[ghc-hetmet.git] / docs / installing.vsgml

<!doctype linuxdoc system>
<article>

<title>Building and Installing the Glasgow Functional Programming Tools Suite
Version 4.02
<author>The GHC Team,
Department of Computing Science,
University of Glasgow,
Glasgow, Scotland,
G12 8QQ.

Email: @glasgow-haskell-{users,bugs}@@dcs.gla.ac.uk@
<date>April 1998</date>

<abstract>

This guide is intended for people who want to install or modify
programs from the Glasgow @fptools@ suite (as distinct from those
who merely want to <em/run/ them).

</abstract>

<toc>

<sect>Getting the Glasgow @fptools@ suite
<label id="sec:getting">
<p>

Building the Glasgow tools <em/can/ be complicated, mostly because
there are so many permutations of what/why/how, e.g., ``Build Happy
with HBC, everything else with GHC, leave out profiling, and test it
all on the `real' NoFib programs.''  Yeeps!

Happily, such complications don't apply to most people.  A few common
``strategies'' serve most purposes.  Pick one and proceed
as suggested:

<descrip>

<tag><idx>Binary distribution</idx>.</tag>
 If your only purpose is to install some of the @fptools@ suite
then the easiest thing to do is to get a binary distribution. In the
binary distribution everything is pre-compiled for your particular
machine architecture and operating system, so all you should have to
do is install the binaries and libraries in suitable places.  Section
<ref id="sec:installing-bin-distrib" name="Installing a Binary
Distribution"> describes how to do this.

A binary distribution may not work for you for two reasons.  First, we
may not have built the suite for the particular architecture/OS
platform you want. That may be due to lack of time and energy (in
which case you can get a source distribution and build from it; see
below).  Alternatively, it may be because we haven't yet ported the
suite to your architecture, in which case you are considerably worse
off.

The second reason a binary distribution may not be what you want is
if you want to read or modify the souce code.

<tag><idx>Source distribution</idx>.</tag> You have a supported
platform, but (a)~you like the warm fuzzy feeling of compiling things
yourself; (b)~you want to build something ``extra''---e.g., a set of
libraries with strictness-analysis turned off; or (c)~you want to hack
on GHC yourself.

A source distribution contains complete sources for one or more
projects in the @fptools@ suite.  Not only that, but the more awkward
machine-independent steps are done for you.  For example, if you don't
have @flex@<ncdx/flex/ you'll find it convenient that the source
distribution contains the result of running @flex@ on the lexical
analyser specification.  If you don't want to alter the lexical
analyser then this saves you having to find and install @flex@. You
will still need a working version of GHC on your machine in order to
compile (most of) the sources, however.

We make source distributions more frequently than binary
distributions; a release that comes with pre-compiled binaries
is considered a major release, i.e., a release that we have some
confidence will work well by having tested it (more) thoroughly.

Source-only distributions are either bugfix releases or snapshots of
current state of development. The release has undergone some testing.
Source releases of 4.xx can be compiled up using 2.10 or later.

<tag/Build GHC from intermediate C @.hc@ files<nidx/hc files/:/ You
need a working GHC to use a source distribution. What if you don't
have a working GHC? Then you have no choice but to ``bootstrap'' up
from the intermediate C (@.hc@) files that we provide.  Building GHC
on an unsupported platform falls into this category.  Please see
Section <ref id="sec:booting-from-C" name="Booting From C">.

Once you have built GHC, you can build the other Glasgow tools with
it.

In theory, you can (could?) build GHC with another Haskell compiler
(e.g., HBC). We haven't tried to do this for ages and it almost
certainly doesn't work any more (for tedious reasons).

<tag/The CVS repository./

We make source distributions slightly more often than binary
distributions; but still infrequently.  If you want more up-to-the
minute (but less tested) source code then you need to get access to
our CVS repository.

All the @fptools@ source code is held in a CVS repository. CVS is a
pretty good source-code control system, and best of all it works over
the network.

The repository holds source code only. It holds no mechanically
generated files at all.  So if you check out a source tree from CVS
you will need to install every utility so that you can build all the
derived files from scratch.

More information about our CVS repository is available at <url
name="The Fptools CVS Cheat Sheet"
url="http://www.dcs.gla.ac.uk/fp/software/ghc/cvs-cheat-sheet.html">.
</descrip>

If you are going to do any building from sources (either from a source
distribution or the CVS repository) then you need to read all of this
manual in detail.

<sect>Things to check before you start typing
<p>

Here's a list of things to check before you get started.
<enum>
<item>
<idx>Disk space needed</idx>: About 30MB (five hamburgers' worth) of disk space
for the most basic binary distribution of GHC; more for some
platforms, e.g., Alphas.  An extra ``bundle'' (e.g., concurrent
Haskell libraries) might take you to 8--10 hamburgers.

You'll need over 100MB (say, 20 hamburgers' worth) if you need to
build the basic stuff from scratch.


All of the above are <em/estimates/ of disk-space needs.(I don't yet
know the disk requirements for the non-GHC tools).

<item>
Use an appropriate machine, compilers, and things.

SPARC boxes, DEC Alphas running OSF/1, and PCs running Linux, FreeBSD,
or Solaris are all fully supported.  MIPS, AIX, Win32 and HP boxes are
in pretty good shape.  Section <ref id="sec:port-info" name="Port Info">
gives the full run-down on ports or lack thereof.

NOTE: as of version 4.00, we lost a few ports.  All of the x86 ports
are working, as is the Sparc/Solaris port, but the rest will need a
little work.  Please contact us if you can provide hardware cycles
and/or porting expertise.

<item> Be sure that the ``pre-supposed'' utilities are installed.
Section <ref id="sec:pre-supposed" name="Installing Pre-Supposed
Utilities"> elaborates.

<item> If you have any problem when building or installing the Glasgow
tools, please check the ``known pitfalls'' (Section <ref
id="sec:build-pitfalls" name="Building Pitfalls">).  Also check the
known bugs page: <url
url="http://www.dcs.gla.ac.uk/fp/software/ghc/ghc-bugs.html">.
<nidx/known bugs/
<nidx/bugs, known/

If you feel there is still some shortcoming in our procedure or
instructions, please report it.

For GHC, please see the bug-reporting section of the User's guide
(separate document), to maximise the usefulness of your report.
<nidx/bugs, reporting/

If in doubt, please send a message to @glasgow-haskell-bugs@@dcs.gla.ac.uk@.
<nidx/bugs, mailing list/
</enum>


<sect>What machines the Glasgow tools run on
<label id="sec:port-info">
<p>
<nidx>ports, GHC</nidx>
<nidx>GHC ports</nidx>
<nidx>supported platforms</nidx>
<nidx>platforms, supported</nidx>

The main question is whether or not the Haskell compiler (GHC) runs on
your platform.

A ``platform'' is a architecture/manufacturer/operating-system
combination, such as @sparc-sun-solaris2@.  Other common ones are
@alpha-dec-osf2@, @hppa1.1-hp-hpux9@, @i386-unknown-linux@,
@i386-unknown-solaris2@, @i386-unknown-freebsd@,
@i386-unknown-cygwin32@, @m68k-sun-sunos4@, @mips-sgi-irix5@,
@sparc-sun-sunos4@, @sparc-sun-solaris2@, @powerpc-ibm-aix@.

Bear in mind that certain ``bundles'', e.g. parallel Haskell, may not
work on all machines for which basic Haskell compiling is supported.

Some libraries may only work on a limited number of platforms; for
example, a sockets library is of no use unless the operating system
supports the underlying BSDisms.

<sect1>What platforms the Haskell compiler (GHC) runs on
<p>
<nidx>fully-supported platforms</nidx>
<nidx>native-code generator</nidx>
<nidx>registerised ports</nidx>
<nidx>unregisterised ports</nidx>

The GHC hierarchy of Porting Goodness: (a)~Best is a native-code
generator; (b)~next best is a ``registerised''
port; (c)~the bare minimum is an ``unregisterised'' port.
(``Unregisterised'' is so terrible that we won't say more about it).

We use Sparcs running Solaris 2.5, x86 boxes running FreeBSD and
Linux, and DEC~Alphas running OSF/1~V2.0, so those are the
``fully-supported'' platforms, unsurprisingly.  All have native-code
generators, for quicker compilations.  The native-code generator for
iX86 platforms (e.g., Linux ELF) is <em/nearly/ working; but is not
turned on by default.

Here's everything that's known about GHC ports.  We identify platforms
by their ``canonical'' CPU/Manufacturer/OS triple.

Note that some ports are fussy about which GCC version you use; or
require GAS; or ...

<descrip>
<tag/alpha-dec-osf1:/
<nidx>alpha-dec-osf1: fully supported</nidx>

(We have OSF/1 V3.0.) Fully supported, including native-code
generator.  We recommend GCC 2.6.x or later.

<tag/sparc-sun-sunos4:/
<nidx>sparc-sun-sunos4: fully supported</nidx>

Fully supported, including native-code generator.

<tag/sparc-sun-solaris2:/ 
<nidx>sparc-sun-solaris2: fully supported</nidx>

Fully supported, including native-code generator.  A couple of quirks,
though: (a)~the profiling libraries are bizarrely huge when compiled
with object splitting; (b)~the default @xargs@<ncdx/xargs/ program is
atrociously bad for building GHC libraries (see Section <ref
id="sec:pre-supposed" name="Installing Pre-Supposed Utilities"> for
details).

