* attila/doc/man1 :

- Un chti'oubli...

alliance/src/nero/doc/man1/Makefile.am

@@ -0,0 +1,5 @@
+
+
+man_MANS = nero.1
+
+EXTRA_DIST = $(man_MANS)

alliance/src/nero/doc/man1/nero.1

@@ -0,0 +1,113 @@
+.\\" auto-generated by docbook2man-spec $Revision: 1.1 $
+.TH "NERO" "1" "02 October 2002" "ASIM/LIP6" "Alliance - nero User's Manual"
+.SH NAME
+nero \- Negotiating Router
+.SH SYNOPSIS
+.sp
+\fBnero\fR [ \fB-h\fR ]  [ \fB-v\fR ]  [ \fB-V\fR ]  [ \fB-c\fR ]  [ \fB-2\fR ]  [ \fB-3\fR ]  [ \fB-4\fR ]  [ \fB-5\fR ]  [ \fB-6\fR ]  [ \fB-L\fR ]  [ \fB-G\fR ] 
+ [ \fB--help\fR ]  [ \fB--verbose\fR ]  [ \fB--very-verbose\fR ]  [ \fB--core-dump\fR ]  [ \fB--local\fR ]  [ \fB--global\fR ] 
+ [ \fB--place \fIplacement\fB\fR ]  \fB\fInetlist\fB\fR \fB\fIlayout\fB\fR
+.SH "DESCRIPTION"
+.PP
+nero is a simple router suited for small
+academic designs. Currently it can process designs of size up to
+4K gates.
+.SS "GLOBAL ROUTING"
+.PP
+A design is considered as \fBbig\fR if it
+contains nets which half perimeter is greater than 800 lambdas.
+Global routing is used on big designs.
+.PP
+In nero, "global routing" means that
+the longuests nets are completly routed in a first step with only
+routing layers numbers 3 & 4. Then the smaller nets are routed with all
+avalaibles layers. This implies that when global routing is
+used, the number of routing layers is forced to at least 4.
+In each step, the nets are routed from the shortest to the
+longuest with the same routing algorithm.
+.SH "INVOCATION"
+.PP
+nero mandatory arguments\ :
+.TP 0.2i
+\(bu
+\fInetlist\fR\ : the name of
+the design, whithout any extention. Please note that unless a
+\fB-p\fR \fIplacement\fR option
+is given, the file holding the placement is expected to have the
+same name as the netlist file (short of the extention).
+.TP 0.2i
+\(bu
+\fIlayout\fR\ : the name of
+the resulting routed layout. Be careful of not giving the same
+name as the netlist, in this case the initial placement will be
+silently overwritten.
+.PP
+.PP
+nero optionnals arguments\ :
+.TP 0.2i
+\(bu
+\fB-h\fR, \fB--help\fR\ :
+print help.
+.TP 0.2i
+\(bu
+\fB-v\fR, \fB--verbose\fR\ :
+be verbose.
+.TP 0.2i
+\(bu
+\fB-V\fR, \fB--very-verbose\fR\ :
+be unbearably verbose.
+.TP 0.2i
+\(bu
+\fB-c\fR, \fB--core-dump\fR\ :
+do not prevent the generation of core dumps in case of a crash.
+Mostly used during the development stage.
+.TP 0.2i
+\(bu
+\fB-2\fR,
+\fB-3\fR,
+\fB-4\fR,
+\fB-5\fR,
+\fB-6\fR\ : set the numbers of layers
+used for routing. The default for small designs is 2 and 4 for big
+ones. When global routing took place, the router will uses at least
+4 routing layers.
+.TP 0.2i
+\(bu
+\fB-L\fR, \fB--local\fR\ :
+turn off the global routing stage, whatever the size of the design.
+Be warned that this will enormously slow down the routing process.
+.TP 0.2i
+\(bu
+\fB-G\fR, \fB--global\fR\ :
+turn on the global routing regardless of the size of the design.
+.TP 0.2i
+\(bu
+\fB-p\fR \fIplacement\fR,
+\fB--place\fR \fIplacement\fR\ :
+specify a name for the placement file different from the netlist
+name.
+.PP
+.SH "EXAMPLE"
+.PP
+Route the netlist amd2901_core.vst, using
+the placement file amd2901_core_p.ap and store the
+result in amd2901_core_r.
+.sp
+.nf
+$ \fBexport MBK_IN_LO=vst\fR
+$ \fBexport MBK_IN_PH=ap\fR
+$ \fBexport MBK_OUT_PH=ap\fR
+$ \fBnero -V --place amd2901_core_p amd2901_core amd2901_core_r\fR
+    
+.sp
+.fi
+.SH "SEE ALSO"
+.PP
+\fBMBK_IN_LO\fR(1),
+\fBMBK_OUT_LO\fR(1),
+\fBMBK_IN_PH\fR(1),
+\fBMBK_OUT_PH\fR(1),
+\fBocp\fR(1),
+\fBocr\fR(1),
+\fBdruc\fR(1),
+\fBcougar\fR(1)