Contents of /xml/htdocs/doc/en/articles/autotools-practices.xml

Parent Directory Parent Directory | Revision Log Revision Log

Revision 1.4 - (show annotations) (download) (as text)
Thu Feb 23 17:52:06 2012 UTC (6 years, 3 months ago) by swift
Branch: MAIN
Changes since 1.3: +3 -3 lines
File MIME type: application/xml
Mark article as... article. Going against policy and not bumping date as that would confuse users. Thanks to Agostino Sarubbo for noticing/notifying

1 <?xml version='1.0' encoding="UTF-8"?>
2 <!DOCTYPE guide SYSTEM "/dtd/guide.dtd">
3 <!-- $Header: /var/cvsroot/gentoo/xml/htdocs/doc/en/articles/autotools-practices.xml,v 1.3 2011/09/04 17:53:41 swift Exp $ -->
5 <guide disclaimer="articles">
6 <title>Best practices with autotools</title>
8 <author title="Author">
9 <mail link="flameeyes@gentoo.org">Diego Pettenò</mail>
10 </author>
12 <abstract>
13 This article covers some of the most common errors people make when using
14 autotools and ways to achieve better results.
15 </abstract>
17 <!-- The content of this document is licensed under the CC-BY-SA license -->
18 <!-- See http://creativecommons.org/licenses/by-sa/2.5 -->
19 <license/>
21 <version>2</version>
22 <date>2005-12-16</date>
24 <chapter>
25 <title>Best practices with autotools</title>
26 <section>
27 <body>
29 <p>
30 The core of GNU's compile chain -- the set of tools used to build GNU software
31 packages -- is the so-called "autotools," a term that refers to the autoconf
32 and automake programs, as well as libtool, autoheader, pkg-config, and
33 sometimes gettext. These tools let you compile GNU software on a wide variety
34 of platforms and Unix and Unix-like operating systems, providing developers a
35 framework to check for the presence of the libraries, functions, and tools that
36 they want to use. While autotools are great in the hands of an experienced
37 developer, they can be quite a handful for the first-time user, and it's not so
38 rare that packages are shipped with working-but-broken autotools support. This
39 article will cover some of the most common errors people make when using
40 autotools and ways to achieve better results.
41 </p>
43 <p>
44 Regardless of anyone's opinion about them, we currently have no valid
45 alternative for autotools. Projects such as Scons are not as portable as
46 autotools, and they don't embody enough knowledge to be useful yet. We have
47 tons of automatic checks with autotools, and a lot of libraries come with an m4
48 library with macros to check for their presence.
49 </p>
51 <p>
52 The basic structure of an autotooled project is simple. Autoconf takes help of
53 an <path>aclocal.m4</path> file (created by aclocal using the m4 libraries on
54 its search path and <path>acinclude.m4</path> file) to parse the
55 <path>configure.ac</path> file (formerly <path>configure.in</path>) and
56 transform it into a "configure" script. For every directory there should exist
57 an <path>Makefile.am</path>, which automake uses to create Makefile.in
58 templates. The Makefile.in templates are then processed and transformed by the
59 <path>configure</path> script into real Makefiles. You can avoid avoid using
60 automake and just write your own Makefile.in files, but this is quite complex,
61 and you lose a few features of autotools.
62 </p>
64 <p>
65 In a <path>configure.ac</path> file you can use macros you define yourself, the
66 default ones provided by autoconf and aclocal, or external macros provided, for
67 instance, by other packages. In such a case aclocal will create the
68 <path>aclocal.m4</path> file adding the library files it finds on the system's
69 library with the defined macros; this is a critical step to have a working
70 autotooled project, as we'll see in a moment.
