1 <!DOCTYPE Article PUBLIC "-//OASIS//DTD DocBook V3.1//EN">
3 <Article id="building-guide">
7 <Title>Building the Glasgow Functional Programming Tools Suite</Title>
8 <Author><OtherName>The GHC Team</OtherName></Author>
9 <Address><Email>glasgow-haskell-{users,bugs}@haskell.org</Email></Address>
10 <PubDate>November 2001</PubDate>
13 <para>The Glasgow fptools suite is a collection of Functional
14 Programming related tools, including the Glasgow Haskell
15 Compiler (GHC). The source code for the whole suite is kept in
16 a single CVS repository and shares a common build and
17 installation system.</para>
19 <para>This guide is intended for people who want to build or
20 modify programs from the Glasgow <Literal>fptools</Literal>
21 suite (as distinct from those who merely want to
22 <Emphasis>run</Emphasis> them). Installation instructions are
23 now provided in the user guide.</para>
25 <para>The bulk of this guide applies to building on Unix
26 systems; see <XRef LinkEnd="winbuild"> for Windows notes.</para>
32 <sect1 id="sec-getting">
33 <title>Getting the sources</title>
35 <para>You can get your hands on the <literal>fptools</literal>
41 <term><indexterm><primary>Source
42 distributions</primary></indexterm>Source distributions</term>
44 <para>You have a supported platform, but (a) you like
45 the warm fuzzy feeling of compiling things yourself;
46 (b) you want to build something ``extra”—e.g., a
47 set of libraries with strictness-analysis turned off; or
48 (c) you want to hack on GHC yourself.</para>
50 <para>A source distribution contains complete sources for
51 one or more projects in the <literal>fptools</literal>
52 suite. Not only that, but the more awkward
53 machine-independent steps are done for you. For example, if
55 <command>happy</command><indexterm><primary>happy</primary></indexterm>
56 you'll find it convenient that the source distribution
57 contains the result of running <command>happy</command> on
58 the parser specifications. If you don't want to alter the
59 parser then this saves you having to find and install
60 <command>happy</command>. You will still need a working
61 version of GHC (preferably version 4.08+) on your machine in
62 order to compile (most of) the sources, however.</para>
67 <term>The CVS repository.</term>
68 <indexterm><primary>CVS repository</primary>
71 <para>We make releases infrequently. If you want more
72 up-to-the minute (but less tested) source code then you need
73 to get access to our CVS repository.</para>
75 <para>All the <literal>fptools</literal> source code is held
76 in a CVS repository. CVS is a pretty good source-code
77 control system, and best of all it works over the
80 <para>The repository holds source code only. It holds no
81 mechanically generated files at all. So if you check out a
82 source tree from CVS you will need to install every utility
83 so that you can build all the derived files from
86 <para>More information about our CVS repository can be found
87 in <xref linkend="sec-cvs">.</para>
92 <para>If you are going to do any building from sources (either
93 from a source distribution or the CVS repository) then you need to
94 read all of this manual in detail.</para>
98 <title>Using the CVS repository</title>
100 <para>We use <ulink url="http://www.cvshome.org/">CVS</ulink> (Concurrent Version System) to keep track of our
101 sources for various software projects. CVS lets several people
102 work on the same software at the same time, allowing changes to be
103 checked in incrementally. </para>
105 <para>This section is a set of guidelines for how to use our CVS
106 repository, and will probably evolve in time. The main thing to
107 remember is that most mistakes can be undone, but if there's
108 anything you're not sure about feel free to bug the local CVS
109 meister (namely Jeff Lewis
110 <email>jlewis@galconn.com</email>). </para>
112 <sect2 id="cvs-access">
113 <title>Getting access to the CVS Repository</title>
115 <para>You can access the repository in one of two ways:
116 read-only (<xref linkend="cvs-read-only">), or read-write (<xref
117 linkend="cvs-read-write">).</para>
119 <sect3 id="cvs-read-only">
120 <title>Remote Read-only CVS Access</title>
122 <para>Read-only access is available to anyone - there's no
123 need to ask us first. With read-only CVS access you can do
124 anything except commit changes to the repository. You can
125 make changes to your local tree, and still use CVS's merge
126 facility to keep your tree up to date, and you can generate
127 patches using 'cvs diff' in order to send to us for
130 <para>To get read-only access to the repository:</para>
134 <para>Make sure that <application>cvs</application> is
135 installed on your machine.</para>
138 <para>Set your <literal>$CVSROOT</literal> environment variable to
139 <literal>:pserver:anoncvs@glass.cse.ogi.edu:/cvs</literal></para>
140 <para>If you set <literal>$CVSROOT</literal> in a shell script, be sure not to
141 have any trailing spaces on that line, otherwise CVS will respond with
142 a perplexing message like
144 /cvs : no such repository
145 </programlisting></para>
148 <para>Run the command</para>
152 <para>The password is simply <literal>cvs</literal>. This
153 sets up a file in your home directory called
154 <literal>.cvspass</literal>, which squirrels away the
155 dummy password, so you only need to do this step once.</para>
159 <para>Now go to <xref linkend="cvs-first">.</para>
164 <sect3 id="cvs-read-write">
165 <title>Remote Read-Write CVS Access</title>
167 <para>We generally supply read-write access to folk doing
168 serious development on some part of the source tree, when
169 going through us would be a pain. If you're developing some
170 feature, or think you have the time and inclination to fix
171 bugs in our sources, feel free to ask for read-write
172 access. There is a certain amount of responsibility that goes
173 with commit privileges; we are more likely to grant you access
174 if you've demonstrated your competence by sending us patches
175 via mail in the past.</para>
177 <para>To get remote read-write CVS access, you need to do the
178 following steps.</para>
182 <para>Make sure that <literal>cvs</literal> and
183 <literal>ssh</literal> are both installed on your
188 <para>Generate a DSA private-key/public-key pair, thus:</para>
192 <para>(<literal>ssh-keygen</literal> comes with
193 <literal>ssh</literal>.) Running <literal>ssh-keygen
194 -d</literal> creates the private and public keys in
195 <literal>$HOME/.ssh/id_dsa</literal> and
196 <literal>$HOME/.ssh/id_dsa.pub</literal> respectively
197 (assuming you accept the standard defaults).</para>
199 <para><literal>ssh-keygen -d</literal> will only work if
200 you have Version 2 <literal>ssh</literal> installed; it
201 will fail harmlessly otherwise. If you only have Version
202 1 you can instead generate an RSA key pair using plain</para>
207 <para>Doing so creates the private and public RSA keys in
208 <literal>$HOME/.ssh/identity</literal> and
209 <literal>$HOME/.ssh/identity.pub</literal>
212 <para>[Deprecated.] Incidentally, you can force a Version
213 2 <literal>ssh</literal> to use the Version 1 protocol by
214 creating <literal>$HOME/config</literal> with the
215 following in it:</para>
223 <para>In both cases, <literal>ssh-keygen</literal> will
224 ask for a <firstterm>passphrase</firstterm>. The
225 passphrase is a password that protects your private key.
226 In response to the 'Enter passphrase' question, you can
230 <para>[Recommended.] Enter a passphrase, which you
231 will quote each time you use CVS.
232 <literal>ssh-agent</literal> makes this entirely
236 <para>[Deprecated.] Just hit return (i.e. use an empty
237 passphrase); then you won't need to quote the
238 passphrase when using CVS. The downside is that
239 anyone who can see into your <literal>.ssh</literal>
240 directory, and thereby get your private key, can mess
241 up the repository. So you must keep the
242 <literal>.ssh</literal> directory with draconian
243 no-access permissions.</para>
249 <emphasis>Windows users: see the notes in <xref linkend="configure-ssh"> about <command>ssh</command> wrinkles!</emphasis>
256 <para>Send a message to to the CVS repository
257 administrator (currently Jeff Lewis
258 <email>jeff@galconn.com</email>), containing:</para>
261 <para>Your desired user-name.</para>
264 <para>Your <literal>.ssh/id_dsa.pub</literal> (or
265 <literal>.ssh/identity.pub</literal>).</para>
268 <para>He will set up your account.</para>
272 <para>Set the following environment variables:</para>
276 <constant>$HOME</constant>: points to your home directory. This is where CVS
277 will look for its <filename>.cvsrc</filename> file.
283 <constant>$CVS_RSH</constant> to <filename>ssh</filename>
285 <para>[Windows users.] Setting your <literal>CVS_RSH</literal> to
286 <literal>ssh</literal> assumes that your CVS client
287 understands how to execute shell script
288 ("#!"s,really), which is what
289 <literal>ssh</literal> is. This may not be the case on
290 Win32 platforms, so in that case set <literal>CVS_RSH</literal> to
291 <literal>ssh1</literal>.</para>
295 <para><literal>$CVSROOT</literal> to
296 <literal>:ext:</literal><replaceable>your-username</replaceable>
297 <literal>@cvs.haskell.org:/home/cvs/root</literal>
298 where <replaceable>your-username</replaceable> is your user name on
299 <literal>cvs.haskell.org</literal>.
301 <para>The <literal>CVSROOT</literal> environment variable will
302 be recorded in the checked-out tree, so you don't need to set
303 this every time. </para>
309 <constant>$CVSEDITOR</constant>: <filename>bin/gnuclient.exe</filename>
310 if you want to use an Emacs buffer for typing in those long commit messages.
316 <constant>$SHELL</constant>: To use bash as the shell in Emacs, you need to
317 set this to point to <filename>bash.exe</filename>.
328 Put the following in <filename>$HOME/.cvsrc</filename>:
339 These are the default options for the specified CVS commands,
340 and represent better defaults than the usual ones. (Feel
341 free to change them.)
345 [Windows users.] Filenames starting with <filename>.</filename> were illegal in
346 the 8.3 DOS filesystem, but that restriction should have
347 been lifted by now (i.e., you're using VFAT or later filesystems.) If
348 you're still having problems creating it, don't worry; <filename>.cvsrc</filename> is entirely
356 <para>[Experts.] Once your account is set up, you can get
357 access from other machines without bothering Jeff, thus:</para>
360 <para>Generate a public/private key pair on the new
364 <para>Use ssh to log in to
365 <literal>cvs.haskell.org</literal>, from your old
369 <para>Add the public key for the new machine to the file
370 <literal>$HOME/ssh/authorized_keys</literal> on
371 <literal>cvs.haskell.org</literal>.
372 (<literal>authorized_keys2</literal>, I think, for Version
376 <para>Make sure that the new version of
377 <literal>authorized_keys</literal> still has 600 file
386 <sect2 id="cvs-first">
387 <title>Checking Out a Source Tree</title>
391 <para>Make sure you set your <literal>CVSROOT</literal>
392 environment variable according to either of the remote
393 methods above. The Approved Way to check out a source tree
394 is as follows:</para>
397 $ cvs checkout fpconfig
400 <para>At this point you have a new directory called
401 <literal>fptools</literal> which contains the basic stuff
402 for the fptools suite, including the configuration files and
403 some other junk. </para>
405 <para>[Windows users.] The following messages appear to be harmless:
407 setsockopt IPTOS_LOWDELAY: Invalid argument
408 setsockopt IPTOS_THROUGHPUT: Invalid argument
413 <para>You can call the fptools directory whatever you like,
414 CVS won't mind: </para>
417 $ mv fptools <replaceable>directory</replaceable>
420 <para> NB: after you've read the CVS manual you might be
421 tempted to try</para>
423 $ cvs checkout -d <replaceable>directory</replaceable> fpconfig
426 <para>instead of checking out <literal>fpconfig</literal>
427 and then renaming it. But this doesn't work, and will
428 result in checking out the entire repository instead of just
429 the <literal>fpconfig</literal> bit.</para>
431 $ cd <replaceable>directory</replaceable>
432 $ cvs checkout ghc hslibs libraries
435 <para>The second command here checks out the relevant
436 modules you want to work on. For a GHC build, for instance,
437 you need at least the <literal>ghc</literal>,
438 <literal>hslibs</literal> and <literal>libraries</literal>
439 modules (for a full list of the projects available, see
440 <xref linkend="projects">).</para>
442 <para>Remember that if you do not have
443 <literal>happy</literal> and/or <literal>Alex</literal>
444 installed, you need to check them out as well.</para>
449 <sect2 id="cvs-committing">
450 <title>Committing Changes</title>
452 <para>This is only if you have read-write access to the
453 repository. For anoncvs users, CVS will issue a "read-only
454 repository" error if you try to commit changes.</para>
458 <para>Build the software, if necessary. Unless you're just
459 working on documentation, you'll probably want to build the
460 software in order to test any changes you make.</para>
464 <para>Make changes. Preferably small ones first.</para>
468 <para>Test them. You can see exactly what changes you've
469 made by using the <literal>cvs diff</literal> command:</para>
473 <para>lists all the changes (using the
474 <literal>diff</literal> command) in and below the current
475 directory. In emacs, <literal>C-c C-v =</literal> runs
476 <literal>cvs diff</literal> on the current buffer and shows
477 you the results.</para>
481 <para>If you changed something in the
482 <literal>fptools/libraries</literal> subdirectories, also run
483 <literal>make html</literal> to check if the documentation can
484 be generated successfully, too.</para>
488 <para>Before checking in a change, you need to update your
495 <para>This pulls in any changes that other people have made,
496 and merges them with yours. If there are any conflicts, CVS
497 will tell you, and you'll have to resolve them before you
498 can check your changes in. The documentation describes what
499 to do in the event of a conflict.</para>
501 <para>It's not always necessary to do a full cvs update
502 before checking in a change, since CVS will always tell you
503 if you try to check in a file that someone else has changed.
504 However, you should still update at regular intervals to
505 avoid making changes that don't work in conjuction with
506 changes that someone else made. Keeping an eye on what goes
507 by on the mailing list can help here.</para>
511 <para>When you're happy that your change isn't going to
512 break anything, check it in. For a one-file change:</para>
515 $ cvs commit <replaceable>filename</replaceable>
518 <para>CVS will then pop up an editor for you to enter a
519 "commit message", this is just a short description
520 of what your change does, and will be kept in the history of
523 <para>If you're using emacs, simply load up the file into a
524 buffer and type <literal>C-x C-q</literal>, and emacs will
525 prompt for a commit message and then check in the file for
528 <para>For a multiple-file change, things are a bit
529 trickier. There are several ways to do this, but this is the
530 way I find easiest. First type the commit message into a
531 temporary file. Then either</para>
534 $ cvs commit -F <replaceable>commit-message</replaceable> <replaceable>file_1</replaceable> .... <replaceable>file_n</replaceable>
537 <para>or, if nothing else has changed in this part of the
541 $ cvs commit -F <replaceable>commit-message</replaceable> <replaceable>directory</replaceable>
544 <para>where <replaceable>directory</replaceable> is a common
545 parent directory for all your changes, and
546 <replaceable>commit-message</replaceable> is the name of the
547 file containing the commit message.</para>
549 <para>Shortly afterwards, you'll get some mail from the
550 relevant mailing list saying which files changed, and giving
551 the commit message. For a multiple-file change, you should
552 still get only <emphasis>one</emphasis> message.</para>
557 <sect2 id="cvs-update">
558 <title>Updating Your Source Tree</title>
560 <para>It can be tempting to cvs update just part of a source
561 tree to bring in some changes that someone else has made, or
562 before committing your own changes. This is NOT RECOMMENDED!
