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 (version 5.x or later) 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 the autoconf package is required.
1310 NB. version 2.13 will no longer work, as of GHC version
1313 <para><command>autoreconf</command> (from the autoconf package)
1314 recursively builds <command>configure</command> scripts from
1315 the corresponding <filename>configure.ac</filename> and
1316 <filename>aclocal.m4</filename> files. If you modify one of
1317 the latter files, you'll need <command>autoreconf</command> to
1318 rebuild the corresponding <filename>configure</filename>.</para>
1323 <term><command>sed</command></term>
1324 <indexterm><primary>pre-supposed: sed</primary></indexterm>
1325 <indexterm><primary>sed, pre-supposed</primary></indexterm>
1327 <para>You need a working <command>sed</command> if you are
1328 going to build from sources. The build-configuration stuff
1329 needs it. GNU sed version 2.0.4 is no good! It has a bug
1330 in it that is tickled by the build-configuration. 2.0.5 is
1331 OK. Others are probably OK too (assuming we don't create too
1332 elaborate configure scripts.)</para>
1337 <para>One <literal>fptools</literal> project is worth a quick note
1338 at this point, because it is useful for all the others:
1339 <literal>glafp-utils</literal> contains several utilities which
1340 aren't particularly Glasgow-ish, but Occasionally Indispensable.
1341 Like <command>lndir</command> for creating symbolic link
1344 <sect2 id="pre-supposed-gph-tools">
1345 <title>Tools for building parallel GHC (GPH)</title>
1349 <term>PVM version 3:</term>
1350 <indexterm><primary>pre-supposed: PVM3 (Parallel Virtual Machine)</primary></indexterm>
1351 <indexterm><primary>PVM3 (Parallel Virtual Machine), pre-supposed</primary></indexterm>
1353 <para>PVM is the Parallel Virtual Machine on which
1354 Parallel Haskell programs run. (You only need this if you
1355 plan to run Parallel Haskell. Concurrent Haskell, which
1356 runs concurrent threads on a uniprocessor doesn't need
1357 it.) Underneath PVM, you can have (for example) a network
1358 of workstations (slow) or a multiprocessor box
1361 <para>The current version of PVM is 3.3.11; we use 3.3.7.
1362 It is readily available on the net; I think I got it from
1363 <literal>research.att.com</literal>, in
1364 <filename>netlib</filename>.</para>
1366 <para>A PVM installation is slightly quirky, but easy to
1367 do. Just follow the <filename>Readme</filename>
1368 instructions.</para>
1373 <term><command>bash</command>:</term>
1374 <indexterm><primary>bash, presupposed (Parallel Haskell only)</primary></indexterm>
1376 <para>Sadly, the <command>gr2ps</command> script, used to
1377 convert “parallelism profiles” to PostScript,
1378 is written in Bash (GNU's Bourne Again shell). This bug
1379 will be fixed (someday).</para>
1385 <sect2 id="pre-supposed-other-tools">
1386 <title>Other useful tools</title>
1391 <indexterm><primary>pre-supposed: flex</primary></indexterm>
1392 <indexterm><primary>flex, pre-supposed</primary></indexterm>
1394 <para>This is a quite-a-bit-better-than-Lex lexer. Used
1395 to build a couple of utilities in
1396 <literal>glafp-utils</literal>. Depending on your
1397 operating system, the supplied <command>lex</command> may
1398 or may not work; you should get the GNU version.</para>
1403 <para>More tools are required if you want to format the documentation
1404 that comes with GHC and other fptools projects. See <xref
1405 linkend="building-docs">.</para>
1409 <sect1 id="sec-building-from-source">
1410 <title>Building from source</title>
1412 <indexterm><primary>Building from source</primary></indexterm>
1413 <indexterm><primary>Source, building from</primary></indexterm>
1415 <para>You've been rash enough to want to build some of the Glasgow
1416 Functional Programming tools (GHC, Happy, nofib, etc.) from
1417 source. You've slurped the source, from the CVS repository or
1418 from a source distribution, and now you're sitting looking at a
1419 huge mound of bits, wondering what to do next.</para>
1421 <para>Gingerly, you type <command>make</command>. Wrong
1424 <para>This rest of this guide is intended for duffers like me, who
1425 aren't really interested in Makefiles and systems configurations,
1426 but who need a mental model of the interlocking pieces so that
1427 they can make them work, extend them consistently when adding new
1428 software, and lay hands on them gently when they don't
1431 <sect2 id="quick-start">
1432 <title>Quick Start</title>
1434 <para>If you are starting from a source distribution, and just
1435 want a completely standard build, then the following should
1438 <screen>$ autoreconf
1444 <para>For GHC, this will do a 2-stage bootstrap build of the
1445 compiler, with profiling libraries, and install the
1448 <para>If you want to do anything at all non-standard, or you
1449 want to do some development, read on...</para>
1452 <sect2 id="sec-source-tree">
1453 <title>Your source tree</title>
1455 <para>The source code is held in your <emphasis>source
1456 tree</emphasis>. The root directory of your source tree
1457 <emphasis>must</emphasis> contain the following directories and
1462 <para><filename>Makefile</filename>: the root
1467 <para><filename>mk/</filename>: the directory that contains
1468 the main Makefile code, shared by all the
1469 <literal>fptools</literal> software.</para>
1473 <para><filename>configure.ac</filename>,
1474 <filename>config.sub</filename>,
1475 <filename>config.guess</filename>: these files support the
1476 configuration process.</para>
1480 <para><filename>install-sh</filename>.</para>
1484 <para>All the other directories are individual
1485 <emphasis>projects</emphasis> of the <literal>fptools</literal>
1486 system—for example, the Glasgow Haskell Compiler
1487 (<literal>ghc</literal>), the Happy parser generator
1488 (<literal>happy</literal>), the <literal>nofib</literal>
1489 benchmark suite, and so on. You can have zero or more of these.
1490 Needless to say, some of them are needed to build others.</para>
1492 <para>The important thing to remember is that even if you want
1493 only one project (<literal>happy</literal>, say), you must have
1494 a source tree whose root directory contains
1495 <filename>Makefile</filename>, <filename>mk/</filename>,
1496 <filename>configure.ac</filename>, and the project(s) you want
1497 (<filename>happy/</filename> in this case). You cannot get by
1498 with just the <filename>happy/</filename> directory.</para>
1502 <title>Build trees</title>
1503 <indexterm><primary>build trees</primary></indexterm>
1504 <indexterm><primary>link trees, for building</primary></indexterm>
1506 <para>If you just want to build the software once on a single
1507 platform, then your source tree can also be your build tree, and
1508 you can skip the rest of this section.</para>
1510 <para>We often want to build multiple versions of our software
1511 for different architectures, or with different options
1512 (e.g. profiling). It's very desirable to share a single copy of
1513 the source code among all these builds.</para>
1515 <para>So for every source tree we have zero or more
1516 <emphasis>build trees</emphasis>. Each build tree is initially
1517 an exact copy of the source tree, except that each file is a
1518 symbolic link to the source file, rather than being a copy of
1519 the source file. There are “standard” Unix
1520 utilities that make such copies, so standard that they go by
1522 <command>lndir</command><indexterm><primary>lndir</primary></indexterm>,
1523 <command>mkshadowdir</command><indexterm><primary>mkshadowdir</primary></indexterm>
1524 are two (If you don't have either, the source distribution
1525 includes sources for the X11
1526 <command>lndir</command>—check out
1527 <filename>fptools/glafp-utils/lndir</filename>). See <Xref
1528 LinkEnd="sec-storysofar"> for a typical invocation.</para>
1530 <para>The build tree does not need to be anywhere near the
1531 source tree in the file system. Indeed, one advantage of
1532 separating the build tree from the source is that the build tree
1533 can be placed in a non-backed-up partition, saving your systems
1534 support people from backing up untold megabytes of
1535 easily-regenerated, and rapidly-changing, gubbins. The golden
1536 rule is that (with a single exception—<XRef
1537 LinkEnd="sec-build-config">) <emphasis>absolutely everything in
1538 the build tree is either a symbolic link to the source tree, or
1539 else is mechanically generated</emphasis>. It should be
1540 perfectly OK for your build tree to vanish overnight; an hour or
1541 two compiling and you're on the road again.</para>
1543 <para>You need to be a bit careful, though, that any new files
1544 you create (if you do any development work) are in the source
1545 tree, not a build tree!</para>
1547 <para>Remember, that the source files in the build tree are
1548 <emphasis>symbolic links</emphasis> to the files in the source
1549 tree. (The build tree soon accumulates lots of built files like
1550 <filename>Foo.o</filename>, as well.) You can
1551 <emphasis>delete</emphasis> a source file from the build tree
1552 without affecting the source tree (though it's an odd thing to
1553 do). On the other hand, if you <emphasis>edit</emphasis> a
1554 source file from the build tree, you'll edit the source-tree
1555 file directly. (You can set up Emacs so that if you edit a
1556 source file from the build tree, Emacs will silently create an
1557 edited copy of the source file in the build tree, leaving the
1558 source file unchanged; but the danger is that you think you've
1559 edited the source file whereas actually all you've done is edit
1560 the build-tree copy. More commonly you do want to edit the
1561 source file.)</para>
1563 <para>Like the source tree, the top level of your build tree
1564 must be (a linked copy of) the root directory of the
1565 <literal>fptools</literal> suite. Inside Makefiles, the root of
1566 your build tree is called
1567 <constant>$(FPTOOLS_TOP)</constant><indexterm><primary>FPTOOLS_TOP</primary></indexterm>.
1568 In the rest of this document path names are relative to
1569 <constant>$(FPTOOLS_TOP)</constant> unless
1570 otherwise stated. For example, the file
1571 <filename>ghc/mk/target.mk</filename> is actually
1572 <filename><constant>$(FPTOOLS_TOP)</constant>/ghc/mk/target.mk</filename>.</para>
1575 <sect2 id="sec-build-config">
1576 <title>Getting the build you want</title>
1578 <para>When you build <literal>fptools</literal> you will be
1579 compiling code on a particular <emphasis>host
1580 platform</emphasis>, to run on a particular <emphasis>target
1581 platform</emphasis> (usually the same as the host
1582 platform)<indexterm><primary>platform</primary></indexterm>.
1583 The difficulty is that there are minor differences between
1584 different platforms; minor, but enough that the code needs to be
1585 a bit different for each. There are some big differences too:
1586 for a different architecture we need to build GHC with a
1587 different native-code generator.</para>
1589 <para>There are also knobs you can turn to control how the
1590 <literal>fptools</literal> software is built. For example, you
1591 might want to build GHC optimised (so that it runs fast) or
1592 unoptimised (so that you can compile it fast after you've
1593 modified it. Or, you might want to compile it with debugging on
1594 (so that extra consistency-checking code gets included) or off.
1597 <para>All of this stuff is called the
1598 <emphasis>configuration</emphasis> of your build. You set the
1599 configuration using a three-step process.</para>
1603 <term>Step 1: get ready for configuration.</term>
1605 <para>NOTE: if you're starting from a source distribution,
1606 rather than CVS sources, you can skip this step.</para>
1608 <para>Change directory to
1609 <constant>$(FPTOOLS_TOP)</constant> and
1610 issue the command</para>
1614 <indexterm><primary>autoreconf</primary></indexterm>
1615 <para>(with no arguments). This GNU program (recursively) converts
1616 <filename><constant>$(FPTOOLS_TOP)</constant>/configure.ac</filename> and
1617 <filename><constant>$(FPTOOLS_TOP)</constant>/aclocal.m4</filename>
1618 to a shell script called
1619 <filename><constant>$(FPTOOLS_TOP)</constant>/configure</filename>.
1620 If <command>autoreconf</command> bleats that it can't write the file <filename>configure</filename>,
1621 then delete the latter and try again. Note that you must use <command>autoreconf</command>,
1622 and not the old <command>autoconf</command>! If you erroneously use the latter, you'll get
1623 a message like "No rule to make target 'mk/config.h.in'".
1626 <para>Some projects, including GHC, have their own configure script.
1627 <command>autoreconf</command> takes care of that, too, so all you have
1628 to do is calling <command>autoreconf</command> in the top-level directory
1629 <constant>$(FPTOOLS_TOP)</constant>.</para>
1631 <para>These steps are completely platform-independent; they just mean
1632 that the human-written files (<filename>configure.ac</filename> and
1633 <filename>aclocal.m4</filename>) can be short, although the resulting
1634 files (the <command>configure</command> shell scripts and the C header
1635 template <filename>mk/config.h.in</filename>) are long.</para>
1640 <term>Step 2: system configuration.</term>
1642 <para>Runs the newly-created <command>configure</command>
1643 script, thus:</para>
1646 ./configure <optional><parameter>args</parameter></optional>
1649 <para><command>configure</command>'s mission is to scurry
1650 round your computer working out what architecture it has,
1651 what operating system, whether it has the
1652 <Function>vfork</Function> system call, where
1653 <command>tar</command> is kept, whether
1654 <command>gcc</command> is available, where various obscure
1655 <literal>#include</literal> files are, whether it's a
1656 leap year, and what the systems manager had for lunch. It
1657 communicates these snippets of information in two
1664 <filename>mk/config.mk.in</filename><indexterm><primary>config.mk.in</primary></indexterm>
1666 <filename>mk/config.mk</filename><indexterm><primary>config.mk</primary></indexterm>,
1667 substituting for things between
1668 “<literal>@</literal>” brackets. So,
1669 “<literal>@HaveGcc@</literal>” will be
1670 replaced by “<literal>YES</literal>” or
1671 “<literal>NO</literal>” depending on what
1672 <command>configure</command> finds.
1673 <filename>mk/config.mk</filename> is included by every
1674 Makefile (directly or indirectly), so the
1675 configuration information is thereby communicated to
1676 all Makefiles.</para>
1680 <para> It translates
1681 <filename>mk/config.h.in</filename><indexterm><primary>config.h.in</primary></indexterm>
1683 <filename>mk/config.h</filename><indexterm><primary>config.h</primary></indexterm>.
1684 The latter is <literal>#include</literal>d by
1685 various C programs, which can thereby make use of
1686 configuration information.</para>
1690 <para><command>configure</command> takes some optional
1691 arguments. Use <literal>./configure --help</literal> to
1692 get a list of the available arguments. Here are some of
1693 the ones you might need:</para>
1697 <term><literal>--with-ghc=<parameter>path</parameter></literal></term>
1698 <indexterm><primary><literal>--with-ghc</literal></primary>
1701 <para>Specifies the path to an installed GHC which
1702 you would like to use. This compiler will be used
1703 for compiling GHC-specific code (eg. GHC itself).