<tag/HP-PA box running HP/UX 9.x:/
<nidx>hppa1.1-hp-hpux: registerised port</nidx>

Works registerised.  No native-code generator.  For GCC, you're best
off with one of the Utah releases of GCC~2.6.3 (`u3' or later), from
@jaguar.cs.utah.edu@.  We think a straight GCC 2.7.x works,
too.

Concurrent/Parallel Haskell probably don't work (yet).
<nidx>hppa1.1-hp-hpux: concurrent---no</nidx>
<nidx>hppa1.1-hp-hpux: parallel---no</nidx>

<tag/i386-*-linux (PCs running Linux---ELF format):/
<nidx>i386-*-linux: registerised port</nidx>

GHC works registerised.  You <em/must/ have GCC 2.7.x or later.  The
iX86 native-code generator is <em/nearly/ there, but it isn't turned
on by default.

Profiling works, and Concurrent Haskell works.
<nidx>i386-*-linux: profiling---yes</nidx>
<nidx>i386-*-linux: concurrent---yes</nidx>
Parallel Haskell probably works.
<nidx>i386-*-linux: parallel---maybe</nidx>

On old Linux a.out systems: should be the same.
<nidx>i386-*-linuxaout: registerised port</nidx>

<tag>i386-*-freebsd (PCs running FreeBSD 2.2 or higher, and
NetBSD/OpenBSD using FreeBSD emulation):</tag> 
<nidx>i386-*-freebsd:registerised port</nidx> 

GHC works registerised. Supports same set of bundles as the above.

<nidx>i386-*-freebsd: profiling---yes</nidx>
<nidx>i386-*-freebsd: concurrent---yes</nidx>
<nidx>i386-*-freebsd: parallel---maybe</nidx>

<tag/i386-unknown-cygwin32:/
<nidx>i386-unknown-cygwin32: fully supported</nidx>

Fully supported under Win95/NT, including a native code
generator. Requires the @cygwin32@ compatibility library and a
healthy collection of GNU tools (i.e., gcc, GNU ld, bash etc.)
Profiling works, so does Concurrent Haskell.  

<nidx>i386-*-cygwin32: profiling---yes</nidx> 
<nidx>i386-*-cygwin32: concurrent---yes</nidx>

<tag/mips-sgi-irix5:/
<nidx>mips-sgi-irix5: registerised port</nidx>

GHC works registerised (no native-code generator).  I suspect any
GCC~2.6.x (or later) is OK.  The GCC that I used was built with
@--with-gnu-as@; turns out that is important!

Concurrent/Parallel Haskell probably don't work (yet).
Profiling might work, but it is untested.
<nidx>mips-sgi-irix5: concurrent---no</nidx>
<nidx>mips-sgi-irix5: parallel---no</nidx>
<nidx>mips-sgi-irix5: profiling---maybe</nidx>

<tag/mips-sgi-irix6:/
<nidx>mips-sgi-irix6: registerised port</nidx>

Thanks to the fine efforts of Tomasz Cholewo <htmlurl
url="mailto:tjchol01@@mecca.spd.louisville.edu"
name="tjchol01@@mecca.spd.louisville.edu">, GHC works registerised (no
native code generator) under IRIX 6.2 and 6.3. Depends on having
specially tweaked version of gcc-2.7.2 around, which can be downloaded
from <url url="http://mecca.spd.louisville.edu/~tjchol01/software/">.

Profiling works, Concurrent/Parallel Haskell might work (AFAIK, untested).
<nidx>mips-sgi-irix6: concurrent---maybe</nidx>
<nidx>mips-sgi-irix6: parallel---maybe</nidx>
<nidx>mips-sgi-irix6: profiling---yes</nidx>

<tag/powerpc-ibm-aix:/
<nidx>powerpc-ibm-aix: registerised port</nidx>
GHC works registerised (no native-code generator..yet).
I suspect 2.7.x is what you need together with this.

Concurrent/Parallel Haskell probably don't work (yet).
Profiling might work, but it is untested.
<nidx>mips-sgi-irix5: concurrent---no</nidx>
<nidx>mips-sgi-irix5: parallel---no</nidx>
<nidx>mips-sgi-irix5: profiling---maybe</nidx>

<tag/m68k-apple-macos7 (Mac, using MPW):/
<nidx>m68k-apple-macos7: historically ported</nidx>
Once upon a time, David Wright in Tasmania has actually
gotten GHC to run on a Macintosh.  Ditto James Thomson here at Glasgow.
You may be able to get Thomson's from here.  (Not sure that it will
excite you to death, but...)

No particularly recent GHC is known to work on a Mac.

<tag/m68k-next-nextstep3:/
<nidx>m68k-next-nextstep3: historically ported</nidx>
Carsten Schultz succeeded with a ``registerised'' port of GHC~0.29.
There's probably a little bit-rot since then, but otherwise it should
still be fine.

Concurrent/Parallel Haskell probably won't work (yet).
<nidx>m68k-next-nextstep3: concurrent---no</nidx>
<nidx>m68k-next-nextstep3: parallel---no</nidx>

<tag/m68k-sun-sunos4 (Sun3):/ <nidx>m68k-sun-sunos4: registerised
port</nidx> GHC 2.0x and 3.0x haven't been tried on a Sun3.  GHC~0.26
worked registerised.  No native-code generator.

Concurrent/Parallel Haskell probably don't work (yet).
<nidx>m68k-sun-sunos4: concurrent---no</nidx>
<nidx>m68k-sun-sunos4: parallel---no</nidx>
</descrip>

<sect1>What machines the other tools run on
<p>

Unless you hear otherwise, the other tools work if GHC works.

Haggis requires Concurrent Haskell to work.
<nidx>Haggis, Concurrent Haskell</nidx>


<sect>Installing from binary distributions
<p>
<label id="sec:installing-bin-distrib">
<nidx>binary installations</nidx>
<nidx>installation, of binaries</nidx>

Installing from binary distributions is easiest, and recommended!
(Why binaries?  Because GHC is a Haskell compiler written in Haskell,
so you've got to ``bootstrap'' it, somehow.  We provide
machine-generated C-files-from-Haskell for this purpose, but it's
really quite a pain to use them.  If you must build GHC from its
sources, using a binary-distributed GHC to do so is a sensible way to
proceed. For the other @fptools@ programs, many are written in Haskell,
so binary distributions allow you to install them without having a Haskell compiler.)


<sect1>Bundle structure<p>
<nidx>bundles of binary stuff</nidx>

Binary distributions come in ``bundles,'' one bundle per file called
@<bundle>-<platform>.tar.gz@.  (See Section <ref
id="sec:port-info" name="Porting Information"> for what a platform
is.)  Suppose that you untar a binary-distribution bundle, thus:

<tscreen><verb>
  % cd /your/scratch/space
  % gunzip < ghc-x.xx-sun-sparc-solaris2.tar.gz | tar xvf -
</verb></tscreen>

Then you should find a single directory, @fptools@, with the following
structure:

<nidx>binary distribution, layout</nidx>
<nidx>directory layout (binary distributions)</nidx>
<descrip>

<tag>@Makefile.in@</tag> the raw material from which the @Makefile@
will be made (Section <ref id="sec:install" name="Installation">).

<tag>@configure@</tag> the configuration script (Section <ref
id="sec:install" name="Installing">).

<tag>@README@</tag> Contains this file summary.

<tag>@INSTALL@</tag> Contains this description of how to install
the bundle.

<tag>@ANNOUNCE@</tag> The announcement message for the bundle.

<tag>@NEWS@</tag> release notes for the bundle -- a longer version
of @ANNOUNCE@.  For GHC, the release notes are contained in the User
Guide and this file isn't present.

<tag>@bin/<platform>@</tag> contains platform-specific executable
files to be invoked directly by the user.  These are the files that
must end up in your path.

<tag>@lib/<platform>/@</tag> contains platform-specific support
files for the installation.  Typically there is a subdirectory for
each @fptools@ project, whose name is the name of the project with its
version number.  For example, for GHC there would be a sub-directory
@ghc-x.xx@/ where @x.xx@ is the version number of GHC in the bundle.

These sub-directories have the following general structure:

<descrip>
<tag>@libHS.a@ etc:</tag> supporting library archives.
<tag>@ghc-iface.prl@ etc:</tag> support scripts.
<tag>@import/@</tag> <idx>Interface files</idx> (@.hi@) for the prelude.
<tag>@include/@</tag> A few C @#include@ files.
</descrip>

<tag>@share/@</tag> contains platform-independent support files
for the installation.  Again, there is a sub-directory for each
@fptools@ project.

<tag>@info/@</tag> contains Emacs info documentation files (one
sub-directory per project).

<tag>@html/@</tag> contains HTML documentation files (one
sub-directory per project).

<tag>@man/@</tag> contains Unix manual pages.

</descrip>

This structure is designed so that you can unpack multiple bundles
(including ones from different releases or platforms) into a single
@fptools@ directory<footnote>this doesn't work at the
moment</footnote>:

<tscreen><verb>
  % cd /your/scratch/space
  % gunzip < ghc-x.xx-sun-sparc-solaris2.tar.gz | tar xvf -
  % gunzip < happy-x.xx-sun-sparc-sunos4.tar.gz | tar xvf -
</verb></tscreen>

When you do multiple unpacks like this, the top level @Makefile@,
@README@, and @INSTALL@ get overwritten each time.
That's fine -- they should be the same.  Likewise, the
@ANNOUNCE-<bundle>@ and @NEWS-<bundle>@
files will be duplicated across multiple platforms, so they will be
harmlessly overwritten when you do multiple unpacks.  Finally, the
@share/@ stuff will get harmlessly overwritten when you do
multiple unpacks for one bundle on different platforms.