563 Quite often changes in one part of the tree are dependent on
564 changes in another part of the tree (the
565 <literal>mk/*.mk</literal> files are a good example where
566 problems crop up quite often). Having an inconsistent tree is a
567 major cause of headaches. </para>
569 <para>So, to avoid a lot of hassle, follow this recipe for
570 updating your tree:</para>
574 $ cvs update -P 2>&1 | tee log</screen>
576 <para>Look at the log file, and fix any conflicts (denoted by a
577 <quote>C</quote> in the first column). New directories may have
578 appeared in the repository; CVS doesn't check these out by
579 default, so to get new directories you have to explicitly do
581 $ cvs update -d</screen>
582 in each project subdirectory. Don't do this at the top level,
583 because then <emphasis>all</emphasis> the projects will be
586 <para>If you're using multiple build trees, then for every build
587 tree you have pointing at this source tree, you need to update
588 the links in case any new files have appeared: </para>
591 $ cd <replaceable>build-tree</replaceable>
592 $ lndir <replaceable>source-tree</replaceable>
595 <para>Some files might have been removed, so you need to remove
596 the links pointing to these non-existent files:</para>
599 $ find . -xtype l -exec rm '{}' \;
602 <para>To be <emphasis>really</emphasis> safe, you should do
605 <screen>$ gmake all</screen>
607 <para>from the top-level, to update the dependencies and build
608 any changed files. </para>
611 <sect2 id="cvs-tags">
612 <title>GHC Tag Policy</title>
614 <para>If you want to check out a particular version of GHC,
615 you'll need to know how we tag versions in the repository. The
616 policy (as of 4.04) is:</para>
620 <para>The tree is branched before every major release. The
621 branch tag is <literal>ghc-x-xx-branch</literal>, where
622 <literal>x-xx</literal> is the version number of the release
623 with the <literal>'.'</literal> replaced by a
624 <literal>'-'</literal>. For example, the 4.04 release lives
625 on <literal>ghc-4-04-branch</literal>.</para>
629 <para>The release itself is tagged with
630 <literal>ghc-x-xx</literal> (on the branch). eg. 4.06 is
631 called <literal>ghc-4-06</literal>.</para>
635 <para>We didn't always follow these guidelines, so to see
636 what tags there are for previous versions, do <literal>cvs
637 log</literal> on a file that's been around for a while (like
638 <literal>fptools/ghc/README</literal>).</para>
642 <para>So, to check out a fresh GHC 4.06 tree you would
646 $ cvs co -r ghc-4-06 fpconfig
648 $ cvs co -r ghc-4-06 ghc hslibs
652 <sect2 id="cvs-hints">
653 <title>General Hints</title>
657 <para>As a general rule: commit changes in small units,
658 preferably addressing one issue or implementing a single
659 feature. Provide a descriptive log message so that the
660 repository records exactly which changes were required to
661 implement a given feature/fix a bug. I've found this
662 <emphasis>very</emphasis> useful in the past for finding out
663 when a particular bug was introduced: you can just wind back
664 the CVS tree until the bug disappears.</para>
668 <para>Keep the sources at least *buildable* at any given
669 time. No doubt bugs will creep in, but it's quite easy to
670 ensure that any change made at least leaves the tree in a
671 buildable state. We do nightly builds of GHC to keep an eye
672 on what things work/don't work each day and how we're doing
673 in relation to previous verions. This idea is truely wrecked
674 if the compiler won't build in the first place!</para>
678 <para>To check out extra bits into an already-checked-out
679 tree, use the following procedure. Suppose you have a
680 checked-out fptools tree containing just ghc, and you want
681 to add nofib to it:</para>
692 $ cvs update -d nofib
695 <para>(the -d flag tells update to create a new
696 directory). If you just want part of the nofib suite, you
701 $ cvs checkout nofib/spectral
704 <para>This works because <literal>nofib</literal> is a
705 module in its own right, and spectral is a subdirectory of
706 the nofib module. The path argument to checkout must always
707 start with a module name. There's no equivalent form of this
708 command using <literal>update</literal>.</para>
714 <sect1 id="projects">
715 <title>What projects are there?</title>
717 <para>The <literal>fptools</literal> suite consists of several
718 <firstterm>projects</firstterm>, most of which can be downloaded,
719 built and installed individually. Each project corresponds to a
720 subdirectory in the source tree, and if checking out from CVS then
721 each project can be checked out individually by sitting in the top
722 level of your source tree and typing <command>cvs checkout
723 <replaceable>project</replaceable></command>.</para>
725 <para>Here is a list of the projects currently available:</para>
729 <term><literal>alex</literal></term>
730 <indexterm><primary><literal>alex</literal></primary>
731 <secondary>project</secondary></indexterm>
734 url="http://www.haskell.org/alex/">Alex</ulink> lexical
735 analyser generator for Haskell.</para>
740 <term><literal>ghc</literal></term>
741 <indexterm><primary><literal>ghc</literal></primary>
742 <secondary>project</secondary></indexterm>
744 <para>The <ulink url="http://www.haskell.org/ghc/">Glasgow
745 Haskell Compiler</ulink> (minus libraries). Absolutely
746 required for building GHC.</para>
751 <term><literal>glafp-utils</literal></term>
752 <indexterm><primary><literal>glafp-utils</literal></primary><secondary>project</secondary></indexterm>
754 <para>Utility programs, some of which are used by the
755 build/installation system. Required for pretty much
761 <term><literal>greencard</literal></term>
762 <indexterm><primary><literal>greencard</literal></primary><secondary>project</secondary></indexterm>
765 url="http://www.haskell.org/greencard/">GreenCard</ulink>
766 system for generating Haskell foreign function
772 <term><literal>haggis</literal></term>
773 <indexterm><primary><literal>haggis</literal></primary><secondary>project</secondary></indexterm>
776 url="http://www.dcs.gla.ac.uk/fp/software/haggis/">Haggis</ulink>
777 Haskell GUI framework.</para>
782 <term><literal>haddock</literal></term>
783 <indexterm><primary><literal>haddock</literal></primary><secondary>project</secondary></indexterm>
786 url="http://www.haskell.org/haddock/">Haddock</ulink>
787 documentation tool.</para>
792 <term><literal>happy</literal></term>
793 <indexterm><primary><literal>happy</literal></primary><secondary>project</secondary></indexterm>
796 url="http://www.haskell.org/happy/">Happy</ulink> Parser
802 <term><literal>hdirect</literal></term>
803 <indexterm><primary><literal>hdirect</literal></primary><secondary>project</secondary></indexterm>
806 url="http://www.haskell.org/hdirect/">H/Direct</ulink>
807 Haskell interoperability tool.</para>
812 <term><literal>hood</literal></term>
813 <indexterm><primary><literal>hood</literal></primary><secondary>project</secondary></indexterm>
815 <para>The <ulink url="http://www.haskell.org/hood/">Haskell
816 Object Observation Debugger</ulink>.</para>
821 <term><literal>hslibs</literal></term>
822 <indexterm><primary><literal>hslibs</literal></primary><secondary>project</secondary></indexterm>
824 <para>Supplemental libraries for GHC
825 (<emphasis>required</emphasis> for building GHC).</para>
830 <term><literal>libraries</literal></term>
831 <indexterm><primary><literal></literal></primary><secondary>project</secondary></indexterm>
833 <para>Hierarchical Haskell library suite
834 (<emphasis>required</emphasis> for building GHC).</para>
839 <term><literal>mhms</literal></term>
840 <indexterm><primary><literal></literal></primary><secondary>project</secondary></indexterm>
842 <para>The Modular Haskell Metric System.</para>
847 <term><literal>nofib</literal></term>
848 <indexterm><primary><literal>nofib</literal></primary><secondary>project</secondary></indexterm>
850 <para>The NoFib suite: A collection of Haskell programs used
851 primarily for benchmarking.</para>
856 <term><literal>testsuite</literal></term>
857 <indexterm><primary><literal>testsuite</literal></primary><secondary>project</secondary></indexterm>
859 <para>A testing framework, including GHC's regression test
865 <para>So, to build GHC you need at least the
866 <literal>ghc</literal>, <literal>libraries</literal> and
867 <literal>hslibs</literal> projects (a GHC source distribution will
868 already include the bits you need).</para>
871 <sect1 id="sec-build-checks">
872 <title>Things to check before you start</title>
874 <para>Here's a list of things to check before you get
880 <indexterm><primary>Disk space needed</primary></indexterm>
881 <para>Disk space needed: from about 100Mb for a basic GHC
882 build, up to probably 500Mb for a GHC build with everything
883 included (libraries built several different ways,
888 <para>Use an appropriate machine / operating system. <xref
889 linkend="sec-port-info"> lists the supported platforms; if
890 yours isn't amongst these then you can try porting GHC (see
891 <xref linkend="sec-porting-ghc">).</para>
895 <para>Be sure that the “pre-supposed” utilities are
896 installed. <Xref LinkEnd="sec-pre-supposed">
901 <para>If you have any problem when building or installing the
902 Glasgow tools, please check the “known pitfalls” (<Xref
903 LinkEnd="sec-build-pitfalls">). Also check the FAQ for the
904 version you're building, which is part of the User's Guide and
905 available on the <ulink URL="http://www.haskell.org/ghc/" >GHC web
908 <indexterm><primary>bugs</primary><secondary>known</secondary></indexterm>
910 <para>If you feel there is still some shortcoming in our
911 procedure or instructions, please report it.</para>
913 <para>For GHC, please see the <ulink
914 url="http://www.haskell.org/ghc/docs/latest/set/bug-reporting.html">bug-reporting
915 section of the GHC Users' Guide</ulink>, to maximise the
916 usefulness of your report.</para>
918 <indexterm><primary>bugs</primary><secondary>seporting</secondary></indexterm>
919 <para>If in doubt, please send a message to
920 <email>glasgow-haskell-bugs@haskell.org</email>.
921 <indexterm><primary>bugs</primary><secondary>mailing
922 list</secondary></indexterm></para>
927 <sect1 id="sec-port-info">
928 <title>What machines the Glasgow tools run on</title>
930 <indexterm><primary>ports</primary><secondary>GHC</secondary></indexterm>
931 <indexterm><primary>GHC</primary><secondary>ports</secondary></indexterm>
932 <indexterm><primary>platforms</primary><secondary>supported</secondary></indexterm>
934 <para>The main question is whether or not the Haskell compiler
935 (GHC) runs on your platform.</para>
937 <para>A “platform” is a
938 architecture/manufacturer/operating-system combination, such as
939 <literal>sparc-sun-solaris2</literal>. Other common ones are
940 <literal>alpha-dec-osf2</literal>,
941 <literal>hppa1.1-hp-hpux9</literal>,
942 <literal>i386-unknown-linux</literal>,
943 <literal>i386-unknown-solaris2</literal>,
944 <literal>i386-unknown-freebsd</literal>,
945 <literal>i386-unknown-cygwin32</literal>,
946 <literal>m68k-sun-sunos4</literal>,
947 <literal>mips-sgi-irix5</literal>,
948 <literal>sparc-sun-sunos4</literal>,
949 <literal>sparc-sun-solaris2</literal>,
950 <literal>powerpc-ibm-aix</literal>.</para>
952 <para>Some libraries may only work on a limited number of
953 platforms; for example, a sockets library is of no use unless the
954 operating system supports the underlying BSDisms.</para>
957 <title>What platforms the Haskell compiler (GHC) runs on</title>
959 <indexterm><primary>fully-supported platforms</primary></indexterm>
960 <indexterm><primary>native-code generator</primary></indexterm>
961 <indexterm><primary>registerised ports</primary></indexterm>
962 <indexterm><primary>unregisterised ports</primary></indexterm>
964 <para>The GHC hierarchy of Porting Goodness: (a) Best is a
965 native-code generator; (b) next best is a
966 “registerised” port; (c) the bare minimum is an
967 “unregisterised” port.
968 (“Unregisterised” is so terrible that we won't say
969 more about it).</para>
971 <para>We use Sparcs running Solaris 2.7 and x86 boxes running
972 FreeBSD and Linux, so those are the best supported platforms,
973 unsurprisingly.</para>
975 <para>Here's everything that's known about GHC ports. We
976 identify platforms by their “canonical”
977 CPU/Manufacturer/OS triple.</para>
981 <term>alpha-dec-{osf,linux,freebsd,openbsd,netbsd}:</term>
982 <indexterm><primary>alpha-dec-osf</primary></indexterm>
983 <indexterm><primary>alpha-dec-linux</primary></indexterm>
984 <indexterm><primary>alpha-dec-freebsd</primary></indexterm>
985 <indexterm><primary>alpha-dec-openbsd</primary></indexterm>
986 <indexterm><primary>alpha-dec-netbsd</primary></indexterm>
989 <para>The OSF port is currently working (as of GHC version
990 5.02.1) and well supported. The native code generator is
991 currently non-working. Other operating systems will
992 require some minor porting.</para>
997 <term>sparc-sun-sunos4</term>
998 <indexterm><primary>sparc-sun-sunos4</primary></indexterm>
1000 <para>Probably works with minor tweaks, hasn't been tested
1006 <term>sparc-sun-solaris2</term>
1007 <indexterm><primary>sparc-sun-solaris2</primary></indexterm>
1009 <para>Fully supported (at least for Solaris 2.7 and 2.6),
1010 including native-code generator.</para>
1015 <term>sparc-unknown-openbsd</term>
1016 <indexterm><primary>sparc-unknown-openbsd</primary></indexterm>
1018 <para>Supported, including native-code generator. The
1019 same should also be true of NetBSD</para>
1024 <term>hppa1.1-hp-hpux (HP-PA boxes running HPUX 9.x)</term>
1025 <indexterm><primary>hppa1.1-hp-hpux</primary></indexterm>
1027 <para>A registerised port is available for version 4.08,
1028 but GHC hasn't been built on that platform since (as far
1029 as we know). No native-code generator.</para>
1034 <term>i386-unknown-linux (PCs running Linux, ELF binary format)</term>
1035 <indexterm><primary>i386-*-linux</primary></indexterm>
1037 <para>GHC works registerised and has a native code
1038 generator. You <Emphasis>must</Emphasis> have GCC 2.7.x
1039 or later. NOTE about <literal>glibc</literal> versions:
1040 GHC binaries built on a system running <literal>glibc
1041 2.0</literal> won't work on a system running
1042 <literal>glibc 2.1</literal>, and vice versa. In general,
1043 don't expect compatibility between
1044 <literal>glibc</literal> versions, even if the shared
1045 library version hasn't changed.</para>
1050 <term>i386-unknown-freebsd (PCs running FreeBSD 2.2 or
1052 <indexterm><primary>i386-unknown-freebsd</primary></indexterm>
1054 <para>GHC works registerised. Pre-built packages are
1055 available in the native package format, so if you just
1056 need binaries you're better off just installing the
1057 package (it might even be on your installation
1063 <term>i386-unknown-openbsd (PCs running OpenBSD)</term>
1064 <indexterm><primary>i386-unknown-openbsd</primary></indexterm>
1066 <para>Supported, with native code generator. Packages are
1067 available through the ports system in the native package
1073 <term>i386-unknown-netbsd (PCs running NetBSD and
1075 <indexterm><primary>i386-unknown-netbsd</primary></indexterm>
1077 <para>Will require some minor porting effort, but should
1078 work registerised.</para>
1083 <term>i386-unknown-mingw32 (PCs running Windows)</term>
1084 <indexterm><primary>i386-unknown-mingw32</primary></indexterm>
1086 <para>Fully supported under Win9x, WinNT, Win2k, and
1087 WinXP. Includes a native code generator. Building from
1088 source requires a recent <ulink
1089 url="http://www.cygwin.com/">Cygwin</ulink> distribution
1090 to be installed.</para>
1095 <term>ia64-unknown-linux</term>
1096 <indexterm><primary>ia64-unknown-linux</primary></indexterm>
1098 <para>Supported, except there is no native code
1104 <term>x86_64-unknown-linux</term>
1105 <indexterm><primary>x86_64-unknown-linux</primary></indexterm>
1107 <para>GHC currently works unregisterised. A registerised
1108 port is in progress.</para>
1113 <term>mips-sgi-irix5</term>
1114 <indexterm><primary>mips-sgi-irix[5-6]</primary></indexterm>
1116 <para>Port has worked in the past, but hasn't been tested
1117 for some time (and will certainly have rotted in various
1118 ways). As usual, we don't have access to machines and
1119 there hasn't been an overwhelming demand for this port,
1120 but feel free to get in touch.</para>
1125 <term>mips64-sgi-irix6</term>
1126 <indexterm><primary>mips-sgi-irix6</primary></indexterm>
1128 <para>GHC currently works unregisterised.</para>
1133 <term>powerpc-ibm-aix</term>
1134 <indexterm><primary>powerpc-ibm-aix</primary></indexterm>
1136 <para>Port currently doesn't work, needs some minimal
1137 porting effort. As usual, we don't have access to
1138 machines and there hasn't been an overwhelming demand for
1139 this port, but feel free to get in touch.</para>
1144 <term>powerpc-apple-darwin</term>
1145 <indexterm><primary>powerpc-apple-darwin</primary></indexterm>
1147 <para>Supported registerised. Native code generator is
1148 almost working.</para>
1153 <term>powerpc-apple-linux</term>
1154 <indexterm><primary>powerpc-apple-linux</primary></indexterm>
1156 <para>Not supported (yet).</para>
1161 <para>Various other systems have had GHC ported to them in the
1162 distant past, including various Motorola 68k boxes. The 68k
1163 support still remains, but porting to one of these systems will
1164 certainly be a non-trivial task.</para>
1168 <title>What machines the other tools run on</title>
1170 <para>Unless you hear otherwise, the other tools work if GHC
1176 <sect1 id="sec-pre-supposed">
1177 <title>Installing pre-supposed utilities</title>
1179 <indexterm><primary>pre-supposed utilities</primary></indexterm>
1180 <indexterm><primary>utilities, pre-supposed</primary></indexterm>
1182 <para>Here are the gory details about some utility programs you
1183 may need; <command>perl</command>, <command>gcc</command> and
1184 <command>happy</command> are the only important
1185 ones. (PVM<indexterm><primary>PVM</primary></indexterm> is
1186 important if you're going for Parallel Haskell.) The
1187 <command>configure</command><indexterm><primary>configure</primary></indexterm>
1188 script will tell you if you are missing something.</para>
1194 <indexterm><primary>pre-supposed: GHC</primary></indexterm>
1195 <indexterm><primary>GHC, pre-supposed</primary></indexterm>
1197 <para>GHC is required to build many of the tools, including
1198 GHC itself. If you need to port GHC to your platform
1199 because there isn't a binary distribution of GHC available,
1200 then see <xref linkend="sec-porting-ghc">.</para>
1202 <para>Which version of GHC you need will depend on the
1203 packages you intend to build. GHC itself will normally
1204 build using one of several older versions of itself - check
1205 the announcement or release notes for details.</para>
1211 <indexterm><primary>pre-supposed: Perl</primary></indexterm>
1212 <indexterm><primary>Perl, pre-supposed</primary></indexterm>
1214 <para><emphasis>You have to have Perl to proceed!</emphasis>
1215 Perl version 5 at least is required. GHC has been known to
1216 tickle bugs in Perl, so if you find that Perl crashes when
1217 running GHC try updating (or downgrading) your Perl
1218 installation. Versions of Perl that we use and are known to
1219 be fairly stable are 5.005 and 5.6.1.</para>
1221 <para>For Win32 platforms, you should use the binary
1222 supplied in the InstallShield (copy it to
1223 <filename>/bin</filename>). The Cygwin-supplied Perl seems
1226 <para>Perl should be put somewhere so that it can be invoked
1227 by the <literal>#!</literal> script-invoking
1228 mechanism. The full pathname may need to be less than 32
1229 characters long on some systems.</para>
1234 <term>GNU C (<command>gcc</command>)</term>
1235 <indexterm><primary>pre-supposed: GCC (GNU C
1236 compiler)</primary></indexterm> <indexterm><primary>GCC (GNU C
1237 compiler), pre-supposed</primary></indexterm>
1239 <para>We recommend using GCC version 2.95.2 on all
1240 platforms. Failing that, version 2.7.2 is stable on most
1241 platforms. Earlier versions of GCC can be assumed not to
1242 work, and versions in between 2.7.2 and 2.95.2 (including
1243 <command>egcs</command>) have varying degrees of stability
1244 depending on the platform.</para>
1246 <para>GCC 3.2 is currently known to have problems building
1247 GHC on Sparc, but is stable on x86.</para>
1249 <para>If your GCC dies with “internal error” on
1250 some GHC source file, please let us know, so we can report
1251 it and get things improved. (Exception: on x86
1252 boxes—you may need to fiddle with GHC's
1253 <option>-monly-N-regs</option> option; see the User's
1259 <term>GNU Make</term>
1260 <indexterm><primary>make</primary><secondary>GNU</secondary>
1263 <para>The fptools build system makes heavy use of features
1264 specific to GNU <command>make</command>, so you must have
1265 this installed in order to build any of the fptools
1272 <indexterm><primary>Happy</primary></indexterm>
1274 <para>Happy is a parser generator tool for Haskell, and is
1275 used to generate GHC's parsers. Happy is written in
1276 Haskell, and is a project in the CVS repository
1277 (<literal>fptools/happy</literal>). It can be built from
1278 source, but bear in mind that you'll need GHC installed in
1279 order to build it. To avoid the chicken/egg problem,
1280 install a binary distribution of either Happy or GHC to get
1281 started. Happy distributions are available from <ulink
1282 url="http://www.haskell.org/happy/">Happy's Web
1283 Page</ulink>.</para>
1289 <indexterm><primary>Alex</primary></indexterm>
1291 <para>Alex is a lexical-analyser generator for Haskell,
1292 which GHC uses to generate its lexer. Like Happy, Alex is
1293 written in Haskell and is a project in the CVS repository.
1294 Alex distributions are available from <ulink
1295 url="http://www.haskell.org/alex/">Alex's Web
1296 Page</ulink>.</para>
1301 <term>Autoconf</term>
1302 <indexterm><primary>pre-supposed: Autoconf</primary></indexterm>
1303 <indexterm><primary>Autoconf, pre-supposed</primary></indexterm>
1305 <para>GNU Autoconf is needed if you intend to build from the
1306 CVS sources, it is <emphasis>not</emphasis> needed if you
1307 just intend to build a standard source distribution.</para>
1309 <para>Version 2.52 or later of autoconf is required.
1310 NB. vesrion 2.13 will no longer work, as of GHC version
1313 <para>Autoconf builds the <command>configure</command>
1314 script from <filename>configure.ac</filename> and
1315 <filename>aclocal.m4</filename>. If you modify either of
1316 these files, you'll need <command>autoconf</command> to
1317 rebuild <filename>configure</filename>.</para>
1322 <term><command>sed</command></term>
1323 <indexterm><primary>pre-supposed: sed</primary></indexterm>
1324 <indexterm><primary>sed, pre-supposed</primary></indexterm>
1326 <para>You need a working <command>sed</command> if you are
1327 going to build from sources. The build-configuration stuff
1328 needs it. GNU sed version 2.0.4 is no good! It has a bug
1329 in it that is tickled by the build-configuration. 2.0.5 is
1330 OK. Others are probably OK too (assuming we don't create too
1331 elaborate configure scripts.)</para>
1336 <para>One <literal>fptools</literal> project is worth a quick note
1337 at this point, because it is useful for all the others:
1338 <literal>glafp-utils</literal> contains several utilities which
1339 aren't particularly Glasgow-ish, but Occasionally Indispensable.