71 </p>
73 <p>
74 A <path>Makefile.am</path> is mainly a declaration of intents: you can fill
75 some targets variables with the name of the targets you want to build. These
76 variables are structured in a format like placetoinstall_TYPEOFTARGET. The
77 place is the location in a hierarchical Unix filesystem (bin, lib, include,
78 ...), a non-used keyword that can be defined with an arbitrary path (using the
79 keyworddir variable), or the special keyword noinst that marks the targets that
80 need not to be installed (for example private headers, or static libraries used
81 during build). After naming the target, you can use the name (replacing dots
82 with underscores) as the prefix for the variables that affects its build. In
83 this way you can provide special CFLAGS, LDFLAGS, and LDADD variables used
84 during the build of a single target, instead of changing them for all the
85 targets. You can also use variables collected during configure phase, if you
86 passed them to the AC_SUBST macro in <path>configure.ac</path>, so that they
87 are replaced inside makefiles. Also, though defining CFLAGS and LDFLAGS on a
88 per-target basis seems useful, adding static flags in Makefile.am is a bad
89 thing for portability, as you can't tell if the compiler you're using supports
90 them, or if you really need them (-ldl put in LDFLAGS is a good example of a
91 flag needed on Linux but not on FreeBSD); in such cases you should use
92 <path>configure.ac</path> to add these flags.
93 </p>
95 <p>
96 The most commonly used macros in configure.ac are AC_CHECK_HEADERS,
97 AC_CHECK_FUNCS, and AC_CHECK_LIB, which test for the presence of, respectively,
98 some header files, some library functions, and a given library (with a specific
99 function in it). They are important for portability as they provides a way to
100 check which headers are present and which are not (for example system headers
101 that have different locations in different operating systems), and to check
102 whether a function is present in the system library (asprintf() is missing in
103 OpenBSD for example, while it's present on GNU C library and FreeBSD), and
104 finally to check for the presence of some third-party library or to see if a
105 specific link to a library is needed to get some functions (for example
106 dlopen() function is in libdl library on GNU systems, while it's provided by
107 the system's C library on FreeBSD).
108 </p>
110 <p>
111 Along with testing for the presence or absence of functions or headers (and
112 sometimes libraries) you usually need to change the code's path (for example to
113 avoid the use of missing functions, or to define a drop-in replacement for
114 them). Autoconf is commonly coupled with another tool, autoheader, which
115 creates a <path>config.h.in</path> template, used by configure script to create
116 a <path>config.h</path> header in which are defined a few preprocessor macros
117 in form of HAVE_givenfunction or HAVE_givenheader_H which can be tested with
118 #ifdef/#ifndef directives inside a C or C++ source file to change the code
119 according to the features present.
120 </p>
122 <p>
123 Here are some practices to keep in mind to help you use autotools to create the
124 most portable code possible.
125 </p>
127 <p>
128 <b>The config.h header file should be considered to be an internal header
129 file</b>, so it should be used just by the single package in which it's
130 created. You should avoid editing the <path>config.h.in</path> template to add
131 your own code there, as this requires you to manually update it according to
132 the <path>configure.ac</path> you're writing.
133 </p>
135 <p>
136 Unfortunately a few projects, such as Net-SNMP, export this header file with
137 other libraries' headers, which requires any projects that use their libraries
138 to include them (or provide their own copy of the internal Net-SNMP
139 structures). This is a bad thing, as the autotools structure of a library
140 project should be invisible to software using it (which might not use autotools
141 at all). Also, changes in autotools behavior are anything but rare, so you can
142 have two identical checks with different results due to changes in the way they
143 are executed. If you need to define your own wrappers or replacements in case
144 something is not in the environment you're compiling for, you should do that in
145 private headers that do not get installed (declared as noinst_HEADERS in
146 <path>Makefile.am</path> files).
147 </p>
149 <p>
150 <b>Always provide the m4 files you used</b>. As autotools have been in use for
151 years, many packages (for example libraries) that can be reused by other
152 programs provide an m4 library file in <path>/usr/share/aclocal</path> that
153 makes it possible to check for their presence (for example using the -config
154 scripts) with a simple macro call. These files are used by aclocal to create
155 the <path>aclocal.m4</path> file, and they usually are present on the
156 developers' systems where aclocal is executed to create the release, but when
157 they are for optional dependencies, they can be missing on users' systems.