<sect2>Installing<p>
<label id="sec:install">

OK, so let's assume that you have unpacked your chosen bundles into a
scratch directory @fptools@. What next? Well, you will at least need
to run the @configure@<ncdx/configure/ script by changing your
directory to @fptools@ and typing @./configure@.  That should convert
@Makefile.in@ to @Makefile@.

<nidx/installing in-place/
<nidx/in-place installation/

You can now either start using the tools <em/in-situ/ without going
through any installation process, just type @make in-place@ to set the
tools up for this. You'll also want to add the path which @make@ will
now echo to your @PATH@ environment variable. This option is useful if
you simply want to try out the package and/or you don't have the
necessary priviledges (or inclination) to properly install the tools
locally. Note that if you do decide to install the package `properly'
at a later date, you have to go through the installation steps that
follows.

To install an @fptools@ package, you'll have to do the following:

<enum>
<item> Edit the @Makefile@ and check the settings of the following variables:

<nidx/directories, installation/
<nidx/installation directories/

<descrip>
<tag>@platform@</tag> the platform you are going to install for.

<tag>@bindir@</tag> the directory in which to install user-invokable
binaries.

<tag>@libdir@</tag> the directory in which to install
platform-dependent support files.

<tag>@datadir@</tag> the directory in which to install
platform-independent support files.

<tag>@infodir@</tag> the directory in which to install Emacs info
files.

<tag>@htmldir@</tag> the directory in which to install HTML
documentation.

<tag>@dvidir@</tag> the directory in which to install DVI
documentation.
</descrip>

The values for these variables can be set through invocation of the
@configure@<ncdx/configure/ script that comes with the distribution,
but doing an optical diff to see if the values match your expectations
is always a Good Idea.

<em>Instead of running @configure@, it is perfectly OK to copy
@Makefile.in@ to @Makefile@ and set all these variables
directly yourself.  But do it right!</em>

<item>Run @make install@.  This <em/ should/ work with ordinary Unix
@make@ -- no need for fancy stuff like GNU @make@. 

<item>@rehash@ (t?csh or zsh users), so your shell will see the new
stuff in your bin directory.

<item> Once done, test your ``installation'' as suggested in Section
<ref id="sec:GHC-test" name="Testing GHC">.  Be sure to use a @-v@
option, so you can see exactly what pathnames it's using.

If things don't work as expected, check the list of know pitfalls in
Section <ref id="sec:build-pitfalls" name="Building Pitfalls">.
</enum>

<nidx/link, installed as ghc/
When installing the user-invokable binaries, this installation
procedure will install GHC as @ghc-x.xx@ where @x.xx@ is the version
number of GHC.  It will also make a link (in the binary installation
directory) from @ghc@ to @ghc-x.xx@.  If you install multiple versions
of GHC then the last one ``wins'', and ``@ghc@'' will invoke the last
one installed.  You can change this manually if you want.  But
regardless, @ghc-x.xx@ should always invoke GHC version @x.xx@.

<sect1>What bundles there are
<p>

<nidx/bundles, binary/
There are plenty of ``non-basic'' GHC bundles.  The files for them are
called @ghc-x.xx-<bundle>-<platform>.tar.gz@, where
the @<platform>@ is as above, and @<bundle>@ is one
of these:

<descrip>

<tag>@prof@:</tag>  Profiling with cost-centres.  You probably want this.
<nidx/profiling bundles/
<nidx/bundles, profiling/

<tag>@conc@:</tag> Concurrent Haskell features.  You may want this.
<nidx/concurrent bundles/
<nidx/bundles, concurrent/

<tag>@par@:</tag> Parallel Haskell features (sits on top of PVM).
You'll want this if you're into that kind of thing.
<nidx/parallel bundles/
<nidx/bundles, parallel/

<tag>@gran@:</tag> The ``GranSim'' parallel-Haskell simulator
(hmm... mainly for implementors).
<nidx/bundles, gransim/
<nidx/gransim bundles/

<tag>@ticky@:</tag> ``Ticky-ticky'' profiling; very detailed
information about ``what happened when I ran this program''---really
for implementors.
<nidx/bundles, ticky-ticky/
<nidx/ticky-ticky bundles/

<tag>@prof-conc@:</tag> Cost-centre profiling for Concurrent Haskell.
<nidx/bundles, profiled-concurrent/
<nidx/profiled-concurrent bundles/

<tag>@prof-ticky@:</tag>  Ticky-ticky profiling for Concurrent Haskell.
<nidx/bundles, profiled-ticky/
<nidx/ticky-concurrent bundles/
</descrip>

One likely scenario is that you will grab <em/three/ binary
bundles---basic, profiling, and concurrent.  We don't usually make the
rest, although you can build them yourself from a source distribution.

<sect1>Testing that GHC seems to be working
<label id="sec:GHC-test">
<p>
<nidx>testing a new GHC</nidx>

The way to do this is, of course, to compile and run <em/this/ program
(in a file @Main.hs@):

<tscreen><verb>
main = putStr "Hello, world!\n"
</verb></tscreen>

Compile the program, using the @-v@ (verbose) flag to verify that
libraries, etc., are being found properly:
<tscreen><verb>
% ghc -v -o hello Main.hs
</verb></tscreen>

Now run it:
<tscreen><verb>
% ./hello
Hello, world!
</verb></tscreen>

Some simple-but-profitable tests are to compile and run the notorious
@nfib@<ncdx/nfib/ program, using different numeric types.  Start with
@nfib :: Int -> Int@, and then try @Integer@, @Float@, @Double@,
@Rational@ and perhaps the overloaded version.  Code for this is
distributed in @ghc/misc/examples/nfib/@ in a source distribution.

For more information on how to ``drive'' GHC, either do @ghc -help@ or
consult the User's Guide (distributed in several pre-compiled formats
with a binary distribution, or in source form in
@ghc/docs/users_guide@ in a source distribution).

<sect>Installing pre-supposed utilities
<label id="sec:pre-supposed">
<nidx>pre-supposed utilities</nidx>
<nidx>utilities, pre-supposed</nidx>
<p>

Here are the gory details about some utility programs you may need;
@perl@ and @gcc@ are the only important ones. (<idx/PVM/ is important
if you're going for Parallel Haskell.)  The <tt><cdx/configure/</tt>
script will tell you if you are missing something.

<descrip>
<tag>Perl:</tag>
<nidx>pre-supposed: Perl</nidx>
<nidx>Perl, pre-supposed</nidx>
<em/You have to have Perl to proceed!/ Perl is a language quite good
for doing shell-scripty tasks that involve lots of text processing.
It is pretty easy to install.

Perl~5 is required.  For Win32 platforms, we strongly suggest you pick
up a port of Perl~5 for @cygwin32@, as the common Hip/ActiveWare port
of Perl is not Cool Enough for our purposes.

Perl should be put somewhere so that it can be invoked by the @#!@
script-invoking mechanism. (I believe @/usr/bin/perl@ is preferred;
we use @/usr/local/bin/perl@ at Glasgow.)  The full pathname should
be less than 32 characters long.

<tag>GNU C (@gcc@):</tag>
<nidx>pre-supposed: GCC (GNU C compiler)</nidx>
<nidx>GCC (GNU C compiler), pre-supposed</nidx>
Version 2.7.2.x is known to work.

If your GCC dies with ``internal error'' on some GHC source file,
please let us know, so we can report it and get things improved.
(Exception: on @iX86@ boxes---you may need to fiddle with GHC's
@-monly-N-regs@ option; see the User's Guide)

<nidx/EGCS/ EGCS (the Enhanced GNU Compiler Suite) may or may not
work, we haven't tested it fully yet.

<tag>PVM version 3:</tag>
<nidx>pre-supposed: PVM3 (Parallel Virtual Machine)</nidx>
<nidx>PVM3 (Parallel Virtual Machine), pre-supposed</nidx>

PVM is the Parallel Virtual Machine on which Parallel Haskell programs
run.  (You only need this if you plan to run Parallel Haskell.
Concurent Haskell, which runs concurrent threads on a uniprocessor
doesn't need it.)  Underneath PVM, you can have (for example) a
network of workstations (slow) or a multiprocessor box (faster).

The current version of PVM is 3.3.11; we use 3.3.7.  It is readily
available on the net; I think I got it from @research.att.com@, in
@netlib@.

A PVM installation is slightly quirky, but easy to do.  Just follow
the @Readme@ instructions.

<tag>@xargs@ on Solaris2:</tag>
<nidx>xargs, presupposed (Solaris only)</nidx>
<nidx>Solaris: alternative xargs</nidx>
The GHC libraries are put together with something like:
<tscreen><verb>
find bunch-of-dirs -name '*.o' -print | xargs ar q ...
</verb></tscreen>
Unfortunately the Solaris @xargs@ (the shell-script equivalent
of @map@) only ``bites off'' the @.o@ files a few at a
time---with near-infinite rebuilding of the symbol table in
the @.a@ file.

The best solution is to install a sane @xargs@ from the GNU
findutils distribution.  You can unpack, build, and install the GNU
version in the time the Solaris @xargs@ mangles just one GHC
library.

<tag>Autoconf:</tag>
<nidx>pre-supposed: Autoconf</nidx>
<nidx>Autoconf, pre-supposed</nidx>

GNU Autoconf is used to build the @configure@ script from
@configure.in@ in a source distribution.  If you modify
@configure.in@, you'll need @autoconf@ to regenerate @configure@.