1340 Like <command>lndir</command> for creating symbolic link
1343 <sect2 id="pre-supposed-gph-tools">
1344 <title>Tools for building parallel GHC (GPH)</title>
1348 <term>PVM version 3:</term>
1349 <indexterm><primary>pre-supposed: PVM3 (Parallel Virtual Machine)</primary></indexterm>
1350 <indexterm><primary>PVM3 (Parallel Virtual Machine), pre-supposed</primary></indexterm>
1352 <para>PVM is the Parallel Virtual Machine on which
1353 Parallel Haskell programs run. (You only need this if you
1354 plan to run Parallel Haskell. Concurrent Haskell, which
1355 runs concurrent threads on a uniprocessor doesn't need
1356 it.) Underneath PVM, you can have (for example) a network
1357 of workstations (slow) or a multiprocessor box
1360 <para>The current version of PVM is 3.3.11; we use 3.3.7.
1361 It is readily available on the net; I think I got it from
1362 <literal>research.att.com</literal>, in
1363 <filename>netlib</filename>.</para>
1365 <para>A PVM installation is slightly quirky, but easy to
1366 do. Just follow the <filename>Readme</filename>
1367 instructions.</para>
1372 <term><command>bash</command>:</term>
1373 <indexterm><primary>bash, presupposed (Parallel Haskell only)</primary></indexterm>
1375 <para>Sadly, the <command>gr2ps</command> script, used to
1376 convert “parallelism profiles” to PostScript,
1377 is written in Bash (GNU's Bourne Again shell). This bug
1378 will be fixed (someday).</para>
1384 <sect2 id="pre-supposed-other-tools">
1385 <title>Other useful tools</title>
1390 <indexterm><primary>pre-supposed: flex</primary></indexterm>
1391 <indexterm><primary>flex, pre-supposed</primary></indexterm>
1393 <para>This is a quite-a-bit-better-than-Lex lexer. Used
1394 to build a couple of utilities in
1395 <literal>glafp-utils</literal>. Depending on your
1396 operating system, the supplied <command>lex</command> may
1397 or may not work; you should get the GNU version.</para>
1402 <para>More tools are required if you want to format the documentation
1403 that comes with GHC and other fptools projects. See <xref
1404 linkend="building-docs">.</para>
1408 <sect1 id="sec-building-from-source">
1409 <title>Building from source</title>
1411 <indexterm><primary>Building from source</primary></indexterm>
1412 <indexterm><primary>Source, building from</primary></indexterm>
1414 <para>You've been rash enough to want to build some of the Glasgow
1415 Functional Programming tools (GHC, Happy, nofib, etc.) from
1416 source. You've slurped the source, from the CVS repository or
1417 from a source distribution, and now you're sitting looking at a
1418 huge mound of bits, wondering what to do next.</para>
1420 <para>Gingerly, you type <command>make</command>. Wrong
1423 <para>This rest of this guide is intended for duffers like me, who
1424 aren't really interested in Makefiles and systems configurations,
1425 but who need a mental model of the interlocking pieces so that
1426 they can make them work, extend them consistently when adding new
1427 software, and lay hands on them gently when they don't
1430 <sect2 id="quick-start">
1431 <title>Quick Start</title>
1433 <para>If you are starting from a source distribution, and just
1434 want a completely standard build, then the following should
1437 <screen>$ ./configure
1442 <para>For GHC, this will do a 2-stage bootstrap build of the
1443 compiler, with profiling libraries, and install the
1446 <para>If you want to do anything at all non-standard, or you
1447 want to do some development, read on...</para>
1450 <sect2 id="sec-source-tree">
1451 <title>Your source tree</title>
1453 <para>The source code is held in your <emphasis>source
1454 tree</emphasis>. The root directory of your source tree
1455 <emphasis>must</emphasis> contain the following directories and
1460 <para><filename>Makefile</filename>: the root
1465 <para><filename>mk/</filename>: the directory that contains
1466 the main Makefile code, shared by all the
1467 <literal>fptools</literal> software.</para>
1471 <para><filename>configure.ac</filename>,
1472 <filename>config.sub</filename>,
1473 <filename>config.guess</filename>: these files support the
1474 configuration process.</para>
1478 <para><filename>install-sh</filename>.</para>
1482 <para>All the other directories are individual
1483 <emphasis>projects</emphasis> of the <literal>fptools</literal>
1484 system—for example, the Glasgow Haskell Compiler
1485 (<literal>ghc</literal>), the Happy parser generator
1486 (<literal>happy</literal>), the <literal>nofib</literal>
1487 benchmark suite, and so on. You can have zero or more of these.
1488 Needless to say, some of them are needed to build others.</para>
1490 <para>The important thing to remember is that even if you want
1491 only one project (<literal>happy</literal>, say), you must have
1492 a source tree whose root directory contains
1493 <filename>Makefile</filename>, <filename>mk/</filename>,
1494 <filename>configure.ac</filename>, and the project(s) you want
1495 (<filename>happy/</filename> in this case). You cannot get by
1496 with just the <filename>happy/</filename> directory.</para>
1500 <title>Build trees</title>
1501 <indexterm><primary>build trees</primary></indexterm>
1502 <indexterm><primary>link trees, for building</primary></indexterm>
1504 <para>If you just want to build the software once on a single
1505 platform, then your source tree can also be your build tree, and
1506 you can skip the rest of this section.</para>
1508 <para>We often want to build multiple versions of our software
1509 for different architectures, or with different options
1510 (e.g. profiling). It's very desirable to share a single copy of
1511 the source code among all these builds.</para>
1513 <para>So for every source tree we have zero or more
1514 <emphasis>build trees</emphasis>. Each build tree is initially
1515 an exact copy of the source tree, except that each file is a
1516 symbolic link to the source file, rather than being a copy of
1517 the source file. There are “standard” Unix
1518 utilities that make such copies, so standard that they go by
1520 <command>lndir</command><indexterm><primary>lndir</primary></indexterm>,
1521 <command>mkshadowdir</command><indexterm><primary>mkshadowdir</primary></indexterm>
1522 are two (If you don't have either, the source distribution
1523 includes sources for the X11
1524 <command>lndir</command>—check out
1525 <filename>fptools/glafp-utils/lndir</filename>). See <Xref
1526 LinkEnd="sec-storysofar"> for a typical invocation.</para>
1528 <para>The build tree does not need to be anywhere near the
1529 source tree in the file system. Indeed, one advantage of
1530 separating the build tree from the source is that the build tree
1531 can be placed in a non-backed-up partition, saving your systems
1532 support people from backing up untold megabytes of
1533 easily-regenerated, and rapidly-changing, gubbins. The golden
1534 rule is that (with a single exception—<XRef
1535 LinkEnd="sec-build-config">) <emphasis>absolutely everything in
1536 the build tree is either a symbolic link to the source tree, or
1537 else is mechanically generated</emphasis>. It should be
1538 perfectly OK for your build tree to vanish overnight; an hour or
1539 two compiling and you're on the road again.</para>
1541 <para>You need to be a bit careful, though, that any new files
1542 you create (if you do any development work) are in the source
1543 tree, not a build tree!</para>
1545 <para>Remember, that the source files in the build tree are
1546 <emphasis>symbolic links</emphasis> to the files in the source
1547 tree. (The build tree soon accumulates lots of built files like
1548 <filename>Foo.o</filename>, as well.) You can
1549 <emphasis>delete</emphasis> a source file from the build tree
1550 without affecting the source tree (though it's an odd thing to
1551 do). On the other hand, if you <emphasis>edit</emphasis> a
1552 source file from the build tree, you'll edit the source-tree
1553 file directly. (You can set up Emacs so that if you edit a
1554 source file from the build tree, Emacs will silently create an
1555 edited copy of the source file in the build tree, leaving the
1556 source file unchanged; but the danger is that you think you've
1557 edited the source file whereas actually all you've done is edit
1558 the build-tree copy. More commonly you do want to edit the
1559 source file.)</para>
1561 <para>Like the source tree, the top level of your build tree
1562 must be (a linked copy of) the root directory of the
1563 <literal>fptools</literal> suite. Inside Makefiles, the root of
1564 your build tree is called
1565 <constant>$(FPTOOLS_TOP)</constant><indexterm><primary>FPTOOLS_TOP</primary></indexterm>.
1566 In the rest of this document path names are relative to
1567 <constant>$(FPTOOLS_TOP)</constant> unless
1568 otherwise stated. For example, the file
1569 <filename>ghc/mk/target.mk</filename> is actually
1570 <filename><constant>$(FPTOOLS_TOP)</constant>/ghc/mk/target.mk</filename>.</para>
1573 <sect2 id="sec-build-config">
1574 <title>Getting the build you want</title>
1576 <para>When you build <literal>fptools</literal> you will be
1577 compiling code on a particular <emphasis>host
1578 platform</emphasis>, to run on a particular <emphasis>target
1579 platform</emphasis> (usually the same as the host
1580 platform)<indexterm><primary>platform</primary></indexterm>.
1581 The difficulty is that there are minor differences between
1582 different platforms; minor, but enough that the code needs to be
1583 a bit different for each. There are some big differences too:
1584 for a different architecture we need to build GHC with a
1585 different native-code generator.</para>
1587 <para>There are also knobs you can turn to control how the
1588 <literal>fptools</literal> software is built. For example, you
1589 might want to build GHC optimised (so that it runs fast) or
1590 unoptimised (so that you can compile it fast after you've
1591 modified it. Or, you might want to compile it with debugging on
1592 (so that extra consistency-checking code gets included) or off.
1595 <para>All of this stuff is called the
1596 <emphasis>configuration</emphasis> of your build. You set the
1597 configuration using a three-step process.</para>
1601 <term>Step 1: get ready for configuration.</term>
1603 <para>NOTE: if you're starting from a source distribution,
1604 rather than CVS sources, you can skip this step.</para>
1606 <para>Change directory to
1607 <constant>$(FPTOOLS_TOP)</constant> and
1609 <command>autoconf</command><indexterm><primary>autoconf</primary></indexterm>
1610 (with no arguments). This GNU program converts
1611 <filename><constant>$(FPTOOLS_TOP)</constant>/configure.ac</filename>
1612 to a shell script called
1613 <filename><constant>$(FPTOOLS_TOP)</constant>/configure</filename>.
1616 <para>Some projects, including GHC, have their own
1617 configure script. If there's an
1618 <constant>$(FPTOOLS_TOP)/<project>/configure.ac</constant>,
1619 then you need to run <command>autoconf</command> in that
1620 directory too.</para>
1622 <para>Both these steps are completely
1623 platform-independent; they just mean that the
1624 human-written file (<filename>configure.ac</filename>) can
1625 be short, although the resulting shell script,
1626 <command>configure</command>, and
1627 <filename>mk/config.h.in</filename>, are long.</para>
1632 <term>Step 2: system configuration.</term>
1634 <para>Runs the newly-created <command>configure</command>
1635 script, thus:</para>
1638 ./configure <optional><parameter>args</parameter></optional>
1641 <para><command>configure</command>'s mission is to scurry
1642 round your computer working out what architecture it has,
1643 what operating system, whether it has the
1644 <Function>vfork</Function> system call, where
1645 <command>tar</command> is kept, whether
1646 <command>gcc</command> is available, where various obscure
1647 <literal>#include</literal> files are, whether it's a
1648 leap year, and what the systems manager had for lunch. It
1649 communicates these snippets of information in two
1656 <filename>mk/config.mk.in</filename><indexterm><primary>config.mk.in</primary></indexterm>
1658 <filename>mk/config.mk</filename><indexterm><primary>config.mk</primary></indexterm>,
1659 substituting for things between
1660 “<literal>@</literal>” brackets. So,
1661 “<literal>@HaveGcc@</literal>” will be
1662 replaced by “<literal>YES</literal>” or
1663 “<literal>NO</literal>” depending on what
1664 <command>configure</command> finds.
1665 <filename>mk/config.mk</filename> is included by every
1666 Makefile (directly or indirectly), so the
1667 configuration information is thereby communicated to
1668 all Makefiles.</para>
1672 <para> It translates
1673 <filename>mk/config.h.in</filename><indexterm><primary>config.h.in</primary></indexterm>
1675 <filename>mk/config.h</filename><indexterm><primary>config.h</primary></indexterm>.
1676 The latter is <literal>#include</literal>d by
1677 various C programs, which can thereby make use of
1678 configuration information.</para>
1682 <para><command>configure</command> takes some optional
1683 arguments. Use <literal>./configure --help</literal> to
1684 get a list of the available arguments. Here are some of
1685 the ones you might need:</para>
1689 <term><literal>--with-ghc=<parameter>path</parameter></literal></term>
1690 <indexterm><primary><literal>--with-ghc</literal></primary>
1693 <para>Specifies the path to an installed GHC which
1694 you would like to use. This compiler will be used
1695 for compiling GHC-specific code (eg. GHC itself).
1696 This option <emphasis>cannot</emphasis> be specified
1697 using <filename>build.mk</filename> (see later),
1698 because <command>configure</command> needs to
1699 auto-detect the version of GHC you're using. The
1700 default is to look for a compiler named
1701 <literal>ghc</literal> in your path.</para>
1706 <term><literal>--with-hc=<parameter>path</parameter></literal></term>
1707 <indexterm><primary><literal>--with-hc</literal></primary>
1710 <para>Specifies the path to any installed Haskell
1711 compiler. This compiler will be used for compiling
1712 generic Haskell code. The default is to use
1713 <literal>ghc</literal>.</para>
1718 <term><literal>--with-gcc=<parameter>path</parameter></literal></term>
1719 <indexterm><primary><literal>--with-gcc</literal></primary>
1722 <para>Specifies the path to the installed GCC. This
1723 compiler will be used to compile all C files,
1724 <emphasis>except</emphasis> any generated by the
1725 installed Haskell compiler, which will have its own
1726 idea of which C compiler (if any) to use. The
1727 default is to use <literal>gcc</literal>.</para>
1732 <para><command>configure</command> caches the results of
1733 its run in <filename>config.cache</filename>. Quite often
1734 you don't want that; you're running
1735 <command>configure</command> a second time because
1736 something has changed. In that case, simply delete
1737 <filename>config.cache</filename>.</para>
1742 <term>Step 3: build configuration.</term>
1744 <para>Next, you say how this build of
1745 <literal>fptools</literal> is to differ from the standard
1746 defaults by creating a new file
1747 <filename>mk/build.mk</filename><indexterm><primary>build.mk</primary></indexterm>
1748 <emphasis>in the build tree</emphasis>. This file is the
1749 one and only file you edit in the build tree, precisely
1750 because it says how this build differs from the source.
1751 (Just in case your build tree does die, you might want to
1752 keep a private directory of <filename>build.mk</filename>
1753 files, and use a symbolic link in each build tree to point
1754 to the appropriate one.) So
1755 <filename>mk/build.mk</filename> never exists in the
1756 source tree—you create one in each build tree from
1757 the template. We'll discuss what to put in it
1763 <para>And that's it for configuration. Simple, eh?</para>
1765 <para>What do you put in your build-specific configuration file
1766 <filename>mk/build.mk</filename>? <emphasis>For almost all
1767 purposes all you will do is put make variable definitions that
1768 override those in</emphasis>
1769 <filename>mk/config.mk.in</filename>. The whole point of
1770 <filename>mk/config.mk.in</filename>—and its derived
1771 counterpart <filename>mk/config.mk</filename>—is to define
1772 the build configuration. It is heavily commented, as you will
1773 see if you look at it. So generally, what you do is look at
1774 <filename>mk/config.mk.in</filename>, and add definitions in
1775 <filename>mk/build.mk</filename> that override any of the
1776 <filename>config.mk</filename> definitions that you want to
1777 change. (The override occurs because the main boilerplate file,
1778 <filename>mk/boilerplate.mk</filename><indexterm><primary>boilerplate.mk</primary></indexterm>,
1779 includes <filename>build.mk</filename> after
1780 <filename>config.mk</filename>.)</para>
1782 <para>For your convenience, there's a file called <filename>build.mk.sample</filename>
1783 that can serve as a starting point for your <filename>build.mk</filename>.</para>
1785 <para>For example, <filename>config.mk.in</filename> contains
1786 the definition:</para>
1789 GhcHcOpts=-O -Rghc-timing
1792 <para>The accompanying comment explains that this is the list of
1793 flags passed to GHC when building GHC itself. For doing
1794 development, it is wise to add <literal>-DDEBUG</literal>, to
1795 enable debugging code. So you would add the following to
1796 <filename>build.mk</filename>:</para>
1798 <para>or, if you prefer,</para>
1801 GhcHcOpts += -DDEBUG
1804 <para>GNU <command>make</command> allows existing definitions to
1805 have new text appended using the “<literal>+=</literal>”
1806 operator, which is quite a convenient feature.)</para>
1808 <para>If you want to remove the <literal>-O</literal> as well (a
1809 good idea when developing, because the turn-around cycle gets a
1810 lot quicker), you can just override
1811 <literal>GhcLibHcOpts</literal> altogether:</para>
1814 GhcHcOpts=-DDEBUG -Rghc-timing
1817 <para>When reading <filename>config.mk.in</filename>, remember
1818 that anything between “@...@” signs is going to be substituted
1819 by <command>configure</command> later. You
1820 <emphasis>can</emphasis> override the resulting definition if
1821 you want, but you need to be a bit surer what you are doing.