1704 This option <emphasis>cannot</emphasis> be specified
1705 using <filename>build.mk</filename> (see later),
1706 because <command>configure</command> needs to
1707 auto-detect the version of GHC you're using. The
1708 default is to look for a compiler named
1709 <literal>ghc</literal> in your path.</para>
1714 <term><literal>--with-hc=<parameter>path</parameter></literal></term>
1715 <indexterm><primary><literal>--with-hc</literal></primary>
1718 <para>Specifies the path to any installed Haskell
1719 compiler. This compiler will be used for compiling
1720 generic Haskell code. The default is to use
1721 <literal>ghc</literal>.</para>
1726 <term><literal>--with-gcc=<parameter>path</parameter></literal></term>
1727 <indexterm><primary><literal>--with-gcc</literal></primary>
1730 <para>Specifies the path to the installed GCC. This
1731 compiler will be used to compile all C files,
1732 <emphasis>except</emphasis> any generated by the
1733 installed Haskell compiler, which will have its own
1734 idea of which C compiler (if any) to use. The
1735 default is to use <literal>gcc</literal>.</para>
1743 <term>Step 3: build configuration.</term>
1745 <para>Next, you say how this build of
1746 <literal>fptools</literal> is to differ from the standard
1747 defaults by creating a new file
1748 <filename>mk/build.mk</filename><indexterm><primary>build.mk</primary></indexterm>
1749 <emphasis>in the build tree</emphasis>. This file is the
1750 one and only file you edit in the build tree, precisely
1751 because it says how this build differs from the source.
1752 (Just in case your build tree does die, you might want to
1753 keep a private directory of <filename>build.mk</filename>
1754 files, and use a symbolic link in each build tree to point
1755 to the appropriate one.) So
1756 <filename>mk/build.mk</filename> never exists in the
1757 source tree—you create one in each build tree from
1758 the template. We'll discuss what to put in it
1764 <para>And that's it for configuration. Simple, eh?</para>
1766 <para>What do you put in your build-specific configuration file
1767 <filename>mk/build.mk</filename>? <emphasis>For almost all
1768 purposes all you will do is put make variable definitions that
1769 override those in</emphasis>
1770 <filename>mk/config.mk.in</filename>. The whole point of
1771 <filename>mk/config.mk.in</filename>—and its derived
1772 counterpart <filename>mk/config.mk</filename>—is to define
1773 the build configuration. It is heavily commented, as you will
1774 see if you look at it. So generally, what you do is look at
1775 <filename>mk/config.mk.in</filename>, and add definitions in
1776 <filename>mk/build.mk</filename> that override any of the
1777 <filename>config.mk</filename> definitions that you want to
1778 change. (The override occurs because the main boilerplate file,
1779 <filename>mk/boilerplate.mk</filename><indexterm><primary>boilerplate.mk</primary></indexterm>,
1780 includes <filename>build.mk</filename> after
1781 <filename>config.mk</filename>.)</para>
1783 <para>For your convenience, there's a file called <filename>build.mk.sample</filename>
1784 that can serve as a starting point for your <filename>build.mk</filename>.</para>
1786 <para>For example, <filename>config.mk.in</filename> contains
1787 the definition:</para>
1790 GhcHcOpts=-O -Rghc-timing
1793 <para>The accompanying comment explains that this is the list of
1794 flags passed to GHC when building GHC itself. For doing
1795 development, it is wise to add <literal>-DDEBUG</literal>, to
1796 enable debugging code. So you would add the following to
1797 <filename>build.mk</filename>:</para>
1799 <para>or, if you prefer,</para>
1802 GhcHcOpts += -DDEBUG
1805 <para>GNU <command>make</command> allows existing definitions to
1806 have new text appended using the “<literal>+=</literal>”
1807 operator, which is quite a convenient feature.)</para>
1809 <para>If you want to remove the <literal>-O</literal> as well (a
1810 good idea when developing, because the turn-around cycle gets a
1811 lot quicker), you can just override
1812 <literal>GhcLibHcOpts</literal> altogether:</para>
1815 GhcHcOpts=-DDEBUG -Rghc-timing
1818 <para>When reading <filename>config.mk.in</filename>, remember
1819 that anything between “@...@” signs is going to be substituted
1820 by <command>configure</command> later. You
1821 <emphasis>can</emphasis> override the resulting definition if
1822 you want, but you need to be a bit surer what you are doing.
1823 For example, there's a line that says:</para>
1829 <para>This defines the Make variables <constant>TAR</constant>
1830 to the pathname for a <command>tar</command> that
1831 <command>configure</command> finds somewhere. If you have your
1832 own pet <command>tar</command> you want to use instead, that's
1833 fine. Just add this line to <filename>mk/build.mk</filename>:</para>
1839 <para>You do not <emphasis>have</emphasis> to have a
1840 <filename>mk/build.mk</filename> file at all; if you don't,
1841 you'll get all the default settings from
1842 <filename>mk/config.mk.in</filename>.</para>
1844 <para>You can also use <filename>build.mk</filename> to override
1845 anything that <command>configure</command> got wrong. One place
1846 where this happens often is with the definition of
1847 <constant>FPTOOLS_TOP_ABS</constant>: this
1848 variable is supposed to be the canonical path to the top of your
1849 source tree, but if your system uses an automounter then the
1850 correct directory is hard to find automatically. If you find
1851 that <command>configure</command> has got it wrong, just put the
1852 correct definition in <filename>build.mk</filename>.</para>
1856 <sect2 id="sec-storysofar">
1857 <title>The story so far</title>
1859 <para>Let's summarise the steps you need to carry to get
1860 yourself a fully-configured build tree from scratch.</para>
1864 <para> Get your source tree from somewhere (CVS repository
1865 or source distribution). Say you call the root directory
1866 <filename>myfptools</filename> (it does not have to be
1867 called <filename>fptools</filename>). Make sure that you
1868 have the essential files (see <XRef
1869 LinkEnd="sec-source-tree">).</para>
1874 <para>(Optional) Use <command>lndir</command> or
1875 <command>mkshadowdir</command> to create a build tree.</para>
1879 $ mkshadowdir . /scratch/joe-bloggs/myfptools-sun4
1882 <para>(N.B. <command>mkshadowdir</command>'s first argument
1883 is taken relative to its second.) You probably want to give
1884 the build tree a name that suggests its main defining
1885 characteristic (in your mind at least), in case you later
1890 <para>Change directory to the build tree. Everything is
1891 going to happen there now.</para>
1894 $ cd /scratch/joe-bloggs/myfptools-sun4
1900 <para>Prepare for system configuration:</para>
1906 <para>(You can skip this step if you are starting from a
1907 source distribution, and you already have
1908 <filename>configure</filename> and
1909 <filename>mk/config.h.in</filename>.)</para>
1913 <para>Do system configuration:</para>
1919 <para>Don't forget to check whether you need to add any
1920 arguments to <literal>configure</literal>; for example, a
1921 common requirement is to specify which GHC to use with
1922 <option>--with-ghc=<replaceable>ghc</replaceable></option>.</para>
1926 <para>Create the file <filename>mk/build.mk</filename>,
1927 adding definitions for your desired configuration
1936 <para>You can make subsequent changes to
1937 <filename>mk/build.mk</filename> as often as you like. You do
1938 not have to run any further configuration programs to make these
1939 changes take effect. In theory you should, however, say
1940 <command>gmake clean</command>, <command>gmake all</command>,
1941 because configuration option changes could affect
1942 anything—but in practice you are likely to know what's
1947 <title>Making things</title>
1949 <para>At this point you have made yourself a fully-configured
1950 build tree, so you are ready to start building real
1953 <para>The first thing you need to know is that <emphasis>you
1954 must use GNU <command>make</command>, usually called
1955 <command>gmake</command>, not standard Unix
1956 <command>make</command></emphasis>. If you use standard Unix
1957 <command>make</command> you will get all sorts of error messages
1958 (but no damage) because the <literal>fptools</literal>
1959 <command>Makefiles</command> use GNU <command>make</command>'s
1960 facilities extensively.</para>
1962 <para>To just build the whole thing, <command>cd</command> to
1963 the top of your <literal>fptools</literal> tree and type
1964 <command>gmake</command>. This will prepare the tree and build
1965 the various projects in the correct order.</para>
1968 <sect2 id="sec-bootstrapping">
1969 <title>Bootstrapping GHC</title>
1971 <para>GHC requires a 2-stage bootstrap in order to provide
1972 full functionality, including GHCi. By a 2-stage bootstrap, we
1973 mean that the compiler is built once using the installed GHC,
1974 and then again using the compiler built in the first stage. You
1975 can also build a stage 3 compiler, but this normally isn't
1976 necessary except to verify that the stage 2 compiler is working
1979 <para>Note that when doing a bootstrap, the stage 1 compiler
1980 must be built, followed by the runtime system and libraries, and
1981 then the stage 2 compiler. The correct ordering is implemented
1982 by the top-level fptools <filename>Makefile</filename>, so if
1983 you want everything to work automatically it's best to start
1984 <command>make</command> from the top of the tree. When building
1985 GHC, the top-level fptools <filename>Makefile</filename> is set
1986 up to do a 2-stage bootstrap by default (when you say
1987 <command>make</command>). Some other targets it supports
1994 <para>Build everything as normal, including the stage 1
2002 <para>Build the stage 2 compiler only.</para>
2009 <para>Build the stage 3 compiler only.</para>
2014 <term>bootstrap</term> <term>bootstrap2</term>
2016 <para>Build stage 1 followed by stage 2.</para>
2021 <term>bootstrap3</term>
2023 <para>Build stages 1, 2 and 3.</para>
2028 <term>install</term>
2030 <para>Install everything, including the compiler built in
2031 stage 2. To override the stage, say <literal>make install
2032 stage=<replaceable>n</replaceable></literal> where
2033 <replaceable>n</replaceable> is the stage to install.</para>
2038 <para>The top-level <filename>Makefile</filename> also arranges
2039 to do the appropriate <literal>make boot</literal> steps (see
2040 below) before actually building anything.</para>
2042 <para>The <literal>stage1</literal>, <literal>stage2</literal>
2043 and <literal>stage3</literal> targets also work in the
2044 <literal>ghc/compiler</literal> directory, but don't forget that
2045 each stage requires its own <literal>make boot</literal> step:
2046 for example, you must do</para>
2048 <screen>$ make boot stage=2</screen>
2050 <para>before <literal>make stage2</literal> in
2051 <literal>ghc/compiler</literal>.</para>
2054 <sect2 id="sec-standard-targets">
2055 <title>Standard Targets</title>
2056 <indexterm><primary>targets, standard makefile</primary></indexterm>
2057 <indexterm><primary>makefile targets</primary></indexterm>
2059 <para>In any directory you should be able to make the following:</para>
2063 <term><literal>boot</literal></term>
2065 <para>does the one-off preparation required to get ready
2066 for the real work. Notably, it does <command>gmake
2067 depend</command> in all directories that contain programs.
2068 It also builds the necessary tools for compilation to
2071 <para>Invoking the <literal>boot</literal> target
2072 explicitly is not normally necessary. From the top-level
2073 <literal>fptools</literal> directory, invoking
2074 <literal>gmake</literal> causes <literal>gmake boot
2075 all</literal> to be invoked in each of the project
2076 subdirectories, in the order specified by
2077 <literal>$(AllTargets)</literal> in
2078 <literal>config.mk</literal>.</para>
2080 <para>If you're working in a subdirectory somewhere and
2081 need to update the dependencies, <literal>gmake
2082 boot</literal> is a good way to do it.</para>
2087 <term><literal>all</literal></term>
2089 <para>makes all the final target(s) for this Makefile.
2090 Depending on which directory you are in a “final
2091 target” may be an executable program, a library
2092 archive, a shell script, or a Postscript file. Typing
2093 <command>gmake</command> alone is generally the same as
2094 typing <command>gmake all</command>.</para>
2099 <term><literal>install</literal></term>
2101 <para>installs the things built by <literal>all</literal>
2102 (except for the documentation). Where does it install
2103 them? That is specified by
2104 <filename>mk/config.mk.in</filename>; you can override it
2105 in <filename>mk/build.mk</filename>, or by running
2106 <command>configure</command> with command-line arguments
2107 like <literal>--bindir=/home/simonpj/bin</literal>; see
2108 <literal>./configure --help</literal> for the full
2114 <term><literal>install-docs</literal></term>
2116 <para>installs the documentation. Otherwise behaves just
2117 like <literal>install</literal>.</para>
2122 <term><literal>uninstall</literal></term>
2124 <para>reverses the effect of
2125 <literal>install</literal>.</para>
2130 <term><literal>clean</literal></term>
2132 <para>Delete all files from the current directory that are
2133 normally created by building the program. Don't delete
2134 the files that record the configuration, or files
2135 generated by <command>gmake boot</command>. Also preserve
2136 files that could be made by building, but normally aren't
2137 because the distribution comes with them.</para>
2142 <term><literal>distclean</literal></term>
2144 <para>Delete all files from the current directory that are
2145 created by configuring or building the program. If you
2146 have unpacked the source and built the program without
2147 creating any other files, <literal>make
2148 distclean</literal> should leave only the files that were
2149 in the distribution.</para>
2154 <term><literal>mostlyclean</literal></term>
2156 <para>Like <literal>clean</literal>, but may refrain from
2157 deleting a few files that people normally don't want to
2163 <term><literal>maintainer-clean</literal></term>
2165 <para>Delete everything from the current directory that
2166 can be reconstructed with this Makefile. This typically
2167 includes everything deleted by
2168 <literal>distclean</literal>, plus more: C source files
2169 produced by Bison, tags tables, Info files, and so
2172 <para>One exception, however: <literal>make
2173 maintainer-clean</literal> should not delete
2174 <filename>configure</filename> even if
2175 <filename>configure</filename> can be remade using a rule
2176 in the <filename>Makefile</filename>. More generally,
2177 <literal>make maintainer-clean</literal> should not delete
2178 anything that needs to exist in order to run
2179 <filename>configure</filename> and then begin to build the
2185 <term><literal>check</literal></term>
2187 <para>run the test suite.</para>
2192 <para>All of these standard targets automatically recurse into
2193 sub-directories. Certain other standard targets do not:</para>
2197 <term><literal>configure</literal></term>
2199 <para>is only available in the root directory
2200 <constant>$(FPTOOLS_TOP)</constant>; it has
2201 been discussed in <XRef
2202 LinkEnd="sec-build-config">.</para>
2207 <term><literal>depend</literal></term>
2209 <para>make a <filename>.depend</filename> file in each
2210 directory that needs it. This <filename>.depend</filename>
2211 file contains mechanically-generated dependency
2212 information; for example, suppose a directory contains a
2213 Haskell source module <filename>Foo.lhs</filename> which
2214 imports another module <literal>Baz</literal>. Then the
2215 generated <filename>.depend</filename> file will contain
2216 the dependency:</para>
2222 <para>which says that the object file
2223 <filename>Foo.o</filename> depends on the interface file
2224 <filename>Baz.hi</filename> generated by compiling module
2225 <literal>Baz</literal>. The <filename>.depend</filename>
2226 file is automatically included by every Makefile.</para>
2231 <term><literal>binary-dist</literal></term>
2233 <para>make a binary distribution. This is the target we
2234 use to build the binary distributions of GHC and
2240 <term><literal>dist</literal></term>
2242 <para>make a source distribution. Note that this target
2243 does “make distclean” as part of its work;
2244 don't use it if you want to keep what you've built.</para>
2249 <para>Most <filename>Makefile</filename>s have targets other
2250 than these. You can discover them by looking in the
2251 <filename>Makefile</filename> itself.</para>
2255 <title>Using a project from the build tree</title>
2257 <para>If you want to build GHC (say) and just use it direct from
2258 the build tree without doing <literal>make install</literal>
2259 first, you can run the in-place driver script:
2260 <filename>ghc/compiler/ghc-inplace</filename>.</para>
2262 <para> Do <emphasis>NOT</emphasis> use
2263 <filename>ghc/compiler/ghc</filename>, or
2264 <filename>ghc/compiler/ghc-6.xx</filename>, as these are the
2265 scripts intended for installation, and contain hard-wired paths
2266 to the installed libraries, rather than the libraries in the
2269 <para>Happy can similarly be run from the build tree, using
2270 <filename>happy/src/happy-inplace</filename>, and similarly for
2271 Alex and Haddock.</para>
2275 <title>Fast Making</title>
2277 <indexterm><primary>fastmake</primary></indexterm>
2278 <indexterm><primary>dependencies, omitting</primary></indexterm>
2279 <indexterm><primary>FAST, makefile variable</primary></indexterm>
2281 <para>Sometimes the dependencies get in the way: if you've made
2282 a small change to one file, and you're absolutely sure that it
2283 won't affect anything else, but you know that
2284 <command>make</command> is going to rebuild everything anyway,
2285 the following hack may be useful:</para>
2291 <para>This tells the make system to ignore dependencies and just
2292 build what you tell it to. In other words, it's equivalent to
2293 temporarily removing the <filename>.depend</filename> file in
2294 the current directory (where <command>mkdependHS</command> and
2295 friends store their dependency information).</para>
2297 <para>A bit of history: GHC used to come with a
2298 <command>fastmake</command> script that did the above job, but
2299 GNU make provides the features we need to do it without
2300 resorting to a script. Also, we've found that fastmaking is
2301 less useful since the advent of GHC's recompilation checker (see
2302 the User's Guide section on "Separate Compilation").</para>
2306 <sect1 id="sec-makefile-arch">
2307 <title>The <filename>Makefile</filename> architecture</title>
2308 <indexterm><primary>makefile architecture</primary></indexterm>
2310 <para><command>make</command> is great if everything
2311 works—you type <command>gmake install</command> and lo! the
2312 right things get compiled and installed in the right places. Our
2313 goal is to make this happen often, but somehow it often doesn't;
2314 instead some weird error message eventually emerges from the
2315 bowels of a directory you didn't know existed.</para>
2317 <para>The purpose of this section is to give you a road-map to
2318 help you figure out what is going right and what is going
2322 <title>Debugging</title>
2324 <para>Debugging <filename>Makefile</filename>s is something of a
2325 black art, but here's a couple of tricks that we find
2326 particularly useful. The following command allows you to see
2327 the contents of any make variable in the context of the current
2328 <filename>Makefile</filename>:</para>
2330 <screen>$ make show VALUE=HS_SRCS</screen>
2332 <para>where you can replace <literal>HS_SRCS</literal> with the
2333 name of any variable you wish to see the value of.</para>
2335 <para>GNU make has a <option>-d</option> option which generates
2336 a dump of the decision procedure used to arrive at a conclusion
2337 about which files should be recompiled. Sometimes useful for
2338 tracking down problems with superfluous or missing
2339 recompilations.</para>
2343 <title>A small project</title>
2345 <para>To get started, let us look at the
2346 <filename>Makefile</filename> for an imaginary small
2347 <literal>fptools</literal> project, <literal>small</literal>.