158 While this is usually not a problem, because users rarely executes aclocal,
159 it's a problem for source distributions, such as Gentoo, where sometimes you
160 need to patch a Makefile.am or the <path>configure.ac</path> and then re-run
161 autoconf without having all the optional dependencies installed (or having
162 different versions, which can be incompatible or bugged, of the same m4 file).
163 </p>
165 <p>
166 To avoid this problem, you should create an m4 subdirectory in your package's
167 directory and then put there the m4 library files you are using. You must then
168 call aclocal with aclocal -I m4 options to search in that directory before the
169 system library. You can then choose whether to put that directory under
170 revision control (CVS, SVN, or whatever else you are using) or just create it
171 for the releases. The latter case is the bare minimum requirement for a
172 package. It minimizes the amount of revision-controlled code and ensures that
173 you're always using the latest m4 version, but has the drawback that anyone who
174 checks out your repository won't be able to execute autoconf without having to
175 look on a release tarball to take the m4 from (and that might not work, as you
176 can have updated the configure.ac to suit a newer macro or added more
177 dependencies). On the other hand, putting the m4 directory under revision
178 control sometimes tempts the developers to change the macros to suit their
179 needs. Although this seems logical, as the m4 files are under your revision
180 control, it will upset many package maintainers, as sometimes new versions of
181 m4 files fix bugs or support newer options and installation paths (for example
182 multilib setups), and having the m4 files modified makes it impossible to just
183 replace them with updated versions. It also mean that when you're going to
184 update an m4 file you must redo the modification against the original.
185 </p>
187 <p>
188 m4 files are always a problem to work with. They must replicate almost the same
189 code from library to library (depending on the way you need to provide
190 CFLAGS/LDFLAGS: with tests or with a -config script). To avoid this problem,
191 the GNOME and FreeDesktop projects developed a tool called pkg-config, which
192 provides both an executable binary and an m4 file to include in configure.ac
193 files, and lets developers check for the presence of a given library (and/or
194 package), provided that the package itself installed a pkg-config .pc data
195 file. This approach simplifies the work of maintaining configure.ac scripts,
196 and requires a lot less time to be processed during execution of configure
197 script, as it uses the information provided by the installed package itself
198 instead of just trying if it's present. On the other hand, this approach means
199 that an error the developers make concerning a dependency can break the user
200 program, as they just hardcode the compiler and linker flags in the data file
201 and the configure script doesn't actually check whether the library works.
202 Fortunately, this doesn't happen too often.
203 </p>
205 <p>
206 To create the configure file, you need PKG_CHECK_MODULES, contained in the
207 <path>pkg.m4</path> library. You should add that file to your m4 directory. If
208 pkg-config dependency is mandatory (as the tool is run by the configure script)
209 you can't be sure that the m4 file you are using is the same as one on users'
210 systems, nor you can be sure that it does not include extra bugs, as it can be
211 older than yours.
212 </p>
214 <p>
215 <b>Always check for the libraries you're going to link to</b>, if you have them
216 as mandatory dependencies. Usually autoconf macros or pkg-config data files
217 define prerequisite libraries that you need to successfully link to your
218 library. Also, some functions that are in extra libraries in some systems
219 (like dlopen() in libdl on Linux and Mac OS X) can be in the libc of another
220 system (the same function is in libc on FreeBSD). In these cases you need to
221 check whether the function can be found without linking to anything, or if you
222 need to use a specific library (for example to avoid linking to a non-existent
223 libdl that would fail where it's not needed).
