A physicist, an engineer, and a computer scientist were discussing the nature of God. ``Surely a Physicist,'' said the physicist, ``because early in the Creation, God made Light; and you know, Maxwell's equations, the dual nature of electromagnetic waves, the relativistic consequences...'' ``An Engineer!,'' said the engineer, ``because before making Light, God split the Chaos into Land and Water; it takes a hell of an engineer to handle that big amount of mud, and orderly separation of solids from liquids...'' The computer scientist shouted: ``And the Chaos, where do you think it was coming from, hmm?'' ---Anonymous
Autoconf is a tool for producing shell scripts that automatically configure software source code packages to adapt to many kinds of UNIX-like systems. The configuration scripts produced by Autoconf are independent of Autoconf when they are run, so their users do not need to have Autoconf.
The configuration scripts produced by Autoconf require no manual user intervention when run; they do not normally even need an argument specifying the system type. Instead, they individually test for the presence of each feature that the software package they are for might need. (Before each check, they print a one-line message stating what they are checking for, so the user doesn't get too bored while waiting for the script to finish.) As a result, they deal well with systems that are hybrids or customized from the more common UNIX variants. There is no need to maintain files that list the features supported by each release of each variant of UNIX.
For each software package that Autoconf is used with, it creates a configuration script from a template file that lists the system features that the package needs or can use. After the shell code to recognize and respond to a system feature has been written, Autoconf allows it to be shared by many software packages that can use (or need) that feature. If it later turns out that the shell code needs adjustment for some reason, it needs to be changed in only one place; all of the configuration scripts can be regenerated automatically to take advantage of the updated code.
The Metaconfig package is similar in purpose to Autoconf, but the scripts it produces require manual user intervention, which is quite inconvenient when configuring large source trees. Unlike Metaconfig scripts, Autoconf scripts can support cross-compiling, if some care is taken in writing them.
Autoconf does not solve all problems related to making portable software packages--for a more complete solution, it should be used in concert with other GNU build tools like Automake and Libtool. These other tools take on jobs like the creation of a portable, recursive `Makefile' with all of the standard targets, linking of shared libraries, and so on. See section The GNU build system, for more information.
Autoconf imposes some restrictions on the names of macros used with
#if in C programs (see section Preprocessor Symbol Index).
Autoconf requires GNU M4 in order to generate the scripts. It uses features that some UNIX versions of M4, including GNU M4 1.3, do not have. You must use version 1.4 or later of GNU M4.
See section Upgrading From Version 1, for information about upgrading from version 1. See section History of Autoconf, for the story of Autoconf's development. See section Questions About Autoconf, for answers to some common questions about Autoconf.
See the @href{http://www.gnu.org/software/autoconf/autoconf.html, Autoconf web page} for up-to-date information, details on the mailing lists, pointers to a list of known bugs, etc.
Mail suggestions to the Autoconf mailing list.
Bug reports should be preferably submitted to the @href{http://sources.redhat.com/cgi-bin/gnatsweb.pl?database=autoconf, Autoconf Gnats database}, or sent to the Autoconf Bugs mailing list. If possible, first check that your bug is not already solved in current development versions, and that it has not been reported yet. Be sure to include all the needed information and a short `configure.ac' that demonstrates the problem.
Autoconf's development tree is accessible via CVS; see the Autoconf web page for details. There is also a @href{http://subversions.gnu.org/cgi-bin/cvsweb/autoconf/, CVSweb interface to the Autoconf development tree}. Patches relative to the current CVS version can be sent for review to the Autoconf Patches mailing list.
Because of its mission, Autoconf includes only a set of often-used macros that have already demonstrated their usefulness. Nevertheless, if you wish to share your macros, or find existing ones, see the @href{http://www.gnu.org/software/ac-archive/, Autoconf Macro Archive}, which is kindly run by Peter Simons.
Autoconf solves an important problem--reliable discovery of system-specific build and runtime information--but this is only one piece of the puzzle for the development of portable software. To this end, the GNU project has developed a suite of integrated utilities to finish the job Autoconf started: the GNU build system, whose most important components are Autoconf, Automake, and Libtool. In this chapter, we introduce you to those tools, point you to sources of more information, and try to convince you to use the entire GNU build system for your software.
The ubiquity of make means that a Makefile is almost the
only viable way to distribute automatic build rules for software, but
one quickly runs into make's numerous limitations. Its lack of
support for automatic dependency tracking, recursive builds in
subdirectories, reliable timestamps (e.g. for network filesystems), and
so on, mean that developers must painfully (and often incorrectly)
reinvent the wheel for each project. Portability is non-trivial, thanks
to the quirks of make on many systems. On top of all this is the
manual labor required to implement the many standard targets that users
have come to expect (make install, make distclean,
make uninstall, etc.). Since you are, of course, using Autoconf,
you also have to insert repetitive code in your Makefile.in to
recognize @CC@, @CFLAGS@, and other substitutions
provided by @command{configure}. Into this mess steps Automake.
Automake allows you to specify your build needs in a Makefile.am
file with a vastly simpler and more powerful syntax than that of a plain
Makefile, and then generates a portable Makefile.in for
use with Autoconf. For example, the Makefile.am to build and
install a simple "Hello world" program might look like:
bin_PROGRAMS = hello hello_SOURCES = hello.c
The resulting Makefile.in (~400 lines) automatically supports all
the standard targets, the substitutions provided by Autoconf, automatic
dependency tracking, VPATH building, and so on. make will
build the hello program, and make install will install it
in `/usr/local/bin' (or whatever prefix was given to
@command{configure}, if not `/usr/local').
Automake may require that additional tools be present on the
developer's machine. For example, the Makefile.in that
the developer works with may not be portable (e.g. it might use special
features of your compiler to automatically generate dependency
information). Running make dist, however, produces a
`hello-1.0.tar.gz' package (or whatever the program/version is)
with a Makefile.in that will work on any system.
The benefits of Automake increase for larger packages (especially ones with subdirectories), but even for small programs the added convenience and portability can be substantial. And that's not all...
Very often, one wants to build not only programs, but libraries, so that other programs can benefit from the fruits of your labor. Ideally, one would like to produce shared (dynamically-linked) libraries, which can be used by multiple programs without duplication on disk or in memory and can be updated independently of the linked programs. Producing shared libraries portably, however, is the stuff of nightmares--each system has its own incompatible tools, compiler flags, and magic incantations. Fortunately, GNU provides a solution: Libtool.
Libtool handles all the requirements of building shared libraries for
you, and at this time seems to be the only way to do so with any
portability. It also handles many other headaches, such as: the
interaction of Makefile rules with the variable suffixes of
shared libraries, linking reliably to shared libraries before they are
installed by the superuser, and supplying a consistent versioning system
(so that different versions of a library can be installed or upgraded
without breaking binary compatibility). Although Libtool, like
Autoconf, can be used on its own, it is most simply utilized in
conjunction with Automake--there, Libtool is used automatically
whenever shared libraries are needed, and you need not know its syntax.
Developers who are used to the simplicity of make for small
projects on a single system might be daunted at the prospect of learning
to use Automake and Autoconf. As your software is distributed to more
and more users, however, you will otherwise quickly find yourself
putting lots of effort into reinventing the services that the GNU build
tools provide, and making the same mistakes that they once made and
overcame. (Besides, since you're already learning Autoconf, Automake
will be a piece of cake.)
There are a number of places that you can go to for more information on the GNU build tools.
The configuration scripts that Autoconf produces are by convention called @command{configure}. When run, @command{configure} creates several files, replacing configuration parameters in them with appropriate values. The files that @command{configure} creates are:
#define directives (see section Configuration Header Files);
To create a @command{configure} script with Autoconf, you need to write an
Autoconf input file `configure.ac' (or `configure.in') and run
@command{autoconf} on it. If you write your own feature tests to
supplement those that come with Autoconf, you might also write files
called `aclocal.m4' and `acsite.m4'. If you use a C header
file to contain #define directives, you might also run
@command{autoheader}, and you will distribute the generated file
`config.h.in' with the package.
Here is a diagram showing how the files that can be used in configuration are produced. Programs that are executed are suffixed by `*'. Optional files are enclosed in square brackets (`[]'). @command{autoconf} and @command{autoheader} also read the installed Autoconf macro files (by reading `autoconf.m4').
Files used in preparing a software package for distribution:
your source files --> [autoscan*] --> [configure.scan] --> configure.ac
configure.ac --.
| .------> autoconf* -----> configure
[aclocal.m4] --+---+
| `-----> [autoheader*] --> [config.h.in]
[acsite.m4] ---'
Makefile.in -------------------------------> Makefile.in
Files used in configuring a software package:
.-------------> [config.cache]
configure* ------------+-------------> config.log
|
[config.h.in] -. v .-> [config.h] -.
+--> config.status* -+ +--> make*
Makefile.in ---' `-> Makefile ---'
To produce a @command{configure} script for a software package, create a file called `configure.ac' that contains invocations of the Autoconf macros that test the system features your package needs or can use. Autoconf macros already exist to check for many features; see section Existing Tests, for their descriptions. For most other features, you can use Autoconf template macros to produce custom checks; see section Writing Tests, for information about them. For especially tricky or specialized features, `configure.ac' might need to contain some hand-crafted shell commands; see section Portable Shell Programming. The @command{autoscan} program can give you a good start in writing `configure.ac' (see section Using @command{autoscan} to Create @file{configure.ac}, for more information).
Previous versions of Autoconf promoted the name `configure.in', which is somewhat ambiguous (the tool needed to produce this file is not described by its extension), and introduces a slight confusion with `config.h.in' and so on (for which `.in' means "to be processed by @command{configure}"). Using `configure.ac' is now preferred.
Just as for any other computer language, in order to properly program `configure.ac' in Autoconf you must understand what problem the language tries to address and how it does so.
The problem Autoconf addresses is that the world is a mess. After all, you are using Autoconf in order to have your package compile easily on all sorts of different systems, some of them being extremely hostile. Autoconf itself bears the price for these differences: @command{configure} must run on all those systems, and thus @command{configure} must limit itself to their lowest common denominator of features.
Naturally, you might then think of shell scripts; who needs @command{autoconf}? A set of properly written shell functions is enough to make it easy to write @command{configure} scripts by hand. Sigh! Unfortunately, shell functions do not belong to the least common denominator; therefore, where you would like to define a function and use it ten times, you would instead need to copy its body ten times.
So, what is really needed is some kind of compiler, @command{autoconf}, that takes an Autoconf program, `configure.ac', and transforms it into a portable shell script, @command{configure}.
How does @command{autoconf} perform this task?
There are two obvious possibilities: creating a brand new language or
extending an existing one. The former option is very attractive: all
sorts of optimizations could easily be implemented in the compiler and
many rigorous checks could be performed on the Autoconf program
(e.g. rejecting any non-portable construct). Alternatively, you can
extend an existing language, such as the sh (Bourne shell)
language.