2348 Each project in <literal>fptools</literal> has its own directory
2349 in <constant>FPTOOLS_TOP</constant>, so the
2350 <literal>small</literal> project will have its own directory
2351 <constant>FPOOLS_TOP/small/</constant>. Inside the
2352 <filename>small/</filename> directory there will be a
2353 <filename>Makefile</filename>, looking something like
2356 <indexterm><primary>Makefile, minimal</primary></indexterm>
2359 # Makefile for fptools project "small"
2362 include $(TOP)/mk/boilerplate.mk
2364 SRCS = $(wildcard *.lhs) $(wildcard *.c)
2367 include $(TOP)/target.mk
2370 <para>this <filename>Makefile</filename> has three
2375 <para>The first section includes
2378 One of the most important
2379 features of GNU <command>make</command> that we use is the ability for a <filename>Makefile</filename> to
2380 include another named file, very like <command>cpp</command>'s <literal>#include</literal>
2385 a file of “boilerplate” code from the level
2386 above (which in this case will be
2387 <filename><constant>FPTOOLS_TOP</constant>/mk/boilerplate.mk</filename><indexterm><primary>boilerplate.mk</primary></indexterm>).
2388 As its name suggests, <filename>boilerplate.mk</filename>
2389 consists of a large quantity of standard
2390 <filename>Makefile</filename> code. We discuss this
2391 boilerplate in more detail in <XRef LinkEnd="sec-boiler">.
2392 <indexterm><primary>include, directive in
2393 Makefiles</primary></indexterm> <indexterm><primary>Makefile
2394 inclusion</primary></indexterm></para>
2396 <para>Before the <literal>include</literal> statement, you
2397 must define the <command>make</command> variable
2398 <constant>TOP</constant><indexterm><primary>TOP</primary></indexterm>
2399 to be the directory containing the <filename>mk</filename>
2400 directory in which the <filename>boilerplate.mk</filename>
2401 file is. It is <emphasis>not</emphasis> OK to simply say</para>
2404 include ../mk/boilerplate.mk # NO NO NO
2408 <para>Why? Because the <filename>boilerplate.mk</filename>
2409 file needs to know where it is, so that it can, in turn,
2410 <literal>include</literal> other files. (Unfortunately,
2411 when an <literal>include</literal>d file does an
2412 <literal>include</literal>, the filename is treated relative
2413 to the directory in which <command>gmake</command> is being
2414 run, not the directory in which the
2415 <literal>include</literal>d sits.) In general,
2416 <emphasis>every file <filename>foo.mk</filename> assumes
2418 <filename><constant>$(TOP)</constant>/mk/foo.mk</filename>
2419 refers to itself.</emphasis> It is up to the
2420 <filename>Makefile</filename> doing the
2421 <literal>include</literal> to ensure this is the case.</para>
2423 <para>Files intended for inclusion in other
2424 <filename>Makefile</filename>s are written to have the
2425 following property: <emphasis>after
2426 <filename>foo.mk</filename> is <literal>include</literal>d,
2427 it leaves <constant>TOP</constant> containing the same value
2428 as it had just before the <literal>include</literal>
2429 statement</emphasis>. In our example, this invariant
2430 guarantees that the <literal>include</literal> for
2431 <filename>target.mk</filename> will look in the same
2432 directory as that for <filename>boilerplate.mk</filename>.</para>
2436 <para> The second section defines the following standard
2437 <command>make</command> variables:
2438 <constant>SRCS</constant><indexterm><primary>SRCS</primary></indexterm>
2439 (the source files from which is to be built), and
2440 <constant>HS_PROG</constant><indexterm><primary>HS_PROG</primary></indexterm>
2441 (the executable binary to be built). We will discuss in
2442 more detail what the “standard variables” are,
2443 and how they affect what happens, in <XRef
2444 LinkEnd="sec-targets">.</para>
2446 <para>The definition for <constant>SRCS</constant> uses the
2447 useful GNU <command>make</command> construct
2448 <literal>$(wildcard $pat$)</literal><indexterm><primary>wildcard</primary></indexterm>,
2449 which expands to a list of all the files matching the
2450 pattern <literal>pat</literal> in the current directory. In
2451 this example, <constant>SRCS</constant> is set to the list
2452 of all the <filename>.lhs</filename> and
2453 <filename>.c</filename> files in the directory. (Let's
2454 suppose there is one of each, <filename>Foo.lhs</filename>
2455 and <filename>Baz.c</filename>.)</para>
2459 <para>The last section includes a second file of standard
2461 <filename>target.mk</filename><indexterm><primary>target.mk</primary></indexterm>.
2462 It contains the rules that tell <command>gmake</command> how
2463 to make the standard targets (<Xref
2464 LinkEnd="sec-standard-targets">). Why, you ask, can't this
2465 standard code be part of
2466 <filename>boilerplate.mk</filename>? Good question. We
2467 discuss the reason later, in <Xref
2468 LinkEnd="sec-boiler-arch">.</para>
2470 <para>You do not <emphasis>have</emphasis> to
2471 <literal>include</literal> the
2472 <filename>target.mk</filename> file. Instead, you can write
2473 rules of your own for all the standard targets. Usually,
2474 though, you will find quite a big payoff from using the
2475 canned rules in <filename>target.mk</filename>; the price
2476 tag is that you have to understand what canned rules get
2477 enabled, and what they do (<Xref
2478 LinkEnd="sec-targets">).</para>
2482 <para>In our example <filename>Makefile</filename>, most of the
2483 work is done by the two <literal>include</literal>d files. When
2484 you say <command>gmake all</command>, the following things
2489 <para><command>gmake</command> figures out that the object
2490 files are <filename>Foo.o</filename> and
2491 <filename>Baz.o</filename>.</para>
2495 <para>It uses a boilerplate pattern rule to compile
2496 <filename>Foo.lhs</filename> to <filename>Foo.o</filename>
2497 using a Haskell compiler. (Which one? That is set in the
2498 build configuration.)</para>
2502 <para>It uses another standard pattern rule to compile
2503 <filename>Baz.c</filename> to <filename>Baz.o</filename>,
2504 using a C compiler. (Ditto.)</para>
2508 <para>It links the resulting <filename>.o</filename> files
2509 together to make <literal>small</literal>, using the Haskell
2510 compiler to do the link step. (Why not use
2511 <command>ld</command>? Because the Haskell compiler knows
2512 what standard libraries to link in. How did
2513 <command>gmake</command> know to use the Haskell compiler to
2514 do the link, rather than the C compiler? Because we set the
2515 variable <constant>HS_PROG</constant> rather than
2516 <constant>C_PROG</constant>.)</para>
2520 <para>All <filename>Makefile</filename>s should follow the above
2521 three-section format.</para>
2525 <title>A larger project</title>
2527 <para>Larger projects are usually structured into a number of
2528 sub-directories, each of which has its own
2529 <filename>Makefile</filename>. (In very large projects, this
2530 sub-structure might be iterated recursively, though that is
2531 rare.) To give you the idea, here's part of the directory
2532 structure for the (rather large) GHC project:</para>
2542 ...source files for documentation...
2545 ...source files for driver...
2548 parser/...source files for parser...
2549 renamer/...source files for renamer...
2553 <para>The sub-directories <filename>docs</filename>,
2554 <filename>driver</filename>, <filename>compiler</filename>, and
2555 so on, each contains a sub-component of GHC, and each has its
2556 own <filename>Makefile</filename>. There must also be a
2557 <filename>Makefile</filename> in
2558 <filename><constant>$(FPTOOLS_TOP)</constant>/ghc</filename>.
2559 It does most of its work by recursively invoking
2560 <command>gmake</command> on the <filename>Makefile</filename>s
2561 in the sub-directories. We say that
2562 <filename>ghc/Makefile</filename> is a <emphasis>non-leaf
2563 <filename>Makefile</filename></emphasis>, because it does little
2564 except organise its children, while the
2565 <filename>Makefile</filename>s in the sub-directories are all
2566 <emphasis>leaf <filename>Makefile</filename>s</emphasis>. (In
2567 principle the sub-directories might themselves contain a
2568 non-leaf <filename>Makefile</filename> and several
2569 sub-sub-directories, but that does not happen in GHC.)</para>
2571 <para>The <filename>Makefile</filename> in
2572 <filename>ghc/compiler</filename> is considered a leaf
2573 <filename>Makefile</filename> even though the
2574 <filename>ghc/compiler</filename> has sub-directories, because
2575 these sub-directories do not themselves have
2576 <filename>Makefile</filename>s in them. They are just used to
2577 structure the collection of modules that make up GHC, but all
2578 are managed by the single <filename>Makefile</filename> in
2579 <filename>ghc/compiler</filename>.</para>
2581 <para>You will notice that <filename>ghc/</filename> also
2582 contains a directory <filename>ghc/mk/</filename>. It contains
2583 GHC-specific <filename>Makefile</filename> boilerplate code.
2584 More precisely:</para>
2588 <para><filename>ghc/mk/boilerplate.mk</filename> is included
2589 at the top of <filename>ghc/Makefile</filename>, and of all
2590 the leaf <filename>Makefile</filename>s in the
2591 sub-directories. It in turn <literal>include</literal>s the
2592 main boilerplate file
2593 <filename>mk/boilerplate.mk</filename>.</para>
2597 <para><filename>ghc/mk/target.mk</filename> is
2598 <literal>include</literal>d at the bottom of
2599 <filename>ghc/Makefile</filename>, and of all the leaf
2600 <filename>Makefile</filename>s in the sub-directories. It
2601 in turn <literal>include</literal>s the file
2602 <filename>mk/target.mk</filename>.</para>
2606 <para>So these two files are the place to look for GHC-wide
2607 customisation of the standard boilerplate.</para>
2610 <sect2 id="sec-boiler-arch">
2611 <title>Boilerplate architecture</title>
2612 <indexterm><primary>boilerplate architecture</primary></indexterm>
2614 <para>Every <filename>Makefile</filename> includes a
2615 <filename>boilerplate.mk</filename><indexterm><primary>boilerplate.mk</primary></indexterm>
2616 file at the top, and
2617 <filename>target.mk</filename><indexterm><primary>target.mk</primary></indexterm>
2618 file at the bottom. In this section we discuss what is in these
2619 files, and why there have to be two of them. In general:</para>
2623 <para><filename>boilerplate.mk</filename> consists of:</para>
2627 <para><emphasis>Definitions of millions of
2628 <command>make</command> variables</emphasis> that
2629 collectively specify the build configuration. Examples:
2630 <constant>HC_OPTS</constant><indexterm><primary>HC_OPTS</primary></indexterm>,
2631 the options to feed to the Haskell compiler;
2632 <constant>NoFibSubDirs</constant><indexterm><primary>NoFibSubDirs</primary></indexterm>,
2633 the sub-directories to enable within the
2634 <literal>nofib</literal> project;
2635 <constant>GhcWithHc</constant><indexterm><primary>GhcWithHc</primary></indexterm>,
2636 the name of the Haskell compiler to use when compiling
2637 GHC in the <literal>ghc</literal> project.</para>
2641 <para><emphasis>Standard pattern rules</emphasis> that
2642 tell <command>gmake</command> how to construct one file
2643 from another.</para>
2647 <para><filename>boilerplate.mk</filename> needs to be
2648 <literal>include</literal>d at the <emphasis>top</emphasis>
2649 of each <filename>Makefile</filename>, so that the user can
2650 replace the boilerplate definitions or pattern rules by
2651 simply giving a new definition or pattern rule in the
2652 <filename>Makefile</filename>. <command>gmake</command>
2653 simply takes the last definition as the definitive one.</para>
2655 <para>Instead of <emphasis>replacing</emphasis> boilerplate
2656 definitions, it is also quite common to
2657 <emphasis>augment</emphasis> them. For example, a
2658 <filename>Makefile</filename> might say:</para>
2664 <para>thereby adding “<option>-O</option>” to
2666 <constant>SRC_HC_OPTS</constant><indexterm><primary>SRC_HC_OPTS</primary></indexterm>.</para>
2670 <para><filename>target.mk</filename> contains
2671 <command>make</command> rules for the standard targets
2672 described in <Xref LinkEnd="sec-standard-targets">. These
2673 rules are selectively included, depending on the setting of
2674 certain <command>make</command> variables. These variables
2675 are usually set in the middle section of the
2676 <filename>Makefile</filename> between the two
2677 <literal>include</literal>s.</para>
2679 <para><filename>target.mk</filename> must be included at the
2680 end (rather than being part of
2681 <filename>boilerplate.mk</filename>) for several tiresome
2687 <para><command>gmake</command> commits target and
2688 dependency lists earlier than it should. For example,
2689 <FIlename>target.mk</FIlename> has a rule that looks
2693 $(HS_PROG) : $(OBJS)
2694 $(HC) $(LD_OPTS) $< -o $@
2697 <para>If this rule was in
2698 <filename>boilerplate.mk</filename> then
2699 <constant>$(HS_PROG)</constant><indexterm><primary>HS_PROG</primary></indexterm>
2701 <constant>$(OBJS)</constant><indexterm><primary>OBJS</primary></indexterm>
2702 would not have their final values at the moment
2703 <command>gmake</command> encountered the rule. Alas,
2704 <command>gmake</command> takes a snapshot of their
2705 current values, and wires that snapshot into the rule.