<tag>@bash@ (Parallel Haskell only):</tag>
<nidx>bash, presupposed (Parallel Haskell only)</nidx>
Sadly, the @gr2ps@ script, used to convert ``parallelism profiles''
to PostScript, is written in Bash (GNU's Bourne Again shell).
This bug will be fixed (someday).

<tag>Flex:</tag> 
<nidx>pre-supposed: flex</nidx> 
<nidx>flex, pre-supposed</nidx>

This is a quite-a-bit-better-than-Lex lexer.  Used to build GHC's
lexer, and a couple of utilities in @glafp-utils@.  On our machines,
the version in @/bin@ doesn't work; you need the GNU version.  Find
out by saying @flex --version@ (our current version is 2.5.4, but
maybe earlier ones will work).  If it doesn't know about the
@--version@ flag, it ain't the right @flex@.

<tag>Yacc:</tag>
<nidx>pre-supposed: non-worthless Yacc</nidx>
<nidx>Yacc, pre-supposed</nidx>

If you mess with the Haskell parser, you'll need a Yacc that can cope.
The unbundled @/usr/lang/yacc@ is OK; the GNU @bison@ is OK; Berkeley
yacc, @byacc@, is not OK.

<tag>@sed@</tag>
<nidx>pre-supposed: sed</nidx>
<nidx>sed, pre-supposed</nidx>

You need a working @sed@ if you are going to build from sources.  The
build-configuration stuff needs it.  GNU sed version 2.0.4 is no good!
It has a bug in it that is tickled by the build-configuration.  2.0.5
is ok. Others are probably ok too (assuming we don't create too
elaborate configure scripts..)

</descrip>

One @fptools@ project is worth a quick note at this point, because it
is useful for all the others: @glafp-utils@ contains several utilities
which aren't particularly Glasgow-ish, but Occasionally Indispensable.
Like @lndir@ for creating symbolic link trees.

<sect1> Tools for building the Documentation
<label id="pre-supposed-doc-tools">
<p>

The following additional tools are required if you want to format the
documentation that comes with the @fptools@ projects:

<descrip>
<tag>SGML-Tools:</tag>
<nidx>pre-supposed: SGML-Tools</nidx>
<nidx>SGML-Tools, pre-supposed</nidx>

All our documentation is written in SGML, using the LinuxDoc DTD that
comes with the SGML-Tools, which is the most shrink-wrapped SGML suite
that we could find.  Should unpack and build painlessly on most
architectures, and you can use it to generate HTML, Info, LaTeX (and
hence DVI and Postscript), Groff, and plain text output from any
LinuxDoc source file (including this manual).  Sources are available
from <url name="The SGML-Tools Web Page"
url="http://www.sgmltools.org/">

<tag>TeX:</tag>
<nidx>pre-supposed: TeX</nidx>
<nidx>TeX, pre-supposed</nidx>
A decent TeX distribution is required if you want to produce printable
documentation.  We recomment teTeX, which includes just about
everything you need.
</descrip>


<sect>Building from source
<label id="sec:building-from-source">
<nidx>Building from source</nidx>
<nidx>Source, building from</nidx>
<p>

You've been rash enough to want to build some of
the Glasgow Functional Programming tools (GHC, Happy,
nofib, etc) from source.  You've slurped the source,
from the CVS repository or from a source distribution, and
now you're sitting looking at a huge mound of bits, wondering
what to do next.

Gingerly, you type @make@.  Wrong already!

This rest of this guide is intended for duffers like me, who aren't
really interested in Makefiles and systems configurations, but who
need a mental model of the interlocking pieces so that they can make
them work, extend them consistently when adding new software, and lay
hands on them gently when they don't work.

<sect1>Your source tree
<label id="sec:source-tree">
<p>

The source code is held in your <em/source tree/.
The root directory of your source tree <em/must/
contain the following directories and files:

<itemize>
<item> @Makefile@: the root Makefile.
<item> @mk/@: the directory that contains the
main Makefile code, shared by all the
@fptools@ software.
<item> @configure.in@, @config.sub@, @config.guess@:
these files support the configuration process.
<item> @install-sh@.
</itemize>

All the other directories are individual <em/projects/ of the
@fptools@ system --- for example, the Glasgow Haskell Compiler
(@ghc@), the Happy parser generator (@happy@), the @nofib@ benchmark
suite, and so on.  You can have zero or more of these.  Needless to
say, some of them are needed to build others.

The important thing to remember is that even if you want only one
project (@happy@, say), you must have a source tree whose root
directory contains @Makefile@, @mk/@, @configure.in@, and the
project(s) you want (@happy/@ in this case).  You cannot get by with
just the @happy/@ directory.

<sect1>Build trees
<nidx/build trees/
<nidx/link trees, for building/
<p>

While you can build a system in the source tree, we don't recommend it.
We often want to build multiple versions of our software
for different architectures, or with different options (e.g. profiling).
It's very desirable to share a single copy of the source code among
all these builds.

So for every source tree we have zero or more <em/build trees/.  Each
build tree is initially an exact copy of the source tree, except that
each file is a symbolic link to the source file, rather than being a
copy of the source file.  There are ``standard'' Unix utilities that
make such copies, so standard that they go by different names:
@lndir@<ncdx/lndir/, @mkshadowdir@<ncdx/mkshadowdir/ are two (If you
don't have either, the source distribution includes sources for the
@X11@ @lndir@ --- check out @fptools/glafp-utils/lndir@ ).

The build tree does not need to be anywhere near the source tree in
the file system.  Indeed, one advantage of separating the build tree
from the source is that the build tree can be placed in a
non-backed-up partition, saving your systems support people from
backing up untold megabytes of easily-regenerated, and
rapidly-changing, gubbins.  The golden rule is that (with a single
exception -- Section~<ref id="sec:build-config" name="Build
Configuration"> <em/absolutely everything in the build tree is either
a symbolic link to the source tree, or else is mechanically
generated/.  It should be perfectly OK for your build tree to vanish
overnight; an hour or two compiling and you're on the road again.

You need to be a bit careful, though, that any new files you create
(if you do any development work) are in the source tree, not a build tree!

Remember, that the source files in the build tree are <em/symbolic
links/ to the files in the source tree.  (The build tree soon
accumulates lots of built files like @Foo.o@, as well.)  You
can <em/delete/ a source file from the build tree without affecting
the source tree (though it's an odd thing to do).  On the other hand,
if you <em/edit/ a source file from the build tree, you'll edit the
source-tree file directly.  (You can set up Emacs so that if you edit
a source file from the build tree, Emacs will silently create an
edited copy of the source file in the build tree, leaving the source
file unchanged; but the danger is that you think you've edited the
source file whereas actually all you've done is edit the build-tree
copy.  More commonly you do want to edit the source file.)

Like the source tree, the top level of your build tree must be (a
linked copy of) the root directory of the @fptools@ suite.  Inside
Makefiles, the root of your build tree is called
@$(FPTOOLS_TOP)@<ncdx/FPTOOLS_TOP/.  In the rest of this document path
names are relative to @$(FPTOOLS_TOP)@ unless otherwise stated.  For
example, the file @ghc/mk/target.mk@ is actually
@$(FPTOOLS_TOP)/ghc/mk/target.mk@.


<sect1>Getting the build you want
<label id="sec:build-config">
<p>

When you build @fptools@ you will be compiling code on a particular
<em/host platform/, to run on a particular <em/target platform/
(usually the same as the host platform)<nidx>platform</nidx>.  The
difficulty is that there are minor differences between different
platforms; minor, but enough that the code needs to be a bit different
for each.  There are some big differences too: for a different
architecture we need to build GHC with a different native-code
generator.

There are also knobs you can turn to control how the @fptools@
software is built.  For example, you might want to build GHC optimised
(so that it runs fast) or unoptimised (so that you can compile it fast
after you've modified it.  Or, you might want to compile it with
debugging on (so that extra consistency-checking code gets included)
or off.  And so on.

All of this stuff is called the <em/configuration/ of your build.
You set the configuration using an exciting three-step process.
<descrip>

<tag>Step 1: get ready for configuration.</tag> Change directory to
@$(FPTOOLS_TOP)@ and issue the command @autoconf@<ncdx/autoconf/ (with
no arguments). This GNU program converts @$(FPTOOLS_TOP)/configure.in@
to a shell script called @$(FPTOOLS_TOP)/configure@.

Both these steps are completely platform-independent; they just mean
that the human-written file (@configure.in@) can be short, although
the resulting shell script, @configure@, and @mk/config.h.in@, are
long.

In case you don't have @autoconf@ we distribute the results,
@configure@, and @mk/config.h.in@, with the source distribution.  They
aren't kept in the repository, though.

<tag>Step 2: system configuration.</tag>
Runs the newly-created @configure@ script, thus:
<tscreen><verb>
  ./configure
</verb></tscreen>
@configure@'s mission is to scurry round your computer working out
what architecture it has, what operating system, whether it has the
@vfork@ system call, where @yacc@ is kept, whether @gcc@ is available,
where various obscure @#include@ files are, whether it's a leap year,
and what the systems manager had for lunch.  It communicates these
snippets of information in two ways:

<itemize>

<item> It translates @mk/config.mk.in@<ncdx/config.mk.in/ to
@mk/config.mk@<ncdx/config.mk/, substituting for things between
``@@@@@@@@}'' brackets.  So, ``@@HaveGcc@@'' will be replaced by
``@YES@'' or ``@NO@'' depending on what @configure@ finds.
@mk/config.mk@ is included by every Makefile (directly or indirectly),
so the configuration information is thereby communicated to all
Makefiles.