1822 For example, there's a line that says:</para>
1828 <para>This defines the Make variables <constant>TAR</constant>
1829 to the pathname for a <command>tar</command> that
1830 <command>configure</command> finds somewhere. If you have your
1831 own pet <command>tar</command> you want to use instead, that's
1832 fine. Just add this line to <filename>mk/build.mk</filename>:</para>
1838 <para>You do not <emphasis>have</emphasis> to have a
1839 <filename>mk/build.mk</filename> file at all; if you don't,
1840 you'll get all the default settings from
1841 <filename>mk/config.mk.in</filename>.</para>
1843 <para>You can also use <filename>build.mk</filename> to override
1844 anything that <command>configure</command> got wrong. One place
1845 where this happens often is with the definition of
1846 <constant>FPTOOLS_TOP_ABS</constant>: this
1847 variable is supposed to be the canonical path to the top of your
1848 source tree, but if your system uses an automounter then the
1849 correct directory is hard to find automatically. If you find
1850 that <command>configure</command> has got it wrong, just put the
1851 correct definition in <filename>build.mk</filename>.</para>
1855 <sect2 id="sec-storysofar">
1856 <title>The story so far</title>
1858 <para>Let's summarise the steps you need to carry to get
1859 yourself a fully-configured build tree from scratch.</para>
1863 <para> Get your source tree from somewhere (CVS repository
1864 or source distribution). Say you call the root directory
1865 <filename>myfptools</filename> (it does not have to be
1866 called <filename>fptools</filename>). Make sure that you
1867 have the essential files (see <XRef
1868 LinkEnd="sec-source-tree">).</para>
1873 <para>(Optional) Use <command>lndir</command> or
1874 <command>mkshadowdir</command> to create a build tree.</para>
1878 $ mkshadowdir . /scratch/joe-bloggs/myfptools-sun4
1881 <para>(N.B. <command>mkshadowdir</command>'s first argument
1882 is taken relative to its second.) You probably want to give
1883 the build tree a name that suggests its main defining
1884 characteristic (in your mind at least), in case you later
1889 <para>Change directory to the build tree. Everything is
1890 going to happen there now.</para>
1893 $ cd /scratch/joe-bloggs/myfptools-sun4
1899 <para>Prepare for system configuration:</para>
1905 <para>(You can skip this step if you are starting from a
1906 source distribution, and you already have
1907 <filename>configure</filename> and
1908 <filename>mk/config.h.in</filename>.)</para>
1910 <para>Some projects, including GHC itself, have their own
1911 configure scripts, so it is necessary to run autoconf again
1912 in the appropriate subdirectories. eg:</para>
1915 $ (cd ghc; autoconf)
1920 <para>Do system configuration:</para>
1926 <para>Don't forget to check whether you need to add any
1927 arguments to <literal>configure</literal>; for example, a
1928 common requirement is to specify which GHC to use with
1929 <option>--with-ghc=<replaceable>ghc</replaceable></option>.</para>
1933 <para>Create the file <filename>mk/build.mk</filename>,
1934 adding definitions for your desired configuration
1943 <para>You can make subsequent changes to
1944 <filename>mk/build.mk</filename> as often as you like. You do
1945 not have to run any further configuration programs to make these
1946 changes take effect. In theory you should, however, say
1947 <command>gmake clean</command>, <command>gmake all</command>,
1948 because configuration option changes could affect
1949 anything—but in practice you are likely to know what's
1954 <title>Making things</title>
1956 <para>At this point you have made yourself a fully-configured
1957 build tree, so you are ready to start building real
1960 <para>The first thing you need to know is that <emphasis>you
1961 must use GNU <command>make</command>, usually called
1962 <command>gmake</command>, not standard Unix
1963 <command>make</command></emphasis>. If you use standard Unix
1964 <command>make</command> you will get all sorts of error messages
1965 (but no damage) because the <literal>fptools</literal>
1966 <command>Makefiles</command> use GNU <command>make</command>'s
1967 facilities extensively.</para>
1969 <para>To just build the whole thing, <command>cd</command> to
1970 the top of your <literal>fptools</literal> tree and type
1971 <command>gmake</command>. This will prepare the tree and build
1972 the various projects in the correct order.</para>
1975 <sect2 id="sec-bootstrapping">
1976 <title>Bootstrapping GHC</title>
1978 <para>GHC requires a 2-stage bootstrap in order to provide
1979 full functionality, including GHCi. By a 2-stage bootstrap, we
1980 mean that the compiler is built once using the installed GHC,
1981 and then again using the compiler built in the first stage. You
1982 can also build a stage 3 compiler, but this normally isn't
1983 necessary except to verify that the stage 2 compiler is working
1986 <para>Note that when doing a bootstrap, the stage 1 compiler
1987 must be built, followed by the runtime system and libraries, and
1988 then the stage 2 compiler. The correct ordering is implemented
1989 by the top-level fptools <filename>Makefile</filename>, so if
1990 you want everything to work automatically it's best to start
1991 <command>make</command> from the top of the tree. When building
1992 GHC, the top-level fptools <filename>Makefile</filename> is set
1993 up to do a 2-stage bootstrap by default (when you say
1994 <command>make</command>). Some other targets it supports
2001 <para>Build everything as normal, including the stage 1
2009 <para>Build the stage 2 compiler only.</para>
2016 <para>Build the stage 3 compiler only.</para>
2021 <term>bootstrap</term> <term>bootstrap2</term>
2023 <para>Build stage 1 followed by stage 2.</para>
2028 <term>bootstrap3</term>
2030 <para>Build stages 1, 2 and 3.</para>
2035 <term>install</term>
2037 <para>Install everything, including the compiler built in
2038 stage 2. To override the stage, say <literal>make install
2039 stage=<replaceable>n</replaceable></literal> where
2040 <replaceable>n</replaceable> is the stage to install.</para>
2045 <para>The top-level <filename>Makefile</filename> also arranges
2046 to do the appropriate <literal>make boot</literal> steps (see
2047 below) before actually building anything.</para>
2049 <para>The <literal>stage1</literal>, <literal>stage2</literal>
2050 and <literal>stage3</literal> targets also work in the
2051 <literal>ghc/compiler</literal> directory, but don't forget that
2052 each stage requires its own <literal>make boot</literal> step:
2053 for example, you must do</para>
2055 <screen>$ make boot stage=2</screen>
2057 <para>before <literal>make stage2</literal> in
2058 <literal>ghc/compiler</literal>.</para>
2061 <sect2 id="sec-standard-targets">
2062 <title>Standard Targets</title>
2063 <indexterm><primary>targets, standard makefile</primary></indexterm>
2064 <indexterm><primary>makefile targets</primary></indexterm>
2066 <para>In any directory you should be able to make the following:</para>
2070 <term><literal>boot</literal></term>
2072 <para>does the one-off preparation required to get ready
2073 for the real work. Notably, it does <command>gmake
2074 depend</command> in all directories that contain programs.
2075 It also builds the necessary tools for compilation to
2078 <para>Invoking the <literal>boot</literal> target
2079 explicitly is not normally necessary. From the top-level
2080 <literal>fptools</literal> directory, invoking
2081 <literal>gmake</literal> causes <literal>gmake boot
2082 all</literal> to be invoked in each of the project
2083 subdirectories, in the order specified by
2084 <literal>$(AllTargets)</literal> in
2085 <literal>config.mk</literal>.</para>
2087 <para>If you're working in a subdirectory somewhere and
2088 need to update the dependencies, <literal>gmake
2089 boot</literal> is a good way to do it.</para>
2094 <term><literal>all</literal></term>
2096 <para>makes all the final target(s) for this Makefile.
2097 Depending on which directory you are in a “final
2098 target” may be an executable program, a library
2099 archive, a shell script, or a Postscript file. Typing
2100 <command>gmake</command> alone is generally the same as
2101 typing <command>gmake all</command>.</para>
2106 <term><literal>install</literal></term>
2108 <para>installs the things built by <literal>all</literal>
2109 (except for the documentation). Where does it install
2110 them? That is specified by
2111 <filename>mk/config.mk.in</filename>; you can override it
2112 in <filename>mk/build.mk</filename>, or by running
2113 <command>configure</command> with command-line arguments
2114 like <literal>--bindir=/home/simonpj/bin</literal>; see
2115 <literal>./configure --help</literal> for the full
2121 <term><literal>install-docs</literal></term>
2123 <para>installs the documentation. Otherwise behaves just
2124 like <literal>install</literal>.</para>
2129 <term><literal>uninstall</literal></term>
2131 <para>reverses the effect of
2132 <literal>install</literal>.</para>
2137 <term><literal>clean</literal></term>
2139 <para>Delete all files from the current directory that are
2140 normally created by building the program. Don't delete
2141 the files that record the configuration, or files
2142 generated by <command>gmake boot</command>. Also preserve
2143 files that could be made by building, but normally aren't
2144 because the distribution comes with them.</para>
2149 <term><literal>distclean</literal></term>
2151 <para>Delete all files from the current directory that are
2152 created by configuring or building the program. If you
2153 have unpacked the source and built the program without
2154 creating any other files, <literal>make
2155 distclean</literal> should leave only the files that were
2156 in the distribution.</para>
2161 <term><literal>mostlyclean</literal></term>
2163 <para>Like <literal>clean</literal>, but may refrain from
2164 deleting a few files that people normally don't want to
2170 <term><literal>maintainer-clean</literal></term>
2172 <para>Delete everything from the current directory that
2173 can be reconstructed with this Makefile. This typically
2174 includes everything deleted by
2175 <literal>distclean</literal>, plus more: C source files
2176 produced by Bison, tags tables, Info files, and so
2179 <para>One exception, however: <literal>make
2180 maintainer-clean</literal> should not delete
2181 <filename>configure</filename> even if
2182 <filename>configure</filename> can be remade using a rule
2183 in the <filename>Makefile</filename>. More generally,
2184 <literal>make maintainer-clean</literal> should not delete
2185 anything that needs to exist in order to run
2186 <filename>configure</filename> and then begin to build the
2192 <term><literal>check</literal></term>
2194 <para>run the test suite.</para>
2199 <para>All of these standard targets automatically recurse into
2200 sub-directories. Certain other standard targets do not:</para>
2204 <term><literal>configure</literal></term>
2206 <para>is only available in the root directory
2207 <constant>$(FPTOOLS_TOP)</constant>; it has
2208 been discussed in <XRef
2209 LinkEnd="sec-build-config">.</para>
2214 <term><literal>depend</literal></term>
2216 <para>make a <filename>.depend</filename> file in each
2217 directory that needs it. This <filename>.depend</filename>
2218 file contains mechanically-generated dependency
2219 information; for example, suppose a directory contains a
2220 Haskell source module <filename>Foo.lhs</filename> which
2221 imports another module <literal>Baz</literal>. Then the
2222 generated <filename>.depend</filename> file will contain
2223 the dependency:</para>
2229 <para>which says that the object file
2230 <filename>Foo.o</filename> depends on the interface file
2231 <filename>Baz.hi</filename> generated by compiling module
2232 <literal>Baz</literal>. The <filename>.depend</filename>
2233 file is automatically included by every Makefile.</para>
2238 <term><literal>binary-dist</literal></term>
2240 <para>make a binary distribution. This is the target we
2241 use to build the binary distributions of GHC and
2247 <term><literal>dist</literal></term>
2249 <para>make a source distribution. Note that this target
2250 does “make distclean” as part of its work;
2251 don't use it if you want to keep what you've built.</para>
2256 <para>Most <filename>Makefile</filename>s have targets other
2257 than these. You can discover them by looking in the
2258 <filename>Makefile</filename> itself.</para>
2262 <title>Using a project from the build tree</title>
2264 <para>If you want to build GHC (say) and just use it direct from
2265 the build tree without doing <literal>make install</literal>
2266 first, you can run the in-place driver script:
2267 <filename>ghc/compiler/ghc-inplace</filename>.</para>
2269 <para> Do <emphasis>NOT</emphasis> use
2270 <filename>ghc/compiler/ghc</filename>, or
2271 <filename>ghc/compiler/ghc-6.xx</filename>, as these are the
2272 scripts intended for installation, and contain hard-wired paths
2273 to the installed libraries, rather than the libraries in the
2276 <para>Happy can similarly be run from the build tree, using
2277 <filename>happy/src/happy-inplace</filename>, and similarly for
2278 Alex and Haddock.</para>
2282 <title>Fast Making</title>
2284 <indexterm><primary>fastmake</primary></indexterm>
2285 <indexterm><primary>dependencies, omitting</primary></indexterm>
2286 <indexterm><primary>FAST, makefile variable</primary></indexterm>
2288 <para>Sometimes the dependencies get in the way: if you've made
2289 a small change to one file, and you're absolutely sure that it
2290 won't affect anything else, but you know that
2291 <command>make</command> is going to rebuild everything anyway,
2292 the following hack may be useful:</para>
2298 <para>This tells the make system to ignore dependencies and just
2299 build what you tell it to. In other words, it's equivalent to
2300 temporarily removing the <filename>.depend</filename> file in
2301 the current directory (where <command>mkdependHS</command> and
2302 friends store their dependency information).</para>
2304 <para>A bit of history: GHC used to come with a
2305 <command>fastmake</command> script that did the above job, but
2306 GNU make provides the features we need to do it without
2307 resorting to a script. Also, we've found that fastmaking is
2308 less useful since the advent of GHC's recompilation checker (see
2309 the User's Guide section on "Separate Compilation").</para>
2313 <sect1 id="sec-makefile-arch">
2314 <title>The <filename>Makefile</filename> architecture</title>
2315 <indexterm><primary>makefile architecture</primary></indexterm>
2317 <para><command>make</command> is great if everything
2318 works—you type <command>gmake install</command> and lo! the
2319 right things get compiled and installed in the right places. Our
2320 goal is to make this happen often, but somehow it often doesn't;
2321 instead some weird error message eventually emerges from the
2322 bowels of a directory you didn't know existed.</para>
2324 <para>The purpose of this section is to give you a road-map to
2325 help you figure out what is going right and what is going
2329 <title>Debugging</title>
2331 <para>Debugging <filename>Makefile</filename>s is something of a
2332 black art, but here's a couple of tricks that we find
2333 particularly useful. The following command allows you to see
2334 the contents of any make variable in the context of the current
2335 <filename>Makefile</filename>:</para>
2337 <screen>$ make show VALUE=HS_SRCS</screen>
2339 <para>where you can replace <literal>HS_SRCS</literal> with the
2340 name of any variable you wish to see the value of.</para>
2342 <para>GNU make has a <option>-d</option> option which generates
2343 a dump of the decision procedure used to arrive at a conclusion
2344 about which files should be recompiled. Sometimes useful for
2345 tracking down problems with superfluous or missing
2346 recompilations.</para>
2350 <title>A small project</title>
2352 <para>To get started, let us look at the
2353 <filename>Makefile</filename> for an imaginary small
2354 <literal>fptools</literal> project, <literal>small</literal>.
2355 Each project in <literal>fptools</literal> has its own directory
2356 in <constant>FPTOOLS_TOP</constant>, so the
2357 <literal>small</literal> project will have its own directory
2358 <constant>FPOOLS_TOP/small/</constant>. Inside the
2359 <filename>small/</filename> directory there will be a
2360 <filename>Makefile</filename>, looking something like
2363 <indexterm><primary>Makefile, minimal</primary></indexterm>
2366 # Makefile for fptools project "small"
2369 include $(TOP)/mk/boilerplate.mk
2371 SRCS = $(wildcard *.lhs) $(wildcard *.c)
2374 include $(TOP)/target.mk
2377 <para>this <filename>Makefile</filename> has three
2382 <para>The first section includes
2385 One of the most important
2386 features of GNU <command>make</command> that we use is the ability for a <filename>Makefile</filename> to
2387 include another named file, very like <command>cpp</command>'s <literal>#include</literal>
2392 a file of “boilerplate” code from the level
2393 above (which in this case will be
2394 <filename><constant>FPTOOLS_TOP</constant>/mk/boilerplate.mk</filename><indexterm><primary>boilerplate.mk</primary></indexterm>).
2395 As its name suggests, <filename>boilerplate.mk</filename>
2396 consists of a large quantity of standard
2397 <filename>Makefile</filename> code. We discuss this
2398 boilerplate in more detail in <XRef LinkEnd="sec-boiler">.
2399 <indexterm><primary>include, directive in
2400 Makefiles</primary></indexterm> <indexterm><primary>Makefile
2401 inclusion</primary></indexterm></para>
2403 <para>Before the <literal>include</literal> statement, you
2404 must define the <command>make</command> variable
2405 <constant>TOP</constant><indexterm><primary>TOP</primary></indexterm>
2406 to be the directory containing the <filename>mk</filename>
2407 directory in which the <filename>boilerplate.mk</filename>
2408 file is. It is <emphasis>not</emphasis> OK to simply say</para>
2411 include ../mk/boilerplate.mk # NO NO NO
2415 <para>Why? Because the <filename>boilerplate.mk</filename>
2416 file needs to know where it is, so that it can, in turn,
2417 <literal>include</literal> other files. (Unfortunately,
2418 when an <literal>include</literal>d file does an
2419 <literal>include</literal>, the filename is treated relative
2420 to the directory in which <command>gmake</command> is being
2421 run, not the directory in which the
2422 <literal>include</literal>d sits.) In general,
2423 <emphasis>every file <filename>foo.mk</filename> assumes
2425 <filename><constant>$(TOP)</constant>/mk/foo.mk</filename>
2426 refers to itself.</emphasis> It is up to the
2427 <filename>Makefile</filename> doing the
2428 <literal>include</literal> to ensure this is the case.</para>
2430 <para>Files intended for inclusion in other
2431 <filename>Makefile</filename>s are written to have the
2432 following property: <emphasis>after
2433 <filename>foo.mk</filename> is <literal>include</literal>d,
2434 it leaves <constant>TOP</constant> containing the same value
2435 as it had just before the <literal>include</literal>
2436 statement</emphasis>. In our example, this invariant
2437 guarantees that the <literal>include</literal> for
2438 <filename>target.mk</filename> will look in the same
2439 directory as that for <filename>boilerplate.mk</filename>.</para>
2443 <para> The second section defines the following standard
2444 <command>make</command> variables:
2445 <constant>SRCS</constant><indexterm><primary>SRCS</primary></indexterm>
2446 (the source files from which is to be built), and
2447 <constant>HS_PROG</constant><indexterm><primary>HS_PROG</primary></indexterm>
2448 (the executable binary to be built). We will discuss in
2449 more detail what the “standard variables” are,
2450 and how they affect what happens, in <XRef
2451 LinkEnd="sec-targets">.</para>
2453 <para>The definition for <constant>SRCS</constant> uses the
2454 useful GNU <command>make</command> construct
2455 <literal>$(wildcard $pat$)</literal><indexterm><primary>wildcard</primary></indexterm>,
2456 which expands to a list of all the files matching the
2457 pattern <literal>pat</literal> in the current directory. In
2458 this example, <constant>SRCS</constant> is set to the list
2459 of all the <filename>.lhs</filename> and
2460 <filename>.c</filename> files in the directory. (Let's
2461 suppose there is one of each, <filename>Foo.lhs</filename>
2462 and <filename>Baz.c</filename>.)</para>
2466 <para>The last section includes a second file of standard
2468 <filename>target.mk</filename><indexterm><primary>target.mk</primary></indexterm>.