2706 (In contrast, the commands executed when the rule
2707 “fires” are only substituted at the moment
2708 of firing.) So, the rule must follow the definitions
2709 given in the <filename>Makefile</filename> itself.</para>
2713 <para>Unlike pattern rules, ordinary rules cannot be
2714 overriden or replaced by subsequent rules for the same
2715 target (at least, not without an error message).
2716 Including ordinary rules in
2717 <filename>boilerplate.mk</filename> would prevent the
2718 user from writing rules for specific targets in specific
2723 <para>There are a couple of other reasons I've
2724 forgotten, but it doesn't matter too much.</para>
2731 <sect2 id="sec-boiler">
2732 <title>The main <filename>mk/boilerplate.mk</filename> file</title>
2733 <indexterm><primary>boilerplate.mk</primary></indexterm>
2735 <para>If you look at
2736 <filename><constant>$(FPTOOLS_TOP)</constant>/mk/boilerplate.mk</filename>
2737 you will find that it consists of the following sections, each
2738 held in a separate file:</para>
2742 <term><filename>config.mk</filename></term>
2743 <indexterm><primary>config.mk</primary></indexterm>
2745 <para>is the build configuration file we discussed at
2746 length in <Xref LinkEnd="sec-build-config">.</para>
2751 <term><filename>paths.mk</filename></term>
2752 <indexterm><primary>paths.mk</primary></indexterm>
2754 <para>defines <command>make</command> variables for
2755 pathnames and file lists. This file contains code for
2756 automatically compiling lists of source files and deriving
2757 lists of object files from those. The results can be
2758 overriden in the <filename>Makefile</filename>, but in
2759 most cases the automatic setup should do the right
2762 <para>The following variables may be set in the
2763 <filename>Makefile</filename> to affect how the automatic
2764 source file search is done:</para>
2768 <term><literal>ALL_DIRS</literal></term>
2769 <indexterm><primary><literal>ALL_DIRS</literal></primary>
2772 <para>Set to a list of directories to search in
2773 addition to the current directory for source
2779 <term><literal>EXCLUDE_SRCS</literal></term>
2780 <indexterm><primary><literal>EXCLUDE_SRCS</literal></primary>
2783 <para>Set to a list of source files (relative to the
2784 current directory) to omit from the automatic
2785 search. The source searching machinery is clever
2786 enough to know that if you exclude a source file
2787 from which other sources are derived, then the
2788 derived sources should also be excluded. For
2789 example, if you set <literal>EXCLUDED_SRCS</literal>
2790 to include <filename>Foo.y</filename>, then
2791 <filename>Foo.hs</filename> will also be
2797 <term><literal>EXTRA_SRCS</literal></term>
2798 <indexterm><primary><literal>EXCLUDE_SRCS</literal></primary>
2801 <para>Set to a list of extra source files (perhaps
2802 in directories not listed in
2803 <literal>ALL_DIRS</literal>) that should be
2809 <para>The results of the automatic source file search are
2810 placed in the following make variables:</para>
2814 <term><literal>SRCS</literal></term>
2815 <indexterm><primary><literal>SRCS</literal></primary></indexterm>
2817 <para>All source files found, sorted and without
2818 duplicates, including those which might not exist
2819 yet but will be derived from other existing sources.
2820 <literal>SRCS</literal> <emphasis>can</emphasis> be
2821 overriden if necessary, in which case the variables
2822 below will follow suit.</para>
2827 <term><literal>HS_SRCS</literal></term>
2828 <indexterm><primary><literal>HS_SRCS</literal></primary></indexterm>
2830 <para>all Haskell source files in the current
2831 directory, including those derived from other source
2832 files (eg. Happy sources also give rise to Haskell
2838 <term><literal>HS_OBJS</literal></term>
2839 <indexterm><primary><literal>HS_OBJS</literal></primary></indexterm>
2841 <para>Object files derived from
2842 <literal>HS_SRCS</literal>.</para>
2847 <term><literal>HS_IFACES</literal></term>
2848 <indexterm><primary><literal>HS_IFACES</literal></primary></indexterm>
2850 <para>Interface files (<literal>.hi</literal> files)
2851 derived from <literal>HS_SRCS</literal>.</para>
2856 <term><literal>C_SRCS</literal></term>
2857 <indexterm><primary><literal>C_SRCS</literal></primary></indexterm>
2859 <para>All C source files found.</para>
2864 <term><literal>C_OBJS</literal></term>
2865 <indexterm><primary><literal>C_OBJS</literal></primary></indexterm>
2867 <para>Object files derived from
2868 <literal>C_SRCS</literal>.</para>
2873 <term><literal>SCRIPT_SRCS</literal></term>
2874 <indexterm><primary><literal>SCRIPT_SRCS</literal></primary></indexterm>
2876 <para>All script source files found
2877 (<literal>.lprl</literal> files).</para>
2882 <term><literal>SCRIPT_OBJS</literal></term>
2883 <indexterm><primary><literal>SCRIPT_OBJS</literal></primary></indexterm>
2885 <para><quote>object</quote> files derived from
2886 <literal>SCRIPT_SRCS</literal>
2887 (<literal>.prl</literal> files).</para>
2892 <term><literal>HSC_SRCS</literal></term>
2893 <indexterm><primary><literal>HSC_SRCS</literal></primary></indexterm>
2895 <para>All <literal>hsc2hs</literal> source files
2896 (<literal>.hsc</literal> files).</para>
2901 <term><literal>HAPPY_SRCS</literal></term>
2902 <indexterm><primary><literal>HAPPY_SRCS</literal></primary></indexterm>
2904 <para>All <literal>happy</literal> source files
2905 (<literal>.y</literal> or <literal>.hy</literal> files).</para>
2910 <term><literal>OBJS</literal></term>
2911 <indexterm><primary>OBJS</primary></indexterm>
2913 <para>the concatenation of
2914 <literal>$(HS_OBJS)</literal>,
2915 <literal>$(C_OBJS)</literal>, and
2916 <literal>$(SCRIPT_OBJS)</literal>.</para>
2921 <para>Any or all of these definitions can easily be
2922 overriden by giving new definitions in your
2923 <filename>Makefile</filename>.</para>
2925 <para>What, exactly, does <filename>paths.mk</filename>
2926 consider a <quote>source file</quote> to be? It's based
2927 on the file's suffix (e.g. <filename>.hs</filename>,
2928 <filename>.lhs</filename>, <filename>.c</filename>,
2929 <filename>.hy</filename>, etc), but this is the kind of
2930 detail that changes, so rather than enumerate the source
2931 suffices here the best thing to do is to look in
2932 <filename>paths.mk</filename>.</para>
2937 <term><filename>opts.mk</filename></term>
2938 <indexterm><primary>opts.mk</primary></indexterm>
2940 <para>defines <command>make</command> variables for option
2941 strings to pass to each program. For example, it defines
2942 <constant>HC_OPTS</constant><indexterm><primary>HC_OPTS</primary></indexterm>,
2943 the option strings to pass to the Haskell compiler. See
2944 <Xref LinkEnd="sec-suffix">.</para>
2949 <term><filename>suffix.mk</filename></term>
2950 <indexterm><primary>suffix.mk</primary></indexterm>
2952 <para>defines standard pattern rules—see <Xref
2953 LinkEnd="sec-suffix">.</para>
2958 <para>Any of the variables and pattern rules defined by the
2959 boilerplate file can easily be overridden in any particular
2960 <filename>Makefile</filename>, because the boilerplate
2961 <literal>include</literal> comes first. Definitions after this
2962 <literal>include</literal> directive simply override the default
2963 ones in <filename>boilerplate.mk</filename>.</para>
2966 <sect2 id="sec-suffix">
2967 <title>Pattern rules and options</title>
2968 <indexterm><primary>Pattern rules</primary></indexterm>
2971 <filename>suffix.mk</filename><indexterm><primary>suffix.mk</primary></indexterm>
2972 defines standard <emphasis>pattern rules</emphasis> that say how
2973 to build one kind of file from another, for example, how to
2974 build a <filename>.o</filename> file from a
2975 <filename>.c</filename> file. (GNU <command>make</command>'s
2976 <emphasis>pattern rules</emphasis> are more powerful and easier
2977 to use than Unix <command>make</command>'s <emphasis>suffix
2978 rules</emphasis>.)</para>
2980 <para>Almost all the rules look something like this:</para>
2985 $(CC) $(CC_OPTS) -c $< -o $@
2988 <para>Here's how to understand the rule. It says that
2989 <emphasis>something</emphasis><filename>.o</filename> (say
2990 <filename>Foo.o</filename>) can be built from
2991 <emphasis>something</emphasis><filename>.c</filename>
2992 (<filename>Foo.c</filename>), by invoking the C compiler (path
2993 name held in <constant>$(CC)</constant>), passing to it
2994 the options <constant>$(CC_OPTS)</constant> and
2995 the rule's dependent file of the rule
2996 <literal>$<</literal> (<filename>Foo.c</filename> in
2997 this case), and putting the result in the rule's target
2998 <literal>$@</literal> (<filename>Foo.o</filename> in this
3001 <para>Every program is held in a <command>make</command>
3002 variable defined in <filename>mk/config.mk</filename>—look
3003 in <filename>mk/config.mk</filename> for the complete list. One
3004 important one is the Haskell compiler, which is called
3005 <constant>$(HC)</constant>.</para>
3007 <para>Every program's options are are held in a
3008 <command>make</command> variables called
3009 <constant><prog>_OPTS</constant>. the
3010 <constant><prog>_OPTS</constant> variables are
3011 defined in <filename>mk/opts.mk</filename>. Almost all of them
3012 are defined like this:</para>
3015 CC_OPTS = $(SRC_CC_OPTS) $(WAY$(_way)_CC_OPTS) $($*_CC_OPTS) $(EXTRA_CC_OPTS)
3018 <para>The four variables from which
3019 <constant>CC_OPTS</constant> is built have the following
3024 <term><constant>SRC_CC_OPTS</constant><indexterm><primary>SRC_CC_OPTS</primary></indexterm>:</term>
3026 <para>options passed to all C compilations.</para>
3031 <term><constant>WAY_<way>_CC_OPTS</constant>:</term>
3033 <para>options passed to C compilations for way
3034 <literal><way></literal>. For example,
3035 <constant>WAY_mp_CC_OPTS</constant>
3036 gives options to pass to the C compiler when compiling way
3037 <literal>mp</literal>. The variable
3038 <constant>WAY_CC_OPTS</constant> holds
3039 options to pass to the C compiler when compiling the
3040 standard way. (<Xref LinkEnd="sec-ways"> dicusses
3041 multi-way compilation.)</para>
3046 <term><constant><module>_CC_OPTS</constant>:</term>
3048 <para>options to pass to the C compiler that are specific
3049 to module <literal><module></literal>. For example,
3050 <constant>SMap_CC_OPTS</constant> gives the
3051 specific options to pass to the C compiler when compiling
3052 <filename>SMap.c</filename>.</para>
3057 <term><constant>EXTRA_CC_OPTS</constant><indexterm><primary>EXTRA_CC_OPTS</primary></indexterm>:</term>
3059 <para>extra options to pass to all C compilations. This
3060 is intended for command line use, thus:</para>
3063 gmake libHS.a EXTRA_CC_OPTS="-v"
3070 <sect2 id="sec-targets">
3071 <title>The main <filename>mk/target.mk</filename> file</title>
3072 <indexterm><primary>target.mk</primary></indexterm>
3074 <para><filename>target.mk</filename> contains canned rules for
3075 all the standard targets described in <Xref
3076 LinkEnd="sec-standard-targets">. It is complicated by the fact
3077 that you don't want all of these rules to be active in every
3078 <filename>Makefile</filename>. Rather than have a plethora of
3079 tiny files which you can include selectively, there is a single
3080 file, <filename>target.mk</filename>, which selectively includes
3081 rules based on whether you have defined certain variables in
3082 your <filename>Makefile</filename>. This section explains what
3083 rules you get, what variables control them, and what the rules
3084 do. Hopefully, you will also get enough of an idea of what is
3085 supposed to happen that you can read and understand any weird
3086 special cases yourself.</para>
3090 <term><constant>HS_PROG</constant><indexterm><primary>HS_PROG</primary></indexterm>.</term>
3092 <para>If <constant>HS_PROG</constant> is defined,
3093 you get rules with the following targets:</para>
3097 <term><filename>HS_PROG</filename><indexterm><primary>HS_PROG</primary></indexterm></term>
3099 <para>itself. This rule links
3100 <constant>$(OBJS)</constant> with the Haskell
3101 runtime system to get an executable called
3102 <constant>$(HS_PROG)</constant>.</para>
3107 <term><literal>install</literal><indexterm><primary>install</primary></indexterm></term>
3110 <constant>$(HS_PROG)</constant> in
3111 <constant>$(bindir)</constant>.</para>
3120 <term><constant>C_PROG</constant><indexterm><primary>C_PROG</primary></indexterm></term>
3122 <para>is similar to <constant>HS_PROG</constant>,
3123 except that the link step links
3124 <constant>$(C_OBJS)</constant> with the C
3125 runtime system.</para>
3130 <term><constant>LIBRARY</constant><indexterm><primary>LIBRARY</primary></indexterm></term>
3132 <para>is similar to <constant>HS_PROG</constant>,
3133 except that it links
3134 <constant>$(LIB_OBJS)</constant> to make the
3135 library archive <constant>$(LIBRARY)</constant>,
3136 and <literal>install</literal> installs it in
3137 <constant>$(libdir)</constant>.</para>
3142 <term><constant>LIB_DATA</constant><indexterm><primary>LIB_DATA</primary></indexterm></term>
3144 <para>…</para>
3149 <term><constant>LIB_EXEC</constant><indexterm><primary>LIB_EXEC</primary></indexterm></term>
3151 <para>…</para>
3156 <term><constant>HS_SRCS</constant><indexterm><primary>HS_SRCS</primary></indexterm>, <constant>C_SRCS</constant><indexterm><primary>C_SRCS</primary></indexterm>.</term>
3158 <para>If <constant>HS_SRCS</constant> is defined
3159 and non-empty, a rule for the target
3160 <literal>depend</literal> is included, which generates
3161 dependency information for Haskell programs. Similarly
3162 for <constant>C_SRCS</constant>.</para>
3167 <para>All of these rules are “double-colon” rules,
3171 install :: $(HS_PROG)
3172 ...how to install it...