<item> It translates @mk/config.h.in@<ncdx/config.h.in/ to
@mk/config.h@<ncdx/config.h/.  The latter is @#include@d by various C
programs, which can thereby make use of configuration information.

</itemize>

@configure@ caches the results of its run in @config.cache@.  Quite
often you don't want that; you're running @configure@ a second time
because something has changed.  In that case, simply delete
@config.cache@.

<tag>Step 3: build configuration.</tag>

 Next, you say how this build of @fptools@ is to differ from the
standard defaults by creating a new file @mk/build.mk@<ncdx/build.mk/
<em/in the build tree/.  This file is the one and only file you edit
in the build tree, precisely because it says how this build differs
from the source.  (Just in case your build tree does die, you might
want to keep a private directory of @build.mk@ files, and use a
symbolic link in each build tree to point to the appropriate one.)  So
@mk/build.mk@ never exists in the source tree --- you create one in
each build tree from the template.  We'll discuss what to put in it
shortly.  

</descrip>

And that's it for configuration. Simple, eh?

What do you put in your build-specific configuration file
@mk/build.mk@?  <em/For almost all purposes all you will do is put
make variable definitions that override those in/ @mk/config.mk.in@.
The whole point of @mk/config.mk.in@ --- and its derived counterpart
@mk/config.mk@ --- is to define the build configuration. It is heavily
commented, as you will see if you look at it.  So generally, what you
do is look at @mk/config.mk.in@, and add definitions in @mk/build.mk@
that override any of the @config.mk@ definitions that you want to
change.  (The override occurs because the main boilerplate file,
@mk/boilerplate.mk@<ncdx/boilerplate.mk/, includes @build.mk@ after
@config.mk@.)

For example, @config.mk.in@ contains the definition:

<tscreen><verb>
  ProjectsToBuild = glafp-utils ghc
</verb></tscreen>

The accompanying comment explains that this is the list of enabled
projects; that is, if (after configuring) you type @gmake all@ in
@FPTOOLS_TOP@ four specified projects will be made.  If you want to
add @green-card@, you can add this line to @build.mk@:

<tscreen><verb>
  ProjectsToBuild += green-card
</verb></tscreen>

or, if you prefer,

<tscreen><verb>
  ProjectsToBuild = glafp-utils ghc green-card
</verb></tscreen>

(GNU @make@ allows existing definitions to have new text appended
using the ``@+=@'' operator, which is quite a convenient feature.)

When reading @config.mk.in@, remember that anything between
``@@...@@'' signs is going to be substituted by @configure@
later.  You <em/can/ override the resulting definition if you want,
but you need to be a bit surer what you are doing.  For example,
there's a line that says: 

<tscreen><verb>
  YACC = @YaccCmd@
</verb></tscreen>

This defines the Make variables @YACC@ to the pathname for a Yacc that
@configure@ finds somewhere.  If you have your own pet Yacc you want
to use instead, that's fine. Just add this line to @mk/build.mk@:

<tscreen><verb>
  YACC = myyacc
</verb></tscreen>

You do not <em/have/ to have a @mk/build.mk@ file at all; if you
don't, you'll get all the default settings from @mk/config.mk.in@.

You can also use @build.mk@ to override anything that @configure@ got
wrong.  One place where this happens often is with the definition of
@FPTOOLS_TOP_ABS@: this variable is supposed to be the canonical path
to the top of your source tree, but if your system uses an automounter
then the correct directory is hard to find automatically.  If you find
that @configure@ has got it wrong, just put the correct definition in
@build.mk@.

<sect1>The story so far
<p>

Let's summarise the steps you need to carry to get yourself
a fully-configured build tree from scratch.

<enum>

<item> Get your source tree from somewhere (CVS repository or source
distribution).  Say you call the root directory @myfptools@ (it
does not have to be called @fptools@).  Make sure that you have
the essential files (see Section~<ref id="sec:source-tree"
name="Source Tree">).

<item> Use @lndir@ or @mkshadowdir@ to create a build tree.
<tscreen><verb>
    cd myfptools
    mkshadowdir . /scratch/joe-bloggs/myfptools-sun4
</verb></tscreen>
You probably want to give the build tree a name that
suggests its main defining characteristic (in your mind at least),
in case you later add others.

<item> Change directory to the build tree.  Everything is going
to happen there now.
<tscreen><verb>
    cd /scratch/joe-bloggs/myfptools-sun4
</verb></tscreen>
<item> Prepare for system configuration:
<tscreen><verb>
    autoconf
</verb></tscreen>
(You can skip this step if you are starting from a source distribution,
and you already have @configure@ and @mk/config.h.in@.)

<item> Do system configuration:
<tscreen><verb>
    ./configure
</verb></tscreen>

<item> Create the file @mk/build.mk@, 
adding definitions for your desired configuration options.
<tscreen><verb>
    emacs mk/build.mk
</verb></tscreen>
</enum>
You can make subsequent changes to @mk/build.mk@ as often 
as you like.  You do not have to run any further configuration 
programs to make these changes take effect.
In theory you should, however, say @gmake clean@, @gmake all@,
because configuration option changes could affect anything --- but in practice you are likely to know what's affected.

<sect1>Making things
<p>

At this point you have made yourself a fully-configured build tree,
so you are ready to start building real things.

The first thing you need to know is that 
<em/you must use GNU @make@, usually called @gmake@, not standard Unix @make@/.
If you use standard Unix @make@ you will get all sorts of error messages
(but no damage) because the @fptools@ @Makefiles@ use GNU @make@'s facilities
extensively.

<sect1>Standard Targets
<label id="sec:standard-targets">
<nidx/targets, standard makefile/
<nidx/makefile targets/
<p>

In any directory you should be able to make the following:
<descrip>

<tag>@boot@:</tag> 

does the one-off preparation required to get ready for the real work.
Notably, it does @gmake depend@ in all directories that contain
programs.  But @boot@ does more.  For example, you can't do @gmake
depend@ in a directory of C program until you have converted the
literate @.lh@ header files into standard @.h@ header files.
Similarly, you can't convert a literate file to illiterate form until
you have built the @unlit@ tool.  @boot@ takes care of these
inter-directory dependencies.

You should say @gmake boot@ right after configuring your build tree,
but note that this is a one-off, i.e., there's no need to re-do
@gmake boot@ if you should re-configure your build tree at a later
stage (no harm caused if you do though).

<tag>@all@:</tag> makes all the final target(s) for this Makefile.
Depending on which directory you are in a ``final target'' may be an
executable program, a library archive, a shell script, or a Postscript
file.  Typing @gmake@ alone is generally the same as typing @gmake
all@.

<tag>@install@:</tag> installs the things built by @all@.  Where does it
install them?  That is specified by @mk/config.mk.in@; you can 
override it in @mk/build.mk@.

<tag>@uninstall@:</tag> reverses the effect of @install@.

<tag>@clean@:</tag> remove all easily-rebuilt files.

<tag>@veryclean@:</tag> remove all files that can be rebuilt at all.
There's a danger here that you may remove a file that needs a more
obscure utility to rebuild it (especially if you started from a source
distribution).

<tag>@check@:</tag> run the test suite.

</descrip>

All of these standard targets automatically recurse into
sub-directories.  Certain other standard targets do not:

<descrip>

<tag>@configure@:</tag> is only available in the root directory
@$(FPTOOLS_TOP)@; it has been discussed in Section~<ref
id="sec:build-config" name="Build Configuration">.

<tag>@depend@:</tag> make a @.depend@ file in each directory that needs
it. This @.depend@ file contains mechanically-generated dependency
information; for example, suppose a directory contains a Haskell 
source module @Foo.lhs@ which imports another module @Baz@.
Then the generated @.depend@ file will contain the dependency:

<tscreen><verb>
  Foo.o : Baz.hi
</verb></tscreen>

which says that the object file @Foo.o@ depends on the interface file
@Baz.hi@ generated by compiling module @Baz@.  The @.depend@ file is
automatically included by every Makefile.

<tag>@binary-dist@:</tag> make a binary distribution.  This is the
target we use to build the binary distributions of GHC and Happy.

<tag>@dist@:</tag> make a source distribution.  You must be in a
linked buid tree to make this target.
</descrip>

Most @Makefiles@ have targets other than these.  You can find
this out by looking in the @Makefile@ itself.

<sect1>Fast Making
<ncdx/fastmake/
<nidx/dependencies, omitting/
<nidx/FAST, makefile variable/
<p>

Sometimes the dependencies get in the way: if you've made a small
change to one file, and you're absolutely sure that it won't affect
anything else, but you know that @make@ is going to rebuid everything
anyway, the following hack may be useful:

<tscreen> <verb>
gmake FAST=YES
</verb> </tscreen>

This tells the make system to ignore dependencies and just build what
you tell it to.  In other words, it's equivalent to temporarily
removing the @.depend@ file in the current directory (where
@mkdependHS@ and friends store their dependency information).

A bit of history: GHC used to come with a @fastmake@ script that did
the above job, but GNU make provides the features we need to do it
without resorting to a script.  Also, we've found that fastmaking is
less useful since the advent of GHC's recompilation checker (see the
User's Guide section on "Separate Compilation").

<sect>The @Makefile@ architecture
<nidx/makefile architecture/
<p>

@make@ is great if everything works --- you type @gmake install@ and,
lo, the right things get compiled and installed in the right places.
Our goal is to make this happen often, but somehow it often doesn't;
instead some wierd error message eventually emerges from the bowels of
a directory you didn't know existed.