224 </p>
226 <p>
227 <b>Be careful with GNU extensions</b>. One of the things that makes portability
228 a big pain is the use of extension functions, which are provided by GNU libc
229 but aren't present on other C libraries like BSD's or uClibc. When you use such
230 functions, you should always provide a "drop-in replacement," a function that
231 can provide the same functionality as the library function, maybe with less
232 performance or security, which can be used when the extension function is not
233 present on system's C library. Those functions must be protected by a #ifdef
234 HAVE_function ... #endif block, so that they don't get duplicated when they are
235 already present. Make sure that these functions are not exported by the library
236 to the external users; they should be declared inside an internal header, to
237 avoid breaking other libraries that may be doing similar tricks.
238 </p>
240 <p>
241 <b>Avoid compiling OS-specific code when not needed</b>. When a program
242 optionally supports specific libraries or specific operating systems, it's not
243 rare to have entire source files that are specific to that code path. To avoid
244 compiling them when they're not needed, use the AM_CONDITIONAL macro inside a
245 configure.ac file. This automake macro (usable only if you're using automake to
246 build the project) allows you to define if .. endif blocks inside a Makefile.am
247 file, inside which you can set special variables. You can, for example, add a
248 "platformsrcs" variable that you set to the right source file for the platform
249 to build for, then use in a _SOURCES variable.
250 </p>
252 <p>
253 However, there are two common errors developers make when using AM_CONDITIONAL.
254 The first is the use of AM_CONDITIONAL in an already conditional branch (for
255 example under an info or in a case switch), which leads to automake complaining
256 about a conditional defined only conditionally (AM_CONDITIONAL must be called
257 on global scope, out of every if block, so you must define a variable to
258 contain the status of the conditions and then test against when calling the
259 AM_CONDITIONAL). The other one is that you can't change the targets' variables
260 directly, and you must define "commodity" variables, whose results empty out of
261 the conditional, to add or remove source files and targets.
262 </p>
264 <p>
265 Many projects, to avoid compiling code for specific code paths, add the entire
266 files in #ifdef ... #endif preprocessor conditionals. While this usually works,
267 it makes the code ugly and error-prone, as a single statement out of the
268 conditional block can be compiled where the source file is not needed. It also
269 misleads users sometimes, as the source files seem to be compiled in situations
270 where they don't make sense.
271 </p>
273 <p>
274 <b>Be smart in looking for operating system or hardware platform</b>. Sometimes
275 you need to search for a specific operating system or hardware platform. The
276 right way to do this depends on where you need to know this. If you must know
277 it to enable extra tests on configure, or you must add extra targets on
278 makefiles, you must do the check in configure.ac. On the other hand, if the
279 difference must be known in a source file, for example to enable an optional
280 asm-coded function, you should rely directly on the compiler/preprocessor, so
281 you should use #ifdef directives with the default macros enabled on the target
282 platform (for example __linux__, __i386__, _ARC_PPC, __sparc__, _FreeBSD_ and
283 __APPLE__).
284 </p>
286 <p>
287 <b>Don't run commands in configure.ac</b>. If you need to check for hardware or
288 operating system in a configure.ac, you should avoid using the uname command,
289 despite this being one of the most common way to do such a test. This is
290 actually an error, as this breaks crosscompilation. Autotools supports
291 crosscompile projects from one machine to another using hosts definitions:
292 strings in the form "hardware-vendor-os" (actually, "hardware-vendor-os-libc"
293 when GNU libc is used), such as i686-pc-linux-gnu and
294 x86_64-unknown-freebsd5.4. CHOST is the host definition for the system you're
295 compiling the software for, CBUILD is the host definition for the system you're
296 compiling on; when CHOST and CBUILD differ, you're crosscompiling.
297 </p>
299 <p>
300 In the examples above, the first host definition shows an x86-like system, with
301 a pentium2-equivalent (or later) processor, running a Linux kernel with a GNU
302 libc (usually this refers to a GNU/Linux system). The second refers to an AMD64
303 system with a FreeBSD 5.4 operating system. (For a GNU/kFreeBSD system, which
304 uses FreeBSD kernel and GNU libc, the host definition is
305 hw-unknown-freebsd-gnu, while for a Gentoo/FreeBSD, using FreeBSD's kernel and
306 libc, but with Gentoo framework, the host definition is hw-gentoo-freebsd5.4.)