3175 <para>GNU <command>make</command> treats double-colon rules as
3176 separate entities. If there are several double-colon rules for
3177 the same target it takes each in turn and fires it if its
3178 dependencies say to do so. This means that you can, for
3179 example, define both <constant>HS_PROG</constant> and
3180 <constant>LIBRARY</constant>, which will generate two rules for
3181 <literal>install</literal>. When you type <command>gmake
3182 install</command> both rules will be fired, and both the program
3183 and the library will be installed, just as you wanted.</para>
3186 <sect2 id="sec-subdirs">
3187 <title>Recursion</title>
3188 <indexterm><primary>recursion, in makefiles</primary></indexterm>
3189 <indexterm><primary>Makefile, recursing into subdirectories</primary></indexterm>
3191 <para>In leaf <filename>Makefile</filename>s the variable
3192 <constant>SUBDIRS</constant><indexterm><primary>SUBDIRS</primary></indexterm>
3193 is undefined. In non-leaf <filename>Makefile</filename>s,
3194 <constant>SUBDIRS</constant> is set to the list of
3195 sub-directories that contain subordinate
3196 <filename>Makefile</filename>s. <emphasis>It is up to you to
3197 set <constant>SUBDIRS</constant> in the
3198 <filename>Makefile</filename>.</emphasis> There is no automation
3199 here—<constant>SUBDIRS</constant> is too important to
3202 <para>When <constant>SUBDIRS</constant> is defined,
3203 <filename>target.mk</filename> includes a rather neat rule for
3204 the standard targets (<Xref LinkEnd="sec-standard-targets"> that
3205 simply invokes <command>make</command> recursively in each of
3206 the sub-directories.</para>
3208 <para><emphasis>These recursive invocations are guaranteed to
3209 occur in the order in which the list of directories is specified
3210 in <constant>SUBDIRS</constant>. </emphasis>This guarantee can
3211 be important. For example, when you say <command>gmake
3212 boot</command> it can be important that the recursive invocation
3213 of <command>make boot</command> is done in one sub-directory
3214 (the include files, say) before another (the source files).
3215 Generally, put the most independent sub-directory first, and the
3216 most dependent last.</para>
3219 <sect2 id="sec-ways">
3220 <title>Way management</title>
3221 <indexterm><primary>way management</primary></indexterm>
3223 <para>We sometimes want to build essentially the same system in
3224 several different “ways”. For example, we want to build GHC's
3225 <literal>Prelude</literal> libraries with and without profiling,
3226 so that there is an appropriately-built library archive to link
3227 with when the user compiles his program. It would be possible
3228 to have a completely separate build tree for each such “way”,
3229 but it would be horribly bureaucratic, especially since often
3230 only parts of the build tree need to be constructed in multiple
3234 <filename>target.mk</filename><indexterm><primary>target.mk</primary></indexterm>
3235 contains some clever magic to allow you to build several
3236 versions of a system; and to control locally how many versions
3237 are built and how they differ. This section explains the
3240 <para>The files for a particular way are distinguished by
3241 munging the suffix. The <quote>normal way</quote> is always
3242 built, and its files have the standard suffices
3243 <filename>.o</filename>, <filename>.hi</filename>, and so on.
3244 In addition, you can build one or more extra ways, each
3245 distinguished by a <emphasis>way tag</emphasis>. The object
3246 files and interface files for one of these extra ways are
3247 distinguished by their suffix. For example, way
3248 <literal>mp</literal> has files
3249 <filename>.mp_o</filename> and
3250 <filename>.mp_hi</filename>. Library archives have their
3251 way tag the other side of the dot, for boring reasons; thus,
3252 <filename>libHS_mp.a</filename>.</para>
3254 <para>A <command>make</command> variable called
3255 <constant>way</constant> holds the current way tag.
3256 <emphasis><constant>way</constant> is only ever set on the
3257 command line of <command>gmake</command></emphasis> (usually in
3258 a recursive invocation of <command>gmake</command> by the
3259 system). It is never set inside a
3260 <filename>Makefile</filename>. So it is a global constant for
3261 any one invocation of <command>gmake</command>. Two other
3262 <command>make</command> variables,
3263 <constant>way_</constant> and
3264 <constant>_way</constant> are immediately derived from
3265 <constant>$(way)</constant> and never altered. If
3266 <constant>way</constant> is not set, then neither are
3267 <constant>way_</constant> and
3268 <constant>_way</constant>, and the invocation of
3269 <command>make</command> will build the <quote>normal
3270 way</quote>. If <constant>way</constant> is set, then the other
3271 two variables are set in sympathy. For example, if
3272 <constant>$(way)</constant> is “<literal>mp</literal>”,
3273 then <constant>way_</constant> is set to
3274 “<literal>mp_</literal>” and
3275 <constant>_way</constant> is set to
3276 “<literal>_mp</literal>”. These three variables are
3277 then used when constructing file names.</para>
3279 <para>So how does <command>make</command> ever get recursively
3280 invoked with <constant>way</constant> set? There are two ways
3281 in which this happens:</para>
3285 <para>For some (but not all) of the standard targets, when
3286 in a leaf sub-directory, <command>make</command> is
3287 recursively invoked for each way tag in
3288 <constant>$(WAYS)</constant>. You set
3289 <constant>WAYS</constant> in the
3290 <filename>Makefile</filename> to the list of way tags you
3291 want these targets built for. The mechanism here is very
3292 much like the recursive invocation of
3293 <command>make</command> in sub-directories (<Xref
3294 LinkEnd="sec-subdirs">). It is up to you to set
3295 <constant>WAYS</constant> in your
3296 <filename>Makefile</filename>; this is how you control what
3297 ways will get built.</para>
3301 <para>For a useful collection of targets (such as
3302 <filename>libHS_mp.a</filename>,
3303 <filename>Foo.mp_o</filename>) there is a rule which
3304 recursively invokes <command>make</command> to make the
3305 specified target, setting the <constant>way</constant>
3306 variable. So if you say <command>gmake
3307 Foo.mp_o</command> you should see a recursive
3308 invocation <command>gmake Foo.mp_o way=mp</command>,
3309 and <emphasis>in this recursive invocation the pattern rule
3310 for compiling a Haskell file into a <filename>.o</filename>
3311 file will match</emphasis>. The key pattern rules (in
3312 <filename>suffix.mk</filename>) look like this:
3316 $(HC) $(HC_OPTS) $< -o $@
3323 <para>You can invoke <command>make</command> with a
3324 particular <literal>way</literal> setting yourself, in order
3325 to build files related to a particular
3326 <literal>way</literal> in the current directory. eg.
3332 will build files for the profiling way only in the current
3339 <title>When the canned rule isn't right</title>
3341 <para>Sometimes the canned rule just doesn't do the right thing.
3342 For example, in the <literal>nofib</literal> suite we want the
3343 link step to print out timing information. The thing to do here
3344 is <emphasis>not</emphasis> to define
3345 <constant>HS_PROG</constant> or
3346 <constant>C_PROG</constant>, and instead define a special
3347 purpose rule in your own <filename>Makefile</filename>. By
3348 using different variable names you will avoid the canned rules
3349 being included, and conflicting with yours.</para>
3353 <sect1 id="building-docs">
3354 <title>Building the documentation</title>
3356 <sect2 id="pre-supposed-doc-tools">
3357 <title>Tools for building the Documentation</title>
3359 <para>The following additional tools are required if you want to
3360 format the documentation that comes with the
3361 <literal>fptools</literal> projects:</para>
3365 <term>DocBook</term>
3366 <indexterm><primary>pre-supposed: DocBook</primary></indexterm>
3367 <indexterm><primary>DocBook, pre-supposed</primary></indexterm>
3369 <para>Much of our documentation is written in SGML, using
3370 the DocBook DTD. Instructions on installing and
3371 configuring the DocBook tools are below.</para>
3377 <indexterm><primary>pre-supposed: TeX</primary></indexterm>
3378 <indexterm><primary>TeX, pre-supposed</primary></indexterm>
3380 <para>A decent TeX distribution is required if you want to
3381 produce printable documentation. We recomment teTeX,
3382 which includes just about everything you need.</para>
3387 <term>Haddock</term>
3388 <indexterm><primary>Haddock</primary>
3391 <para>Haddock is a Haskell documentation tool that we use
3392 for automatically generating documentation from the
3393 library source code. It is an <literal>fptools</literal>
3394 project in itself. To build documentation for the
3395 libraries (<literal>fptools/libraries</literal>) you
3396 should check out and build Haddock in
3397 <literal>fptools/haddock</literal>. Haddock requires GHC
3405 <title>Installing the DocBook tools</title>
3408 <title>Installing the DocBook tools on Linux</title>
3410 <para>If you're on a recent RedHat system (7.0+), you probably
3411 have working DocBook tools already installed. The configure
3412 script should detect your setup and you're away.</para>
3414 <para>If you don't have DocBook tools installed, and you are
3415 using a system that can handle RedHat RPM packages, you can
3416 probably use the <ULink
3417 URL="http://sourceware.cygnus.com/docbook-tools/">Cygnus
3418 DocBook tools</ULink>, which is the most shrink-wrapped SGML
3419 suite that we could find. You need all the RPMs except for
3420 psgml (i.e. <Filename>docbook</Filename>,
3421 <Filename>jade</Filename>, <Filename>jadetex</Filename>,
3422 <Filename>sgmlcommon</Filename> and
3423 <Filename>stylesheets</Filename>). Note that most of these
3424 RPMs are architecture neutral, so are likely to be found in a
3425 <Filename>noarch</Filename> directory. The SuSE RPMs also
3426 work; the RedHat ones <Emphasis>don't</Emphasis> in RedHat 6.2
3427 (7.0 and later should be OK), but they are easy to fix: just
3429 <Filename>/usr/lib/sgml/stylesheets/nwalsh-modular/lib/dblib.dsl</Filename>
3430 to <Filename>/usr/lib/sgml/lib/dblib.dsl</Filename>. </para>
3434 <title>Installing DocBook on FreeBSD</title>
3436 <para>On FreeBSD systems, the easiest way to get DocBook up
3437 and running is to install it from the ports tree or a
3438 pre-compiled package (packages are available from your local
3439 FreeBSD mirror site).</para>
3441 <para>To use the ports tree, do this:
3443 $ cd /usr/ports/textproc/docproj
3446 This installs the FreeBSD documentation project tools, which
3447 includes everything needed to format the GHC
3448 documentation.</para>
3452 <title>Installing from binaries on Windows</title>
3454 <Para>It's a good idea to use Norman Walsh's <ULink
3455 URL="http://nwalsh.com/docbook/dsssl/doc/install.html">installation
3456 notes</ULink> as a guide. You should get version 3.1 of
3457 DocBook, and note that his file <Filename>test.sgm</Filename>
3458 won't work, as it needs version 3.0. You should unpack Jade
3459 into <Filename>\Jade</Filename>, along with the entities,
3460 DocBook into <Filename>\docbook</Filename>, and the DocBook
3461 stylesheets into <Filename>\docbook\stylesheets</Filename> (so
3462 they actually end up in
3463 <Filename>\docbook\stylesheets\docbook</Filename>).</para>
3468 <title>Installing the DocBook tools from source</title>
3473 <para>Install <ULink
3474 URL="http://openjade.sourceforge.net/">OpenJade</ULink>
3475 (Windows binaries are available as well as sources). If you
3476 want DVI, PS, or PDF then install JadeTeX from the
3477 <Filename>dsssl</Filename> subdirectory. (If you get the
3481 ! LaTeX Error: Unknown option implicit=false' for package hyperref'.
3484 your version of <Command>hyperref</Command> is out of date;
3485 download it from CTAN
3486 (<Filename>macros/latex/contrib/supported/hyperref</Filename>),
3487 and make it, ensuring that you have first removed or renamed
3488 your old copy. If you start getting file not found errors
3489 when making the test for <Command>hyperref</Command>, you
3490 can abort at that point and proceed straight to
3491 <Command>make install</Command>, or enter them as
3492 <Filename>../</Filename><Emphasis>filename</Emphasis>.)</para>
3494 <para>Make links from <Filename>virtex</Filename> to
3495 <Filename>jadetex</Filename> and
3496 <Filename>pdfvirtex</Filename> to
3497 <Filename>pdfjadetex</Filename> (otherwise DVI, PostScript
3498 and PDF output will not work). Copy
3499 <Filename>dsssl/*.{dtd,dsl}</Filename> and
3500 <Filename>catalog</Filename> to
3501 <Filename>/usr/[local/]lib/sgml</Filename>.</para>
3505 <title>DocBook and the DocBook stylesheets</title>
3507 <para>Get a Zip of <ULink
3508 URL="http://www.oasis-open.org/docbook/sgml/3.1/index.html">DocBook</ULink>
3509 and install the contents in
3510 <Filename>/usr/[local/]/lib/sgml</Filename>.</para>
3512 <para>Get the <ULink
3513 URL="http://nwalsh.com/docbook/dsssl/">DocBook
3514 stylesheets</ULink> and install in
3515 <Filename>/usr/[local/]lib/sgml/stylesheets</Filename>
3516 (thereby creating a subdirectory docbook). For indexing,
3517 copy or link <Filename>collateindex.pl</Filename> from the
3518 DocBook stylesheets archive in <Filename>bin</Filename> into
3519 a directory on your <Constant>PATH</Constant>.</para>
3521 <para>Download the <ULink
3522 URL="http://www.oasis-open.org/cover/ISOEnts.zip">ISO
3523 entities</ULink> into
3524 <Filename>/usr/[local/]lib/sgml</Filename>.</para>
3530 <title>Configuring the DocBook tools</title>
3532 <Para>Once the DocBook tools are installed, the configure script
3533 will detect them and set up the build system accordingly. If you
3534 have a system that isn't supported, let us know, and we'll try
3539 <title>Remaining problems</title>
3541 <para>If you install from source, you'll get a pile of warnings
3544 <Screen>DTDDECL catalog entries are not supported</Screen>
3546 every time you build anything. These can safely be ignored, but
3547 if you find them tedious you can get rid of them by removing all
3548 the <Constant>DTDDECL</Constant> entries from
3549 <Filename>docbook.cat</Filename>.</para>
3553 <title>Building the documentation</title>
3555 <para>To build documentation in a certain format, you can
3556 say, for example,</para>
3562 <para>to build HTML documentation below the current directory.