Autoconf does the latter: it is a layer on top of sh. It was
therefore most convenient to implement @command{autoconf} as a macro
expander: a program that repeatedly performs macro expansions on
text input, replacing macro calls with macro bodies and producing a pure
sh script in the end. Instead of implementing a dedicated
Autoconf macro expander, it is natural to use an existing
general-purpose macro language, such as M4, and implement the extensions
as a set of M4 macros.
The Autoconf language is very different from many other computer languages because it treats actual code the same as plain text. Whereas in C, for instance, data and instructions have very different syntactic status, in Autoconf their status is rigorously the same. Therefore, we need a means to distinguish literal strings from text to be expanded: quotation.
When calling macros that take arguments, there must not be any blank space between the macro name and the open parenthesis. Arguments should be enclosed within the M4 quote characters `[' and `]', and be separated by commas. Any leading spaces in arguments are ignored, unless they are quoted. You may safely leave out the quotes when the argument is simple text, but always quote complex arguments such as other macro calls. This rule applies recursively for every macro call, including macros called from other macros.
For instance:
AC_CHECK_HEADER([stdio.h],
[AC_DEFINE([HAVE_STDIO_H])],
[AC_MSG_ERROR([Sorry, can't do anything for you])])
is quoted properly. You may safely simplify its quotation to:
AC_CHECK_HEADER(stdio.h,
[AC_DEFINE(HAVE_STDIO_H)],
[AC_MSG_ERROR([Sorry, can't do anything for you])])
Notice that the argument of AC_MSG_ERROR is still quoted;
otherwise, its comma would have been interpreted as an argument separator.
The following example is wrong and dangerous, as it is underquoted:
AC_CHECK_HEADER(stdio.h,
AC_DEFINE(HAVE_STDIO_H),
AC_MSG_ERROR([Sorry, can't do anything for you]))
In other cases, you may have to use text that also resembles a macro call. You must quote that text even when it is not passed as a macro argument:
echo "Hard rock was here! --[AC_DC]"
which will result in
echo "Hard rock was here! --AC_DC"
When you use the same text in a macro argument, you must therefore have an extra quotation level (since one is stripped away by the macro substitution). In general, then, it is a good idea to use double quoting for all literal string arguments:
AC_MSG_WARN([[AC_DC stinks --Iron Maiden]])
You are now able to understand one of the constructs of Autoconf that has been continually misunderstood... The rule of thumb is that whenever you expect macro expansion, expect quote expansion; i.e., expect one level of quotes to be lost. For instance:
AC_COMPILE_IFELSE([char b[10];],, [AC_MSG_ERROR([you lose])])
is incorrect: here, the first argument of AC_COMPILE_IFELSE is
`char b[10];' and will be expanded once, which results in
`char b10;'. (There was an idiom common in Autoconf's past to
address this issue via the M4 changequote primitive, but do not
use it!) Let's take a closer look: the author meant the first argument
to be understood as a literal, and therefore it must be quoted twice:
AC_COMPILE_IFELSE([[char b[10];]],, [AC_MSG_ERROR([you lose])])
Voil`a, you actually produce `char b[10];' this time!
The careful reader will notice that, according to these guidelines, the
"properly" quoted AC_CHECK_HEADER example above is actually
lacking three pairs of quotes! Nevertheless, for the sake of readability,
double quotation of literals is used only where needed in this manual.
Some macros take optional arguments, which this documentation represents as @ovar{arg} (not to be confused with the quote characters). You may just leave them empty, or use `[]' to make the emptiness of the argument explicit, or you may simply omit the trailing commas. The three lines below are equivalent:
AC_CHECK_HEADERS(stdio.h, [], [], []) AC_CHECK_HEADERS(stdio.h,,,) AC_CHECK_HEADERS(stdio.h)
It is best to put each macro call on its own line in `configure.ac'. Most of the macros don't add extra newlines; they rely on the newline after the macro call to terminate the commands. This approach makes the generated @command{configure} script a little easier to read by not inserting lots of blank lines. It is generally safe to set shell variables on the same line as a macro call, because the shell allows assignments without intervening newlines.
You can include comments in `configure.ac' files by starting them with the `#'. For example, it is helpful to begin `configure.ac' files with a line like this:
# Process this file with autoconf to produce a configure script.
The order in which `configure.ac' calls the Autoconf macros is not
important, with a few exceptions. Every `configure.ac' must
contain a call to AC_INIT before the checks, and a call to
AC_OUTPUT at the end (see section Outputting Files). Additionally, some macros
rely on other macros having been called first, because they check
previously set values of some variables to decide what to do. These
macros are noted in the individual descriptions (see section Existing Tests), and they also warn you when @command{configure} is created if they
are called out of order.
To encourage consistency, here is a suggested order for calling the Autoconf macros. Generally speaking, the things near the end of this list are those that could depend on things earlier in it. For example, library functions could be affected by types and libraries.
Autoconf requirementsAC_INIT(package, version, bug-report-address)information on the package checks for programs checks for libraries checks for header files checks for types checks for structures checks for compiler characteristics checks for library functions checks for system servicesAC_CONFIG_FILES([file...])AC_OUTPUT
The @command{autoscan} program can help you create and/or maintain a `configure.ac' file for a software package. @command{autoscan} examines source files in the directory tree rooted at a directory given as a command line argument, or the current directory if none is given. It searches the source files for common portability problems and creates a file `configure.scan' which is a preliminary `configure.ac' for that package, and checks a possibly existing `configure.ac' for completeness.
When using @command{autoscan} to create a `configure.ac', you
should manually examine `configure.scan' before renaming it to
`configure.ac'; it will probably need some adjustments.
Occasionally, @command{autoscan} outputs a macro in the wrong order
relative to another macro, so that @command{autoconf} produces a warning;
you need to move such macros manually. Also, if you want the package to
use a configuration header file, you must add a call to
AC_CONFIG_HEADERS (see section Configuration Header Files). You might
also have to change or add some #if directives to your program in
order to make it work with Autoconf (see section Using @command{ifnames} to List Conditionals, for
information about a program that can help with that job).
When using @command{autoscan} to maintain a `configure.ac', simply consider adding its suggestions. The file `autoscan.log' will contain detailed information on why a macro is requested.
@command{autoscan} uses several data files (installed along with Autoconf) to determine which macros to output when it finds particular symbols in a package's source files. These data files all have the same format: each line consists of a symbol, whitespace, and the Autoconf macro to output if that symbol is encountered. Lines starting with `#' are comments.
@command{autoscan} accepts the following options:
@command{ifnames} can help you write `configure.ac' for a software package. It prints the identifiers that the package already uses in C preprocessor conditionals. If a package has already been set up to have some portability, @command{ifnames} can thus help you figure out what its @command{configure} needs to check for. It may help fill in some gaps in a `configure.ac' generated by @command{autoscan} (see section Using @command{autoscan} to Create @file{configure.ac}).
@command{ifnames} scans all of the C source files named on the command line
(or the standard input, if none are given) and writes to the standard
output a sorted list of all the identifiers that appear in those files
in #if, #elif, #ifdef, or #ifndef
directives. It prints each identifier on a line, followed by a
space-separated list of the files in which that identifier occurs.
@command{ifnames} accepts the following options:
To create @command{configure} from `configure.ac', run the
@command{autoconf} program with no arguments. @command{autoconf} processes
`configure.ac' with the m4 macro processor, using the
Autoconf macros. If you give @command{autoconf} an argument, it reads that
file instead of `configure.ac' and writes the configuration script
to the standard output instead of to @command{configure}. If you give
@command{autoconf} the argument @option{-}, it reads from the standard
input instead of `configure.ac' and writes the configuration script
to the standard output.
The Autoconf macros are defined in several files. Some of the files are distributed with Autoconf; @command{autoconf} reads them first. Then it looks for the optional file `acsite.m4' in the directory that contains the distributed Autoconf macro files, and for the optional file `aclocal.m4' in the current directory. Those files can contain your site's or the package's own Autoconf macro definitions (see section Writing Autoconf Macros, for more information). If a macro is defined in more than one of the files that @command{autoconf} reads, the last definition it reads overrides the earlier ones.
@command{autoconf} accepts the following options:
AC_DIAGNOSE, for a comprehensive list of categories. Special
values include:
WARNINGS, a comma separated list of categories, is
honored. @command{autoconf -W category} will actually
behave as if you had run:
autoconf --warnings=syntax,$WARNINGS,categoryIf you want to disable @command{autoconf}'s defaults and
WARNINGS,
but (for example) enable the warnings about obsolete constructs, you
would use @option{-W none,obsolete}.
@command{autoconf} displays a back trace for errors, but not for
warnings; if you want them, just pass @option{-W error}. For instance,
on this `configure.ac':
AC_DEFUN([INNER], [AC_TRY_RUN([exit (0)])]) AC_DEFUN([OUTER], [INNER]) AC_INIT OUTERyou get:
$ autoconf -Wcross configure.ac:8: warning: AC_TRY_RUN called without default \ to allow cross compiling $ autoconf -Wcross,error configure.ac:8: error: AC_TRY_RUN called without default \ to allow cross compiling acgeneral.m4:3044: AC_TRY_RUN is expanded from... configure.ac:2: INNER is expanded from... configure.ac:5: OUTER is expanded from... configure.ac:8: the top level
AC_DEFUN definitions). This
results in a noticeable speedup, but can be disabled by this option.
It is often necessary to check the content of a `configure.ac' file, but parsing it yourself is extremely fragile and error-prone. It is suggested that you rely upon @option{--trace} to scan `configure.ac'.
The format of @option{--trace} can use the following special escapes:
For instance, to find the list of variables that are substituted, use:
$ autoconf -t AC_SUBST configure.ac:2:AC_SUBST:ECHO_C configure.ac:2:AC_SUBST:ECHO_N configure.ac:2:AC_SUBST:ECHO_T More traces deleted
The example below highlights the difference between `$@', `$*', and $%.
$ cat configure.ac AC_DEFINE(This, is, [an [example]]) $ autoconf -t 'AC_DEFINE:@: $@ *: $* $: $%' @: [This],[is],[an [example]] *: This,is,an [example] $: This:is:an [example]
The format gives you a lot of freedom:
$ autoconf -t 'AC_SUBST:$$ac_subst{"$1"} = "$f:$l";'
$ac_subst{"ECHO_C"} = "configure.ac:2";
$ac_subst{"ECHO_N"} = "configure.ac:2";
$ac_subst{"ECHO_T"} = "configure.ac:2";
More traces deleted
A long separator can be used to improve the readability of complex structures, and to ease its parsing (for instance when no single character is suitable as a separator)):
$ autoconf -t 'AM_MISSING_PROG:${|:::::|}*'
ACLOCAL|:::::|aclocal|:::::|$missing_dir
AUTOCONF|:::::|autoconf|:::::|$missing_dir
AUTOMAKE|:::::|automake|:::::|$missing_dir
More traces deleted
Installing the various components of the GNU Build System can be tedious: running @command{gettextize}, @command{automake} etc. in each directory. It may be needed either because some tools such as @command{automake} have been updated on your system, or because some of the sources such as `configure.ac' have been updated, or finally, simply in order to install the GNU Build System in a fresh tree.