2469 It contains the rules that tell <command>gmake</command> how
2470 to make the standard targets (<Xref
2471 LinkEnd="sec-standard-targets">). Why, you ask, can't this
2472 standard code be part of
2473 <filename>boilerplate.mk</filename>? Good question. We
2474 discuss the reason later, in <Xref
2475 LinkEnd="sec-boiler-arch">.</para>
2477 <para>You do not <emphasis>have</emphasis> to
2478 <literal>include</literal> the
2479 <filename>target.mk</filename> file. Instead, you can write
2480 rules of your own for all the standard targets. Usually,
2481 though, you will find quite a big payoff from using the
2482 canned rules in <filename>target.mk</filename>; the price
2483 tag is that you have to understand what canned rules get
2484 enabled, and what they do (<Xref
2485 LinkEnd="sec-targets">).</para>
2489 <para>In our example <filename>Makefile</filename>, most of the
2490 work is done by the two <literal>include</literal>d files. When
2491 you say <command>gmake all</command>, the following things
2496 <para><command>gmake</command> figures out that the object
2497 files are <filename>Foo.o</filename> and
2498 <filename>Baz.o</filename>.</para>
2502 <para>It uses a boilerplate pattern rule to compile
2503 <filename>Foo.lhs</filename> to <filename>Foo.o</filename>
2504 using a Haskell compiler. (Which one? That is set in the
2505 build configuration.)</para>
2509 <para>It uses another standard pattern rule to compile
2510 <filename>Baz.c</filename> to <filename>Baz.o</filename>,
2511 using a C compiler. (Ditto.)</para>
2515 <para>It links the resulting <filename>.o</filename> files
2516 together to make <literal>small</literal>, using the Haskell
2517 compiler to do the link step. (Why not use
2518 <command>ld</command>? Because the Haskell compiler knows
2519 what standard libraries to link in. How did
2520 <command>gmake</command> know to use the Haskell compiler to
2521 do the link, rather than the C compiler? Because we set the
2522 variable <constant>HS_PROG</constant> rather than
2523 <constant>C_PROG</constant>.)</para>
2527 <para>All <filename>Makefile</filename>s should follow the above
2528 three-section format.</para>
2532 <title>A larger project</title>
2534 <para>Larger projects are usually structured into a number of
2535 sub-directories, each of which has its own
2536 <filename>Makefile</filename>. (In very large projects, this
2537 sub-structure might be iterated recursively, though that is
2538 rare.) To give you the idea, here's part of the directory
2539 structure for the (rather large) GHC project:</para>
2549 ...source files for documentation...
2552 ...source files for driver...
2555 parser/...source files for parser...
2556 renamer/...source files for renamer...
2560 <para>The sub-directories <filename>docs</filename>,
2561 <filename>driver</filename>, <filename>compiler</filename>, and
2562 so on, each contains a sub-component of GHC, and each has its
2563 own <filename>Makefile</filename>. There must also be a
2564 <filename>Makefile</filename> in
2565 <filename><constant>$(FPTOOLS_TOP)</constant>/ghc</filename>.
2566 It does most of its work by recursively invoking
2567 <command>gmake</command> on the <filename>Makefile</filename>s
2568 in the sub-directories. We say that
2569 <filename>ghc/Makefile</filename> is a <emphasis>non-leaf
2570 <filename>Makefile</filename></emphasis>, because it does little
2571 except organise its children, while the
2572 <filename>Makefile</filename>s in the sub-directories are all
2573 <emphasis>leaf <filename>Makefile</filename>s</emphasis>. (In
2574 principle the sub-directories might themselves contain a
2575 non-leaf <filename>Makefile</filename> and several
2576 sub-sub-directories, but that does not happen in GHC.)</para>
2578 <para>The <filename>Makefile</filename> in
2579 <filename>ghc/compiler</filename> is considered a leaf
2580 <filename>Makefile</filename> even though the
2581 <filename>ghc/compiler</filename> has sub-directories, because
2582 these sub-directories do not themselves have
2583 <filename>Makefile</filename>s in them. They are just used to
2584 structure the collection of modules that make up GHC, but all
2585 are managed by the single <filename>Makefile</filename> in
2586 <filename>ghc/compiler</filename>.</para>
2588 <para>You will notice that <filename>ghc/</filename> also
2589 contains a directory <filename>ghc/mk/</filename>. It contains
2590 GHC-specific <filename>Makefile</filename> boilerplate code.
2591 More precisely:</para>
2595 <para><filename>ghc/mk/boilerplate.mk</filename> is included
2596 at the top of <filename>ghc/Makefile</filename>, and of all
2597 the leaf <filename>Makefile</filename>s in the
2598 sub-directories. It in turn <literal>include</literal>s the
2599 main boilerplate file
2600 <filename>mk/boilerplate.mk</filename>.</para>
2604 <para><filename>ghc/mk/target.mk</filename> is
2605 <literal>include</literal>d at the bottom of
2606 <filename>ghc/Makefile</filename>, and of all the leaf
2607 <filename>Makefile</filename>s in the sub-directories. It
2608 in turn <literal>include</literal>s the file
2609 <filename>mk/target.mk</filename>.</para>
2613 <para>So these two files are the place to look for GHC-wide
2614 customisation of the standard boilerplate.</para>
2617 <sect2 id="sec-boiler-arch">
2618 <title>Boilerplate architecture</title>
2619 <indexterm><primary>boilerplate architecture</primary></indexterm>
2621 <para>Every <filename>Makefile</filename> includes a
2622 <filename>boilerplate.mk</filename><indexterm><primary>boilerplate.mk</primary></indexterm>
2623 file at the top, and
2624 <filename>target.mk</filename><indexterm><primary>target.mk</primary></indexterm>
2625 file at the bottom. In this section we discuss what is in these
2626 files, and why there have to be two of them. In general:</para>
2630 <para><filename>boilerplate.mk</filename> consists of:</para>
2634 <para><emphasis>Definitions of millions of
2635 <command>make</command> variables</emphasis> that
2636 collectively specify the build configuration. Examples:
2637 <constant>HC_OPTS</constant><indexterm><primary>HC_OPTS</primary></indexterm>,
2638 the options to feed to the Haskell compiler;
2639 <constant>NoFibSubDirs</constant><indexterm><primary>NoFibSubDirs</primary></indexterm>,
2640 the sub-directories to enable within the
2641 <literal>nofib</literal> project;
2642 <constant>GhcWithHc</constant><indexterm><primary>GhcWithHc</primary></indexterm>,
2643 the name of the Haskell compiler to use when compiling
2644 GHC in the <literal>ghc</literal> project.</para>
2648 <para><emphasis>Standard pattern rules</emphasis> that
2649 tell <command>gmake</command> how to construct one file
2650 from another.</para>
2654 <para><filename>boilerplate.mk</filename> needs to be
2655 <literal>include</literal>d at the <emphasis>top</emphasis>
2656 of each <filename>Makefile</filename>, so that the user can
2657 replace the boilerplate definitions or pattern rules by
2658 simply giving a new definition or pattern rule in the
2659 <filename>Makefile</filename>. <command>gmake</command>
2660 simply takes the last definition as the definitive one.</para>
2662 <para>Instead of <emphasis>replacing</emphasis> boilerplate
2663 definitions, it is also quite common to
2664 <emphasis>augment</emphasis> them. For example, a
2665 <filename>Makefile</filename> might say:</para>
2671 <para>thereby adding “<option>-O</option>” to
2673 <constant>SRC_HC_OPTS</constant><indexterm><primary>SRC_HC_OPTS</primary></indexterm>.</para>
2677 <para><filename>target.mk</filename> contains
2678 <command>make</command> rules for the standard targets
2679 described in <Xref LinkEnd="sec-standard-targets">. These
2680 rules are selectively included, depending on the setting of
2681 certain <command>make</command> variables. These variables
2682 are usually set in the middle section of the
2683 <filename>Makefile</filename> between the two
2684 <literal>include</literal>s.</para>
2686 <para><filename>target.mk</filename> must be included at the
2687 end (rather than being part of
2688 <filename>boilerplate.mk</filename>) for several tiresome
2694 <para><command>gmake</command> commits target and
2695 dependency lists earlier than it should. For example,
2696 <FIlename>target.mk</FIlename> has a rule that looks
2700 $(HS_PROG) : $(OBJS)
2701 $(HC) $(LD_OPTS) $< -o $@
2704 <para>If this rule was in
2705 <filename>boilerplate.mk</filename> then
2706 <constant>$(HS_PROG)</constant><indexterm><primary>HS_PROG</primary></indexterm>
2708 <constant>$(OBJS)</constant><indexterm><primary>OBJS</primary></indexterm>
2709 would not have their final values at the moment
2710 <command>gmake</command> encountered the rule. Alas,
2711 <command>gmake</command> takes a snapshot of their
2712 current values, and wires that snapshot into the rule.
2713 (In contrast, the commands executed when the rule
2714 “fires” are only substituted at the moment
2715 of firing.) So, the rule must follow the definitions
2716 given in the <filename>Makefile</filename> itself.</para>
2720 <para>Unlike pattern rules, ordinary rules cannot be
2721 overriden or replaced by subsequent rules for the same
2722 target (at least, not without an error message).
2723 Including ordinary rules in
2724 <filename>boilerplate.mk</filename> would prevent the
2725 user from writing rules for specific targets in specific
2730 <para>There are a couple of other reasons I've
2731 forgotten, but it doesn't matter too much.</para>
2738 <sect2 id="sec-boiler">
2739 <title>The main <filename>mk/boilerplate.mk</filename> file</title>
2740 <indexterm><primary>boilerplate.mk</primary></indexterm>
2742 <para>If you look at
2743 <filename><constant>$(FPTOOLS_TOP)</constant>/mk/boilerplate.mk</filename>
2744 you will find that it consists of the following sections, each
2745 held in a separate file:</para>
2749 <term><filename>config.mk</filename></term>
2750 <indexterm><primary>config.mk</primary></indexterm>
2752 <para>is the build configuration file we discussed at
2753 length in <Xref LinkEnd="sec-build-config">.</para>
2758 <term><filename>paths.mk</filename></term>
2759 <indexterm><primary>paths.mk</primary></indexterm>
2761 <para>defines <command>make</command> variables for
2762 pathnames and file lists. This file contains code for
2763 automatically compiling lists of source files and deriving
2764 lists of object files from those. The results can be
2765 overriden in the <filename>Makefile</filename>, but in
2766 most cases the automatic setup should do the right
2769 <para>The following variables may be set in the
2770 <filename>Makefile</filename> to affect how the automatic
2771 source file search is done:</para>
2775 <term><literal>ALL_DIRS</literal></term>
2776 <indexterm><primary><literal>ALL_DIRS</literal></primary>
2779 <para>Set to a list of directories to search in
2780 addition to the current directory for source
2786 <term><literal>EXCLUDE_SRCS</literal></term>
2787 <indexterm><primary><literal>EXCLUDE_SRCS</literal></primary>
2790 <para>Set to a list of source files (relative to the
2791 current directory) to omit from the automatic
2792 search. The source searching machinery is clever
2793 enough to know that if you exclude a source file
2794 from which other sources are derived, then the
2795 derived sources should also be excluded. For
2796 example, if you set <literal>EXCLUDED_SRCS</literal>
2797 to include <filename>Foo.y</filename>, then
2798 <filename>Foo.hs</filename> will also be
2804 <term><literal>EXTRA_SRCS</literal></term>
2805 <indexterm><primary><literal>EXCLUDE_SRCS</literal></primary>
2808 <para>Set to a list of extra source files (perhaps
2809 in directories not listed in
2810 <literal>ALL_DIRS</literal>) that should be
2816 <para>The results of the automatic source file search are
2817 placed in the following make variables:</para>
2821 <term><literal>SRCS</literal></term>
2822 <indexterm><primary><literal>SRCS</literal></primary></indexterm>
2824 <para>All source files found, sorted and without
2825 duplicates, including those which might not exist
2826 yet but will be derived from other existing sources.
2827 <literal>SRCS</literal> <emphasis>can</emphasis> be
2828 overriden if necessary, in which case the variables
2829 below will follow suit.</para>
2834 <term><literal>HS_SRCS</literal></term>
2835 <indexterm><primary><literal>HS_SRCS</literal></primary></indexterm>
2837 <para>all Haskell source files in the current
2838 directory, including those derived from other source
2839 files (eg. Happy sources also give rise to Haskell
2845 <term><literal>HS_OBJS</literal></term>
2846 <indexterm><primary><literal>HS_OBJS</literal></primary></indexterm>
2848 <para>Object files derived from
2849 <literal>HS_SRCS</literal>.</para>
2854 <term><literal>HS_IFACES</literal></term>
2855 <indexterm><primary><literal>HS_IFACES</literal></primary></indexterm>
2857 <para>Interface files (<literal>.hi</literal> files)
2858 derived from <literal>HS_SRCS</literal>.</para>
2863 <term><literal>C_SRCS</literal></term>
2864 <indexterm><primary><literal>C_SRCS</literal></primary></indexterm>
2866 <para>All C source files found.</para>
2871 <term><literal>C_OBJS</literal></term>
2872 <indexterm><primary><literal>C_OBJS</literal></primary></indexterm>
2874 <para>Object files derived from
2875 <literal>C_SRCS</literal>.</para>
2880 <term><literal>SCRIPT_SRCS</literal></term>
2881 <indexterm><primary><literal>SCRIPT_SRCS</literal></primary></indexterm>
2883 <para>All script source files found
2884 (<literal>.lprl</literal> files).</para>
2889 <term><literal>SCRIPT_OBJS</literal></term>
2890 <indexterm><primary><literal>SCRIPT_OBJS</literal></primary></indexterm>
2892 <para><quote>object</quote> files derived from
2893 <literal>SCRIPT_SRCS</literal>
2894 (<literal>.prl</literal> files).</para>
2899 <term><literal>HSC_SRCS</literal></term>
2900 <indexterm><primary><literal>HSC_SRCS</literal></primary></indexterm>
2902 <para>All <literal>hsc2hs</literal> source files
2903 (<literal>.hsc</literal> files).</para>
2908 <term><literal>HAPPY_SRCS</literal></term>
2909 <indexterm><primary><literal>HAPPY_SRCS</literal></primary></indexterm>
2911 <para>All <literal>happy</literal> source files
2912 (<literal>.y</literal> or <literal>.hy</literal> files).</para>
2917 <term><literal>OBJS</literal></term>
2918 <indexterm><primary>OBJS</primary></indexterm>
2920 <para>the concatenation of
2921 <literal>$(HS_OBJS)</literal>,
2922 <literal>$(C_OBJS)</literal>, and
2923 <literal>$(SCRIPT_OBJS)</literal>.</para>
2928 <para>Any or all of these definitions can easily be
2929 overriden by giving new definitions in your
2930 <filename>Makefile</filename>.</para>
2932 <para>What, exactly, does <filename>paths.mk</filename>
2933 consider a <quote>source file</quote> to be? It's based
2934 on the file's suffix (e.g. <filename>.hs</filename>,
2935 <filename>.lhs</filename>, <filename>.c</filename>,
2936 <filename>.hy</filename>, etc), but this is the kind of
2937 detail that changes, so rather than enumerate the source
2938 suffices here the best thing to do is to look in
2939 <filename>paths.mk</filename>.</para>
2944 <term><filename>opts.mk</filename></term>
2945 <indexterm><primary>opts.mk</primary></indexterm>
2947 <para>defines <command>make</command> variables for option
2948 strings to pass to each program. For example, it defines
2949 <constant>HC_OPTS</constant><indexterm><primary>HC_OPTS</primary></indexterm>,
2950 the option strings to pass to the Haskell compiler. See
2951 <Xref LinkEnd="sec-suffix">.</para>
2956 <term><filename>suffix.mk</filename></term>
2957 <indexterm><primary>suffix.mk</primary></indexterm>
2959 <para>defines standard pattern rules—see <Xref
2960 LinkEnd="sec-suffix">.</para>
2965 <para>Any of the variables and pattern rules defined by the
2966 boilerplate file can easily be overridden in any particular
2967 <filename>Makefile</filename>, because the boilerplate
2968 <literal>include</literal> comes first. Definitions after this
2969 <literal>include</literal> directive simply override the default
2970 ones in <filename>boilerplate.mk</filename>.</para>
2973 <sect2 id="sec-suffix">
2974 <title>Pattern rules and options</title>
2975 <indexterm><primary>Pattern rules</primary></indexterm>
2978 <filename>suffix.mk</filename><indexterm><primary>suffix.mk</primary></indexterm>
2979 defines standard <emphasis>pattern rules</emphasis> that say how
2980 to build one kind of file from another, for example, how to
2981 build a <filename>.o</filename> file from a
2982 <filename>.c</filename> file. (GNU <command>make</command>'s
2983 <emphasis>pattern rules</emphasis> are more powerful and easier
2984 to use than Unix <command>make</command>'s <emphasis>suffix
2985 rules</emphasis>.)</para>
2987 <para>Almost all the rules look something like this:</para>
2992 $(CC) $(CC_OPTS) -c $< -o $@
2995 <para>Here's how to understand the rule. It says that
2996 <emphasis>something</emphasis><filename>.o</filename> (say
2997 <filename>Foo.o</filename>) can be built from
2998 <emphasis>something</emphasis><filename>.c</filename>
2999 (<filename>Foo.c</filename>), by invoking the C compiler (path
3000 name held in <constant>$(CC)</constant>), passing to it
3001 the options <constant>$(CC_OPTS)</constant> and
3002 the rule's dependent file of the rule
3003 <literal>$<</literal> (<filename>Foo.c</filename> in
3004 this case), and putting the result in the rule's target
3005 <literal>$@</literal> (<filename>Foo.o</filename> in this
3008 <para>Every program is held in a <command>make</command>
3009 variable defined in <filename>mk/config.mk</filename>—look
3010 in <filename>mk/config.mk</filename> for the complete list. One
3011 important one is the Haskell compiler, which is called
3012 <constant>$(HC)</constant>.</para>
3014 <para>Every program's options are are held in a
3015 <command>make</command> variables called
3016 <constant><prog>_OPTS</constant>. the
3017 <constant><prog>_OPTS</constant> variables are
3018 defined in <filename>mk/opts.mk</filename>. Almost all of them
3019 are defined like this:</para>
3022 CC_OPTS = $(SRC_CC_OPTS) $(WAY$(_way)_CC_OPTS) $($*_CC_OPTS) $(EXTRA_CC_OPTS)
3025 <para>The four variables from which
3026 <constant>CC_OPTS</constant> is built have the following
3031 <term><constant>SRC_CC_OPTS</constant><indexterm><primary>SRC_CC_OPTS</primary></indexterm>:</term>
3033 <para>options passed to all C compilations.</para>
3038 <term><constant>WAY_<way>_CC_OPTS</constant>:</term>
3040 <para>options passed to C compilations for way
3041 <literal><way></literal>. For example,
3042 <constant>WAY_mp_CC_OPTS</constant>
3043 gives options to pass to the C compiler when compiling way
3044 <literal>mp</literal>. The variable
3045 <constant>WAY_CC_OPTS</constant> holds
3046 options to pass to the C compiler when compiling the
3047 standard way. (<Xref LinkEnd="sec-ways"> dicusses
3048 multi-way compilation.)</para>
3053 <term><constant><module>_CC_OPTS</constant>:</term>
3055 <para>options to pass to the C compiler that are specific
3056 to module <literal><module></literal>. For example,
3057 <constant>SMap_CC_OPTS</constant> gives the
3058 specific options to pass to the C compiler when compiling
3059 <filename>SMap.c</filename>.</para>
3064 <term><constant>EXTRA_CC_OPTS</constant><indexterm><primary>EXTRA_CC_OPTS</primary></indexterm>:</term>
3066 <para>extra options to pass to all C compilations. This
3067 is intended for command line use, thus:</para>
3070 gmake libHS.a EXTRA_CC_OPTS="-v"
3077 <sect2 id="sec-targets">
3078 <title>The main <filename>mk/target.mk</filename> file</title>
3079 <indexterm><primary>target.mk</primary></indexterm>
3081 <para><filename>target.mk</filename> contains canned rules for
3082 all the standard targets described in <Xref
3083 LinkEnd="sec-standard-targets">. It is complicated by the fact
3084 that you don't want all of these rules to be active in every
3085 <filename>Makefile</filename>. Rather than have a plethora of
3086 tiny files which you can include selectively, there is a single
3087 file, <filename>target.mk</filename>, which selectively includes
3088 rules based on whether you have defined certain variables in
3089 your <filename>Makefile</filename>. This section explains what
3090 rules you get, what variables control them, and what the rules
3091 do. Hopefully, you will also get enough of an idea of what is
3092 supposed to happen that you can read and understand any weird
3093 special cases yourself.</para>
3097 <term><constant>HS_PROG</constant><indexterm><primary>HS_PROG</primary></indexterm>.</term>
3099 <para>If <constant>HS_PROG</constant> is defined,
3100 you get rules with the following targets:</para>
3104 <term><filename>HS_PROG</filename><indexterm><primary>HS_PROG</primary></indexterm></term>
3106 <para>itself. This rule links
3107 <constant>$(OBJS)</constant> with the Haskell
3108 runtime system to get an executable called
3109 <constant>$(HS_PROG)</constant>.</para>
3114 <term><literal>install</literal><indexterm><primary>install</primary></indexterm></term>
3117 <constant>$(HS_PROG)</constant> in
3118 <constant>$(bindir)</constant>.</para>
3127 <term><constant>C_PROG</constant><indexterm><primary>C_PROG</primary></indexterm></term>
3129 <para>is similar to <constant>HS_PROG</constant>,
3130 except that the link step links
3131 <constant>$(C_OBJS)</constant> with the C
3132 runtime system.</para>
3137 <term><constant>LIBRARY</constant><indexterm><primary>LIBRARY</primary></indexterm></term>
3139 <para>is similar to <constant>HS_PROG</constant>,
3140 except that it links
3141 <constant>$(LIB_OBJS)</constant> to make the
3142 library archive <constant>$(LIBRARY)</constant>,
3143 and <literal>install</literal> installs it in
3144 <constant>$(libdir)</constant>.</para>
3149 <term><constant>LIB_DATA</constant><indexterm><primary>LIB_DATA</primary></indexterm></term>
3151 <para>…</para>
3156 <term><constant>LIB_EXEC</constant><indexterm><primary>LIB_EXEC</primary></indexterm></term>
3158 <para>…</para>
3163 <term><constant>HS_SRCS</constant><indexterm><primary>HS_SRCS</primary></indexterm>, <constant>C_SRCS</constant><indexterm><primary>C_SRCS</primary></indexterm>.</term>
3165 <para>If <constant>HS_SRCS</constant> is defined
3166 and non-empty, a rule for the target
3167 <literal>depend</literal> is included, which generates
3168 dependency information for Haskell programs. Similarly
3169 for <constant>C_SRCS</constant>.</para>
3174 <para>All of these rules are “double-colon” rules,
3178 install :: $(HS_PROG)
3179 ...how to install it...