3563 The available formats are: <literal>dvi</literal>,
3564 <literal>ps</literal>, <literal>pdf</literal>,
3565 <literal>html</literal>, and <literal>rtf</literal>. Note that
3566 not all documentation can be built in all of these formats: HTML
3567 documentation is generally supported everywhere, and DocBook
3568 documentation might support the other formats (depending on what
3569 other tools you have installed).</para>
3571 <para>All of these targets are recursive; that is, saying
3572 <literal>make html</literal> will make HTML docs for all the
3573 documents recursively below the current directory.</para>
3575 <para>Because there are many different formats that the DocBook
3576 documentation can be generated in, you have to select which ones
3577 you want by setting the <literal>SGMLDocWays</literal> variable
3578 to a list of them. For example, in
3579 <filename>build.mk</filename> you might have a line:</para>
3582 SGMLDocWays = html ps
3585 <para>This will cause the documentation to be built in the requested
3586 formats as part of the main build (the default is not to build
3587 any documentation at all).</para>
3591 <title>Installing the documentation</title>
3593 <para>To install the documentation, use:</para>
3599 <para>This will install the documentation into
3600 <literal>$(datadir)</literal> (which defaults to
3601 <literal>$(prefix)/share</literal>). The exception is HTML
3602 documentation, which goes into
3603 <literal>$(datadir)/html</literal>, to keep things tidy.</para>
3605 <para>Note that unless you set <literal>$(SGMLDocWays)</literal>
3606 to a list of formats, the <literal>install-docs</literal> target
3607 won't do anything for SGML documentation.</para>
3613 <sect1 id="sec-porting-ghc">
3614 <title>Porting GHC</title>
3616 <para>This section describes how to port GHC to a currenly
3617 unsupported platform. There are two distinct
3618 possibilities:</para>
3622 <para>The hardware architecture for your system is already
3623 supported by GHC, but you're running an OS that isn't
3624 supported (or perhaps has been supported in the past, but
3625 currently isn't). This is the easiest type of porting job,
3626 but it still requires some careful bootstrapping. Proceed to
3627 <xref linkend="sec-booting-from-hc">.</para>
3631 <para>Your system's hardware architecture isn't supported by
3632 GHC. This will be a more difficult port (though by comparison
3633 perhaps not as difficult as porting gcc). Proceed to <xref
3634 linkend="unregisterised-porting">.</para>
3638 <sect2 id="sec-booting-from-hc">
3639 <title>Booting/porting from C (<filename>.hc</filename>) files</title>
3641 <indexterm><primary>building GHC from .hc files</primary></indexterm>
3642 <indexterm><primary>booting GHC from .hc files</primary></indexterm>
3643 <indexterm><primary>porting GHC</primary></indexterm>
3645 <para>Bootstrapping GHC on a system without GHC already
3646 installed is achieved by taking the intermediate C files (known
3647 as HC files) from a GHC compilation on a supported system to the
3648 target machine, and compiling them using gcc to get a working
3651 <para><emphasis>NOTE: GHC versions 5.xx were hard to bootstrap
3652 from C. We recommend using GHC 6.0.1 or
3653 later.</emphasis></para>
3655 <para>HC files are platform-dependent, so you have to get a set
3656 that were generated on similar hardware. There may be some
3657 supplied on the GHC download page, otherwise you'll have to
3658 compile some up yourself, or start from
3659 <emphasis>unregisterised</emphasis> HC files - see <xref
3660 linkend="unregisterised-porting">.</para>
3662 <para>The following steps should result in a working GHC build
3663 with full libraries:</para>
3667 <para>Unpack the HC files on top of a fresh source tree
3668 (make sure the source tree version matches the version of
3669 the HC files <emphasis>exactly</emphasis>!). This will
3670 place matching <filename>.hc</filename> files next to the
3671 corresponding Haskell source (<filename>.hs</filename> or
3672 <filename>.lhs</filename>) in the compiler subdirectory
3673 <filename>ghc/compiler</filename> and in the libraries
3674 (subdirectories of <filename>hslibs</filename> and
3675 <literal>libraries</literal>).</para>
3679 <para>The actual build process is fully automated by the
3680 <filename>hc-build</filename> script located in the
3681 <filename>distrib</filename> directory. If you eventually
3682 want to install GHC into the directory
3683 <replaceable>dir</replaceable>, the following
3684 command will execute the whole build process (it won't
3685 install yet):</para>
3688 foo% distrib/hc-build --prefix=<replaceable>dir</replaceable>
3690 <indexterm><primary>--hc-build</primary></indexterm>
3692 <para>By default, the installation directory is
3693 <filename>/usr/local</filename>. If that is what you want,
3694 you may omit the argument to <filename>hc-build</filename>.
3695 Generally, any option given to <filename>hc-build</filename>
3696 is passed through to the configuration script
3697 <filename>configure</filename>. If
3698 <filename>hc-build</filename> successfully completes the
3699 build process, you can install the resulting system, as
3709 <sect2 id="unregisterised-porting">
3710 <title>Porting GHC to a new architecture</title>
3712 <para>The first step in porting to a new architecture is to get
3713 an <firstterm>unregisterised</firstterm> build working. An
3714 unregisterised build is one that compiles via vanilla C only.
3715 By contrast, a registerised build uses the following
3716 architecture-specific hacks for speed:</para>
3720 <para>Global register variables: certain abstract machine
3721 <quote>registers</quote> are mapped to real machine
3722 registers, depending on how many machine registers are
3724 <filename>ghc/includes/MachRegs.h</filename>).</para>
3728 <para>Assembly-mangling: when compiling via C, we feed the
3729 assembly generated by gcc though a Perl script known as the
3730 <firstterm>mangler</firstterm> (see
3731 <filename>ghc/driver/mangler/ghc-asm.lprl</filename>). The
3732 mangler rearranges the assembly to support tail-calls and
3733 various other optimisations.</para>
3737 <para>In an unregisterised build, neither of these hacks are
3738 used — the idea is that the C code generated by the
3739 compiler should compile using gcc only. The lack of these
3740 optimisations costs about a factor of two in performance, but
3741 since unregisterised compilation is usually just a step on the
3742 way to a full registerised port, we don't mind too much.</para>
3744 <para>Notes on GHC portability in general: we've tried to stick
3745 to writing portable code in most parts of the system, so it
3746 should compile on any POSIXish system with gcc, but in our
3747 experience most systems differ from the standards in one way or
3748 another. Deal with any problems as they arise - if you get
3749 stuck, ask the experts on
3750 <email>glasgow-haskell-users@haskell.org</email>.</para>
3752 <para>Lots of useful information about the innards of GHC is
3753 available in the <ulink
3754 url="http://www.cse.unsw.edu.au/~chak/haskell/ghc/comm/">GHC
3755 Commentary</ulink>, which might be helpful if you run into some
3756 code which needs tweaking for your system.</para>
3759 <title>Cross-compiling to produce an unregisterised GHC</title>
3761 <para>In this section, we explain how to bootstrap GHC on a
3762 new platform, using unregisterised intermediate C files. We
3763 haven't put a great deal of effort into automating this
3764 process, for two reasons: it is done very rarely, and the
3765 process usually requires human intervention to cope with minor
3766 porting issues anyway.</para>
3768 <para>The following step-by-step instructions should result in
3769 a fully working, albeit unregisterised, GHC. Firstly, you
3770 need a machine that already has a working GHC (we'll call this
3771 the <firstterm>host</firstterm> machine), in order to
3772 cross-compile the intermediate C files that we will use to
3773 bootstrap the compiler on the <firstterm>target</firstterm>
3778 <para>On the target machine:</para>
3782 <para>Unpack a source tree (preferably a released
3783 version). We will call the path to the root of this
3784 tree <replaceable>T</replaceable>.</para>
3789 $ cd <replaceable>T</replaceable>
3790 $ ./configure --enable-hc-boot --enable-hc-boot-unregisterised
3793 <para>You might need to update
3794 <filename>configure.in</filename> to recognise the new
3795 architecture, and re-generate
3796 <filename>configure</filename> with
3797 <literal>autoreconf</literal>.</para>
3802 $ cd <replaceable>T</replaceable>/ghc/includes
3810 <para>On the host machine:</para>
3814 <para>Unpack a source tree (same released version). Call
3815 this directory <replaceable>H</replaceable>.</para>
3820 $ cd <replaceable>H</replaceable>
3827 <filename><replaceable>H</replaceable>/mk/build.mk</filename>,
3828 with the following contents:</para>
3831 GhcUnregisterised = YES
3832 GhcLibHcOpts = -O -H32m -keep-hc-files
3835 GhcWithNativeCodeGen = NO
3836 GhcWithInterpreter = NO
3837 GhcStage1HcOpts = -O -H32m -fasm
3838 GhcStage2HcOpts = -O -fvia-C -keep-hc-files
3844 <filename><replaceable>H</replaceable>/mk/config.mk</filename>:</para>
3847 <para>change <literal>TARGETPLATFORM</literal>
3848 appropriately, and set the variables involving
3849 <literal>TARGET</literal> to the correct values for
3850 the target platform. This step is necessary because
3851 currently <literal>configure</literal> doesn't cope
3852 with specifying different values for the
3853 <literal>--host</literal> and
3854 <literal>--target</literal> flags.</para>
3857 <para>copy <literal>LeadingUnderscore</literal>
3858 setting from target.</para>
3865 <filename><replaceable>T</replaceable>/ghc/includes/config.h</filename>
3867 <filename><replaceable>H</replaceable>/ghc/includes</filename>.
3868 Note that we are building on the host machine, using the
3869 target machine's <literal>config.h</literal> file. This
3870 is so that the intermediate C files generated here will
3871 be suitable for compiling on the target system.</para>
3876 <para>Touch <literal>config.h</literal>, just to make
3877 sure it doesn't get replaced during the build:</para>
3879 $ touch <replaceable>H</replaceable>/ghc/includes/config.h</screen>
3883 <para>Now build the compiler:</para>
3885 $ cd <replaceable>H</replaceable>/glafp-utils && make boot && make
3886 $ cd <replaceable>H</replaceable>/ghc && make boot && make
3888 <para>Don't worry if the build falls over in the RTS, we
3889 don't need the RTS yet.</para>
3894 $ cd <replaceable>H</replaceable>/libraries
3895 $& make boot && make
3901 $ cd <replaceable>H</replaceable>/ghc
3902 $ make boot stage=2 && make stage=2
3908 $ cd <replaceable>H</replaceable>/ghc/utils
3910 $ make -k HC=<replaceable>H</replaceable>/ghc/compiler/stage1/ghc-inplace \
3911 EXTRA_HC_OPTS='-O -fvia-C -keep-hc-files'
3917 $ cd <replaceable>H</replaceable>
3918 $ make hc-file-bundle Project=Ghc
3924 <filename><replaceable>H</replaceable>/*-hc.tar.gz</filename>
3925 to <filename><replaceable>T</replaceable>/..</filename>.</para>
3931 <para>On the target machine:</para>
3933 <para>At this stage we simply need to bootstrap a compiler
3934 from the intermediate C files we generated above. The
3935 process of bootstrapping from C files is automated by the
3936 script in <literal>distrib/hc-build</literal>, and is
3937 described in <xref linkend="sec-booting-from-hc">.</para>
3940 $ ./distrib/hc-build --enable-hc-boot-unregisterised
3943 <para>However, since this is a bootstrap on a new machine,
3944 the automated process might not run to completion the
3945 first time. For that reason, you might want to treat the
3946 <literal>hc-build</literal> script as a list of
3947 instructions to follow, rather than as a fully automated
3948 script. This way you'll be able to restart the process
3949 part-way through if you need to fix anything on the
3952 <para>Don't bother with running
3953 <literal>make install</literal> in the newly
3954 bootstrapped tree; just use the compiler in that tree to
3955 build a fresh compiler from scratch, this time without
3956 booting from C files. Before doing this, you might want
3957 to check that the bootstrapped compiler is generating
3958 working binaries:</para>
3962 main = putStrLn "Hello World!\n"
3964 $ <replaceable>T</replaceable>/ghc/compiler/ghc-inplace hello.hs -o hello
3969 <para>Once you have the unregisterised compiler up and
3970 running, you can use it to start a registerised port. The
3971 following sections describe the various parts of the
3972 system that will need architecture-specific tweaks in
3973 order to get a registerised build going.</para>
3980 <title>Porting the RTS</title>
3982 <para>The following files need architecture-specific code for a
3983 registerised build:</para>
3987 <term><filename>ghc/includes/MachRegs.h</filename></term>
3988 <indexterm><primary><filename>MachRegs.h</filename></primary>
3991 <para>Defines the STG-register to machine-register
3992 mapping. You need to know your platform's C calling
3993 convention, and which registers are generally available
3994 for mapping to global register variables. There are
3995 plenty of useful comments in this file.</para>
3999 <term><filename>ghc/includes/TailCalls.h</filename></term>
4000 <indexterm><primary><filename>TailCalls.h</filename></primary>
4003 <para>Macros that cooperate with the mangler (see <xref
4004 linkend="sec-mangler">) to make proper tail-calls
4009 <term><filename>ghc/rts/Adjustor.c</filename></term>
4010 <indexterm><primary><filename>Adjustor.c</filename></primary>
4014 <literal>foreign import "wrapper"</literal>
4016 <literal>foreign export dynamic</literal>).
4017 Not essential for getting GHC bootstrapped, so this file
4018 can be deferred until later if necessary.</para>
4022 <term><filename>ghc/rts/StgCRun.c</filename></term>
4023 <indexterm><primary><filename>StgCRun.c</filename></primary>
4026 <para>The little assembly layer between the C world and
4027 the Haskell world. See the comments and code for the
4028 other architectures in this file for pointers.</para>
4032 <term><filename>ghc/rts/MBlock.h</filename></term>
4033 <term><filename>ghc/rts/MBlock.c</filename></term>
4034 <indexterm><primary><filename>MBlock.h</filename></primary>
4036 <indexterm><primary><filename>MBlock.c</filename></primary>
4039 <para>These files are really OS-specific rather than
4040 architecture-specific. In <filename>MBlock.h</filename>
4041 is specified the absolute location at which the RTS
4042 should try to allocate memory on your platform (try to
4043 find an area which doesn't conflict with code or dynamic
4044 libraries). In <filename>Mblock.c</filename> you might
4045 need to tweak the call to <literal>mmap()</literal> for
4052 <sect3 id="sec-mangler">
4053 <title>The mangler</title>
4055 <para>The mangler is an evil Perl-script that rearranges the
4056 assembly code output from gcc to do two main things:</para>
4060 <para>Remove function prologues and epilogues, and all
4061 movement of the C stack pointer. This is to support
4062 tail-calls: every code block in Haskell code ends in an
4063 explicit jump, so we don't want the C-stack overflowing
4064 while we're jumping around between code blocks.</para>
4067 <para>Move the <firstterm>info table</firstterm> for a
4068 closure next to the entry code for that closure. In
4069 unregisterised code, info tables contain a pointer to the
4070 entry code, but in registerised compilation we arrange
4071 that the info table is shoved right up against the entry
4072 code, and addressed backwards from the entry code pointer
4073 (this saves a word in the info table and an extra
4074 indirection when jumping to the closure entry
4079 <para>The mangler is abstracted to a certain extent over some
4080 architecture-specific things such as the particular assembler
4081 directives used to herald symbols. Take a look at the
4082 definitions for other architectures and use these as a
4083 starting point.</para>
4087 <title>The native code generator</title>
4089 <para>The native code generator isn't essential to getting a
4090 registerised build going, but it's a desirable thing to have
4091 because it can cut compilation times in half. The native code
4092 generator is described in some detail in the <ulink
4093 url="http://www.cse.unsw.edu.au/~chak/haskell/ghc/comm/">GHC
4094 commentary</ulink>.</para>
4100 <para>To support GHCi, you need to port the dynamic linker
4101 (<filename>fptools/ghc/rts/Linker.c</filename>). The linker
4102 currently supports the ELF and PEi386 object file formats - if
4103 your platform uses one of these then things will be
4104 significantly easier. The majority of Unix platforms use the
4105 ELF format these days. Even so, there are some
4106 machine-specific parts of the ELF linker: for example, the
4107 code for resolving particular relocation types is
4108 machine-specific, so some porting of this code to your
4109 architecture will probaly be necessary.</para>
4111 <para>If your system uses a different object file format, then
4112 you have to write a linker — good luck!</para>
4118 <sect1 id="sec-build-pitfalls">
4119 <title>Known pitfalls in building Glasgow Haskell
4121 <indexterm><primary>problems, building</primary></indexterm>
4122 <indexterm><primary>pitfalls, in building</primary></indexterm>
4123 <indexterm><primary>building pitfalls</primary></indexterm></title>
4126 WARNINGS about pitfalls and known “problems”:
4135 One difficulty that comes up from time to time is running out of space
4136 in <literal>TMPDIR</literal>. (It is impossible for the configuration stuff to
4137 compensate for the vagaries of different sysadmin approaches to temp
4139 <indexterm><primary>tmp, running out of space in</primary></indexterm>
4141 The quickest way around it is <command>setenv TMPDIR /usr/tmp</command><indexterm><primary>TMPDIR</primary></indexterm> or
4142 even <command>setenv TMPDIR .</command> (or the equivalent incantation with your shell
4145 The best way around it is to say
4148 export TMPDIR=<dir>
4151 in your <filename>build.mk</filename> file.