It runs @command{autoconf}, @command{autoheader}, @command{aclocal}, @command{automake}, @command{libtoolize}, and @command{gettextize} (when appropriate) repeatedly to update the GNU Build System in specified directories, and their subdirectories (see section Configuring Other Packages in Subdirectories). By default, it only remakes those files that are older than their sources.
If you install a new version of some tools, you can make @command{autoreconf} remake all of the files by giving it the @option{--force} option.
See section Automatic Remaking, for `Makefile' rules to automatically remake @command{configure} scripts when their source files change. That method handles the timestamps of configuration header templates properly, but does not pass @option{--autoconf-dir=dir} or @option{--localdir=dir}.
@command{autoreconf} accepts the following options:
--add-missing in @command{automake}.
Autoconf-generated @command{configure} scripts need some information about how to initialize, such as how to find the package's source files; and about the output files to produce. The following sections describe initialization and the creation of output files.
Every @command{configure} script must call AC_INIT before doing
anything else. The only other required macro is AC_OUTPUT
(see section Outputting Files).
Set the name of the package and its version. These are typically used in @option{--version} support, including that of @command{configure}. The optional argument bug-report should be the email to which users should send bug reports. The package tarname differs from package: the latter designates the full package name (e.g., `GNU Autoconf'), while the latter is meant for distribution tar ball names (e.g., `autoconf'). It defaults to package once `GNU ' strip, lower cased, and all non alphanumeric character mapped onto `-'.
It is preferable that these arguments be static, i.e., there should not
be any shell computation, but they can be computed by M4. The following
M4 macros (e.g., AC_PACKAGE_NAME), output variables (e.g.,
PACKAGE_NAME), and preprocessor symbols (e.g.,
PACKAGE_NAME) are then defined:
AC_PACKAGE_NAME, PACKAGE_NAME
AC_PACKAGE_TARNAME, PACKAGE_TARNAME
AC_PACKAGE_VERSION, PACKAGE_VERSION
AC_PACKAGE_STRING, PACKAGE_STRING
AC_PACKAGE_BUGREPORT, PACKAGE_BUGREPORT
The following macros manage version numbers for @command{configure} scripts. Using them is optional.
AC_PREREQ(2.53)
This macro is the only macro that may be used before AC_INIT, but
for consistency, you are invited not to do so.
The copyright-notice will show up in both the head of @command{configure} and in `configure --version'.
cvs changing it when you check in
@command{configure}. That way, you can determine easily which revision of
`configure.ac' a particular @command{configure} corresponds to.
For example, this line in `configure.ac':
AC_REVISION($Revision: 1.1 $)
produces this in @command{configure}:
#! /bin/sh # From configure.ac Revision: 1.30
Packages that do manual configuration or use the install program
might need to tell @command{configure} where to find some other shell
scripts by calling AC_CONFIG_AUX_DIR, though the default places
it looks are correct for most cases.
AC_PROG_INSTALL does not automatically require distributing the
other auxiliary files. It checks for `install.sh' also, but that
name is obsolete because some @command{make} have a rule that creates
`install' from it if there is no `Makefile'.
Every Autoconf script, e.g., `configure.ac', should finish by
calling AC_OUTPUT. It is the macro that generates
`config.status', which will create the `Makefile's and any
other files resulting from configuration. The only required macro is
AC_INIT (see section Finding @command{configure} Input).
`config.status' will take all the configuration actions: all the
output files (see section Creating Configuration Files, macro
AC_CONFIG_FILES), header files (see section Configuration Header Files,
macro AC_CONFIG_HEADERS), commands (see section Running Arbitrary Configuration Commands, macro AC_CONFIG_COMMANDS), links (see
section Creating Configuration Links, macro AC_CONFIG_LINKS), subdirectories
to configure (see section Configuring Other Packages in Subdirectories, macro AC_CONFIG_SUBDIRS)
are honored.
Historically, the usage of AC_OUTPUT was somewhat different.
See section Obsolete Macros, for a description of the arguments that
AC_OUTPUT used to support.
If you run make on subdirectories, you should run it using the
make variable MAKE. Most versions of make set
MAKE to the name of the make program plus any options it
was given. (But many do not include in it the values of any variables
set on the command line, so those are not passed on automatically.)
Some old versions of make do not set this variable. The
following macro allows you to use it even with those versions.
make predefines the variable MAKE, define output
variable SET_MAKE to be empty. Otherwise, define SET_MAKE
to contain `MAKE=make'. Calls AC_SUBST for SET_MAKE.
To use this macro, place a line like this in each `Makefile.in'
that runs MAKE on other directories:
@SET_MAKE@
`configure' is designed so that it appears to do everything itself, but there is actually a hidden slave: `config.status'. `configure' is in charge of examining your system, but it is `config.status' that actually takes the proper actions based on the results of `configure'. The most typical task of `config.status' is to instantiate files.
This section describes the common behavior of the four standard
instantiating macros: AC_CONFIG_FILES, AC_CONFIG_HEADERS,
AC_CONFIG_COMMANDS and AC_CONFIG_LINKS. They all
have this prototype:
AC_CONFIG_FOOS(tag..., [commands], [init-cmds])
where the arguments are:
... && my_foos="$my_foos fooo" ... && my_foos="$my_foos foooo" AC_CONFIG_FOOS($my_foos)and use this instead:
... && AC_CONFIG_FOOS(fooo) ... && AC_CONFIG_FOOS(foooo)The macros
AC_CONFIG_FILES and AC_CONFIG_HEADERS use
special tags: they may have the form `output' or
`output:inputs'. The file output is instantiated
from its templates, inputs (defaulting to `output.in').
For instance
`AC_CONFIG_FILES(Makefile:boiler/top.mk:boiler/bot.mk)' asks for
the creation of `Makefile' that will be the expansion of the
output variables in the concatenation of `boiler/top.mk' and
`boiler/bot.mk'.
The special value `-' might be used to denote the standard output
when used in output, or the standard input when used in the
inputs. You most probably don't need to use this in
`configure.ac', but it is convenient when using the command line
interface of `./config.status', see section Recreating a Configuration,
for more details.
The inputs may be absolute or relative filenames. In the latter
case they are first looked for in the build tree, and then in the source
tree.
srcdir
ac_top_srcdir
ac_top_builddir
ac_srcdir
AC_CONFIG_COMMANDS([deep/dir/out:in/in.in], [...], [...])with @option{--srcdir=../package} produces the following values:
# Argument of --srcdir srcdir='../package' # Reversing deep/dir ac_top_builddir='../../' # Concatenation of $ac_top_builddir and srcdir ac_top_srcdir='../../../package' # Concatenation of $ac_top_srcdir and deep/dir ac_srcdir='../../../package/deep/dir'independently of `in/in.in'.
var. init-cmds
is typically used by `configure' to give `config.status' some
variables it needs to run the commands.
You should be extremely cautious in your variable names: all the
init-cmds share the same name space and may overwrite each other
in unpredictable ways. Sorry...
All these macros can be called multiple times, with different tags, of course!
Be sure to read the previous section, section Taking Configuration Actions.
AC_OUTPUT create each `file' by copying an input
file (by default `file.in'), substituting the output variable
values.
This macro is one of the instantiating macros, see section Taking Configuration Actions. See section Substitutions in Makefiles, for more information on using
output variables. See section Setting Output Variables, for more information
on creating them. This macro creates the directory that the file is in
if it doesn't exist. Usually, `Makefile's are created this way,
but other files, such as `.gdbinit', can be specified as well.
Typical calls to AC_CONFIG_FILES look like this:
AC_CONFIG_FILES([Makefile src/Makefile man/Makefile X/Imakefile]) AC_CONFIG_FILES([autoconf], [chmod +x autoconf])
You can override an input file name by appending to file a colon-separated list of input files. Examples:
AC_CONFIG_FILES([Makefile:boiler/top.mk:boiler/bot.mk]
[lib/Makefile:boiler/lib.mk])
Doing this allows you to keep your file names acceptable to MS-DOS, or to prepend and/or append boilerplate to the file.
Each subdirectory in a distribution that contains something to be
compiled or installed should come with a file `Makefile.in', from
which @command{configure} will create a `Makefile' in that directory.
To create a `Makefile', @command{configure} performs a simple variable
substitution, replacing occurrences of `@variable@' in
`Makefile.in' with the value that @command{configure} has determined
for that variable. Variables that are substituted into output files in
this way are called output variables. They are ordinary shell
variables that are set in @command{configure}. To make @command{configure}
substitute a particular variable into the output files, the macro
AC_SUBST must be called with that variable name as an argument.
Any occurrences of `@variable@' for other variables are
left unchanged. See section Setting Output Variables, for more information
on creating output variables with AC_SUBST.
A software package that uses a @command{configure} script should be distributed with a file `Makefile.in', but no `Makefile'; that way, the user has to properly configure the package for the local system before compiling it.
See section `Makefile Conventions' in The GNU Coding Standards, for more information on what to put in `Makefile's.
Some output variables are preset by the Autoconf macros. Some of the
Autoconf macros set additional output variables, which are mentioned in
the descriptions for those macros. See section Output Variable Index, for a
complete list of output variables. See section Installation Directory Variables, for the list of the preset ones related to installation
directories. Below are listed the other preset ones. They all are
precious variables (see section Setting Output Variables,
AC_ARG_VAR).
AC_PROG_CC (or empty if you don't). @command{configure}
uses this variable when compiling programs to test for C features.
AC_OUTPUT adds a comment line containing this variable to the top
of every `Makefile' it creates. For other files, you should
reference this variable in a comment at the top of each input file. For
example, an input shell script should begin like this:
#! /bin/sh # @configure_input@
The presence of that line also reminds people editing the file that it needs to be processed by @command{configure} in order to be used.
AC_PROG_CXX (or empty if you don't).
@command{configure} uses this variable when compiling programs to test for
C++ features.
AC_CONFIG_HEADERS
is called, @command{configure} replaces `@DEFS@' with
@option{-DHAVE_CONFIG_H} instead (see section Configuration Header Files). This
variable is not defined while @command{configure} is performing its tests,
only when creating the output files. See section Setting Output Variables, for
how to check the results of previous tests.
echo for
question-answer message pairs? These variables provide a way:
echo $ECHO_N "And the winner is... $ECHO_C"
sleep 100000000000
echo "${ECHO_T}dead."
Some old and uncommon echo implementations offer no means to
achieve this, in which case ECHO_T is set to tab. You might not
want to use it.
AC_PROG_F77 (or empty if you don't).
@command{configure} uses this variable when compiling programs to test for
Fortran 77 features.