3182 <para>GNU <command>make</command> treats double-colon rules as
3183 separate entities. If there are several double-colon rules for
3184 the same target it takes each in turn and fires it if its
3185 dependencies say to do so. This means that you can, for
3186 example, define both <constant>HS_PROG</constant> and
3187 <constant>LIBRARY</constant>, which will generate two rules for
3188 <literal>install</literal>. When you type <command>gmake
3189 install</command> both rules will be fired, and both the program
3190 and the library will be installed, just as you wanted.</para>
3193 <sect2 id="sec-subdirs">
3194 <title>Recursion</title>
3195 <indexterm><primary>recursion, in makefiles</primary></indexterm>
3196 <indexterm><primary>Makefile, recursing into subdirectories</primary></indexterm>
3198 <para>In leaf <filename>Makefile</filename>s the variable
3199 <constant>SUBDIRS</constant><indexterm><primary>SUBDIRS</primary></indexterm>
3200 is undefined. In non-leaf <filename>Makefile</filename>s,
3201 <constant>SUBDIRS</constant> is set to the list of
3202 sub-directories that contain subordinate
3203 <filename>Makefile</filename>s. <emphasis>It is up to you to
3204 set <constant>SUBDIRS</constant> in the
3205 <filename>Makefile</filename>.</emphasis> There is no automation
3206 here—<constant>SUBDIRS</constant> is too important to
3209 <para>When <constant>SUBDIRS</constant> is defined,
3210 <filename>target.mk</filename> includes a rather neat rule for
3211 the standard targets (<Xref LinkEnd="sec-standard-targets"> that
3212 simply invokes <command>make</command> recursively in each of
3213 the sub-directories.</para>
3215 <para><emphasis>These recursive invocations are guaranteed to
3216 occur in the order in which the list of directories is specified
3217 in <constant>SUBDIRS</constant>. </emphasis>This guarantee can
3218 be important. For example, when you say <command>gmake
3219 boot</command> it can be important that the recursive invocation
3220 of <command>make boot</command> is done in one sub-directory
3221 (the include files, say) before another (the source files).
3222 Generally, put the most independent sub-directory first, and the
3223 most dependent last.</para>
3226 <sect2 id="sec-ways">
3227 <title>Way management</title>
3228 <indexterm><primary>way management</primary></indexterm>
3230 <para>We sometimes want to build essentially the same system in
3231 several different “ways”. For example, we want to build GHC's
3232 <literal>Prelude</literal> libraries with and without profiling,
3233 so that there is an appropriately-built library archive to link
3234 with when the user compiles his program. It would be possible
3235 to have a completely separate build tree for each such “way”,
3236 but it would be horribly bureaucratic, especially since often
3237 only parts of the build tree need to be constructed in multiple
3241 <filename>target.mk</filename><indexterm><primary>target.mk</primary></indexterm>
3242 contains some clever magic to allow you to build several
3243 versions of a system; and to control locally how many versions
3244 are built and how they differ. This section explains the
3247 <para>The files for a particular way are distinguished by
3248 munging the suffix. The <quote>normal way</quote> is always
3249 built, and its files have the standard suffices
3250 <filename>.o</filename>, <filename>.hi</filename>, and so on.
3251 In addition, you can build one or more extra ways, each
3252 distinguished by a <emphasis>way tag</emphasis>. The object
3253 files and interface files for one of these extra ways are
3254 distinguished by their suffix. For example, way
3255 <literal>mp</literal> has files
3256 <filename>.mp_o</filename> and
3257 <filename>.mp_hi</filename>. Library archives have their
3258 way tag the other side of the dot, for boring reasons; thus,
3259 <filename>libHS_mp.a</filename>.</para>
3261 <para>A <command>make</command> variable called
3262 <constant>way</constant> holds the current way tag.
3263 <emphasis><constant>way</constant> is only ever set on the
3264 command line of <command>gmake</command></emphasis> (usually in
3265 a recursive invocation of <command>gmake</command> by the
3266 system). It is never set inside a
3267 <filename>Makefile</filename>. So it is a global constant for
3268 any one invocation of <command>gmake</command>. Two other
3269 <command>make</command> variables,
3270 <constant>way_</constant> and
3271 <constant>_way</constant> are immediately derived from
3272 <constant>$(way)</constant> and never altered. If
3273 <constant>way</constant> is not set, then neither are
3274 <constant>way_</constant> and
3275 <constant>_way</constant>, and the invocation of
3276 <command>make</command> will build the <quote>normal
3277 way</quote>. If <constant>way</constant> is set, then the other
3278 two variables are set in sympathy. For example, if
3279 <constant>$(way)</constant> is “<literal>mp</literal>”,
3280 then <constant>way_</constant> is set to
3281 “<literal>mp_</literal>” and
3282 <constant>_way</constant> is set to
3283 “<literal>_mp</literal>”. These three variables are
3284 then used when constructing file names.</para>
3286 <para>So how does <command>make</command> ever get recursively
3287 invoked with <constant>way</constant> set? There are two ways
3288 in which this happens:</para>
3292 <para>For some (but not all) of the standard targets, when
3293 in a leaf sub-directory, <command>make</command> is
3294 recursively invoked for each way tag in
3295 <constant>$(WAYS)</constant>. You set
3296 <constant>WAYS</constant> in the
3297 <filename>Makefile</filename> to the list of way tags you
3298 want these targets built for. The mechanism here is very
3299 much like the recursive invocation of
3300 <command>make</command> in sub-directories (<Xref
3301 LinkEnd="sec-subdirs">). It is up to you to set
3302 <constant>WAYS</constant> in your
3303 <filename>Makefile</filename>; this is how you control what
3304 ways will get built.</para>
3308 <para>For a useful collection of targets (such as
3309 <filename>libHS_mp.a</filename>,
3310 <filename>Foo.mp_o</filename>) there is a rule which
3311 recursively invokes <command>make</command> to make the
3312 specified target, setting the <constant>way</constant>
3313 variable. So if you say <command>gmake
3314 Foo.mp_o</command> you should see a recursive
3315 invocation <command>gmake Foo.mp_o way=mp</command>,
3316 and <emphasis>in this recursive invocation the pattern rule
3317 for compiling a Haskell file into a <filename>.o</filename>
3318 file will match</emphasis>. The key pattern rules (in
3319 <filename>suffix.mk</filename>) look like this:
3323 $(HC) $(HC_OPTS) $< -o $@
3330 <para>You can invoke <command>make</command> with a
3331 particular <literal>way</literal> setting yourself, in order
3332 to build files related to a particular
3333 <literal>way</literal> in the current directory. eg.
3339 will build files for the profiling way only in the current
3346 <title>When the canned rule isn't right</title>
3348 <para>Sometimes the canned rule just doesn't do the right thing.
3349 For example, in the <literal>nofib</literal> suite we want the
3350 link step to print out timing information. The thing to do here
3351 is <emphasis>not</emphasis> to define
3352 <constant>HS_PROG</constant> or
3353 <constant>C_PROG</constant>, and instead define a special
3354 purpose rule in your own <filename>Makefile</filename>. By
3355 using different variable names you will avoid the canned rules
3356 being included, and conflicting with yours.</para>
3360 <sect1 id="building-docs">
3361 <title>Building the documentation</title>
3363 <sect2 id="pre-supposed-doc-tools">
3364 <title>Tools for building the Documentation</title>
3366 <para>The following additional tools are required if you want to
3367 format the documentation that comes with the
3368 <literal>fptools</literal> projects:</para>
3372 <term>DocBook</term>
3373 <indexterm><primary>pre-supposed: DocBook</primary></indexterm>
3374 <indexterm><primary>DocBook, pre-supposed</primary></indexterm>
3376 <para>Much of our documentation is written in SGML, using
3377 the DocBook DTD. Instructions on installing and
3378 configuring the DocBook tools are below.</para>
3384 <indexterm><primary>pre-supposed: TeX</primary></indexterm>
3385 <indexterm><primary>TeX, pre-supposed</primary></indexterm>
3387 <para>A decent TeX distribution is required if you want to
3388 produce printable documentation. We recomment teTeX,
3389 which includes just about everything you need.</para>
3394 <term>Haddock</term>
3395 <indexterm><primary>Haddock</primary>
3398 <para>Haddock is a Haskell documentation tool that we use
3399 for automatically generating documentation from the
3400 library source code. It is an <literal>fptools</literal>
3401 project in itself. To build documentation for the
3402 libraries (<literal>fptools/libraries</literal>) you
3403 should check out and build Haddock in
3404 <literal>fptools/haddock</literal>. Haddock requires GHC
3412 <title>Installing the DocBook tools</title>
3415 <title>Installing the DocBook tools on Linux</title>
3417 <para>If you're on a recent RedHat system (7.0+), you probably
3418 have working DocBook tools already installed. The configure
3419 script should detect your setup and you're away.</para>
3421 <para>If you don't have DocBook tools installed, and you are
3422 using a system that can handle RedHat RPM packages, you can
3423 probably use the <ULink
3424 URL="http://sourceware.cygnus.com/docbook-tools/">Cygnus
3425 DocBook tools</ULink>, which is the most shrink-wrapped SGML
3426 suite that we could find. You need all the RPMs except for
3427 psgml (i.e. <Filename>docbook</Filename>,
3428 <Filename>jade</Filename>, <Filename>jadetex</Filename>,
3429 <Filename>sgmlcommon</Filename> and
3430 <Filename>stylesheets</Filename>). Note that most of these
3431 RPMs are architecture neutral, so are likely to be found in a
3432 <Filename>noarch</Filename> directory. The SuSE RPMs also
3433 work; the RedHat ones <Emphasis>don't</Emphasis> in RedHat 6.2
3434 (7.0 and later should be OK), but they are easy to fix: just
3436 <Filename>/usr/lib/sgml/stylesheets/nwalsh-modular/lib/dblib.dsl</Filename>
3437 to <Filename>/usr/lib/sgml/lib/dblib.dsl</Filename>. </para>
3441 <title>Installing DocBook on FreeBSD</title>
3443 <para>On FreeBSD systems, the easiest way to get DocBook up
3444 and running is to install it from the ports tree or a
3445 pre-compiled package (packages are available from your local
3446 FreeBSD mirror site).</para>
3448 <para>To use the ports tree, do this:
3450 $ cd /usr/ports/textproc/docproj
3453 This installs the FreeBSD documentation project tools, which
3454 includes everything needed to format the GHC
3455 documentation.</para>
3459 <title>Installing from binaries on Windows</title>
3461 <Para>It's a good idea to use Norman Walsh's <ULink
3462 URL="http://nwalsh.com/docbook/dsssl/doc/install.html">installation
3463 notes</ULink> as a guide. You should get version 3.1 of
3464 DocBook, and note that his file <Filename>test.sgm</Filename>
3465 won't work, as it needs version 3.0. You should unpack Jade
3466 into <Filename>\Jade</Filename>, along with the entities,
3467 DocBook into <Filename>\docbook</Filename>, and the DocBook
3468 stylesheets into <Filename>\docbook\stylesheets</Filename> (so
3469 they actually end up in
3470 <Filename>\docbook\stylesheets\docbook</Filename>).</para>
3475 <title>Installing the DocBook tools from source</title>
3480 <para>Install <ULink
3481 URL="http://openjade.sourceforge.net/">OpenJade</ULink>
3482 (Windows binaries are available as well as sources). If you
3483 want DVI, PS, or PDF then install JadeTeX from the
3484 <Filename>dsssl</Filename> subdirectory. (If you get the
3488 ! LaTeX Error: Unknown option implicit=false' for package hyperref'.
3491 your version of <Command>hyperref</Command> is out of date;
3492 download it from CTAN
3493 (<Filename>macros/latex/contrib/supported/hyperref</Filename>),
3494 and make it, ensuring that you have first removed or renamed
3495 your old copy. If you start getting file not found errors
3496 when making the test for <Command>hyperref</Command>, you
3497 can abort at that point and proceed straight to
3498 <Command>make install</Command>, or enter them as
3499 <Filename>../</Filename><Emphasis>filename</Emphasis>.)</para>
3501 <para>Make links from <Filename>virtex</Filename> to
3502 <Filename>jadetex</Filename> and
3503 <Filename>pdfvirtex</Filename> to
3504 <Filename>pdfjadetex</Filename> (otherwise DVI, PostScript
3505 and PDF output will not work). Copy
3506 <Filename>dsssl/*.{dtd,dsl}</Filename> and
3507 <Filename>catalog</Filename> to
3508 <Filename>/usr/[local/]lib/sgml</Filename>.</para>
3512 <title>DocBook and the DocBook stylesheets</title>
3514 <para>Get a Zip of <ULink
3515 URL="http://www.oasis-open.org/docbook/sgml/3.1/index.html">DocBook</ULink>
3516 and install the contents in
3517 <Filename>/usr/[local/]/lib/sgml</Filename>.</para>
3519 <para>Get the <ULink
3520 URL="http://nwalsh.com/docbook/dsssl/">DocBook
3521 stylesheets</ULink> and install in
3522 <Filename>/usr/[local/]lib/sgml/stylesheets</Filename>
3523 (thereby creating a subdirectory docbook). For indexing,
3524 copy or link <Filename>collateindex.pl</Filename> from the
3525 DocBook stylesheets archive in <Filename>bin</Filename> into
3526 a directory on your <Constant>PATH</Constant>.</para>
3528 <para>Download the <ULink
3529 URL="http://www.oasis-open.org/cover/ISOEnts.zip">ISO
3530 entities</ULink> into
3531 <Filename>/usr/[local/]lib/sgml</Filename>.</para>
3537 <title>Configuring the DocBook tools</title>
3539 <Para>Once the DocBook tools are installed, the configure script
3540 will detect them and set up the build system accordingly. If you
3541 have a system that isn't supported, let us know, and we'll try
3546 <title>Remaining problems</title>
3548 <para>If you install from source, you'll get a pile of warnings
3551 <Screen>DTDDECL catalog entries are not supported</Screen>
3553 every time you build anything. These can safely be ignored, but
3554 if you find them tedious you can get rid of them by removing all
3555 the <Constant>DTDDECL</Constant> entries from
3556 <Filename>docbook.cat</Filename>.</para>
3560 <title>Building the documentation</title>
3562 <para>To build documentation in a certain format, you can
3563 say, for example,</para>
3569 <para>to build HTML documentation below the current directory.
