdiff_t block1, Chris@42: double *in, double *out, MPI_Comm comm, unsigned flags); Chris@42: </pre></div> Chris@42: <a name="index-fftw_005fmpi_005fplan_005fmany_005ftranspose"></a> Chris@42: Chris@42: <p>In this case, one is transposing an <code>n0</code> by <code>n1</code> matrix of Chris@42: <code>howmany</code>-tuples (e.g. <code>howmany = 2</code> for complex numbers). Chris@42: The input is distributed along the <code>n0</code> dimension with block size Chris@42: <code>block0</code>, and the <code>n1</code> by <code>n0</code> output is distributed Chris@42: along the <code>n1</code> dimension with block size <code>block1</code>. If Chris@42: <code>FFTW_MPI_DEFAULT_BLOCK</code> (0) is passed for a block size then FFTW Chris@42: uses its default block size. To get the local size of the data on Chris@42: each process, you should then call <code>fftw_mpi_local_size_many_transposed</code>. Chris@42: <a name="index-FFTW_005fMPI_005fDEFAULT_005fBLOCK-1"></a> Chris@42: <a name="index-fftw_005fmpi_005flocal_005fsize_005fmany_005ftransposed"></a> Chris@42: </p> Chris@42: Chris@42: Chris@42: Chris@42: </body> Chris@42: </html>