LIBS instead. If it is not set
in the environment when @command{configure} runs, the default value is empty.
@command{configure} uses this variable when linking programs to test for
C, C++ and Fortran 77 features.
builddir.
srcbuild.
top_builddir.
srcdir.
srcdir.
top_srcdir.
The following variables specify the directories where the package will be installed, see section `Variables for Installation Directories' in The GNU Coding Standards, for more information. See the end of this section for details on when and how to use these variables.
Most of these variables have values that rely on prefix or
exec_prefix. It is deliberate that the directory output
variables keep them unexpanded: typically `@datadir@' will be
replaced by `${prefix}/share', not `/usr/local/share'.
This behavior is mandated by the GNU coding standards, so that when the user runs:
In order to support these features, it is essential that datadir
remains being defined as `${prefix}/share' to depend upon the
current value of prefix.
A corollary is that you should not use these variables except in
Makefiles. For instance, instead of trying to evaluate datadir
in `configure' and hardcoding it in Makefiles using
e.g. `AC_DEFINE_UNQUOTED(DATADIR, "$datadir")', you should add
`-DDATADIR="$(datadir)"' to your CPPFLAGS.
Similarly you should not rely on AC_OUTPUT_FILES to replace
datadir and friends in your shell scripts and other files, rather
let @command{make} manage their replacement. For instance Autoconf
ships templates of its shell scripts ending with `.sh', and uses
this Makefile snippet:
.sh:
rm -f $@ $@.tmp
sed 's,@datadir\@,$(pkgdatadir),g' $< >$@.tmp
chmod +x $@.tmp
mv $@.tmp $@
Three things are noteworthy:
You can support compiling a software package for several architectures simultaneously from the same copy of the source code. The object files for each architecture are kept in their own directory.
To support doing this, make uses the VPATH variable to
find the files that are in the source directory. GNU make
and most other recent make programs can do this. Older
make programs do not support VPATH; when using them, the
source code must be in the same directory as the object files.
To support VPATH, each `Makefile.in' should contain two
lines that look like:
srcdir = @srcdir@ VPATH = @srcdir@
Do not set VPATH to the value of another variable, for example
`VPATH = $(srcdir)', because some versions of make do not do
variable substitutions on the value of VPATH.
@command{configure} substitutes in the correct value for srcdir when
it produces `Makefile'.
Do not use the make variable $<, which expands to the
file name of the file in the source directory (found with VPATH),
except in implicit rules. (An implicit rule is one such as `.c.o',
which tells how to create a `.o' file from a `.c' file.) Some
versions of make do not set $< in explicit rules; they
expand it to an empty value.
Instead, `Makefile' command lines should always refer to source files by prefixing them with `$(srcdir)/'. For example:
time.info: time.texinfo
$(MAKEINFO) $(srcdir)/time.texinfo
You can put rules like the following in the top-level `Makefile.in' for a package to automatically update the configuration information when you change the configuration files. This example includes all of the optional files, such as `aclocal.m4' and those related to configuration header files. Omit from the `Makefile.in' rules for any of these files that your package does not use.
The `$(srcdir)/' prefix is included because of limitations in the
VPATH mechanism.
The `stamp-' files are necessary because the timestamps of `config.h.in' and `config.h' will not be changed if remaking them does not change their contents. This feature avoids unnecessary recompilation. You should include the file `stamp-h.in' your package's distribution, so @command{make} will consider `config.h.in' up to date. Don't use @command{touch} (see section Limitations of Usual Tools), rather use @command{echo} (using @command{date} would cause needless differences, hence CVS conflicts etc.).
$(srcdir)/configure: configure.ac aclocal.m4
cd $(srcdir) && autoconf
# autoheader might not change config.h.in, so touch a stamp file.
$(srcdir)/config.h.in: stamp-h.in
$(srcdir)/stamp-h.in: configure.ac aclocal.m4
cd $(srcdir) && autoheader
echo timestamp > $(srcdir)/stamp-h.in
config.h: stamp-h
stamp-h: config.h.in config.status
./config.status
Makefile: Makefile.in config.status
./config.status
config.status: configure
./config.status --recheck
(Be careful if you copy these lines directly into your Makefile, as you will need to convert the indented lines to start with the tab character.)
In addition, you should use `AC_CONFIG_FILES([stamp-h], [echo
timestamp > stamp-h])' so `config.status' will ensure that
`config.h' is considered up to date. See section Outputting Files, for more
information about AC_OUTPUT.
See section Recreating a Configuration, for more examples of handling configuration-related dependencies.
When a package tests more than a few C preprocessor symbols, the command
lines to pass @option{-D} options to the compiler can get quite long.
This causes two problems. One is that the make output is hard to
visually scan for errors. More seriously, the command lines can exceed
the length limits of some operating systems. As an alternative to
passing @option{-D} options to the compiler, @command{configure} scripts can
create a C header file containing `#define' directives. The
AC_CONFIG_HEADERS macro selects this kind of output. It should
be called right after AC_INIT.
The package should `#include' the configuration header file before
any other header files, to prevent inconsistencies in declarations (for
example, if it redefines const). Use `#include <config.h>'
instead of `#include "config.h"', and pass the C compiler a
@option{-I.} option (or @option{-I..}; whichever directory contains
`config.h'). That way, even if the source directory is configured
itself (perhaps to make a distribution), other build directories can
also be configured without finding the `config.h' from the source
directory.
AC_OUTPUT create the file(s) in the
whitespace-separated list header containing C preprocessor
#define statements, and replace `@DEFS@' in generated
files with @option{-DHAVE_CONFIG_H} instead of the value of DEFS.
The usual name for header is `config.h'.
If header already exists and its contents are identical to what
AC_OUTPUT would put in it, it is left alone. Doing this allows
some changes in configuration without needlessly causing object files
that depend on the header file to be recompiled.
Usually the input file is named `header.in'; however, you can override the input file name by appending to header, a colon-separated list of input files. Examples:
AC_CONFIG_HEADERS([config.h:config.hin]) AC_CONFIG_HEADERS([defines.h:defs.pre:defines.h.in:defs.post])
Doing this allows you to keep your file names acceptable to MS-DOS, or to prepend and/or append boilerplate to the file.
See section Taking Configuration Actions, for more details on header.
Your distribution should contain a template file that looks as you want
the final header file to look, including comments, with #undef
statements which are used as hooks. For example, suppose your
`configure.ac' makes these calls:
AC_CONFIG_HEADERS([conf.h]) AC_CHECK_HEADERS([unistd.h])
Then you could have code like the following in `conf.h.in'. On systems that have `unistd.h', @command{configure} will `#define' `HAVE_UNISTD_H' to 1. On other systems, the whole line will be commented out (in case the system predefines that symbol).
/* Define as 1 if you have unistd.h. */ #undef HAVE_UNISTD_H
You can then decode the configuration header using the preprocessor directives:
#include <conf.h> #if HAVE_UNISTD_H # include <unistd.h> #else /* We are in trouble. */ #endif
The use of old form templates, with `#define' instead of `#undef' is strongly discouraged.
Since it is a tedious task to keep a template header up to date, you may use @command{autoheader} to generate it, see section Using @command{autoheader} to Create @file{config.h.in}.
The @command{autoheader} program can create a template file of C
`#define' statements for @command{configure} to use. If
`configure.ac' invokes AC_CONFIG_HEADERS(file),
@command{autoheader} creates `file.in'; if multiple file
arguments are given, the first one is used. Otherwise,
@command{autoheader} creates `config.h.in'.
In order to do its job, @command{autoheader} needs you to document all
of the symbols that you might use; i.e., there must be at least one
AC_DEFINE or one AC_DEFINE_UNQUOTED using its third
argument for each symbol (see section Defining C Preprocessor Symbols). An additional
constraint is that the first argument of AC_DEFINE must be a
literal. Note that all symbols defined by Autoconf's built-in tests are
already documented properly; you only need to document those that you
define yourself.
You might wonder why @command{autoheader} is needed: after all, why would @command{configure} need to "patch" a `config.h.in' to produce a `config.h' instead of just creating `config.h' from scratch? Well, when everything rocks, the answer is just that we are wasting our time maintaining @command{autoheader}: generating `config.h' directly is all that is needed. When things go wrong, however, you'll be thankful for the existence of @command{autoheader}.
The fact that the symbols are documented is important in order to
check that `config.h' makes sense. The fact that there is a
well defined list of symbols that should be #define'd (or not) is
also important for people who are porting packages to environments where
@command{configure} cannot be run: they just have to @emph{fill in the
blanks}.
But let's come back to the point: @command{autoheader}'s invocation...
If you give @command{autoheader} an argument, it uses that file instead of `configure.ac' and writes the header file to the standard output instead of to `config.h.in'. If you give @command{autoheader} an argument of @option{-}, it reads the standard input instead of `configure.ac' and writes the header file to the standard output.
@command{autoheader} accepts the following options:
@command{autoheader} scans `configure.ac' and figures out which C
preprocessor symbols it might define. It knows how to generate
templates for symbols defined by AC_CHECK_HEADERS,
AC_CHECK_FUNCS etc., but if you AC_DEFINE any additional
symbol, you must define a template for it. If there are missing
templates, @command{autoheader} fails with an error message.
The simplest way to create a template for a symbol is to supply the description argument to an `AC_DEFINE(symbol)'; see section Defining C Preprocessor Symbols. You may also use one of the following macros.
AC_DEFINE'd.
For example:
AH_VERBATIM([_GNU_SOURCE], [/* Enable GNU extensions on systems that have them. */ #ifndef _GNU_SOURCE # define _GNU_SOURCE #endif])
AC_DEFINE when a
description is given.
For example:
AH_TEMPLATE([CRAY_STACKSEG_END],
[Define to one of _getb67, GETB67, getb67
for Cray-2 and Cray-YMP systems. This
function is required for alloca.c support
on those systems.])
will generate the following template, with the description properly justified.
/* Define to one of _getb67, GETB67, getb67 for Cray-2 and Cray-YMP systems. This function is required for alloca.c support on those systems. */ #undef CRAY_STACKSEG_END
You execute arbitrary commands either before, during and after
`config.status' is run. The three following macros accumulate the
commands to run when they are called multiple times.
AC_CONFIG_COMMANDS replaces the obsolete macro
AC_OUTPUT_COMMANDS, see section Obsolete Macros, for details.
Here is an unrealistic example:
fubar=42
AC_CONFIG_COMMANDS([fubar],
[echo this is extra $fubar, and so on.],
[fubar=$fubar])
Here is a better one:
AC_CONFIG_COMMANDS([time-stamp], [date >time-stamp])
AC_CONFIG_COMMANDS_PRE( [LTLIBOBJS=`echo $LIBOBJS | sed 's/\.o/\.lo/g'` AC_SUBST(LTLIBOBJS)])
You may find it convenient to create links whose destinations depend upon
results of tests. One can use AC_CONFIG_COMMANDS but the
creation of relative symbolic links can be delicate when the package is
built in another directory than its sources.