307 By using $host and $build variables inside a configure.ac script you can enable
308 or disable specific features based on the operating system or on the hardware
309 platform you're compiling to or on.
310 </p>
312 <p>
313 <b>Don't abuse "automagic" dependencies</b>. One of the most useful features of
314 autotools are the automatic checks for the presence of a library, which are
315 often used to automatically enable support for extra dependencies and such.
316 However, abusing this feature makes the build of a package a bit of a problem.
317 While this is quite useful for first-time users, and although most of the
318 projects having complex dependencies (such as multimedia programs like xine and
319 VLC) use a plugin-based framework that allows them to avoid most of the
320 breakages, "automagic" dependencies are a great pain for packagers, especially
321 ones working on source-based distributions such as Gentoo and ports-like
322 frameworks. When you build something with automagical dependencies you enable
323 the functions supported by the libraries found on the system on which the
324 configure script is run. This means that the output binaries might not work on
325 a system that shares the same base packages but misses one extra library, for
326 example. Also, you can't tell the exact dependencies of a package, as some
327 might be optional and not be built when the libraries are not present.
328 </p>
330 <p>
331 To avoid this, autoconf allows you to add --enable/--disable and
332 --with/--without options to configure scripts. With such options you can
333 forcefully enable or disable a specific option (such as the support for an
334 extra library or for a specific feature), and leave the default to automatic
335 tests.
336 </p>
338 <p>
339 Unfortunately, many developers misunderstand the meaning of the two parameters
340 of the functions used to add those options (AC_ARG_ENABLE and AC_ARG_WITH).
341 They represent the code to execute when a parameter is passed and when one is
342 not. Many developers mistakenly think that the two parameters define the code
343 to execute when the feature is enabled and when is disabled. While this usually
344 works when you pass a parameter just to change the default behavior, many
345 source-based distributions pass parameters also to confirm the default
346 behavior, which leads to errors (features explicitly requested missing). Being
347 able to disable optional features if they don't add dependencies (think of OSS
348 audio support on Linux) is always a good thing for users, who can avoid
349 building extra code if they don't plan to use it, and prevents maintainers from
350 doing dirty caching tricks to enable or disable features as their users
351 request.
352 </p>
354 <p>
355 While autotools were a big problem for both developers and maintainers because
356 there are different incompatible versions that do not get along well together
357 (since they install in the same places, with the same names) and which are used
358 in different combinations, the use of autotools saves maintainers from doing
359 all sorts of dirty tricks to compile software. If you look at ebuild from
360 Gentoo's portage, the few that do not use autotools are the more complex ones,
361 as they need to check variables on very different setups (we can or not have
362 NPTL support; we can be on Linux, FreeBSD, or Mac OS X; we can be using GLIBC
363 or another libc; and so on), while autotools usually take care of that on their
364 own. It's also true that many patches applied by maintainers are to fix broken
365 autotools script in upstream sources, but this is just a little problem
366 compared to the chaos of using special build systems that don't work at all
367 with little environmental changes.
368 </p>
370 <p>
371 Autotools can be quite tricky for newcomers, but when you start using them on a
372 daily basis you find it's a lot easier than having to deal with manual
373 makefiles or other strange build tools such as imake or qmake, or even worse,
374 special autotools-like build scripts that try to recognize the system they are
375 building on. Autotools makes it simple to support new OSes and new hardware
376 platforms, and saves maintainers and porters from having to learn how to
377 custom-build a system to fix compilation. By carefully writing a script,
378 developers can support new platforms without any changes at all.
379 </p>
381 </body>
382 </section>
383 </chapter>
384 </guide>

  ViewVC Help
Powered by ViewVC 1.1.20