The purpose of this section is to give you a road-map to help you figure
out what is going right and what is going wrong.

<sect1>A small project
<p>

To get started, let us look at the @Makefile@ for an imaginary small
@fptools@ project, @small@.  Each project in @fptools@ has its own
directory in @FPTOOLS_TOP@, so the @small@ project will have its own
directory @FPOOLS_TOP/small/@.  Inside the @small/@ directory there
will be a @Makefile@, looking something like this:

<nidx/Makefile, minimal/
<tscreen><verb>
  #     Makefile for fptools project "small"

  TOP = ..
  include $(TOP)/mk/boilerplate.mk

  SRCS = $(wildcard *.lhs) $(wildcard *.c)
  HS_PROG = small

  include $(TOP)/target.mk
</verb></tscreen>

This @Makefile@ has three sections:

<enum>

<item> The first section includes<footnote>One of the most important
features of GNU @make@ that we use is the ability for a @Makefile@ to
include another named file, very like @cpp@'s @#include@
directive.</footnote> a file of ``boilerplate'' code from the level
above (which in this case will be
@FPTOOLS_TOP/mk/boilerplate.mk@<ncdx/boilerplate.mk/).  As its name
suggests, @boilerplate.mk@ consists of a large quantity of standard
@Makefile@ code.  We discuss this boilerplate in more detail in
Section~<ref id="sec:boiler" name="Boilerplate">.
<nidx/include, directive in Makefiles/
<nidx/Makefile inclusion/

Before the @include@ statement, you must define the @make@ variable
@TOP@<ncdx/TOP/ to be the directory containing the @mk@ directory in
which the @boilerplate.mk@ file is.  It is <em/not/ OK to simply say

<tscreen><verb>
  include ../mk/boilerplate.mk  # NO NO NO
</verb></tscreen>

Why?  Because the @boilerplate.mk@ file needs to know where it is, so
that it can, in turn, @include@ other files.  (Unfortunately, when an
@include@d file does an @include@, the filename is treated relative to
the directory in which @gmake@ is being run, not the directory in
which the @included@ sits.)  In general, <em>every file @foo.mk@
assumes that @$(TOP)/mk/foo.mk@ refers to itself.</em> It is up to the
@Makefile@ doing the @include@ to ensure this is the case.

Files intended for inclusion in other @Makefile@s are written to have
the following property: <em/after @foo.mk@ is @include@d, it leaves
@TOP@ containing the same value as it had just before the @include@
statement/.  In our example, this invariant guarantees that the
@include@ for @target.mk@ will look in the same directory as that for
@boilerplate.mk@.

<item> The second section defines the following standard @make@
variables: @SRCS@<ncdx/SRCS/ (the source files from which is to be
built), and @HS_PROG@<ncdx/HS_PROG/ (the executable binary to be
built).  We will discuss in more detail what the ``standard
variables'' are, and how they affect what happens, in Section~<ref
id="sec:targets" name="Targets">.

The definition for @SRCS@ uses the useful GNU @make@ construct
@$(wildcard@~$pat$@)@<ncdx/wildcard/, which expands to a list of all
the files matching the pattern @pat@ in the current directory.  In
this example, @SRCS@ is set to the list of all the @.lhs@ and @.c@
files in the directory.  (Let's suppose there is one of each,
@Foo.lhs@ and @Baz.c@.)

<item> The last section includes a second file of standard code,
called @target.mk@<ncdx/target.mk/.  It contains the rules that tell
@gmake@ how to make the standard targets (Section~<ref
id="sec:standard-targets" name="Standard Targets">).  Why, you ask,
can't this standard code be part of @boilerplate.mk@?  Good question.
We discuss the reason later, in Section~<ref id="sec:boiler-arch"
name="Boilerplate Architecture">.

You do not <em/have/ to @include@ the @target.mk@ file.  Instead, you
can write rules of your own for all the standard targets.  Usually,
though, you will find quite a big payoff from using the canned rules
in @target.mk@; the price tag is that you have to understand what
canned rules get enabled, and what they do (Section~<ref
id="sec:targets" name="Targets">.

</enum>

In our example @Makefile@, most of the work is done by the two
@include@d files.  When you say @gmake all@, the following things
happen:

<itemize>

<item> @gmake@ figures out that the object files are @Foo.o@ and
@Baz.o@.

<item> It uses a boilerplate pattern rule to compile @Foo.lhs@ to
@Foo.o@ using a Haskell compiler.  (Which one?  That is set in the
build configuration.)

<item> It uses another standard pattern rule to compile @Baz.c@ to
@Baz.o@, using a C compiler.  (Ditto.)

<item> It links the resulting @.o@ files together to make @small@,
using the Haskell compiler to do the link step.  (Why not use @ld@?
Because the Haskell compiler knows what standard librarise to link in.
How did @gmake@ know to use the Haskell compiler to do the link,
rather than the C compiler?  Because we set the variable @HS_PROG@
rather than @C_PROG@.)

</itemize>

All @Makefile@s should follow the above three-section format.

<sect1>A larger project
<p>

Larger projects are usually structured into a nummber of sub-directories,
each of which has its own @Makefile@.  (In very large projects, this
sub-structure might be iterated recursively, though that is rare.)
To give you the idea, here's part of the directory structure for
the (rather large) @ghc@ project:

<tscreen><verb>
  $(FPTOOLS_TOP)/ghc/
    Makefile

    mk/
      boilerplate.mk
      rules.mk

    docs/
      Makefile
      ...source files for documentation...

    driver/
      Makefile
      ...source files for driver...

    compiler/
      Makefile
      parser/...source files for parser...
      renamer/...source files for renamer...
      ...etc...
</verb></tscreen>

The sub-directories @docs@, @driver@, @compiler@, and so on, each
contains a sub-component of @ghc@, and each has its own @Makefile@.
There must also be a @Makefile@ in @$(FPTOOLS_TOP)/ghc@.  It does most
of its work by recursively invoking @gmake@ on the @Makefile@s in the
sub-directories.  We say that @ghc/Makefile@ is a <em/non-leaf
@Makefile@/, because it does little except organise its children,
while the @Makefile@s in the sub-directories are all <em/leaf
@Makefile@s/.  (In principle the sub-directories might themselves
contain a non-leaf @Makefile@ and several sub-sub-directories, but
that does not happen in @ghc@.)

The @Makefile@ in @ghc/compiler@ is considered a leaf @Makefile@ even
though the @ghc/compiler@ has sub-directories, because these sub-directories
do not themselves have @Makefile@s in them.  They are just used to structure
the collection of modules that make up @ghc@, but all are managed by the
single @Makefile@ in @ghc/compiler@.

You will notice that @ghc/@ also contains a directory @ghc/mk/@.  It
contains @ghc@-specific @Makefile@ boilerplate code.  More precisely:

<itemize> 

<item> @ghc/mk/boilerplate.mk@ is included at the top of
@ghc/Makefile@, and of all the leaf @Makefile@s in the
sub-directories.  It in turn @include@s the main boilerplate file
@mk/boilerplate.mk@.


<item> @ghc/mk/target.mk@ is @include@d at the bottom of
@ghc/Makefile@, and of all the leaf @Makefiles@ in the
sub-directories.  It in turn @include@s the file @mk/target.mk@.

</itemize>

So these two files are the place to look for @ghc@-wide customisation
of the standard boilerplate.

<sect1>Boilerplate architecture
<nidx/boilerplate architecture/
<label id="sec:boiler-arch">
<p>

Every @Makefile@ includes a @boilerplate.mk@<ncdx/boilerplate.mk/ file
at the top, and @target.mk@<ncdx/target.mk/ file at the bottom.  In
this section we discuss what is in these files, and why there have to
be two of them.  In general:

<itemize>

<item> @boilerplate.mk@ consists of:
<itemize>
<item> <em/Definitions of millions of @make@ variables/ that
collectively specify the build configuration.  Examples:
<tt><cdx/HC_OPTS/</tt>, the options to feed to the Haskell compiler;
<tt><cdx/NoFibSubDirs/</tt>, the sub-directories to enable within the
@nofib@ project; <tt><cdx/GhcWithHc/</tt>, the name of the Haskell
compiler to use when compiling @GHC@ in the @ghc@ project.  <item>
<em/Standard pattern rules/ that tell @gmake@ how to construct one
file from another.
</itemize>

@boilerplate.mk@ needs to be @include@d at the <em/top/
of each @Makefile@, so that the user can replace the
boilerplate definitions or pattern rules by simply giving a new
definition or pattern rule in the @Makefile@.  @gmake@
simply takes the last definition as the definitive one.

Instead of <em/replacing/ boilerplate definitions, it is also quite
common to <em/augment/ them. For example, a @Makefile@ might say:

<tscreen><verb>
  SRC_HC_OPTS += -O
</verb></tscreen>

thereby adding ``@-O@'' to the end of <tt><cdx/SRC_HC_OPTS/</tt>.

<item> @target.mk@ contains @make@ rules for the standard
targets described in Section~<ref id="sec:standard-targets"
name="Standard Targets">.  These rules are selectively included,
depending on the setting of certain @make@ variables.  These
variables are usually set in the middle section of the
@Makefile@ between the two @include@s.