AC_OUTPUT link each of the existing files source to
the corresponding link name dest. Makes a symbolic link if
possible, otherwise a hard link. The dest and source names
should be relative to the top level source or build directory. This
macro is one of the instantiating macros, see section Taking Configuration Actions.
For example, this call:
AC_CONFIG_LINKS(host.h:config/$machine.h
object.h:config/$obj_format.h)
creates in the current directory `host.h' as a link to `srcdir/config/$machine.h', and `object.h' as a link to `srcdir/config/$obj_format.h'.
The tempting value `.' for dest is invalid: it makes it impossible for `config.status' to guess the links to establish.
One can then run:
./config.status host.h object.h
to create the links.
In most situations, calling AC_OUTPUT is sufficient to produce
`Makefile's in subdirectories. However, @command{configure} scripts
that control more than one independent package can use
AC_CONFIG_SUBDIRS to run @command{configure} scripts for other
packages in subdirectories.
AC_OUTPUT run @command{configure} in each subdirectory
dir in the given whitespace-separated list. Each dir should
be a literal, i.e., please do not use:
if test "$package_foo_enabled" = yes; then $my_subdirs="$my_subdirs foo" fi AC_CONFIG_SUBDIRS($my_subdirs)
because this prevents `./configure --help=recursive' from
displaying the options of the package foo. Rather, you should
write:
if test "$package_foo_enabled" = yes; then AC_CONFIG_SUBDIRS(foo) fi
If a given dir is not found, an error is reported: if the subdirectory is optional, write:
if test -d $srcdir/foo; then AC_CONFIG_SUBDIRS(foo) fi
If a given dir contains @command{configure.gnu}, it is run instead
of @command{configure}. This is for packages that might use a
non-autoconf script @command{Configure}, which can't be called through a
wrapper @command{configure} since it would be the same file on
case-insensitive filesystems. Likewise, if a dir contains
`configure.ac' but no @command{configure}, the Cygnus
@command{configure} script found by AC_CONFIG_AUX_DIR is used.
The subdirectory @command{configure} scripts are given the same command line options that were given to this @command{configure} script, with minor changes if needed, which include:
$prefix, including if it was
defaulted, and if default values of the top level and of sub directory
`configure' differ.
This macro also sets the output variable subdirs to the list of
directories `dir ...'. `Makefile' rules can use
this variable to determine which subdirectories to recurse into. This
macro may be called multiple times.
By default, @command{configure} sets the prefix for files it installs to `/usr/local'. The user of @command{configure} can select a different prefix using the @option{--prefix} and @option{--exec-prefix} options. There are two ways to change the default: when creating @command{configure}, and when running it.
Some software packages might want to install in a directory besides
`/usr/local' by default. To accomplish that, use the
AC_PREFIX_DEFAULT macro.
It may be convenient for users to have @command{configure} guess the
installation prefix from the location of a related program that they
have already installed. If you wish to do that, you can call
AC_PREFIX_PROGRAM.
PATH, the way the shell does. If program
is found, set the prefix to the parent of the directory containing
program; otherwise leave the prefix specified in
`Makefile.in' unchanged. For example, if program is
gcc and the PATH contains `/usr/local/gnu/bin/gcc',
set the prefix to `/usr/local/gnu'.
These macros test for particular system features that packages might need or want to use. If you need to test for a kind of feature that none of these macros check for, you can probably do it by calling primitive test macros with appropriate arguments (see section Writing Tests).
These tests print messages telling the user which feature they're checking for, and what they find. They cache their results for future @command{configure} runs (see section Caching Results).
Some of these macros set output variables. See section Substitutions in Makefiles, for how to get their values. The phrase "define name" is used below as a shorthand to mean "define C preprocessor symbol name to the value 1". See section Defining C Preprocessor Symbols, for how to get those symbol definitions into your program.
Much effort has been expended to make Autoconf easy to learn. The most obvious way to reach this goal is simply to enforce standard interfaces and behaviors, avoiding exceptions as much as possible. Because of history and inertia, unfortunately, there are still too many exceptions in Autoconf; nevertheless, this section describes some of the common rules.
All the generic macros that AC_DEFINE a symbol as a result of
their test transform their arguments to a standard alphabet.
First, argument is converted to upper case and any asterisks
(`*') are each converted to `P'. Any remaining characters
that are not alphanumeric are converted to underscores.
For instance,
AC_CHECK_TYPES(struct $Expensive*)
will define the symbol `HAVE_STRUCT__EXPENSIVEP' if the check succeeds.
Several tests depend upon a set of header files. Since these headers are not universally available, tests actually have to provide a set of protected includes, such as:
#if TIME_WITH_SYS_TIME # include <sys/time.h> # include <time.h> #else # if HAVE_SYS_TIME_H # include <sys/time.h> # else # include <time.h> # endif #endif
Unless you know exactly what you are doing, you should avoid using unconditional includes, and check the existence of the headers you include beforehand (see section Header Files).
Most generic macros provide the following default set of includes:
#include <stdio.h> #if HAVE_SYS_TYPES_H # include <sys/types.h> #endif #if HAVE_SYS_STAT_H # include <sys/stat.h> #endif #if STDC_HEADERS # include <stdlib.h> # include <stddef.h> #else # if HAVE_STDLIB_H # include <stdlib.h> # endif #endif #if HAVE_STRING_H # if !STDC_HEADERS && HAVE_MEMORY_H # include <memory.h> # endif # include <string.h> #endif #if HAVE_STRINGS_H # include <strings.h> #endif #if HAVE_INTTYPES_H # include <inttypes.h> #else # if HAVE_STDINT_H # include <stdint.h> # endif #endif #if HAVE_UNISTD_H # include <unistd.h> #endif
If the default includes are used, then Autoconf will automatically check
for the presence of these headers and their compatibility, i.e., you
don't need to run AC_HEADERS_STDC, nor check for `stdlib.h'
etc.
These headers are checked for in the same order as they are included.
For instance, on some systems `string.h' and `strings.h' both
exist, but conflict. Then HAVE_STRING_H will be defined, but
HAVE_STRINGS_H won't.
These macros check for the presence or behavior of particular programs. They are used to choose between several alternative programs and to decide what to do once one has been chosen. If there is no macro specifically defined to check for a program you need, and you don't need to check for any special properties of it, then you can use one of the general program-check macros.
These macros check for particular programs--whether they exist, and in some cases whether they support certain features.
gawk, mawk, nawk, and awk, in that
order, and set output variable AWK to the first one that is found.
It tries gawk first because that is reported to be the
best implementation.
INSTALL to the path of a BSD compatible
install program, if one is found in the current PATH.
Otherwise, set INSTALL to `dir/install-sh -c',
checking the directories specified to AC_CONFIG_AUX_DIR (or its
default directories) to determine dir (see section Outputting Files). Also set
the variables INSTALL_PROGRAM and INSTALL_SCRIPT to
`${INSTALL}' and INSTALL_DATA to `${INSTALL} -m 644'.
This macro screens out various instances of install known not to
work. It prefers to find a C program rather than a shell script, for
speed. Instead of `install-sh', it can also use `install.sh',
but that name is obsolete because some make programs have a rule
that creates `install' from it if there is no `Makefile'.
Autoconf comes with a copy of `install-sh' that you can use. If
you use AC_PROG_INSTALL, you must include either
`install-sh' or `install.sh' in your distribution, or
@command{configure} will produce an error message saying it can't find
them--even if the system you're on has a good install program.
This check is a safety measure to prevent you from accidentally leaving
that file out, which would prevent your package from installing on
systems that don't have a BSD-compatible install program.
If you need to use your own installation program because it has features
not found in standard install programs, there is no reason to use
AC_PROG_INSTALL; just put the file name of your program into your
`Makefile.in' files.
flex is found, set output variable LEX to `flex'
and LEXLIB to @option{-lfl}, if that library is in a standard
place. Otherwise set LEX to `lex' and LEXLIB to
@option{-ll}.
Define YYTEXT_POINTER if yytext is a `char *' instead
of a `char []'. Also set output variable LEX_OUTPUT_ROOT to
the base of the file name that the lexer generates; usually
`lex.yy', but sometimes something else. These results vary
according to whether lex or flex is being used.
You are encouraged to use Flex in your sources, since it is both more
pleasant to use than plain Lex and the C source it produces is portable.
In order to ensure portability, however, you must either provide a
function yywrap or, if you don't use it (e.g., your scanner has
no `#include'-like feature), simply include a `%noyywrap'
statement in the scanner's source. Once this done, the scanner is
portable (unless you felt free to use nonportable constructs) and
does not depend on any library. In this case, and in this case only, it
is suggested that you use this Autoconf snippet:
AC_PROG_LEX if test "$LEX" != flex; then LEX="$SHELL $missing_dir/missing flex" AC_SUBST(LEX_OUTPUT_ROOT, lex.yy) AC_SUBST(LEXLIB, '') fi
The shell script @command{missing} can be found in the Automake distribution.
To ensure backward compatibility, Automake's AM_PROG_LEX invokes
(indirectly) this macro twice, which will cause an annoying but benign
"AC_PROG_LEX invoked multiple times" warning. Future versions
of Automake will fix this issue, meanwhile, just ignore this message.
LN_S to `ln -s'; otherwise, if `ln' works, set
LN_S to `ln' and otherwise set it to `cp -p'.
If you make a link a directory other than the current directory, its
meaning depends on whether `ln' or `ln -s' is used. To safely
create links using `$(LN_S)', either find out which form is used
and adjust the arguments, or always invoke ln in the directory
where the link is to be created.
In other words, it does not work to do:
$(LN_S) foo /x/bar
Instead, do:
(cd /x && $(LN_S) foo bar)
RANLIB to `ranlib' if ranlib
is found, and otherwise to `:' (do nothing).
bison is found, set output variable YACC to `bison
-y'. Otherwise, if byacc is found, set YACC to
`byacc'. Otherwise set YACC to `yacc'.
These macros are used to find programs not covered by the "particular"
test macros. If you need to check the behavior of a program as well as
find out whether it is present, you have to write your own test for it
(see section Writing Tests). By default, these macros use the environment
variable PATH. If you need to check for a program that might not
be in the user's PATH, you can pass a modified path to use
instead, like this:
AC_PATH_PROG([INETD], [inetd], [/usr/libexec/inetd],
[$PATH:/usr/libexec:/usr/sbin:/usr/etc:etc])
You are strongly encouraged to declare the variable passed to
AC_CHECK_PROG etc. as precious, See section Setting Output Variables,
AC_ARG_VAR, for more details.
PATH. If
it is found, set variable to value-if-found, otherwise to
value-if-not-found, if given. Always pass over reject (an
absolute file name) even if it is the first found in the search path; in
that case, set variable using the absolute file name of the
prog-to-check-for found that is not reject. If
variable was already set, do nothing. Calls AC_SUBST for
variable.