3570 The available formats are: <literal>dvi</literal>,
3571 <literal>ps</literal>, <literal>pdf</literal>,
3572 <literal>html</literal>, and <literal>rtf</literal>. Note that
3573 not all documentation can be built in all of these formats: HTML
3574 documentation is generally supported everywhere, and DocBook
3575 documentation might support the other formats (depending on what
3576 other tools you have installed).</para>
3578 <para>All of these targets are recursive; that is, saying
3579 <literal>make html</literal> will make HTML docs for all the
3580 documents recursively below the current directory.</para>
3582 <para>Because there are many different formats that the DocBook
3583 documentation can be generated in, you have to select which ones
3584 you want by setting the <literal>SGMLDocWays</literal> variable
3585 to a list of them. For example, in
3586 <filename>build.mk</filename> you might have a line:</para>
3589 SGMLDocWays = html ps
3592 <para>This will cause the documentation to be built in the requested
3593 formats as part of the main build (the default is not to build
3594 any documentation at all).</para>
3598 <title>Installing the documentation</title>
3600 <para>To install the documentation, use:</para>
3606 <para>This will install the documentation into
3607 <literal>$(datadir)</literal> (which defaults to
3608 <literal>$(prefix)/share</literal>). The exception is HTML
3609 documentation, which goes into
3610 <literal>$(datadir)/html</literal>, to keep things tidy.</para>
3612 <para>Note that unless you set <literal>$(SGMLDocWays)</literal>
3613 to a list of formats, the <literal>install-docs</literal> target
3614 won't do anything for SGML documentation.</para>
3620 <sect1 id="sec-porting-ghc">
3621 <title>Porting GHC</title>
3623 <para>This section describes how to port GHC to a currenly
3624 unsupported platform. There are two distinct
3625 possibilities:</para>
3629 <para>The hardware architecture for your system is already
3630 supported by GHC, but you're running an OS that isn't
3631 supported (or perhaps has been supported in the past, but
3632 currently isn't). This is the easiest type of porting job,
3633 but it still requires some careful bootstrapping. Proceed to
3634 <xref linkend="sec-booting-from-hc">.</para>
3638 <para>Your system's hardware architecture isn't supported by
3639 GHC. This will be a more difficult port (though by comparison
3640 perhaps not as difficult as porting gcc). Proceed to <xref
3641 linkend="unregisterised-porting">.</para>
3645 <sect2 id="sec-booting-from-hc">
3646 <title>Booting/porting from C (<filename>.hc</filename>) files</title>
3648 <indexterm><primary>building GHC from .hc files</primary></indexterm>
3649 <indexterm><primary>booting GHC from .hc files</primary></indexterm>
3650 <indexterm><primary>porting GHC</primary></indexterm>
3652 <para>Bootstrapping GHC on a system without GHC already
3653 installed is achieved by taking the intermediate C files (known
3654 as HC files) from a GHC compilation on a supported system to the
3655 target machine, and compiling them using gcc to get a working
3658 <para><emphasis>NOTE: GHC versions 5.xx were hard to bootstrap
3659 from C. We recommend using GHC 6.0.1 or
3660 later.</emphasis></para>
3662 <para>HC files are platform-dependent, so you have to get a set
3663 that were generated on similar hardware. There may be some
3664 supplied on the GHC download page, otherwise you'll have to
3665 compile some up yourself, or start from
3666 <emphasis>unregisterised</emphasis> HC files - see <xref
3667 linkend="unregisterised-porting">.</para>
3669 <para>The following steps should result in a working GHC build
3670 with full libraries:</para>
3674 <para>Unpack the HC files on top of a fresh source tree
3675 (make sure the source tree version matches the version of
3676 the HC files <emphasis>exactly</emphasis>!). This will
3677 place matching <filename>.hc</filename> files next to the
3678 corresponding Haskell source (<filename>.hs</filename> or
3679 <filename>.lhs</filename>) in the compiler subdirectory
3680 <filename>ghc/compiler</filename> and in the libraries
3681 (subdirectories of <filename>hslibs</filename> and
3682 <literal>libraries</literal>).</para>
3686 <para>The actual build process is fully automated by the
3687 <filename>hc-build</filename> script located in the
3688 <filename>distrib</filename> directory. If you eventually
3689 want to install GHC into the directory
3690 <replaceable>dir</replaceable>, the following
3691 command will execute the whole build process (it won't
3692 install yet):</para>
3695 foo% distrib/hc-build --prefix=<replaceable>dir</replaceable>
3697 <indexterm><primary>--hc-build</primary></indexterm>
3699 <para>By default, the installation directory is
3700 <filename>/usr/local</filename>. If that is what you want,
3701 you may omit the argument to <filename>hc-build</filename>.
3702 Generally, any option given to <filename>hc-build</filename>
3703 is passed through to the configuration script
3704 <filename>configure</filename>. If
3705 <filename>hc-build</filename> successfully completes the
3706 build process, you can install the resulting system, as
3716 <sect2 id="unregisterised-porting">
3717 <title>Porting GHC to a new architecture</title>
3719 <para>The first step in porting to a new architecture is to get
3720 an <firstterm>unregisterised</firstterm> build working. An
3721 unregisterised build is one that compiles via vanilla C only.
3722 By contrast, a registerised build uses the following
3723 architecture-specific hacks for speed:</para>
3727 <para>Global register variables: certain abstract machine
3728 <quote>registers</quote> are mapped to real machine
3729 registers, depending on how many machine registers are
3731 <filename>ghc/includes/MachRegs.h</filename>).</para>
3735 <para>Assembly-mangling: when compiling via C, we feed the
3736 assembly generated by gcc though a Perl script known as the
3737 <firstterm>mangler</firstterm> (see
3738 <filename>ghc/driver/mangler/ghc-asm.lprl</filename>). The
3739 mangler rearranges the assembly to support tail-calls and
3740 various other optimisations.</para>
3744 <para>In an unregisterised build, neither of these hacks are
3745 used — the idea is that the C code generated by the
3746 compiler should compile using gcc only. The lack of these
3747 optimisations costs about a factor of two in performance, but
3748 since unregisterised compilation is usually just a step on the
3749 way to a full registerised port, we don't mind too much.</para>
3751 <para>Notes on GHC portability in general: we've tried to stick
3752 to writing portable code in most parts of the system, so it
3753 should compile on any POSIXish system with gcc, but in our
3754 experience most systems differ from the standards in one way or
3755 another. Deal with any problems as they arise - if you get
3756 stuck, ask the experts on
3757 <email>glasgow-haskell-users@haskell.org</email>.</para>
3759 <para>Lots of useful information about the innards of GHC is
3760 available in the <ulink
3761 url="http://www.cse.unsw.edu.au/~chak/haskell/ghc/comm/">GHC
3762 Commentary</ulink>, which might be helpful if you run into some
3763 code which needs tweaking for your system.</para>
3766 <title>Cross-compiling to produce an unregisterised GHC</title>
3768 <para>In this section, we explain how to bootstrap GHC on a
3769 new platform, using unregisterised intermediate C files. We
3770 haven't put a great deal of effort into automating this
3771 process, for two reasons: it is done very rarely, and the
3772 process usually requires human intervention to cope with minor
3773 porting issues anyway.</para>
3775 <para>The following step-by-step instructions should result in
3776 a fully working, albeit unregisterised, GHC. Firstly, you
3777 need a machine that already has a working GHC (we'll call this
3778 the <firstterm>host</firstterm> machine), in order to
3779 cross-compile the intermediate C files that we will use to
3780 bootstrap the compiler on the <firstterm>target</firstterm>
3785 <para>On the target machine:</para>
3789 <para>Unpack a source tree (preferably a released
3790 version). We will call the path to the root of this
3791 tree <replaceable>T</replaceable>.</para>
3796 $ cd <replaceable>T</replaceable>
3797 $ ./configure --enable-hc-boot --enable-hc-boot-unregisterised
3800 <para>You might need to update
3801 <filename>configure.in</filename> to recognise the new
3802 architecture, and re-generate
3803 <filename>configure</filename> with
3804 <literal>autoreconf</literal>.</para>
3809 $ cd <replaceable>T</replaceable>/ghc/includes
3817 <para>On the host machine:</para>
3821 <para>Unpack a source tree (same released version). Call
3822 this directory <replaceable>H</replaceable>.</para>
3827 $ cd <replaceable>H</replaceable>
3834 <filename><replaceable>H</replaceable>/mk/build.mk</filename>,
3835 with the following contents:</para>
3838 GhcUnregisterised = YES
3839 GhcLibHcOpts = -O -H32m -keep-hc-files
3842 GhcWithNativeCodeGen = NO
3843 GhcWithInterpreter = NO
3844 GhcStage1HcOpts = -O -H32m -fasm
3845 GhcStage2HcOpts = -O -fvia-C -keep-hc-files
3851 <filename><replaceable>H</replaceable>/mk/config.mk</filename>:</para>
3854 <para>change <literal>TARGETPLATFORM</literal>
3855 appropriately, and set the variables involving
3856 <literal>TARGET</literal> to the correct values for
3857 the target platform. This step is necessary because
3858 currently <literal>configure</literal> doesn't cope
3859 with specifying different values for the
3860 <literal>--host</literal> and
3861 <literal>--target</literal> flags.</para>
3864 <para>copy <literal>LeadingUnderscore</literal>
3865 setting from target.</para>
3872 <filename><replaceable>T</replaceable>/ghc/includes/config.h</filename>
3874 <filename><replaceable>H</replaceable>/ghc/includes</filename>.
3875 Note that we are building on the host machine, using the
3876 target machine's <literal>config.h</literal> file. This
3877 is so that the intermediate C files generated here will
3878 be suitable for compiling on the target system.</para>
3883 <para>Touch <literal>config.h</literal>, just to make
3884 sure it doesn't get replaced during the build:</para>
3886 $ touch <replaceable>H</replaceable>/ghc/includes/config.h</screen>
3890 <para>Now build the compiler:</para>
3892 $ cd <replaceable>H</replaceable>/glafp-utils && make boot && make
3893 $ cd <replaceable>H</replaceable>/ghc && make boot && make
3895 <para>Don't worry if the build falls over in the RTS, we
3896 don't need the RTS yet.</para>
3901 $ cd <replaceable>H</replaceable>/libraries
3902 $& make boot && make
3908 $ cd <replaceable>H</replaceable>/ghc
3909 $ make boot stage=2 && make stage=2
3915 $ cd <replaceable>H</replaceable>/ghc/utils
3917 $ make -k HC=<replaceable>H</replaceable>/ghc/compiler/stage1/ghc-inplace \
3918 EXTRA_HC_OPTS='-O -fvia-C -keep-hc-files'
3924 $ cd <replaceable>H</replaceable>
3925 $ make hc-file-bundle Project=Ghc
3931 <filename><replaceable>H</replaceable>/*-hc.tar.gz</filename>
3932 to <filename><replaceable>T</replaceable>/..</filename>.</para>
3938 <para>On the target machine:</para>
3940 <para>At this stage we simply need to bootstrap a compiler
3941 from the intermediate C files we generated above. The
3942 process of bootstrapping from C files is automated by the
3943 script in <literal>distrib/hc-build</literal>, and is
3944 described in <xref linkend="sec-booting-from-hc">.</para>
3947 $ ./distrib/hc-build --enable-hc-boot-unregisterised
3950 <para>However, since this is a bootstrap on a new machine,
3951 the automated process might not run to completion the
3952 first time. For that reason, you might want to treat the
3953 <literal>hc-build</literal> script as a list of
3954 instructions to follow, rather than as a fully automated
3955 script. This way you'll be able to restart the process
3956 part-way through if you need to fix anything on the
3959 <para>Don't bother with running
3960 <literal>make install</literal> in the newly
3961 bootstrapped tree; just use the compiler in that tree to
3962 build a fresh compiler from scratch, this time without
3963 booting from C files. Before doing this, you might want
3964 to check that the bootstrapped compiler is generating
3965 working binaries:</para>
3969 main = putStrLn "Hello World!\n"
3971 $ <replaceable>T</replaceable>/ghc/compiler/ghc-inplace hello.hs -o hello
3976 <para>Once you have the unregisterised compiler up and
3977 running, you can use it to start a registerised port. The
3978 following sections describe the various parts of the
3979 system that will need architecture-specific tweaks in
3980 order to get a registerised build going.</para>
3987 <title>Porting the RTS</title>
3989 <para>The following files need architecture-specific code for a
3990 registerised build:</para>
3994 <term><filename>ghc/includes/MachRegs.h</filename></term>
3995 <indexterm><primary><filename>MachRegs.h</filename></primary>
3998 <para>Defines the STG-register to machine-register
3999 mapping. You need to know your platform's C calling
4000 convention, and which registers are generally available
4001 for mapping to global register variables. There are
4002 plenty of useful comments in this file.</para>
4006 <term><filename>ghc/includes/TailCalls.h</filename></term>
4007 <indexterm><primary><filename>TailCalls.h</filename></primary>
4010 <para>Macros that cooperate with the mangler (see <xref
4011 linkend="sec-mangler">) to make proper tail-calls
4016 <term><filename>ghc/rts/Adjustor.c</filename></term>
4017 <indexterm><primary><filename>Adjustor.c</filename></primary>
4021 <literal>foreign import "wrapper"</literal>
4023 <literal>foreign export dynamic</literal>).
4024 Not essential for getting GHC bootstrapped, so this file
4025 can be deferred until later if necessary.</para>
4029 <term><filename>ghc/rts/StgCRun.c</filename></term>
4030 <indexterm><primary><filename>StgCRun.c</filename></primary>
4033 <para>The little assembly layer between the C world and
4034 the Haskell world. See the comments and code for the
4035 other architectures in this file for pointers.</para>
4039 <term><filename>ghc/rts/MBlock.h</filename></term>
4040 <term><filename>ghc/rts/MBlock.c</filename></term>
4041 <indexterm><primary><filename>MBlock.h</filename></primary>
4043 <indexterm><primary><filename>MBlock.c</filename></primary>
4046 <para>These files are really OS-specific rather than
4047 architecture-specific. In <filename>MBlock.h</filename>
4048 is specified the absolute location at which the RTS
4049 should try to allocate memory on your platform (try to
4050 find an area which doesn't conflict with code or dynamic
4051 libraries). In <filename>Mblock.c</filename> you might
4052 need to tweak the call to <literal>mmap()</literal> for
4059 <sect3 id="sec-mangler">
4060 <title>The mangler</title>
4062 <para>The mangler is an evil Perl-script that rearranges the
4063 assembly code output from gcc to do two main things:</para>
4067 <para>Remove function prologues and epilogues, and all
4068 movement of the C stack pointer. This is to support
4069 tail-calls: every code block in Haskell code ends in an
4070 explicit jump, so we don't want the C-stack overflowing
4071 while we're jumping around between code blocks.</para>
4074 <para>Move the <firstterm>info table</firstterm> for a
4075 closure next to the entry code for that closure. In
4076 unregisterised code, info tables contain a pointer to the
4077 entry code, but in registerised compilation we arrange
4078 that the info table is shoved right up against the entry
4079 code, and addressed backwards from the entry code pointer
4080 (this saves a word in the info table and an extra
4081 indirection when jumping to the closure entry
4086 <para>The mangler is abstracted to a certain extent over some
4087 architecture-specific things such as the particular assembler
4088 directives used to herald symbols. Take a look at the
4089 definitions for other architectures and use these as a
4090 starting point.</para>
4094 <title>The native code generator</title>
4096 <para>The native code generator isn't essential to getting a
4097 registerised build going, but it's a desirable thing to have
4098 because it can cut compilation times in half. The native code
4099 generator is described in some detail in the <ulink
4100 url="http://www.cse.unsw.edu.au/~chak/haskell/ghc/comm/">GHC
4101 commentary</ulink>.</para>
4107 <para>To support GHCi, you need to port the dynamic linker
4108 (<filename>fptools/ghc/rts/Linker.c</filename>). The linker
4109 currently supports the ELF and PEi386 object file formats - if
4110 your platform uses one of these then things will be
4111 significantly easier. The majority of Unix platforms use the
4112 ELF format these days. Even so, there are some
4113 machine-specific parts of the ELF linker: for example, the
4114 code for resolving particular relocation types is
4115 machine-specific, so some porting of this code to your
4116 architecture will probaly be necessary.</para>
4118 <para>If your system uses a different object file format, then
4119 you have to write a linker — good luck!</para>
4125 <sect1 id="sec-build-pitfalls">
4126 <title>Known pitfalls in building Glasgow Haskell
4128 <indexterm><primary>problems, building</primary></indexterm>
4129 <indexterm><primary>pitfalls, in building</primary></indexterm>
4130 <indexterm><primary>building pitfalls</primary></indexterm></title>
4133 WARNINGS about pitfalls and known “problems”:
4142 One difficulty that comes up from time to time is running out of space
4143 in <literal>TMPDIR</literal>. (It is impossible for the configuration stuff to
4144 compensate for the vagaries of different sysadmin approaches to temp
4146 <indexterm><primary>tmp, running out of space in</primary></indexterm>
4148 The quickest way around it is <command>setenv TMPDIR /usr/tmp</command><indexterm><primary>TMPDIR</primary></indexterm> or
4149 even <command>setenv TMPDIR .</command> (or the equivalent incantation with your shell
4152 The best way around it is to say
4155 export TMPDIR=<dir>
4158 in your <filename>build.mk</filename> file.
4159 Then GHC and the other <literal>fptools</literal> programs will use the appropriate directory
4168 In compiling some support-code bits, e.g., in <filename>ghc/rts/gmp</filename> and even
4169 in <filename>ghc/lib</filename>, you may get a few C-compiler warnings. We think these
4177 When compiling via C, you'll sometimes get “warning: assignment from
4178 incompatible pointer type” out of GCC. Harmless.
4185 Similarly, <command>ar</command>chiving warning messages like the following are not
4189 ar: filename GlaIOMonad__1_2s.o truncated to GlaIOMonad_
4190 ar: filename GlaIOMonad__2_2s.o truncated to GlaIOMonad_
4200 In compiling the compiler proper (in <filename>compiler/</filename>), you <emphasis>may</emphasis>
4201 get an “Out of heap space” error message. These can vary with the
4202 vagaries of different systems, it seems. The solution is simple:
4209 If you're compiling with GHC 4.00 or later, then the
4210 <emphasis>maximum</emphasis> heap size must have been reached. This
4211 is somewhat unlikely, since the maximum is set to 64M by default.
4212 Anyway, you can raise it with the
4213 <option>-optCrts-M<size></option> flag (add this flag to
4214 <constant><module>_HC_OPTS</constant>
4215 <command>make</command> variable in the appropriate
4216 <filename>Makefile</filename>).
4223 For GHC < 4.00, add a suitable <option>-H</option> flag to the <filename>Makefile</filename>, as
4232 and try again: <command>gmake</command>. (see <Xref LinkEnd="sec-suffix"> for information about
4233 <constant><module>_HC_OPTS</constant>.)
4235 Alternatively, just cut to the chase:
4239 % make EXTRA_HC_OPTS=-optCrts-M128M
4248 If you try to compile some Haskell, and you get errors from GCC about
4249 lots of things from <filename>/usr/include/math.h</filename>, then your GCC was
4250 mis-installed. <command>fixincludes</command> wasn't run when it should've been.
4252 As <command>fixincludes</command> is now automagically run as part of GCC installation,
4253 this bug also suggests that you have an old GCC.
4261 You <emphasis>may</emphasis> need to re-<command>ranlib</command><indexterm><primary>ranlib</primary></indexterm> your libraries (on Sun4s).
4265 % cd $(libdir)/ghc-x.xx/sparc-sun-sunos4
4266 % foreach i ( `find . -name '*.a' -print` ) # or other-shell equiv...
4268 ? # or, on some machines: ar s $i
4273 We'd be interested to know if this is still necessary.
4281 GHC's sources go through <command>cpp</command> before being compiled, and <command>cpp</command> varies
4282 a bit from one Unix to another. One particular gotcha is macro calls
4287 SLIT("Hello, world")
4291 Some <command>cpp</command>s treat the comma inside the string as separating two macro
4292 arguments, so you get
4296 :731: macro `SLIT' used with too many (2) args
4300 Alas, <command>cpp</command> doesn't tell you the offending file!
4302 Workaround: don't put weird things in string args to <command>cpp</command> macros.
4313 <Sect1 id="winbuild"><Title>Notes for building under Windows</Title>
4316 This section summarises how to get the utilities you need on your
4317 Win95/98/NT/2000 machine to use CVS and build GHC. Similar notes for
4318 installing and running GHC may be found in the user guide. In general,
4319 Win95/Win98 behave the same, and WinNT/Win2k behave the same.
4320 You should read the GHC installation guide sections on Windows (in the user
4321 guide) before continuing to read these notes.
4325 <sect2 id="cygwin-and-mingw"><Title>Cygwin and MinGW</Title>
4327 <para> The Windows situation for building GHC is rather confusing. This section
4328 tries to clarify, and to establish terminology.</para>
4330 <sect3 id="ghc-mingw"><title>GHC-mingw</title>
4332 <para> <ulink url="http://www.mingw.org">MinGW (Minimalist GNU for Windows)</ulink>
4333 is a collection of header
4334 files and import libraries that allow one to use <command>gcc</command> and produce
4335 native Win32 programs that do not rely on any third-party DLLs. The
4336 current set of tools include GNU Compiler Collection (<command>gcc</command>), GNU Binary
4337 Utilities (Binutils), GNU debugger (Gdb), GNU make, and a assorted
4340 <para>The GHC that we distribute includes, inside the distribution itself, the MinGW <command>gcc</command>,
4341 <command>as</command>, <command>ld</command>, and a bunch of input/output libraries.