4152 Then GHC and the other <literal>fptools</literal> programs will use the appropriate directory
4161 In compiling some support-code bits, e.g., in <filename>ghc/rts/gmp</filename> and even
4162 in <filename>ghc/lib</filename>, you may get a few C-compiler warnings. We think these
4170 When compiling via C, you'll sometimes get “warning: assignment from
4171 incompatible pointer type” out of GCC. Harmless.
4178 Similarly, <command>ar</command>chiving warning messages like the following are not
4182 ar: filename GlaIOMonad__1_2s.o truncated to GlaIOMonad_
4183 ar: filename GlaIOMonad__2_2s.o truncated to GlaIOMonad_
4193 In compiling the compiler proper (in <filename>compiler/</filename>), you <emphasis>may</emphasis>
4194 get an “Out of heap space” error message. These can vary with the
4195 vagaries of different systems, it seems. The solution is simple:
4202 If you're compiling with GHC 4.00 or later, then the
4203 <emphasis>maximum</emphasis> heap size must have been reached. This
4204 is somewhat unlikely, since the maximum is set to 64M by default.
4205 Anyway, you can raise it with the
4206 <option>-optCrts-M<size></option> flag (add this flag to
4207 <constant><module>_HC_OPTS</constant>
4208 <command>make</command> variable in the appropriate
4209 <filename>Makefile</filename>).
4216 For GHC < 4.00, add a suitable <option>-H</option> flag to the <filename>Makefile</filename>, as
4225 and try again: <command>gmake</command>. (see <Xref LinkEnd="sec-suffix"> for information about
4226 <constant><module>_HC_OPTS</constant>.)
4228 Alternatively, just cut to the chase:
4232 % make EXTRA_HC_OPTS=-optCrts-M128M
4241 If you try to compile some Haskell, and you get errors from GCC about
4242 lots of things from <filename>/usr/include/math.h</filename>, then your GCC was
4243 mis-installed. <command>fixincludes</command> wasn't run when it should've been.
4245 As <command>fixincludes</command> is now automagically run as part of GCC installation,
4246 this bug also suggests that you have an old GCC.
4254 You <emphasis>may</emphasis> need to re-<command>ranlib</command><indexterm><primary>ranlib</primary></indexterm> your libraries (on Sun4s).
4258 % cd $(libdir)/ghc-x.xx/sparc-sun-sunos4
4259 % foreach i ( `find . -name '*.a' -print` ) # or other-shell equiv...
4261 ? # or, on some machines: ar s $i
4266 We'd be interested to know if this is still necessary.
4274 GHC's sources go through <command>cpp</command> before being compiled, and <command>cpp</command> varies
4275 a bit from one Unix to another. One particular gotcha is macro calls
4280 SLIT("Hello, world")
4284 Some <command>cpp</command>s treat the comma inside the string as separating two macro
4285 arguments, so you get
4289 :731: macro `SLIT' used with too many (2) args
4293 Alas, <command>cpp</command> doesn't tell you the offending file!
4295 Workaround: don't put weird things in string args to <command>cpp</command> macros.
4306 <Sect1 id="platforms"><Title>Platforms, scripts, and file names</Title>
4308 GHC is designed both to be built, and to run, on both Unix and Windows. This flexibility
4309 gives rise to a good deal of brain-bending detail, which we have tried to collect in this chapter.
4312 <sect2 id="cygwin-and-mingw"><Title>Windows platforms: Cygwin, MSYS, and MinGW</Title>
4314 <para> The build system is built around Unix-y makefiles. Because it's not native,
4315 the Windows situation for building GHC is particularly confusing. This section
4316 tries to clarify, and to establish terminology.</para>
4318 <sect3 id="ghc-mingw"><title>MinGW</title>
4320 <para> <ulink url="http://www.mingw.org">MinGW (Minimalist GNU for Windows)</ulink>
4321 is a collection of header
4322 files and import libraries that allow one to use <command>gcc</command> and produce
4323 native Win32 programs that do not rely on any third-party DLLs. The
4324 current set of tools include GNU Compiler Collection (<command>gcc</command>), GNU Binary
4325 Utilities (Binutils), GNU debugger (Gdb), GNU make, and a assorted
4329 <para> The down-side of MinGW is that the MinGW libraries do not support anything like the full
4334 <sect3 id="ghc-cygwin"><title>Cygwin and MSYS</title>
4336 <para>You can't use the MinGW to <emphasis>build</emphasis> GHC, because MinGW doesn't have a shell,
4337 or the standard Unix commands such as <command>mv</command>, <command>rm</command>,
4338 <command>ls</command>, nor build-system stuff such as <command>make</command> and <command>cvs</command>.
4339 For that, there are two choices: <ulink url="http://www.cygwin.com">Cygwin</ulink>
4340 and <ulink url="http://www.mingw.org/msys.shtml">MSYS</ulink>:
4344 Cygwin comes with compilation tools (<command>gcc</command>, <command>ld</command> and so on), which
4345 compile code that has access to all of Posix. The price is that the executables must be
4346 dynamically linked with the Cygwin DLL, so that <emphasis>you cannot run a Cywin-compiled program on a machine
4347 that doesn't have Cygwin</emphasis>. Worse, Cygwin is a moving target. The name of the main DLL, <literal>cygwin1.dll</literal>
4348 does not change, but the implementation certainly does. Even the interfaces to functions
4349 it exports seem to change occasionally. </para>
4353 MSYS is a fork of the Cygwin tree, so they
4354 are fundamentally similar. However, MSYS is by design much smaller and simpler. Access to the file system goes
4355 through fewer layers, so MSYS is quite a bit faster too.
4358 <para>Furthermore, MSYS provides no compilation tools; it relies instead on the MinGW tools. These
4359 compile binaries that run with no DLL support, on any Win32 system.
4360 However, MSYS does come with all the make-system tools, such as <command>make</command>, <command>autoconf</command>,
4361 <command>cvs</command>, <command>ssh</command> etc. To get these, you have to download the
4362 MsysDTK (Developer Tool Kit) package, as well as the base MSYS package.
4364 <para>MSYS does have a DLL, but it's only used by MSYS commands (<command>sh</command>, <command>rm</command>,
4365 <command>ssh</command> and so on),
4366 not by programs compiled under MSYS.
4374 <sect3><title>Targeting MinGW</title>
4376 <para>We want GHC to compile programs that work on any Win32 system. Hence:
4379 GHC does invoke a C compiler, assembler, linker and so on, but we ensure that it only
4380 invokes the MinGW tools, not the Cygwin ones. That means that the programs GHC compiles
4381 will work on any system, but it also means that the programs GHC compiles do not have access
4382 to all of Posix. In particular, they cannot import the (Haskell) Posix
4383 library; they have to do
4384 their input output using standard Haskell I/O libraries, or native Win32 bindings.</para>
4385 <para> We will call a GHC that targets MinGW in this way <emphasis>GHC-mingw</emphasis>.</para>
4389 To make the GHC distribution self-contained, the GHC distribution includes the MinGW <command>gcc</command>,
4390 <command>as</command>, <command>ld</command>, and a bunch of input/output libraries.
4393 So <emphasis>GHC targets MinGW</emphasis>, not Cygwin.
4394 It is in principle possible to build a version of GHC, <emphasis>GHC-cygwin</emphasis>,
4395 that targets Cygwin instead. The up-side of GHC-cygwin is
4396 that Haskell programs compiled by GHC-cygwin can import the (Haskell) Posix library.
4397 <emphasis>We do not support GHC-cygwin, however; it is beyond our resources.</emphasis>
4400 <para>While GHC <emphasis>targets</emphasis> MinGW, that says nothing about
4401 how GHC is <emphasis>built</emphasis>. We use both MSYS and Cygwin as build environments for
4402 GHC; both work fine, though MSYS is rather lighter weight.</para>
4404 <para>In your build tree, you build a compiler called <Command>ghc-inplace</Command>. It
4405 uses the <Command>gcc</Command> that you specify using the
4406 <option>--with-gcc</option> flag when you run
4407 <Command>configure</Command> (see below).
4408 The makefiles are careful to use <Command>ghc-inplace</Command> (not <Command>gcc</Command>)
4409 to compile any C files, so that it will in turn invoke the correct <Command>gcc</Command> rather that
4410 whatever one happens to be in your path. However, the makefiles do use whatever <Command>ld</Command>
4411 and <Command>ar</Command> happen to be in your path. This is a bit naughty, but (a) they are only
4412 used to glom together .o files into a bigger .o file, or a .a file,
4413 so they don't ever get libraries (which would be bogus; they might be the wrong libraries), and (b)
4414 Cygwin and MinGW use the same .o file format. So its ok.
4418 <sect3><title> File names </title>
4420 <para>Cygwin, MSYS, and the underlying Windows file system all understand file paths of form <literal>c:/tmp/foo</literal>.
4424 MSYS programs understand <filename>/bin</filename>, <filename>/usr/bin</filename>, and map Windows's lettered drives as
4425 <filename>/c/tmp/foo</filename> etc. The exact mount table is given in the doc subdirectory of the MSYS distribution.
4427 <para> When it invokes a command, the MSYS shell sees whether the invoked binary lives in the MSYS <filename>/bin</filename>
4428 directory. If so, it just invokes it. If not, it assumes the program is no an MSYS program, and walks over the command-line
4429 arguments changing MSYS paths into native-compatible paths. It does this inside sub-arguments and inside quotes. For example,
4434 the MSYS shell will actually call <literal>foogle</literal> with argument <literal>-Bc:/tmp/baz</literal>.
4438 Cygwin programs have a more complicated mount table, and map the lettered drives as <filename>/cygdrive/c/tmp/foo</filename>.
4440 <para>The Cygwin shell does no argument processing when invoking non-Cygwin programs.
4446 <sect3><title>HOST_OS vs TARGET_OS</title>
4449 In the source code you'll find various ifdefs looking like:
4451 #ifdef mingw32_HOST_OS
4457 #ifdef mingw32_TARGET_OS
4461 These macros are set by the configure script (via the file config.h).
4462 Which is which? The criterion is this. In the ifdefs in GHC's source code:
4465 The "host" system is the one on which GHC itself will be run.
4468 The "target" system is the one for which the program compiled by GHC will be run.
4471 For a stage-2 compiler, in which GHCi is available, the "host" and "target" systems must be the same.
4472 So then it doesn't really matter whether you use the HOST_OS or TARGET_OS cpp macros.
4479 <sect2><title>Wrapper scripts</title>
4482 Many programs, including GHC itself and hsc2hs, need to find associated binaries and libraries.
4483 For <emphasis>installed</emphasis> programs, the strategy depends on the platform. We'll use
4484 GHC itself as an example:
4487 On Unix, the command <command>ghc</command> is a shell script, generated by adding installation
4488 paths to the front of the source file <filename>ghc.sh</filename>,
4489 that invokes the real binary, passing "-B<emphasis>path</emphasis>" as an argument to tell <command>ghc</command>
4490 where to find its supporting files.
4494 On vanilla Windows, it turns out to be much harder to make reliable script to be run by the
4495 native Windows shell <command>cmd</command> (e.g. limits on the length
4496 of the command line). So instead we invoke the GHC binary directly, with no -B flag.
4497 GHC uses the Windows <literal>getExecDir</literal> function to find where the executable is,
4498 and from that figures out where the supporting files are.
4501 (You can find the layout of GHC's supporting files in the
4502 section "Layout of installed files" of Section 2 of the GHC user guide.)
4505 Things work differently for <emphasis>in-place</emphasis> execution, where you want to
4506 execute a program that has just been built in a build tree. The difference is that the
4507 layout of the supporting files is different.
4508 In this case, whether on Windows or Unix, we always use a shell script. This works OK
4509 on Windows because the script is executed by MSYS or Cygwin, which don't have the
4510 shortcomings of the native Windows <command>cmd</command> shell.
4517 <Sect1 id="winbuild"><Title>Instructions for building under Windows</Title>
4520 This section gives detailed instructions for how to build
4521 GHC from source on your Windows machine. Similar instructions for
4522 installing and running GHC may be found in the user guide. In general,
4523 Win95/Win98 behave the same, and WinNT/Win2k behave the same.
4526 Make sure you read the preceding section on platforms (<xref linkend="platforms">)
4527 before reading section.
4531 <Sect2><Title>Installing and configuring MSYS</Title>
4534 MSYS is a lightweight alternative to Cygwin.
4535 You don't need MSYS to <emphasis>use</emphasis> GHC,
4536 but you do need it or Cygwin to <emphasis>build</emphasis> GHC.
4537 Here's how to install MSYS.
4540 Go to <ulink url="http://www.mingw.org/download.shtml">http://www.mingw.org/download.shtml</ulink> and
4541 download the following (of course, the version numbers will differ):
4543 <listitem><para>The main MSYS package (binary is sufficient): <literal>MSYS-1.0.9.exe</literal>
4545 <listitem><para>The MSYS developer's toolkit (binary is sufficient): <literal>msysDTK-1.0.1.exe</literal>.
4546 This provides <command>make</command>, <command>autoconf</command>,
4547 <command>ssh</command>, <command>cvs</command> and probably more besides.
4550 Run both executables (in the order given above) to install them. I put them in <literal>c:/msys</literal>
4554 Set the following environment variables
4556 <listitem><para><literal>PATH</literal>: add <literal>c:/msys/1.0/bin</literal> to your path. (Of course, the version number may differ.)
4559 <listitem><para><literal>HOME</literal>: set to your home directory (e.g. <literal>c:/userid</literal>).