PATH. If it is found, set
variable to the name of that program. Otherwise, continue
checking the next program in the list. If none of the programs in the
list are found, set variable to value-if-not-found; if
value-if-not-found is not specified, the value of variable
is not changed. Calls AC_SUBST for variable.
AC_CHECK_PROG, but first looks for prog-to-check-for
with a prefix of the host type as determined by
AC_CANONICAL_HOST, followed by a dash (see section Getting the Canonical System Type).
For example, if the user runs `configure --host=i386-gnu', then
this call:
AC_CHECK_TOOL(RANLIB, ranlib, :)
sets RANLIB to `i386-gnu-ranlib' if that program exists in
PATH, or otherwise to `ranlib' if that program exists in
PATH, or to `:' if neither program exists.
AC_CHECK_TOOL, each of the tools in the list
progs-to-check-for are checked with a prefix of the host type as
determined by AC_CANONICAL_HOST, followed by a dash
(see section Getting the Canonical System Type). If none of the tools can be found with a
prefix, then the first one without a prefix is used. If a tool is found,
set variable to the name of that program. If none of the tools in
the list are found, set variable to value-if-not-found; if
value-if-not-found is not specified, the value of variable
is not changed. Calls AC_SUBST for variable.
AC_CHECK_PROG, but set variable to the entire
path of prog-to-check-for if found.
AC_CHECK_PROGS, but if any of progs-to-check-for
are found, set variable to the entire path of the program
found.
AC_CHECK_TOOL, but set variable to the entire
path of the program if it is found.
You might also need to check for the existence of files. Before using these macros, ask yourself whether a run time test might not be a better solution. Be aware that, like most Autoconf macros, they test a feature of the host machine, and therefore, they die when cross-compiling.
AC_CHECK_FILE once for each file listed in files.
Additionally, defines `HAVE_file' (see section Standard Symbols)
for each file found.
The following macros check for the presence of certain C, C++ or Fortran 77 library archive files.
action-if-found is a list of shell commands to run if the link
with the library succeeds; action-if-not-found is a list of shell
commands to run if the link fails. If action-if-found is not
specified, the default action will prepend @option{-llibrary} to
LIBS and define `HAVE_LIBlibrary' (in all
capitals). This macro is intended to support building of LIBS in
a right-to-left (least-dependent to most-dependent) fashion such that
library dependencies are satisfied as a natural side-effect of
consecutive tests. Some linkers are very sensitive to library ordering
so the order in which LIBS is generated is important to reliable
detection of libraries.
If linking with library results in unresolved symbols that would
be resolved by linking with additional libraries, give those libraries
as the other-libraries argument, separated by spaces:
e.g. @option{-lXt -lX11}. Otherwise, this macro will fail to detect
that library is present, because linking the test program will
always fail with unresolved symbols. The other-libraries argument
should be limited to cases where it is desirable to test for one library
in the presence of another that is not already in LIBS.
AC_TRY_LINK_FUNC first
with no libraries, then for each library listed in search-libs.
Add @option{-llibrary} to LIBS for the first library found
to contain function, and run action-if-found. If the
function is not found, run action-if-not-found.
If linking with library results in unresolved symbols that would be resolved by linking with additional libraries, give those libraries as the other-libraries argument, separated by spaces: e.g. @option{-lXt -lX11}. Otherwise, this macro will fail to detect that function is present, because linking the test program will always fail with unresolved symbols.
The following macros check for particular C library functions. If there is no macro specifically defined to check for a function you need, and you don't need to check for any special properties of it, then you can use one of the general function-check macros.
Most usual functions can either be missing, or be buggy, or be limited on some architectures. This section tries to make an inventory of these portability issues. By definition, this list will always require additions. Please help us keeping it as complete as possible.
snprintf
snprintf and vsnprintf truncate
the output and return the number of bytes that ought to have been
produced. Some older systems return the truncated length (e.g., GNU C
Library 2.0.x or IRIX 6.5), some a negative value (e.g., earlier GNU C
Library versions), and some the buffer length without truncation (e.g.,
32-bit Solaris 7). Also, some buggy older systems ignore the length and
overrun the buffer (e.g., 64-bit Solaris 7).
sprintf
sprintf and vsprintf return the
number of bytes written, but on some old systems (SunOS 4 for
instance) they return the buffer pointer instead.
sscanf
sscanf requires that its
input string is writable (though it doesn't actually change it). This
can be a problem when using @command{gcc} since it normally puts
constant strings in read-only memory
(see section `Incompatibilities' in Using and Porting the GNU Compiler Collection). Apparently in some cases even
having format strings read-only can be a problem.
strnlen
strnlen ("foobar", 0) = 0
strnlen ("foobar", 1) = 3
strnlen ("foobar", 2) = 2
strnlen ("foobar", 3) = 1
strnlen ("foobar", 4) = 0
strnlen ("foobar", 5) = 6
strnlen ("foobar", 6) = 6
strnlen ("foobar", 7) = 6
strnlen ("foobar", 8) = 6
strnlen ("foobar", 9) = 6
unlink
unlink causes the given files to be
removed only after there are no more open file handles for it. Not all
OS's support this behaviour though. So even on systems that provide
unlink, you cannot portably assume it is OK to call it on files
that are open. For example, on Windows 9x and ME, such a call would fail;
on DOS it could even lead to file system corruption, as the file might end
up being written to after the OS has removed it.
va_copy
va_copy for copying
va_list variables. It may be available in older environments
too, though possibly as __va_copy (eg. @command{gcc} in strict
C89 mode). These can be tested with #ifdef. A fallback to
memcpy (&dst, &src, sizeof(va_list)) will give maximum
portability.
va_list
va_list is not necessarily just a pointer. It can be a
struct (eg. @command{gcc} on Alpha), which means NULL is
not portable. Or it can be an array (eg. @command{gcc} in some
PowerPC configurations), which means as a function parameter it can be
effectively call-by-reference and library routines might modify the
value back in the caller (eg. vsnprintf in the GNU C Library
2.1).
>>
>> right shift of a signed type replicates the
high bit, giving a so-called "arithmetic" shift. But care should be
taken since the ISO C standard doesn't require that behaviour. On those
few processors without a native arithmetic shift (for instance Cray
vector systems) zero bits may be shifted in, the same as a shift of an
unsigned type.
These macros check for particular C functions--whether they exist, and in some cases how they respond when given certain arguments.
alloca. Tries to get a builtin version by
checking for `alloca.h' or the predefined C preprocessor macros
__GNUC__ and _AIX. If this macro finds `alloca.h',
it defines HAVE_ALLOCA_H.
If those attempts fail, it looks for the function in the standard C
library. If any of those methods succeed, it defines
HAVE_ALLOCA. Otherwise, it sets the output variable
ALLOCA to `alloca.o' and defines C_ALLOCA (so
programs can periodically call `alloca(0)' to garbage collect).
This variable is separate from LIBOBJS so multiple programs can
share the value of ALLOCA without needing to create an actual
library, in case only some of them use the code in LIBOBJS.
This macro does not try to get alloca from the System V R3
`libPW' or the System V R4 `libucb' because those libraries
contain some incompatible functions that cause trouble. Some versions
do not even contain alloca or contain a buggy version. If you
still want to use their alloca, use ar to extract
`alloca.o' from them instead of compiling `alloca.c'.
Source files that use alloca should start with a piece of code
like the following, to declare it properly. In some versions of AIX,
the declaration of alloca must precede everything else except for
comments and preprocessor directives. The #pragma directive is
indented so that pre-ANSI C compilers will ignore it, rather than
choke on it.
/* AIX requires this to be the first thing in the file. */ #ifndef __GNUC__ # if HAVE_ALLOCA_H # include <alloca.h> # else # ifdef _AIX #pragma alloca # else # ifndef alloca /* predefined by HP cc +Olibcalls */ char *alloca (); # endif # endif # endif #endif
chown function is available and works (in particular, it
should accept @option{-1} for uid and gid), define
HAVE_CHOWN.
closedir function does not return a meaningful value,
define CLOSEDIR_VOID. Otherwise, callers ought to check its
return value for an error indicator.
error_at_line function is not found, require an
AC_LIBOBJ replacement of `error'.
fnmatch function is available and works (unlike the one on
Solaris 2.4), define HAVE_FNMATCH.
fork and vfork functions. If a
working fork is found, define HAVE_WORKING_FORK. This macro
checks whether fork is just a stub by trying to run it.
If `vfork.h' is found, define HAVE_VFORK_H. If a working
vfork is found, define HAVE_WORKING_VFORK. Otherwise,
define vfork to be fork for backward compatibility with
previous versions of @command{autoconf}. This macro checks for several known
errors in implementations of vfork and considers the system to not
have a working vfork if it detects any of them. It is not considered
to be an implementation error if a child's invocation of signal
modifies the parent's signal handler, since child processes rarely change
their signal handlers.
Since this macro defines vfork only for backward compatibility with
previous versions of @command{autoconf} you're encouraged to define it
yourself in new code:
#if !HAVE_WORKING_VFORK # define vfork fork #endif
fseeko function is available, define HAVE_FSEEKO.
Define _LARGEFILE_SOURCE if necessary.
getgroups function is available and works (unlike on
Ultrix 4.3, where `getgroups (0, 0)' always fails), define
HAVE_GETGROUPS. Set GETGROUPS_LIBS to any libraries
needed to get that function. This macro runs AC_TYPE_GETGROUPS.
getloadavg function, define HAVE_GETLOADAVG, and set
GETLOADAVG_LIBS to any libraries needed to get that function.
Also add GETLOADAVG_LIBS to LIBS.
Otherwise, require an AC_LIBOBJ replacement (`getloadavg.c')
of `getloadavg', and possibly define several other C preprocessor
macros and output variables:
C_GETLOADAVG.
SVR4, DGUX, UMAX, or UMAX4_3 if on
those systems.
NLIST_STRUCT.
HAVE_STRUCT_NLIST_N_UN_N_NAME. The obsolete symbol
NLIST_NAME_UNION is still defined, but do not depend upon it.
getloadavg to work. In this case, define
GETLOADAVG_PRIVILEGED, set the output variable NEED_SETGID
to `true' (and otherwise to `false'), and set
KMEM_GROUP to the name of the group that should own the installed
program.
getmntent in the `sun', `seq', and `gen'
libraries, for Irix 4, PTX, and Unixware, respectively. Then, if
getmntent is available, define HAVE_GETMNTENT.
GETPGRP_VOID if it is an error to pass 0 to
getpgrp; this is the POSIX.1 behavior. On older BSD
systems, you must pass 0 to getpgrp, as it takes an argument and
behaves like POSIX.1's getpgid.
#if GETPGRP_VOID pid = getpgrp (); #else pid = getpgrp (0); #endif
This macro does not check whether
getpgrp exists at all; if you need to work in that situation,
first call AC_CHECK_FUNC for getpgrp.
lstat should treat
`link/' the same as `link/.'. However, many older
lstat implementations incorrectly ignore trailing slashes.