@target.mk@ must be included at the end (rather than being part of
@boilerplate.mk@) for several tiresome reasons:

<itemize>
<item> @gmake@ commits target and dependency lists earlier than
it should.  For example, @target.mk@ has a rule that looks like
this: 

<tscreen><verb>
  $(HS_PROG) : $(OBJS)
        $(HC) $(LD_OPTS) $< -o $@
</verb></tscreen>

If this rule was in @boilerplate.mk@ then @$(HS_PROG)@<ncdx/HS_PROG/
and @$(OBJS)@<ncdx/OBJS/ would not have their final values at the
moment @gmake@ encountered the rule.  Alas, @gmake@ takes a snapshot
of their current values, and wires that snapshot into the rule.  (In
contrast, the commands executed when the rule ``fires'' are only
substituted at the moment of firing.)  So, the rule must follow the
definitions given in the @Makefile@ itself.

<item> Unlike pattern rules, ordinary rules cannot be overriden or
replaced by subsequent rules for the same target (at least not without an
error message).  Including ordinary rules in @boilerplate.mk@ would
prevent the user from writing rules for specific targets in specific cases.

<item> There are a couple of other reasons I've forgotten, but it doesn't
matter too much.
</itemize>
</itemize>

<sect1>The main @mk/boilerplate.mk@ file
<label id="sec:boiler">
<ncdx/boilerplate.mk/
<p>

If you look at @$(FPTOOLS_TOP)/mk/boilerplate.mk@ you will find
that it consists of the following sections, each held in a separate
file: 

<descrip> 

<tag><tt><cdx/config.mk/</tt></tag> is the build configuration file we
discussed at length in Section~<ref id="sec:build-config" name="Build
Configuration">.

<tag><tt><cdx/paths.mk/</tt></tag> defines @make@ variables for
pathnames and file lists.  In particular, it gives definitions for:

<descrip>
<tag><tt><cdx/SRCS/</tt>:</tag> all source files in the current directory.
<tag><tt><cdx/HS_SRCS/</tt>:</tag> all Haskell source files in the current directory.
It is derived from @$(SRCS)@, so if you override @SRCS@ with a new value
@HS_SRCS@ will follow suit.
<tag><tt><cdx/C_SRCS/</tt>:</tag> similarly for C source files.
<tag><tt><cdx/HS_OBJS/</tt>:</tag> the @.o@ files derived from @$(HS_SRCS)@.
<tag><tt><cdx/C_OBJS/</tt>:</tag> similarly for @$(C_SRCS)@.
<tag><tt><cdx/OBJS/</tt>:</tag> the concatenation of @$(HS_OBJS)@ and @$(C_OBJS)@.
</descrip>

Any or all of these definitions can easily be overriden by giving new
definitions in your @Makefile@.  For example, if there are things in
the current directory that look like source files but aren't, then
you'll need to set @SRCS@ manually in your @Makefile@.  The other
definitions will then work from this new definition.

What, exactly, does @paths.mk@ consider a ``source file'' to be.  It's
based the file's suffix (e.g. @.hs@, @.lhs@, @.c@, @.lc@, etc), but
this is the kind of detail that changes more rapidly, so rather than
enumerate the source suffices here the best thing to do is to look in
@paths.mk@.

<tag><tt><cdx/opts.mk/</tt></tag> defines @make@ variables for option
strings to pass to each program. For example, it defines
<tt><cdx/HC_OPTS/</tt>, the option strings to pass to the Haskell
compiler.  See Section~<ref id="sec:suffix" name="Pattern Rules and
Options">.

<tag><tt><cdx/suffix.mk/</tt></tag> defines standard pattern rules --
see Section~<ref id="sec:suffix" name="Pattern Rules and Options">.
</descrip>

Any of the variables and pattern rules defined by the boilerplate file
can easily be overridden in any particular @Makefile@, because the
boilerplace @include@ comes first.  Definitions after this @include@
directive simply override the default ones in @boilerplate.mk@.

<sect1>Pattern rules and options
<label id="sec:suffix">
<nidx/Pattern rules/
<p>

The file @suffix.mk@<ncdx/suffix.mk/ defines standard <em/pattern
rules/ that say how to build one kind of file from another, for
example, how to build a @.o@ file from a @.c@ file.  (GNU @make@'s
<em/pattern rules/ are more powerful and easier to use than Unix
@make@'s <em/suffix rules/.)

Almost all the rules look something like this:

<tscreen><verb>
%.o : %.c
        $(RM) $@
        $(CC) $(CC_OPTS) -c $< -o $@
</verb></tscreen>

Here's how to understand the rule.  It says that
<em/something/@.o@ (say @Foo.o@) can be built from
<em/something/@.c@ (@Foo.c@), by invoking the C compiler
(path name held in @$(CC)@), passing to it the options
@$(CC_OPTS)@ and the rule's dependent file of the rule
@$<@ (@Foo.c@ in this case), and putting the result in
the rule's target @$@@@ (@Foo.o@ in this case).

Every program is held in a @make@ variable defined in
@mk/config.mk@ --- look in @mk/config.mk@ for the
complete list.  One important one is the Haskell compiler, which is
called @$(HC)@.

Every programs options are are held in a @make@ variables called
@<prog>_OPTS@.  the @<prog>_OPTS@ variables are defined in
@mk/opts.mk@.  Almost all of them are defined like this:

<tscreen><verb>
  CC_OPTS = $(SRC_CC_OPTS) $(WAY$(_way)_CC_OPTS) $($*_CC_OPTS) $(EXTRA_CC_OPTS)
</verb></tscreen>

The four variables from which @CC_OPTS@ is built have the following meaning:

<descrip>

<tag><tt><cdx/SRC_CC_OPTS/</tt>:</tag> options passed to all C
compilations.

<tag>@WAY_<way>_CC_OPTS@:</tag> options passed to C
compilations for way @<way>@. For example,
@WAY_mp_CC_OPTS@ gives options to pass to the C compiler when
compiling way @mp@.  The variable @WAY_CC_OPTS@ holds
options to pass to the C compiler when compiling the standard way.
(Section~<ref id="sec:ways" name="Ways"> dicusses multi-way
compilation.)  <tag>@<module>_CC_OPTS@:</tag> options to
pass to the C compiler that are specific to module @<module>@.  For example, @SMap_CC_OPTS@ gives the specific options
to pass to the C compiler when compiling @SMap.c@.

<tag><tt><cdx/EXTRA_CC_OPTS/</tt>:</tag> extra options to pass to all
C compilations.  This is intended for command line use, thus;

<tscreen><verb>
  gmake libHS.a EXTRA_CC_OPTS="-v"
</verb></tscreen>
</descrip>

<sect1>The main @mk/target.mk@ file
<label id="sec:targets">
<ncdx/target.mk/
<p>

@target.mk@ contains canned rules for all the standard targets
described in Section~<ref id="sec:standard-targets" name="Standard
Targets">.  It is complicated by the fact that you don't want all of
these rules to be active in every @Makefile@.  Rather than have a
plethora of tiny files which you can include selectively, there is a
single file, @target.mk@, which selectively includes rules based on
whether you have defined certain variables in your @Makefile@.  This
section explains what rules you get, what variables control them, and
what the rules do.  Hopefully, you will also get enough of an idea of
what is supposed to happen that you can read and understand any wierd
special cases yourself.

<descrip>
<tag><tt><cdx/HS_PROG/</tt>.</tag> If @HS_PROG@ is defined, you get
rules with the following targets:
<descrip>
<tag><tt><cdx/HS_PROG/</tt></tag> itself.  This rule links @$(OBJS)@
with the Haskell runtime system to get an executable called
@$(HS_PROG)@.
<tag><tt><cdx/install/</tt></tag> installs @$(HS_PROG)@
in @$(bindir)@ with the execute bit set.
</descrip>

<tag><tt><cdx/C_PROG/</tt></tag> is similar to @HS_PROG@, except that
the link step links @$(C_OBJS)@ with the C runtime system.

<tag><tt><cdx/LIBRARY/</tt></tag> is similar to @HS_PROG@, except that
it links @$(LIB_OBJS)@ to make the library archive @$(LIBRARY)@, and
@install@ installs it in @$(libdir)@, with the execute bit not set.

<tag><tt><cdx/LIB_DATA/</tt></tag> ...
<tag><tt><cdx/LIB_EXEC/</tt></tag> ...

<tag><tt><cdx/HS_SRCS/</tt>, <tt><cdx/C_SRCS/</tt>.</tag> If @HS_SRCS@
is defined and non-empty, a rule for the target @depend@ is included,
which generates dependency information for Haskell programs.
Similarly for @C_SRCS@.
</descrip>

All of these rules are ``double-colon'' rules, thus

<tscreen><verb>
  install :: $(HS_PROG)
        ...how to install it...
</verb></tscreen>

GNU @make@ treats double-colon rules as separate entities.  If there
are several double-colon rules for the same target it takes each in
turn and fires it if its dependencies say to do so.  This means that
you can, for example, define both @HS_PROG@ and @LIBRARY@, which will
generate two rules for @install@.  When you type @gmake install@ both
rules will be fired, and both the program and the library will be
installed, just as you wanted.

<sect1>Recursion
<label id="sec:subdirs">
<nidx/recursion, in makefiles/
<nidx/Makefile, recursing into subdirectories/
<p>

In leaf @Makefiles@ the variable @SUBDIRS@<ncdx/SUBDIRS/ is undefined.
In non-leaf @Makefiles@, @SUBDIRS@ is set to the list of
sub-directories that contain subordinate @Makefile@s.  <em/It is up to
you to set @SUBDIRS@ in the @Makefile@./ There is no automation here
--- @SUBDIRS@ is too important automate.

When @SUBDIRS@ is defined, @target.mk@ includes a rather
neat rule for the standard targets (Section~<ref
id="sec:standard-targets" name="Standard Targets"> that simply invokes
@make@ recursively in each of the sub-directories.