4342 GHC compiles Haskell to C (or to
4343 assembly code), and then invokes these MinGW tools to generate an executable binary.
4344 The resulting binaries can run on any Win32 system.
4346 <para> We will call a GHC that targets MinGW in this way <emphasis>GHC-mingw</emphasis>.</para>
4348 <para> The down-side of GHC-mingw is that the MinGW libraries do not support anything like the full
4349 Posix interface. So programs compiled with GHC-mingw cannot import the (Haskell) Posix
4350 library; they have to do
4351 their input output using standard Haskell I/O libraries, or native Win32 bindings.
4355 <sect3 id="ghc-cygwin"><title>GHC-cygwin</title>
4357 <para>There <emphasis>is</emphasis> a way to get the full Posix interface, which is to use Cygwin.
4358 <ulink url="http://www.cygwin.com">Cygwin</ulink> is a complete Unix simulation that runs on Win32.
4359 Cygwin comes with a shell, and all the usual Unix commands: <command>mv</command>, <command>rm</command>,
4360 <command>ls</command>, plus of course <command>gcc</command>, <command>ld</command> and so on.
4361 A C program compiled with the Cygwin <command>gcc</command> certainly can use all of Posix.
4363 <para>So why doesn't GHC use the Cygwin <command>gcc</command> and libraries? Because
4364 Cygwin comes with a DLL <emphasis>that must be linked with every runnable Cygwin-compiled program</emphasis>.
4365 A program compiled by the Cygwin tools cannot run at all unless Cygwin is installed.
4366 If GHC targeted Cygwin, users would have to install Cygwin just to run the Haskell programs
4367 that GHC compiled; and the Cygwin DLL would have to be in the DLL load path.
4368 Worse, Cygwin is a moving target. The name of the main DLL, <literal>cygwin1.dll</literal>
4369 does not change, but the implementation certainly does. Even the interfaces to functions
4370 it exports seem to change occasionally. So programs compiled by GHC might only run with
4371 particular versions of Cygwin. All of this seems very undesirable.
4374 Nevertheless, it is certainly possible to build a version of GHC that targets Cygwin;
4375 we will call that <emphasis>GHC-cygwin</emphasis>. The up-side of GHC-cygwin is
4376 that Haskell programs compiled by GHC-cygwin can import the (Haskell) Posix library.
4380 <sect3><title>HOST_OS vs TARGET_OS</title>
4383 In the source code you'll find various ifdefs looking like:
4385 #ifdef mingw32_HOST_OS
4391 #ifdef mingw32_TARGET_OS
4395 These macros are set by the configure script (via the file config.h).
4396 Which is which? The criterion is this. In the ifdefs in GHC's source code:
4399 The "host" system is the one on which GHC itself will be run.
4402 The "target" system is the one for which the program compiled by GHC will be run.
4405 For a stage-2 compiler, in which GHCi is available, the "host" and "target" systems must be the same.
4406 So then it doesn't really matter whether you use the HOST_OS or TARGET_OS cpp macros.
4411 <sect3><title>Summary</title>
4413 <para>Notice that "GHC-mingw" means "GHC that <emphasis>targets</emphasis> MinGW". It says nothing about
4414 how that GHC was <emphasis>built</emphasis>. It is entirely possible to have a GHC-mingw that was built
4415 by compiling GHC's Haskell sources with a GHC-cygwin, or vice versa.</para>
4417 <para>We distribute only a GHC-mingw built by a GHC-mingw; supporting
4418 GHC-cygwin too is beyond our resources. The GHC we distribute
4419 therefore does not require Cygwin to run, nor do the programs it
4420 compiles require Cygwin.</para>
4422 <para>The instructions that follow describe how to build GHC-mingw. It is
4423 possible to build GHC-cygwin, but it's not a supported route, and the build system might
4426 <para>In your build tree, you build a compiler called <Command>ghc-inplace</Command>. It
4427 uses the <Command>gcc</Command> that you specify using the
4428 <option>--with-gcc</option> flag when you run
4429 <Command>configure</Command> (see below).
4430 The makefiles are careful to use <Command>ghc-inplace</Command> (not <Command>gcc</Command>)
4431 to compile any C files, so that it will in turn invoke the right <Command>gcc</Command> rather that
4432 whatever one happens to be in your path. However, the makefiles do use whatever <Command>ld</Command>
4433 and <Command>ar</Command> happen to be in your path. This is a bit naughty, but (a) they are only
4434 used to glom together .o files into a bigger .o file, or a .a file,
4435 so they don't ever get libraries (which would be bogus; they might be the wrong libraries), and (b)
4436 Cygwin and Mingw use the same .o file format. So its ok.
4441 <Sect2><Title>Installing and configuring Cygwin</Title>
4443 <para>You don't need Cygwin to <emphasis>use</emphasis> GHC,
4444 but you do need it to <emphasis>build</emphasis> GHC.</para>
4446 <para> Install Cygwin from <ulink url="http://www.cygwin.com/">http://www.cygwin.com/</ulink>.
4447 The installation process is straightforward; we install it in <Filename>c:/cygwin</Filename>.
4448 During the installation dialogue, make sure that you select all of the following:
4451 <command>cvs</command>,
4454 <command>openssh</command>,
4457 <command>autoconf</command>,
4460 <command>automake</command>,
4463 <command>binutils</command> (includes ld and (I think) ar),
4466 <command>gcc</command>,
4469 <command>flex</command>,
4472 <command>make</command>.
4475 If you miss out any of these, strange things will happen to you. To see thse packages,
4476 click on the "View" button in the "Select Packages"
4477 stage of Cygwin's installation dialogue, until the view says "Full". The default view, which is
4478 "Category" isn't very helpful, and the "View" button is rather unobtrousive.
4480 <para> Now set the following user environment variables:
4483 <listitem><para> Add <filename>c:/cygwin/bin</filename> and <filename>c:/cygwin/usr/bin</filename> to your
4484 <constant>PATH</constant></para></listitem>
4488 Set <constant>MAKE_MODE</constant> to <Literal>UNIX</Literal>. If you
4489 don't do this you get very weird messages when you type
4490 <Command>make</Command>, such as:
4492 /c: /c: No such file or directory
4497 <listitem><para> Set <constant>SHELL</constant> to
4498 <Filename>c:/cygwin/bin/sh</Filename>. When you invoke a shell in Emacs, this
4499 <constant>SHELL</constant> is what you get.
4502 <listitem><para> Set <constant>HOME</constant> to point to your
4503 home directory. This is where, for example,
4504 <command>bash</command> will look for your <filename>.bashrc</filename>
4505 file. Ditto <command>emacs</command> looking for <filename>.emacsrc</filename>
4511 There are a few other things to do:
4515 By default, cygwin provides the command shell <filename>ash</filename>
4516 as <filename>sh.exe</filename>. We have often seen build-system problems that
4517 turn out to be due to bugs in <filename>ash</filename>
4519 and length of command lines). On the other hand <filename>bash</filename> seems
4521 So, in <filename>cygwin/bin</filename>
4522 remove the supplied <filename>sh.exe</filename> (or rename it as <filename>ash.exe</filename>),
4523 and copy <filename>bash.exe</filename> to <filename>sh.exe</filename>.
4524 You'll need to do this in Windows Explorer or the Windows <command>cmd</command> shell, because
4525 you can't rename a running program!
4531 Some script files used in the make system start with "<Command>#!/bin/perl</Command>",
4532 (and similarly for <Command>sh</Command>). Notice the hardwired path!
4533 So you need to ensure that your <Filename>/bin</Filename> directory has the following
4536 <listitem> <para><Command>sh</Command></para></listitem>
4537 <listitem> <para><Command>perl</Command></para></listitem>
4538 <listitem> <para><Command>cat</Command></para></listitem>
4540 All these come in Cygwin's <Filename>bin</Filename> directory, which you probably have
4541 installed as <Filename>c:/cygwin/bin</Filename>. By default Cygwin mounts "<Filename>/</Filename>" as
4542 <Filename>c:/cygwin</Filename>, so if you just take the defaults it'll all work ok.
4543 (You can discover where your Cygwin
4544 root directory <Filename>/</Filename> is by typing <Command>mount</Command>.)
4545 Provided <Filename>/bin</Filename> points to the Cygwin <Filename>bin</Filename>
4546 directory, there's no need to copy anything. If not, copy these binaries from the <filename>cygwin/bin</filename>
4547 directory (after fixing the <filename>sh.exe</filename> stuff mentioned in the previous bullet).
4553 <para>Finally, here are some things to be aware of when using Cygwin:
4555 <listitem> <para>Cygwin doesn't deal well with filenames that include
4556 spaces. "<filename>Program Files</filename>" and "<filename>Local files</filename>" are
4560 <listitem> <para> Cygwin implements a symbolic link as a text file with some
4561 magical text in it. So other programs that don't use Cygwin's
4562 I/O libraries won't recognise such files as symlinks.
4563 In particular, programs compiled by GHC are meant to be runnable
4564 without having Cygwin, so they don't use the Cygwin library, so
4565 they don't recognise symlinks.
4569 Win32 has a <command>find</command> command which is not the same as Cygwin's find.
4570 You will probably discover that the Win32 <command>find</command> appears in your <constant>PATH</constant>
4571 before the Cygwin one, because it's in the <emphasis>system</emphasis> <constant>PATH</constant>
4572 environment variable, whereas you have probably modified the <emphasis>user</emphasis> <constant>PATH</constant>
4573 variable. You can always invoke <command>find</command> with an absolute path, or rename it.
4580 <Sect2 id="configure-ssh"><Title>Configuring SSH</Title>
4582 <para><command>ssh</command> comes with Cygwin, provided you remember to ask for it when
4583 you install Cygwin. (If not, the installer lets you update easily.) Look for <command>openssh</command>
4584 (not ssh) in the Cygwin list of applications!</para>
4586 <para>There are several strange things about <command>ssh</command> on Windows that you need to know.
4590 The programs <command>ssh-keygen1</command>, <command>ssh1</command>, and <command>cvs</command>,
4591 seem to lock up <command>bash</command> entirely if they try to get user input (e.g. if
4592 they ask for a password). To solve this, start up <filename>cmd.exe</filename>
4593 and run it as follows:
4595 c:\tmp> set CYGWIN32=tty
4596 c:\tmp> c:/user/local/bin/ssh-keygen1
4601 <command>ssh</command> needs to access your directory <filename>.ssh</filename>, in your home directory.
4602 To determine your home directory <command>ssh</command> first looks in
4603 <filename>c:/cygwin/etc/passwd</filename> (or wherever you have Cygwin installed). If there's an entry
4604 there with your userid, it'll use that entry to determine your home directory, <emphasis>ignoring
4605 the setting of the environment variable $HOME</emphasis>. If the home directory is
4606 bogus, <command>ssh</command> fails horribly. The best way to see what is going on is to say
4608 ssh -v cvs.haskell.org
4610 which makes <command>ssh</command> print out information about its activity.
4612 <para> You can fix this problem, either by correcting the home-directory field in
4613 <filename>c:/cygwin/etc/passwd</filename>, or by simply deleting the entire entry for your userid. If
4614 you do that, <command>ssh</command> uses the $HOME environment variable instead.
4620 <para>To protect your
4621 <literal>.ssh</literal> from access by anyone else,
4622 right-click your <literal>.ssh</literal> directory, and
4623 select <literal>Properties</literal>. If you are not on
4624 the access control list, add yourself, and give yourself
4625 full permissions (the second panel). Remove everyone else
4626 from the access control list. Don't leave them there but
4627 deny them access, because 'they' may be a list that
4628 includes you!</para>
4632 <para>In fact <command>ssh</command> 3.6.1 now seems to <emphasis>require</emphasis>
4633 you to have Unix permissions 600 (read/write for owner only)
4634 on the <literal>.ssh/identity</literal> file, else it
4635 bombs out. For your local C drive, it seems that <literal>chmod 600 identity</literal> works,
4636 but on Windows NT/XP, it doesn't work on a network drive (exact dteails obscure).
4637 The solution seems to be to set the $CYGWIN environment
4638 variable to "<literal>ntsec neta</literal>". The $CYGWIN environment variable is discussed
4639 in <ulink url="http://cygwin.com/cygwin-ug-net/using-cygwinenv.html">the Cygwin User's Guide</ulink>,
4640 and there are more details in <ulink url="http://cygwin.com/faq/faq_4.html#SEC44">the Cygwin FAQ</ulink>.
4647 <Sect2><Title>Other things you need to install</Title>
4649 <para>You have to install the following other things to build GHC:
4653 Install an executable GHC, from <ulink url="http://www.haskell.org/ghc">http://www.haskell.org/ghc</ulink>.
4654 This is what you will use to compile GHC. Add it in your
4655 <constant>PATH</constant>: the installer tells you the path element
4656 you need to add upon completion.
4662 Install an executable Happy, from <ulink url="http://www.haskell.org/happy">http://www.haskell.org/happy</ulink>.
4663 Happy is a parser generator used to compile the Haskell grammar. Add it in your
4664 <constant>PATH</constant>.
4669 <para>Install Alex. This can be done by building from the
4670 source distribution in the usual way. Sources are
4671 available from <ulink
4672 url="http://www.haskell.org/alex">http://www.haskell.org/alex</ulink>.</para>
4676 <para>GHC uses the <emphasis>mingw</emphasis> C compiler to
4677 generate code, so you have to install that (see <xref linkend="cygwin-and-mingw">).
4678 Just pick up a mingw bundle at
4679 <ulink url="http://www.mingw.org/">http://www.mingw.org/</ulink>.
4680 We install it in <filename>c:/mingw</filename>.
4682 <para>Do <emphasis>not</emphasis> add any of the <emphasis>mingw</emphasis> binaries to your path.
4683 They are only going to get used by explicit access (via the --with-gcc flag you
4684 give to <Command>configure</Command> later). If you do add them to your path
4685 you are likely to get into a mess because their names overlap with Cygwin binaries.
4691 <para>We use <command>emacs</command> a lot, so we install that too.
4692 When you are in <filename>fptools/ghc/compiler</filename>, you can use
4693 "<literal>make tags</literal>" to make a TAGS file for emacs. That uses the utility
4694 <filename>fptools/ghc/utils/hasktags/hasktags</filename>, so you need to make that first.
4695 The most convenient way to do this is by going <literal>make boot</literal> in <filename>fptools/ghc</filename>.
4696 The <literal>make tags</literal> command also uses <command>etags</command>, which comes with <command>emacs</command>,
4697 so you will need to add <filename>emacs/bin</filename> to your <literal>PATH</literal>.
4703 <para> Finally, check out a copy of GHC sources from
4704 the CVS repository, following the instructions above (<xref linkend="cvs-access">).
4711 <Sect2><Title>Building GHC</Title>
4714 Now go read the documentation above on building from source (<xref linkend="sec-building-from-source">);
4715 the bullets below only tell
4716 you about Windows-specific wrinkles.</para>
4720 Run <Command>autoconf</Command> both in <filename>fptools</filename>
4721 and in <filename>fptools/ghc</filename>. If you omit the latter step you'll
4722 get an error when you run <filename>./configure</filename>:
4725 creating mk/config.h
4726 mk/config.h is unchanged
4728 running /bin/sh ./configure --cache-file=.././config.cache --srcdir=.
4729 ./configure: ./configure: No such file or directory
4730 configure: error: ./configure failed for ghc
4735 <listitem> <para><command>autoconf</command> seems to create the file <filename>configure</filename>
4736 read-only. So if you need to run autoconf again (which I sometimes do for safety's sake),
4739 /usr/bin/autoconf: cannot create configure: permission denied
4741 Solution: delete <filename>configure</filename> first.
4746 You either need to add <filename>ghc</filename> to your
4747 <constant>PATH</constant> before you invoke
4748 <Command>configure</Command>, or use the <Command>configure</Command>
4749 option <option>--with-ghc=c:/ghc/ghc-some-version/bin/ghc</option>.
4754 If you are paranoid, delete <filename>config.cache</filename> if it exists.
4755 This file occasionally remembers out-of-date configuration information, which
4756 can be really confusing.
4762 After <command>autoconf</command> run <command>./configure</command> in
4763 <filename>fptools/</filename> thus:
4766 ./configure --host=i386-unknown-mingw32 --with-gcc=c:/mingw/bin/gcc
4768 This is the point at which you specify that you are building GHC-mingw
4769 (see <xref linkend="ghc-mingw">). </para>
4771 <para> Both these options are important! It's possible to get into
4772 trouble using the wrong C compiler!</para>
4774 Furthermore, it's <emphasis>very important</emphasis> that you specify a
4775 full MinGW path for <command>gcc</command>, not a Cygwin path, because GHC (which
4776 uses this path to invoke <command>gcc</command>) is a MinGW program and won't
4777 understand a Cygwin path. For example, if you
4778 say <literal>--with-gcc=/mingw/bin/gcc</literal>, it'll be interpreted as
4779 <filename>/cygdrive/c/mingw/bin/gcc</filename>, and GHC will fail the first
4780 time it tries to invoke it. Worse, the failure comes with
4781 no error message whatsoever. GHC simply fails silently when first invoked,
4782 typically leaving you with this:
4784 make[4]: Leaving directory `/cygdrive/e/fptools-stage1/ghc/rts/gmp'
4785 ../../ghc/compiler/ghc-inplace -optc-mno-cygwin -optc-O
4786 -optc-Wall -optc-W -optc-Wstrict-prototypes -optc-Wmissing-prototypes
4787 -optc-Wmissing-declarations -optc-Winline -optc-Waggregate-return
4788 -optc-Wbad-function-cast -optc-Wcast-align -optc-I../includes
4789 -optc-I. -optc-Iparallel -optc-DCOMPILING_RTS
4790 -optc-fomit-frame-pointer -O2 -static
4791 -package-name rts -O -dcore-lint -c Adjustor.c -o Adjustor.o
4792 make[2]: *** [Adjustor.o] Error 1
4793 make[1]: *** [all] Error 1
4794 make[1]: Leaving directory `/cygdrive/e/fptools-stage1/ghc'
4795 make: *** [all] Error 1
4801 If you want to build GHC-cygwin (<xref linkend="ghc-cygwin">)
4802 you'll have to do something more like:
4804 ./configure --with-gcc=...the Cygwin gcc...
4809 <listitem><para> You almost certainly want to set
4813 in your <filename>build.mk</filename> configuration file (see <xref linkend="sec-build-config">).
4814 This tells the build system not to split each library into a myriad of little object files, one
4815 for each function. Doing so reduces binary sizes for statically-linked binaries, but on Windows
4816 it dramatically increases the time taken to build the libraries in the first place.
4820 <listitem><para> Do not attempt to build the documentation.
4821 It needs all kinds of wierd Jade stuff that we haven't worked out for
4822 Win32.</para></listitem>