It is safe to assume that if lstat incorrectly ignores
trailing slashes, then other symbolic-link-aware functions like
unlink and unlink also incorrectly ignore trailing slashes.
If lstat behaves properly, define
LSTAT_FOLLOWS_SLASHED_SYMLINK, otherwise require an
AC_LIBOBJ replacement of lstat.
malloc works correctly (`malloc (0)' returns a valid
pointer), define HAVE_MALLOC.
memcmp function is not available, or does not work on
8-bit data (like the one on SunOS 4.1.3), or fails when comparing 16
bytes or more and with at least one buffer not starting on a 4-byte
boundary (such as the one on NeXT x86 OpenStep), require an
AC_LIBOBJ replacement for `memcmp'.
mktime function is not available, or does not work
correctly, require an AC_LIBOBJ replacement for `mktime'.
mmap function exists and works correctly, define
HAVE_MMAP. Only checks private fixed mapping of already-mapped
memory.
HAVE_OBSTACK, else require an
AC_LIBOBJ replacement for `obstack'.
select function's arguments, and defines those types
in SELECT_TYPE_ARG1, SELECT_TYPE_ARG234, and
SELECT_TYPE_ARG5 respectively. SELECT_TYPE_ARG1 defaults
to `int', SELECT_TYPE_ARG234 defaults to `int *',
and SELECT_TYPE_ARG5 defaults to `struct timeval *'.
setpgrp takes no argument (the POSIX.1 version), define
SETPGRP_VOID. Otherwise, it is the BSD version, which takes
two process IDs as arguments. This macro does not check whether
setpgrp exists at all; if you need to work in that situation,
first call AC_CHECK_FUNC for setpgrp.
stat or lstat have the bug that it
succeeds when given the zero-length file name argument. The stat
and lstat from SunOS 4.1.4 and the Hurd (as of 1998-11-01) do
this.
If it does, then define HAVE_STAT_EMPTY_STRING_BUG (or
HAVE_LSTAT_EMPTY_STRING_BUG) and ask for an AC_LIBOBJ
replacement of it.
setvbuf takes the buffering type as its second argument and
the buffer pointer as the third, instead of the other way around, define
SETVBUF_REVERSED.
strcoll function exists and works correctly, define
HAVE_STRCOLL. This does a bit more than
`AC_CHECK_FUNCS(strcoll)', because some systems have incorrect
definitions of strcoll that should not be used.
strtod function does not exist or doesn't work correctly,
ask for an AC_LIBOBJ replacement of `strtod'. In this case,
because `strtod.c' is likely to need `pow', set the output
variable POW_LIB to the extra library needed.
strerror_r is available, define HAVE_STRERROR_R, and if
it is declared, define HAVE_DECL_STRERROR_R. If it returns a
char * message, define STRERROR_R_CHAR_P; otherwise it
returns an int error number. The Thread-Safe Functions option of
POSIX-200X requires strerror_r to return int, but
many systems (including, for example, version 2.2.4 of the GNU C
Library) return a char * value that is not necessarily equal to
the buffer argument.
strftime in the `intl' library, for SCO UNIX.
Then, if strftime is available, define HAVE_STRFTIME.
strnlen, and ask for its replacement. Some
architectures are know to provide broken versions of strnlen, such
as AIX 4.3.
HAVE_UTIME_NULL.
vprintf is found, define HAVE_VPRINTF. Otherwise, if
_doprnt is found, define HAVE_DOPRNT. (If vprintf
is available, you may assume that vfprintf and vsprintf
are also available.)
These macros are used to find functions not covered by the "particular"
test macros. If the functions might be in libraries other than the
default C library, first call AC_CHECK_LIB for those libraries.
If you need to check the behavior of a function as well as find out
whether it is present, you have to write your own test for
it (see section Writing Tests).
AC_CHECK_FUNCS instead. This macro checks for functions with C
linkage even when AC_LANG(C++) has been called, since C is more
standardized than C++. (see section Language Choice, for more information
about selecting the language for checks.)
HAVE_function (in all capitals) if it is available.
If action-if-found is given, it is additional shell code to
execute when one of the functions is found. You can give it a value of
`break' to break out of the loop on the first match. If
action-if-not-found is given, it is executed when one of the
functions is not found.
Autoconf follows a philosophy that was formed over the years by those who have struggled for portability: isolate the portability issues in specific files, and then program as if you were in a POSIX environment. Some functions may be missing or unfixable, and your package must be ready to replace them.
Use the first three of the following macros to specify a function to be
replaced, and the last one (AC_REPLACE_FUNCS) to check for and
replace the function if needed.
Technically, it adds `function.$ac_objext' to the output
variable LIBOBJS and calls AC_LIBSOURCE for
`function.c'. You should not directly change LIBOBJS,
since this is not traceable.
AC_LIBSOURCE. file must be a literal.
This macro is called automatically from AC_LIBOBJ, but you must
call it explicitly if you pass a shell variable to AC_LIBOBJ. In
that case, since shell variables cannot be traced statically, you must
pass to AC_LIBSOURCE any possible files that the shell variable
might cause AC_LIBOBJ to need. For example, if you want to pass
a variable $foo_or_bar to AC_LIBOBJ that holds either
"foo" or "bar", you should do:
AC_LIBSOURCE(foo.c) AC_LIBSOURCE(bar.c) AC_LIBOBJ($foo_or_bar)
There is usually a way to avoid this, however, and you are encouraged to
simply call AC_LIBOBJ with literal arguments.
Note that this macro replaces the obsolete AC_LIBOBJ_DECL, with
slightly different semantics: the old macro took the function name,
e.g. foo, as its argument rather than the file name.
AC_LIBSOURCE, but accepts one or more files in a
comma-separated M4 list. Thus, the above example might be rewritten:
AC_LIBSOURCES([foo.c, bar.c]) AC_LIBOBJ($foo_or_bar)
AC_CHECK_FUNCS, but uses `AC_LIBOBJ(function)' as
action-if-not-found. You can declare your replacement function by
enclosing the prototype in `#if !HAVE_function'. If the
system has the function, it probably declares it in a header file you
should be including, so you shouldn't redeclare it lest your declaration
conflict.
The following macros check for the presence of certain C header files. If there is no macro specifically defined to check for a header file you need, and you don't need to check for any special properties of it, then you can use one of the general header-file check macros.
These macros check for particular system header files--whether they exist, and in some cases whether they declare certain symbols.
HAVE_DIRENT_H
HAVE_SYS_NDIR_H
HAVE_SYS_DIR_H
HAVE_NDIR_H
The directory-library declarations in your source code should look
something like the following:
#if HAVE_DIRENT_H # include <dirent.h> # define NAMLEN(dirent) strlen((dirent)->d_name) #else # define dirent direct # define NAMLEN(dirent) (dirent)->d_namlen # if HAVE_SYS_NDIR_H # include <sys/ndir.h> # endif # if HAVE_SYS_DIR_H # include <sys/dir.h> # endif # if HAVE_NDIR_H # include <ndir.h> # endif #endifUsing the above declarations, the program would declare variables to be of type
struct dirent, not struct direct, and would access
the length of a directory entry name by passing a pointer to a
struct dirent to the NAMLEN macro.
This macro also checks for the SCO Xenix `dir' and `x' libraries.
major, minor, and
makedev, but `sys/mkdev.h' does, define
MAJOR_IN_MKDEV; otherwise, if `sys/sysmacros.h' does, define
MAJOR_IN_SYSMACROS.
S_ISDIR, S_ISREG et al. defined in
`sys/stat.h' do not work properly (returning false positives),
define STAT_MACROS_BROKEN. This is the case on Tektronix UTekV,
Amdahl UTS and Motorola System V/88.
STDC_HEADERS if the system has ANSI C header files.
Specifically, this macro checks for `stdlib.h', `stdarg.h',
`string.h', and `float.h'; if the system has those, it
probably has the rest of the ANSI C header files. This macro also
checks whether `string.h' declares memchr (and thus
presumably the other mem functions), whether `stdlib.h'
declare free (and thus presumably malloc and other related
functions), and whether the `ctype.h' macros work on characters
with the high bit set, as ANSI C requires.
Use STDC_HEADERS instead of __STDC__ to determine whether
the system has ANSI-compliant header files (and probably C library
functions) because many systems that have GCC do not have ANSI C
header files.
On systems without ANSI C headers, there is so much variation that
it is probably easier to declare the functions you use than to figure
out exactly what the system header files declare. Some systems contain
a mix of functions ANSI and BSD; some are mostly ANSI but
lack `memmove'; some define the BSD functions as macros in
`string.h' or `strings.h'; some have only the BSD
functions but `string.h'; some declare the memory functions in
`memory.h', some in `string.h'; etc. It is probably
sufficient to check for one string function and one memory function; if
the library has the ANSI versions of those then it probably has
most of the others. If you put the following in `configure.ac':
AC_HEADER_STDC AC_CHECK_FUNCS(strchr memcpy)then, in your code, you can put declarations like this:
#if STDC_HEADERS # include <string.h> #else # if !HAVE_STRCHR # define strchr index # define strrchr rindex # endif char *strchr (), *strrchr (); # if !HAVE_MEMCPY # define memcpy(d, s, n) bcopy ((s), (d), (n)) # define memmove(d, s, n) bcopy ((s), (d), (n)) # endif #endifIf you use a function like
memchr, memset, strtok,
or strspn, which have no BSD equivalent, then macros won't
suffice; you must provide an implementation of each function. An easy
way to incorporate your implementations only when needed (since the ones
in system C libraries may be hand optimized) is to, taking memchr
for example, put it in `memchr.c' and use
`AC_REPLACE_FUNCS(memchr)'.
HAVE_SYS_WAIT_H. Incompatibility can occur if `sys/wait.h'
does not exist, or if it uses the old BSD union wait instead
of int to store a status value. If `sys/wait.h' is not
POSIX.1 compatible, then instead of including it, define the
POSIX.1 macros with their usual interpretations. Here is an
example:
#include <sys/types.h> #if HAVE_SYS_WAIT_H # include <sys/wait.h> #endif #ifndef WEXITSTATUS # define WEXITSTATUS(stat_val) ((unsigned)(stat_val) >> 8) #endif #ifndef WIFEXITED # define WIFEXITED(stat_val) (((stat_val) & 255) == 0) #endif
_POSIX_VERSION is defined when `unistd.h' is included on
POSIX.1 systems. If there is no `unistd.h', it is definitely
not a POSIX.1 system. However, some non-POSIX.1 systems do
have `unistd.h'.