<em/These recursive invocations are guaranteed to occur in the order
in which the list of directories is specified in @SUBDIRS@./ This
guarantee can be important.  For example, when you say @gmake boot@ it
can be important that the recursive invocation of @make boot@ is done
in one sub-directory (the include files, say) before another (the
source files).  Generally, put the most independent sub-directory
first, and the most dependent last.

<sect1>Way management
<label id="sec:ways">
<nidx/way management/
<p>

We sometimes want to build essentially the same system in several
different ``ways''.  For example, we want to build @ghc@'s @Prelude@
libraries with and without profiling, with and without concurrency,
and so on, so that there is an appropriately-built library archive to
link with when the user compiles his program.  It would be possible to
have a completely separate build tree for each such ``way'', but it
would be horribly bureaucratic, especially since often only parts of
the build tree need to be constructed in multiple ways.

Instead, the @target.mk@<ncdx/target.mk/ contains some clever magic to
allow you to build several versions of a system; and to control
locally how many versions are built and how they differ.  This section
explains the magic.

The files for a particular way are distinguished by munging the
suffix.  The ``normal way'' is always built, and its files have the
standard suffices @.o@, @.hi@, and so on.  In addition, you can build
one or more extra ways, each distinguished by a <em/way tag/.  The
object files and interface files for one of these extra ways are
distinguished by their suffix.  For example, way @mp@ has files
@.mp_o@ and @.mp_hi@.  Library archives have their way tag the other
side of the dot, for boring reasons; thus, @libHS_mp.a@.

A @make@ variable called @way@ holds the current way tag.  <em/@way@
is only ever set on the command line of a recursive invocation of
@gmake@./ It is never set inside a @Makefile@.  So it is a global
constant for any one invocation of @gmake@.  Two other @make@
variables, @way_@ and @_way@ are immediately derived from @$(way)@ and
never altered.  If @way@ is not set, then neither are @way_@ and
@_way@, and the invocation of @make@ will build the ``normal way''.
If @way@ is set, then the other two variables are set in sympathy.
For example, if @$(way)@ is ``@mp@'', then @way_@ is set to ``@mp_@''
and @_way@ is set to ``@_mp@''.  These three variables are then used
when constructing file names.

So how does @make@ ever get recursively invoked with @way@ set?  There
are two ways in which this happens:

<itemize>

<item> For some (but not all) of the standard targets, when in a leaf
sub-directory, @make@ is recursively invoked for each way tag in
@$(WAYS)@.  You set @WAYS@ to the list of way tags you want these
targets built for.  The mechanism here is very much like the recursive
invocation of @make@ in sub-directories (Section~<ref id="sec:subdirs"
name="Subdirectories">).

It is up to you to set @WAYS@ in your @Makefile@; this is how you
control what ways will get built.  <item> For a useful collection of
targets (such as @libHS_mp.a@, @Foo.mp_o@) there is a rule which
recursively invokes @make@ to make the specified target, setting the
@way@ variable.  So if you say @gmake Foo.mp_o@ you should see a
recursive invocation @gmake Foo.mp_o way=mp@, and <em/in this
recursive invocation the pattern rule for compiling a Haskell file
into a @.o@ file will match/.  The key pattern rules (in @suffix.mk@)
look like this:

<tscreen><verb>
  %.$(way_)o : %.lhs
        $(HC) $(HC_OPTS) $< -o $@
</verb></tscreen>

Neat, eh?
</itemize>


<sect1>When the canned rule isn't right
<p>

Sometimes the canned rule just doesn't do the right thing.  For
example, in the @nofib@ suite we want the link step to print out
timing information.  The thing to do here is <em/not/ to define
@HS_PROG@ or @C_PROG@, and instead define a special purpose rule in
your own @Makefile@.  By using different variable names you will avoid
the canned rules being included, and conflicting with yours.


<sect>Booting/porting from C (@.hc@) files
<label id="sec:booting-from-C">
<nidx>building GHC from .hc files</nidx>
<nidx>booting GHC from .hc files</nidx>
<nidx>porting GHC</nidx>
<p>

This section is for people trying to get GHC going by using the
supplied intermediate C (@.hc@) files.  This would probably be because
no binaries have been provided, or because the machine is not ``fully
supported.''

The intermediate C files are normally made available together with a
source release, please check the announce message for exact directions
of where to find them. If we've haven't made them available or you
can't find them, please ask.

Assuming you've got them, unpack them on top of a fresh source tree.
Then follow the `normal' instructions in Section~<ref
id="sec:building-from-source" name="Buiding From Source"> for setting
up a build tree. When you invoke the configure script, you'll have
to tell the script about your intentions:

<tscreen><verb>
foo% ./configure --enable-hc-boot
</verb></tscreen>
<ncdx/--enable-hc-boot/
<ncdx/--disable-hc-boot/

Assuming it configures OK and you don't need to create @mk/build.mk@
for any other purposes, the next step is to proceed with a @make boot@
followed by @make all@. At the successful completion of @make all@,
you should end up with a binary of the compiler proper,
@ghc/compiler/hsc@, plus archives (but no @.hi@ files!) of the prelude
libraries. To generate the Prelude interface files (and test drive the
bootstrapped compiler), re-run the @configure@ script, but this time
witout the @--enable-hc-boot@ option. After that re-create the
contents of @ghc/lib@:

<tscreen><verb>
foo% ./configure
 ....
foo% cd ghc/lib
foo% make clean
foo% make boot
foo% make all
</verb></tscreen>


That's the mechanics of the boot process, but, of course, if you're
trying to boot on a platform that is not supported and significantly
`different' from any of the supported ones, this is only the start of
the adventure...(ToDo: porting tips - stuff to look out for, etc.)


<sect>Known pitfalls in building Glasgow Haskell
<label id="sec:build-pitfalls">
<nidx>problems, building</nidx>
<nidx>pitfalls, in building</nidx>
<nidx>building pitfalls</nidx>
<p>

WARNINGS about pitfalls and known ``problems'':

<enum>

<item>
One difficulty that comes up from time to time is running out of space
in @/tmp@.  (It is impossible for the configuration stuff to
compensate for the vagaries of different sysadmin approaches re temp
space.)
<nidx/tmp, running out of space in/

The quickest way around it is @setenv TMPDIR /usr/tmp@<ncdx/TMPDIR/ or
even @setenv TMPDIR .@ (or the equivalent incantation with the shell
of your choice).

The best way around it is to say
<tscreen><verb>
export TMPDIR=<dir>
</verb></tscreen>
in your @build.mk@ file.
Then GHC and the other @fptools@ programs will use the appropriate directory
in all cases.


<item>
In compiling some support-code bits, e.g., in @ghc/rts/gmp@ and even
in @ghc/lib@, you may get a few C-compiler warnings.  We think these
are OK.

<item>
When compiling via C, you'll sometimes get ``warning: assignment from
incompatible pointer type'' out of GCC.  Harmless.

<item>
Similarly, @ar@chiving warning messages like the following are not
a problem:
<tscreen><verb>
ar: filename GlaIOMonad__1_2s.o truncated to GlaIOMonad_
ar: filename GlaIOMonad__2_2s.o truncated to GlaIOMonad_
...
</verb></tscreen>

<item> In compiling the compiler proper (in @compiler/@), you <em/may/
get an ``Out of heap space'' error message.  These can vary with the
vagaries of different systems, it seems.  The solution is simple:

<itemize> 

<item> If you're compiling with GHC 4.00 or above, then the
<em/maximum/ heap size must have been reached.  This is somewhat
unlikely, since the maximum is set to 64M by default.  Anyway, you can
raise it with the @-optCrts-M&lt;size&gt;@ flag (add this flag to
@<module>_HC_OPTS@ @make@ variable in the appropriate @Makefile@).

<item> For GHC < 4.00, add a suitable @-H@ flag to the @Makefile@, as
above.

</itemize>

and try again: @gmake@.  (see Section~<ref id="sec:suffix"
name="Pattern Rules and Options"> for information about
@<module>_HC_OPTS@.)

Alternatively, just cut to the chase scene:
<tscreen><verb>
% cd ghc/compiler
% make EXTRA_HC_OPTS=-optCrts-M128M
</verb></tscreen>

<item>
If you try to compile some Haskell, and you get errors from GCC about
lots of things from @/usr/include/math.h@, then your GCC was
mis-installed.  @fixincludes@ wasn't run when it should've been.

As @fixincludes@ is now automagically run as part of GCC installation,
this bug also suggests that you have an old GCC.


<item>
You <em/may/ need to re-@ranlib@<ncdx/ranlib/ your libraries (on Sun4s).

<tscreen><verb>
% cd $(libdir)/ghc-x.xx/sparc-sun-sunos4
% foreach i ( `find . -name '*.a' -print` ) # or other-shell equiv...
?    ranlib $i
?    # or, on some machines: ar s $i
? end
</verb></tscreen>

We'd be interested to know if this is still necessary.


<item>
GHC's sources go through @cpp@ before being compiled, and @cpp@ varies
a bit from one Unix to another.  One particular gotcha is macro calls
like this:

<tscreen><verb>
  SLIT("Hello, world")
</verb></tscreen>

Some @cpp@s treat the comma inside the string as separating two macro
arguments, so you get

<tscreen><verb>
  :731: macro `SLIT' used with too many (2) args
</verb></tscreen>

Alas, @cpp@ doesn't tell you the offending file!

Workaround: don't put wierd things in string args to @cpp@ macros.
</enum>

</article>