4560 This is where, among other things, <command>ssh</command> will look for your <literal>.ssh</literal> directory.
4563 <listitem><para><literal>SHELL</literal>: set to <literal>c:/msys/1.0/bin/sh.exe</literal>
4566 <listitem><para><literal>CVS_RSH</literal>: set to <literal>c:/msys/1.0/bin/ssh.exe</literal>. Only necessary if
4570 <listitem><para><literal>MAKE_MODE</literal>: set to <literal>UNIX</literal>. (I'm not certain this is necessary for MSYS.)
4577 Check that the <literal>CYGWIN</literal> environment variable is <emphasis>not</emphasis> set. It's a bad bug
4578 that MSYS is affected by this, but if you have CYGWIN set to "ntsec ntea", which is right for Cygwin, it
4579 causes the MSYS <command>ssh</command> to bogusly fail complaining that your <filename>.ssh/identity</filename>
4580 file has too-liberal permissinos.
4587 <Sect2><Title>Installing and configuring Cygwin</Title>
4589 <para>You don't need Cygwin to <emphasis>use</emphasis> GHC,
4590 but you do need it or MSYS to <emphasis>build</emphasis> GHC.</para>
4592 <para> Install Cygwin from <ulink url="http://www.cygwin.com/">http://www.cygwin.com/</ulink>.
4593 The installation process is straightforward; we install it in <Filename>c:/cygwin</Filename>.
4594 During the installation dialogue, make sure that you select all of the following:
4597 <command>cvs</command>,
4600 <command>openssh</command>,
4603 <command>autoconf</command>,
4606 <command>binutils</command> (includes ld and (I think) ar),
4609 <command>gcc</command>,
4612 <command>flex</command>,
4615 <command>make</command>.
4618 If you miss out any of these, strange things will happen to you. To see thse packages,
4619 click on the "View" button in the "Select Packages"
4620 stage of Cygwin's installation dialogue, until the view says "Full". The default view, which is
4621 "Category" isn't very helpful, and the "View" button is rather unobtrousive.
4623 <para> Now set the following user environment variables:
4626 <listitem><para> Add <filename>c:/cygwin/bin</filename> and <filename>c:/cygwin/usr/bin</filename> to your
4627 <constant>PATH</constant></para></listitem>
4631 Set <constant>MAKE_MODE</constant> to <Literal>UNIX</Literal>. If you
4632 don't do this you get very weird messages when you type
4633 <Command>make</Command>, such as:
4635 /c: /c: No such file or directory
4640 <listitem><para> Set <constant>SHELL</constant> to
4641 <Filename>c:/cygwin/bin/bash</Filename>. When you invoke a shell in Emacs, this
4642 <constant>SHELL</constant> is what you get.
4645 <listitem><para> Set <constant>HOME</constant> to point to your
4646 home directory. This is where, for example,
4647 <command>bash</command> will look for your <filename>.bashrc</filename>
4648 file. Ditto <command>emacs</command> looking for <filename>.emacsrc</filename>
4654 There are a few other things to do:
4658 By default, cygwin provides the command shell <filename>ash</filename>
4659 as <filename>sh.exe</filename>. We have often seen build-system problems that
4660 turn out to be due to bugs in <filename>ash</filename>
4662 and length of command lines). On the other hand <filename>bash</filename> seems
4664 So, in <filename>cygwin/bin</filename>
4665 remove the supplied <filename>sh.exe</filename> (or rename it as <filename>ash.exe</filename>),
4666 and copy <filename>bash.exe</filename> to <filename>sh.exe</filename>.
4667 You'll need to do this in Windows Explorer or the Windows <command>cmd</command> shell, because
4668 you can't rename a running program!
4674 Some script files used in the make system start with "<Command>#!/bin/perl</Command>",
4675 (and similarly for <Command>sh</Command>). Notice the hardwired path!
4676 So you need to ensure that your <Filename>/bin</Filename> directory has the following
4679 <listitem> <para><Command>sh</Command></para></listitem>
4680 <listitem> <para><Command>perl</Command></para></listitem>
4681 <listitem> <para><Command>cat</Command></para></listitem>
4683 All these come in Cygwin's <Filename>bin</Filename> directory, which you probably have
4684 installed as <Filename>c:/cygwin/bin</Filename>. By default Cygwin mounts "<Filename>/</Filename>" as
4685 <Filename>c:/cygwin</Filename>, so if you just take the defaults it'll all work ok.
4686 (You can discover where your Cygwin
4687 root directory <Filename>/</Filename> is by typing <Command>mount</Command>.)
4688 Provided <Filename>/bin</Filename> points to the Cygwin <Filename>bin</Filename>
4689 directory, there's no need to copy anything. If not, copy these binaries from the <filename>cygwin/bin</filename>
4690 directory (after fixing the <filename>sh.exe</filename> stuff mentioned in the previous bullet).
4696 <para>Finally, here are some things to be aware of when using Cygwin:
4698 <listitem> <para>Cygwin doesn't deal well with filenames that include
4699 spaces. "<filename>Program Files</filename>" and "<filename>Local files</filename>" are
4703 <listitem> <para> Cygwin implements a symbolic link as a text file with some
4704 magical text in it. So other programs that don't use Cygwin's
4705 I/O libraries won't recognise such files as symlinks.
4706 In particular, programs compiled by GHC are meant to be runnable
4707 without having Cygwin, so they don't use the Cygwin library, so
4708 they don't recognise symlinks.
4712 Win32 has a <command>find</command> command which is not the same as Cygwin's find.
4713 You will probably discover that the Win32 <command>find</command> appears in your <constant>PATH</constant>
4714 before the Cygwin one, because it's in the <emphasis>system</emphasis> <constant>PATH</constant>
4715 environment variable, whereas you have probably modified the <emphasis>user</emphasis> <constant>PATH</constant>
4716 variable. You can always invoke <command>find</command> with an absolute path, or rename it.
4724 <Sect2 id="configure-ssh"><Title>Configuring SSH</Title>
4726 <para><command>ssh</command> comes with Cygwin, provided you remember to ask for it when
4727 you install Cygwin. (If not, the installer lets you update easily.) Look for <command>openssh</command>
4728 (not ssh) in the Cygwin list of applications!</para>
4730 <para>There are several strange things about <command>ssh</command> on Windows that you need to know.
4734 The programs <command>ssh-keygen1</command>, <command>ssh1</command>, and <command>cvs</command>,
4735 seem to lock up <command>bash</command> entirely if they try to get user input (e.g. if
4736 they ask for a password). To solve this, start up <filename>cmd.exe</filename>
4737 and run it as follows:
4739 c:\tmp> set CYGWIN32=tty
4740 c:\tmp> c:/user/local/bin/ssh-keygen1
4745 <command>ssh</command> needs to access your directory <filename>.ssh</filename>, in your home directory.
4746 To determine your home directory <command>ssh</command> first looks in
4747 <filename>c:/cygwin/etc/passwd</filename> (or wherever you have Cygwin installed). If there's an entry
4748 there with your userid, it'll use that entry to determine your home directory, <emphasis>ignoring
4749 the setting of the environment variable $HOME</emphasis>. If the home directory is
4750 bogus, <command>ssh</command> fails horribly. The best way to see what is going on is to say
4752 ssh -v cvs.haskell.org
4754 which makes <command>ssh</command> print out information about its activity.
4756 <para> You can fix this problem, either by correcting the home-directory field in
4757 <filename>c:/cygwin/etc/passwd</filename>, or by simply deleting the entire entry for your userid. If
4758 you do that, <command>ssh</command> uses the $HOME environment variable instead.
4764 <para>To protect your
4765 <literal>.ssh</literal> from access by anyone else,
4766 right-click your <literal>.ssh</literal> directory, and
4767 select <literal>Properties</literal>. If you are not on
4768 the access control list, add yourself, and give yourself
4769 full permissions (the second panel). Remove everyone else
4770 from the access control list. Don't leave them there but
4771 deny them access, because 'they' may be a list that
4772 includes you!</para>
4776 <para>In fact <command>ssh</command> 3.6.1 now seems to <emphasis>require</emphasis>
4777 you to have Unix permissions 600 (read/write for owner only)
4778 on the <literal>.ssh/identity</literal> file, else it
4779 bombs out. For your local C drive, it seems that <literal>chmod 600 identity</literal> works,
4780 but on Windows NT/XP, it doesn't work on a network drive (exact dteails obscure).
4781 The solution seems to be to set the $CYGWIN environment
4782 variable to "<literal>ntsec neta</literal>". The $CYGWIN environment variable is discussed
4783 in <ulink url="http://cygwin.com/cygwin-ug-net/using-cygwinenv.html">the Cygwin User's Guide</ulink>,
4784 and there are more details in <ulink url="http://cygwin.com/faq/faq_4.html#SEC44">the Cygwin FAQ</ulink>.
4791 <Sect2><Title>Other things you need to install</Title>
4793 <para>You have to install the following other things to build GHC:
4797 Install an executable GHC, from <ulink url="http://www.haskell.org/ghc">http://www.haskell.org/ghc</ulink>.
4798 This is what you will use to compile GHC. Add it in your
4799 <constant>PATH</constant>: the installer tells you the path element
4800 you need to add upon completion.
4806 Install an executable Happy, from <ulink url="http://www.haskell.org/happy">http://www.haskell.org/happy</ulink>.
4807 Happy is a parser generator used to compile the Haskell grammar. Add it in your
4808 <constant>PATH</constant>.
4813 <para>Install Alex. This can be done by building from the
4814 source distribution in the usual way. Sources are
4815 available from <ulink
4816 url="http://www.haskell.org/alex">http://www.haskell.org/alex</ulink>.</para>
4820 <para>GHC uses the <emphasis>mingw</emphasis> C compiler to
4821 generate code, so you have to install that (see <xref linkend="cygwin-and-mingw">).
4822 Just pick up a mingw bundle at
4823 <ulink url="http://www.mingw.org/">http://www.mingw.org/</ulink>.
4824 We install it in <filename>c:/mingw</filename>.
4826 <para>Do <emphasis>not</emphasis> add any of the <emphasis>mingw</emphasis> binaries to your path.
4827 They are only going to get used by explicit access (via the --with-gcc flag you
4828 give to <Command>configure</Command> later). If you do add them to your path
4829 you are likely to get into a mess because their names overlap with Cygwin binaries.
4835 <para>We use <command>emacs</command> a lot, so we install that too.
4836 When you are in <filename>fptools/ghc/compiler</filename>, you can use
4837 "<literal>make tags</literal>" to make a TAGS file for emacs. That uses the utility
4838 <filename>fptools/ghc/utils/hasktags/hasktags</filename>, so you need to make that first.
4839 The most convenient way to do this is by going <literal>make boot</literal> in <filename>fptools/ghc</filename>.
4840 The <literal>make tags</literal> command also uses <command>etags</command>, which comes with <command>emacs</command>,
4841 so you will need to add <filename>emacs/bin</filename> to your <literal>PATH</literal>.
4847 <para> Finally, check out a copy of GHC sources from
4848 the CVS repository, following the instructions above (<xref linkend="cvs-access">).
4855 <Sect2><Title>Building GHC</Title>
4858 Now go read the documentation above on building from source (<xref linkend="sec-building-from-source">);
4859 the bullets below only tell
4860 you about Windows-specific wrinkles.</para>
4864 If you used <Command>autoconf</Command> instead of <Command>autoreconf</Command>,
4865 you'll get an error when you run <filename>./configure</filename>:
4868 creating mk/config.h
4869 mk/config.h is unchanged
4871 running /bin/sh ./configure --cache-file=.././config.cache --srcdir=.
4872 ./configure: ./configure: No such file or directory
4873 configure: error: ./configure failed for ghc
4878 <listitem> <para><command>autoreconf</command> seems to create the file <filename>configure</filename>
4879 read-only. So if you need to run autoreconf again (which I sometimes do for safety's sake),
4882 /usr/bin/autoconf: cannot create configure: permission denied
4884 Solution: delete <filename>configure</filename> first.
4889 You either need to add <filename>ghc</filename> to your
4890 <constant>PATH</constant> before you invoke
4891 <Command>configure</Command>, or use the <Command>configure</Command>
4892 option <option>--with-ghc=c:/ghc/ghc-some-version/bin/ghc</option>.
4897 If you are paranoid, delete <filename>config.cache</filename> if it exists.
4898 This file occasionally remembers out-of-date configuration information, which
4899 can be really confusing.
4905 After <command>autoreconf</command> run <command>./configure</command> in
4906 <filename>fptools/</filename> thus:
4909 ./configure --host=i386-unknown-mingw32 --with-gcc=c:/mingw/bin/gcc
4911 This is the point at which you specify that you are building GHC-mingw
4912 (see <xref linkend="ghc-mingw">). </para>
4914 <para> Both these options are important! It's possible to get into
4915 trouble using the wrong C compiler!</para>
4917 Furthermore, it's <emphasis>very important</emphasis> that you specify a
4918 full MinGW path for <command>gcc</command>, not a Cygwin path, because GHC (which
4919 uses this path to invoke <command>gcc</command>) is a MinGW program and won't
4920 understand a Cygwin path. For example, if you
4921 say <literal>--with-gcc=/mingw/bin/gcc</literal>, it'll be interpreted as
4922 <filename>/cygdrive/c/mingw/bin/gcc</filename>, and GHC will fail the first
4923 time it tries to invoke it. Worse, the failure comes with
4924 no error message whatsoever. GHC simply fails silently when first invoked,
4925 typically leaving you with this:
4927 make[4]: Leaving directory `/cygdrive/e/fptools-stage1/ghc/rts/gmp'
4928 ../../ghc/compiler/ghc-inplace -optc-mno-cygwin -optc-O
4929 -optc-Wall -optc-W -optc-Wstrict-prototypes -optc-Wmissing-prototypes
4930 -optc-Wmissing-declarations -optc-Winline -optc-Waggregate-return
4931 -optc-Wbad-function-cast -optc-Wcast-align -optc-I../includes
4932 -optc-I. -optc-Iparallel -optc-DCOMPILING_RTS
4933 -optc-fomit-frame-pointer -O2 -static
4934 -package-name rts -O -dcore-lint -c Adjustor.c -o Adjustor.o
4935 make[2]: *** [Adjustor.o] Error 1
4936 make[1]: *** [all] Error 1
4937 make[1]: Leaving directory `/cygdrive/e/fptools-stage1/ghc'
4938 make: *** [all] Error 1
4944 If you want to build GHC-cygwin (<xref linkend="ghc-cygwin">)
4945 you'll have to do something more like:
4947 ./configure --with-gcc=...the Cygwin gcc...
4952 <listitem><para> You almost certainly want to set
4956 in your <filename>build.mk</filename> configuration file (see <xref linkend="sec-build-config">).
4957 This tells the build system not to split each library into a myriad of little object files, one
4958 for each function. Doing so reduces binary sizes for statically-linked binaries, but on Windows
4959 it dramatically increases the time taken to build the libraries in the first place.
4963 <listitem><para> Do not attempt to build the documentation.
4964 It needs all kinds of wierd Jade stuff that we haven't worked out for
4965 Win32.</para></listitem>
projects / ghc-hetmet.git / blob