1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 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 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216
|
.\" Text automatically generated by txt2man
.TH gmap 1 "23 November 2019" "" "Scotch user's manual"
.SH NAME
\fBgmap, gpart \fP- compute static mappings and partitions sequentially
\fB
.SH SYNOPSIS
.nf
.fam C
\fBgmap\fP [\fIoptions\fP] [\fIgfile\fP] [\fItfile\fP] [\fImfile\fP] [\fIlfile\fP]
\fBgpart\fP [\fIoptions\fP] [\fInparts/\fIpwght\fP\fP] [\fIgfile\fP] [\fImfile\fP] [\fIlfile\fP]
.fam T
.fi
.fam T
.fi
.SH DESCRIPTION
The \fBgmap\fP program computes, in a sequential way, a static mapping of a
source graph onto a target graph.
.PP
The \fBgpart\fP program is a simplified interface to \fBgmap\fP, which performs
graph partitioning instead of static mapping. Consequently, the
desired number of parts has to be provided, in lieu of the target
architecture. When using the program for graph clustering, the
number of parts turns into maximum cluster weight.
.PP
The \fB-b\fP and \fB-c\fP options allow the user to set preferences on the
behavior of the mapping strategy which is used by default. The \fB-m\fP
option allows the user to define a custom mapping strategy.
.PP
The \fB-q\fP option turns the programs into graph clustering programs. In
this case, \fBgmap\fP only accepts variable-sized target architectures.
.PP
Source graph file \fIgfile\fP can only be a centralized graph file. For \fBgmap\fP,
the target architecture file \fItfile\fP describes either algorithmically-coded
topologies such as meshes and hypercubes, or decomposition-defined
architectures created by means of the \fBamk_grf\fP(1) program. The resulting
mapping is stored in file \fImfile\fP. Eventual logging information (such
as the one produced by option \fB-v\fP) is sent to file \fIlfile\fP. When file
names are not specified, data is read from standard input and
written to standard output. Standard streams can also be explicitely
represented by a dash '-'.
.PP
When the proper libraries have been included at compile time, \fBgmap\fP
and \fBgpart\fP can directly handle compressed graphs, both as input and
output. A stream is treated as compressed whenever its name is
postfixed with a compressed file extension, such as
in 'brol.grf.bz2' or '-.gz'. The compression formats which can be
supported are the bzip2 format ('.bz2'), the gzip format ('.gz'),
and the lzma format ('.lzma').
.SH OPTIONS
.TP
.B
\fB-b\fP\fIval\fP
Set maximum load imbalance ratio for graph partitioning
or static mapping. When programs are used as clustering
tools, this parameter sets the maximum load imbalance
ratio for recursive bipartitions. Exclusive with
the \fB-m\fP option.
.TP
.B
\fB-c\fP\fIopt\fP
Choose default mapping strategy according to one or
several options among:
.RS
.TP
.B
b
enforce load balance as much as possible.
.TP
.B
q
privilege quality over speed (default).
.TP
.B
s
privilege speed over quality.
.TP
.B
t
enforce safety.
.PP
It is exclusive with the \fB-m\fP option.
.RE
.TP
.B
\fB-h\fP
Display some help.
.TP
.B
\fB-m\fP\fIstrat\fP
Use sequential mapping strategy \fIstrat\fP (see
Scotch user's manual for more information).
.TP
.B
\fB-q\fP
(for \fBgpart\fP)
.TP
.B
\fB-q\fP\fIpwght\fP
(for \fBgmap\fP)
Use the programs as graph clustering tools instead of
static mapping or graph partitioning tools. For \fBgpart\fP,
the number of parts will become the maximum cluster
weight. For \fBgmap\fP, this number \fIpwght\fP has to be passed
after the option.
.TP
.B
\fB-V\fP
Display program version and copyright.
.TP
.B
\fB-v\fP\fIverb\fP
Set verbose mode to \fIverb\fP. It is a set of one of more
characters which can be:
.RS
.TP
.B
m
mapping information.
.TP
.B
s
strategy information.
.TP
.B
t
timing information.
.SH TARGET ARCHITECTURES
Target architectures represent graphs onto which source graphs are
mapped. In order to speed-up the obtainment of target architecture
topological properties during the computation of mappings, some
classical topologies are algorithmically coded into the mapper
itself. These topologies are consequently simply defined by their
code name, followed by their dimensional parameters:
.TP
.B
cmplt \fIdim\fP
unweighted complete graph of size \fIdim\fP.
.TP
.B
cmpltw \fIdim\fP \fIw0\fP \fIw1\fP \.\.\. \fIwdim-1\fP
weighted complete graph of size
size and of respective loads
\fIw0\fP, \fIw1\fP, \.\.\., \fIwdim-1\fP.
.TP
.B
hcub \fIdim\fP
hypercube of dimension \fIdim\fP.
.TP
.B
leaf \fIhgt\fP \fIn0\fP \fIw0\fP \.\.\. \fInhgt-1\fP \fIwhgt-1\fP
tree-leaf graph of height \fIhgt\fP
with (\fIn0\fP times \fIn1\fP times \.\.\. \fInhgt-1\fP)
vertices, with inter-cluster link
weights of \fIw0\fP, \fIw1\fP, \.\.\. \fIwhgt-1\fP.
.TP
.B
mesh2D \fIdimX\fP \fIdimY\fP
2D mesh of \fIdimX\fP times \fIdimY\fP nodes.
.TP
.B
mesh3D \fIdimX\fP \fIdimY\fP \fIdimZ\fP
23 mesh of \fIdimX\fP times \fIdimY\fP times \fIdimZ\fP nodes.
.TP
.B
torus2D \fIdimX\fP \fIdimY\fP
2D torus of \fIdimX\fP times \fIdimY\fP nodes.
.TP
.B
torus3D \fIdimX\fP \fIdimY\fP \fIdimZ\fP
3D torus of \fIdimX\fP times \fIdimY\fP times \fIdimZ\fP nodes.
.PP
Other target topologies can be created from their source graph
description by using the \fBamk_grf\fP(1) command. In this case, the
target description will begin with the code name \fBdeco\fP.
.SH MAPPINGS
Mappings are represented by as many lines as there are vertices in
the source graph. Each of these lines is made of two figures: the
number of the vertex (or its label if source graph vertices are
labeled) and the index of the target vertex to which it has been
assigned. Target vertex indices range from 0 to the number of
vertices in the target architecture (that is, the number of parts)
minus one.
.PP
This block of lines is always preceded by the number of such
lines. In most cases, since full mappings are requested, the number
of lines is equal to the number of vertices in the source graph.
.SH EXAMPLES
Run \fBgpart\fP to compute a partition into 7 parts of graph 'brol.grf' and
save the resulting ordering to file 'brol.map'.
.PP
.nf
.fam C
$ gpart 7 brol.grf brol.map
.fam T
.fi
Run \fBgmap\fP to compute a partition, into 3 parts of respective weights
1, 2 and 4, of graph 'brol.grf' and save the resulting mapping to
file 'brol.map'. The dash '-' standard file name is used so that the
target architecture description is read from the standard input,
through the pipe, as provided by the 'echo' shell command.
.PP
.nf
.fam C
$ echo "cmpltw 3 1 2 4" | gmap brol.grf - brol.map
.fam T
.fi
.SH SEE ALSO
\fBamk_grf\fP(1), \fBacpl\fP(1), \fBgmtst\fP(1), \fBdgmap\fP(1).
.PP
Scotch user's manual.
.SH AUTHOR
Francois Pellegrini <francois.pellegrini@labri.fr>
|