The way to check if the system supports POSIX.1 is:
#if HAVE_UNISTD_H # include <sys/types.h> # include <unistd.h> #endif #ifdef _POSIX_VERSION /* Code for POSIX.1 systems. */ #endif
TIME_WITH_SYS_TIME. On some older systems,
`sys/time.h' includes `time.h', but `time.h' is not
protected against multiple inclusion, so programs should not explicitly
include both files. This macro is useful in programs that use, for
example, struct timeval or struct timezone as well as
struct tm. It is best used in conjunction with
HAVE_SYS_TIME_H, which can be checked for using
AC_CHECK_HEADERS(sys/time.h).
#if TIME_WITH_SYS_TIME # include <sys/time.h> # include <time.h> #else # if HAVE_SYS_TIME_H # include <sys/time.h> # else # include <time.h> # endif #endif
TIOCGWINSZ requires `<sys/ioctl.h>', then
define GWINSZ_IN_SYS_IOCTL. Otherwise TIOCGWINSZ can be
found in `<termios.h>'.
Use:
#if HAVE_TERMIOS_H # include <termios.h> #endif #if GWINSZ_IN_SYS_IOCTL # include <sys/ioctl.h> #endif
These macros are used to find system header files not covered by the "particular" test macros. If you need to check the contents of a header as well as find out whether it is present, you have to write your own test for it (see section Writing Tests).
AC_CHECK_HEADERS
instead.
The meaning of "usable" depends upon the content of includes:
header-filecan be preprocessed without error.
includes #include <header-file>can be compiled without error. You may use
AC_CHECK_HEADER (and AC_CHECK_HEADERS) to check whether
two headers are compatible.
You may pass any kind of dummy content for includes, such as a single space, a comment, to check whether header-file compiles with success.
HAVE_header-file (in all capitals). If action-if-found
is given, it is additional shell code to execute when one of the header
files is found. You can give it a value of `break' to break out of
the loop on the first match. If action-if-not-found is given, it
is executed when one of the header files is not found.
Be sure to read the documentation of AC_CHECK_HEADER to
understand the influence of includes.
The following macros check for the declaration of variables and
functions. If there is no macro specifically defined to check for a
symbol you need, then you can use the general macros (see section Generic Declaration Checks) or, for more complex tests, you may use
AC_TRY_COMPILE (see section Examining Syntax).
The following macros check for certain declarations.
SYS_SIGLIST_DECLARED if the variable sys_siglist
is declared in a system header file, either `signal.h' or
`unistd.h'.
These macros are used to find declarations not covered by the "particular" test macros.
This macro actually tests whether it is valid to use symbol as an r-value, not if it is really declared, because it is much safer to avoid introducing extra declarations when they are not needed.
HAVE_DECL_symbol (in all capitals) to `1' if
symbol is declared, otherwise to `0'. If
action-if-not-found is given, it is additional shell code to
execute when one of the function declarations is needed, otherwise
action-if-found is executed.
This macro uses an m4 list as first argument:
AC_CHECK_DECLS(strdup) AC_CHECK_DECLS([strlen]) AC_CHECK_DECLS([malloc, realloc, calloc, free])
Unlike the other `AC_CHECK_*S' macros, when a symbol is not
declared, HAVE_DECL_symbol is defined to `0' instead
of leaving HAVE_DECL_symbol undeclared. When you are
sure that the check was performed, use
HAVE_DECL_symbol just like any other result of Autoconf:
#if !HAVE_DECL_SYMBOL extern char *symbol; #endif
If the test may have not been performed, however, because it is safer not to declare a symbol than to use a declaration that conflicts with the system's one, you should use:
#if defined HAVE_DECL_MALLOC && !HAVE_DECL_MALLOC char *malloc (size_t *s); #endif
You fall into the second category only in extreme situations: either your files may be used without being configured, or they are used during the configuration. In most cases the traditional approach is enough.
The following macros check for the presence of certain members in C
structures. If there is no macro specifically defined to check for a
member you need, then you can use the general structure-member macro
(see section Generic Structure Checks) or, for more complex tests, you may use
AC_TRY_COMPILE (see section Examining Syntax).
The following macros check for certain structures or structure members.
struct stat contains an st_blksize member, define
HAVE_STRUCT_STAT_ST_BLKSIZE. The former name,
HAVE_ST_BLKSIZE is to be avoided, as its support will cease in
the future. This macro is obsoleted, and should be replaced by
AC_CHECK_MEMBERS([struct stat.st_blksize])
struct stat contains an st_blocks member, define
HAVE_STRUCT STAT_ST_BLOCKS. Otherwise, require an
AC_LIBOBJ replacement of `fileblocks'. The former name,
HAVE_ST_BLOCKS is to be avoided, as its support will cease in the
future.
struct stat contains an st_rdev member, define
HAVE_STRUCT_STAT_ST_RDEV. The former name for this macro,
HAVE_ST_RDEV, is to be avoided as it will cease to be supported
in the future. Actually, even the new macro is obsolete, and should be
replaced by:
AC_CHECK_MEMBERS([struct stat.st_rdev])
struct tm, define
TM_IN_SYS_TIME, which means that including `sys/time.h'
had better define struct tm.
struct tm has a
tm_zone member, define HAVE_STRUCT_TM_TM_ZONE (and the
obsoleted HAVE_TM_ZONE). Otherwise, if the external array
tzname is found, define HAVE_TZNAME.
These macros are used to find structure members not covered by the "particular" test macros.
AC_CHECK_MEMBER(struct passwd.pw_gecos,,
[AC_MSG_ERROR([We need `passwd.pw_gecos'!])],
[#include <pwd.h>])
You can use this macro for sub-members:
AC_CHECK_MEMBER(struct top.middle.bot)
HAVE_aggregate_member (in all
capitals, with spaces and dots replaced by underscores).
This macro uses m4 lists:
AC_CHECK_MEMBERS([struct stat.st_rdev, struct stat.st_blksize])
The following macros check for C types, either builtin or typedefs. If there is no macro specifically defined to check for a type you need, and you don't need to check for any special properties of it, then you can use a general type-check macro.
These macros check for particular C types in `sys/types.h', `stdlib.h' and others, if they exist.
GETGROUPS_T to be whichever of gid_t or int
is the base type of the array argument to getgroups.
signal as returning a pointer to a
function returning void, define RETSIGTYPE to be
void; otherwise, define it to be int.
Define signal handlers as returning type RETSIGTYPE:
RETSIGTYPE
hup_handler ()
{
...
}
These macros are used to check for types not covered by the "particular" test macros.
HAVE_type (in all capitals). If no includes are
specified, the default includes are used (see section Default Includes). If
action-if-found is given, it is additional shell code to execute
when one of the types is found. If action-if-not-found is given,
it is executed when one of the types is not found.
This macro uses m4 lists:
AC_CHECK_TYPES(ptrdiff_t) AC_CHECK_TYPES([unsigned long long, uintmax_t])
Autoconf, up to 2.13, used to provide to another version of
AC_CHECK_TYPE, broken by design. In order to keep backward
compatibility, a simple heuristics, quite safe but not totally, is
implemented. In case of doubt, read the documentation of the former
AC_CHECK_TYPE, see section Obsolete Macros.
All the tests for compilers (AC_PROG_CC, AC_PROG_CXX,
AC_PROG_F77) define the output variable EXEEXT based on
the output of the 1compiler, typically to the empty string if Unix and
`.exe' if Win32 or OS/2.
They also define the output variable OBJEXT based on the
output of the compiler, after .c files have been excluded, typically
to `o' if Unix, `obj' if Win32.
If the compiler being used does not produce executables, they fail. If the executables can't be run, and cross-compilation is not enabled, they fail too. See section Manual Configuration, for more on support for cross compiling.
Some compilers exhibit different behaviors.
int
main (void)
{
static int test_array [sizeof (int) == 4 ? 1 : -1];
test_array [0] = 0
return 0;
}
To our knowledge, there is a single compiler that does not support this
trick: the HP C compilers (the real one, not only the "bundled") on
HP-UX 11.00:
$ cc -c -Ae +O2 +Onolimit conftest.c
cc: "conftest.c": error 1879: Variable-length arrays cannot \
have static storage.
Autoconf works around this problem by casting sizeof (int) to
long before comparing it.
SIZEOF_type (see section Standard Symbols) to be the
size in bytes of type. If `type' is unknown, it gets a size
of 0. If no includes are specified, the default includes are used
(see section Default Includes). If you provide include, make sure to
include `stdio.h' which is required for this macro to run.
This macro now works even when cross-compiling. The unused argument was used when cross-compiling.
For example, the call
AC_CHECK_SIZEOF(int *)
defines SIZEOF_INT_P to be 8 on DEC Alpha AXP systems.
CC is not already set in the
environment, check for gcc and cc, then for other C
compilers. Set output variable CC to the name of the compiler
found.
This macro may, however, be invoked with an optional first argument
which, if specified, must be a space separated list of C compilers to
search for. This just gives the user an opportunity to specify an
alternative search list for the C compiler. For example, if you didn't
like the default order, then you could invoke AC_PROG_CC like
this:
AC_PROG_CC(cl egcs gcc cc)
If using the GNU C compiler, set shell variable GCC to
`yes'. If output variable CFLAGS was not already set, set
it to @option{-g -O2} for the GNU C compiler (@option{-O2} on systems
where GCC does not accept @option{-g}), or @option{-g} for other compilers.
NO_MINUS_C_MINUS_O. This macro actually
tests both the compiler found by AC_PROG_CC, and, if different,
the first cc in the path. The test fails if one fails. This
macro was created for GNU Make to choose the default C compilation
rule.
CC to make it so. This macro tries
various options that select ANSI C on some system or another. It
considers the compiler to be in ANSI C mode if it handles function
prototypes correctly.
If you use this macro, you should check after calling it whether the C
compiler has been set to accept ANSI C; if not, the shell variable
ac_cv_prog_cc_stdc is set to `no'. If you wrote your source
code in ANSI C, you can make an un-ANSIfied copy of it by
using the program ansi2knr, which comes with Automake.
CPP to a command that runs the
C preprocessor. If `$CC -E' doesn't work, `/lib/cpp' is used.
It is only portable to run CPP on files with a `.c'
extension.
If the current language is C (see section Language Choice), many of the
specific test macros use the value of CPP indirectly by calling
AC_TRY_CPP, AC_CHECK_HEADER, AC_EGREP_HEADER, or
AC_EGREP_CPP.
Some preprocessors don't indicate missing include files by the error status. For such preprocessors an internal variable is set that causes other macros to check the standard error from the preprocessor and consider the test failed if any warnings have been reported.
The following macros check for C compiler or machine architecture
features. To check for characteristics not listed here, use
AC_TRY_COMPILE (see section Examining Syntax) or AC_TRY_RUN
(see section Checking Run Time Behavior)
This macro runs a test-case if endianness cannot be determined from the system header files. When cross-compiling the test-case is not run but grep'ed for some magic values. action-if-unknown is executed if the latter case fails to determine the byte sex of the host system.
The default for action-if-true is to define `WORDS_BIGENDIAN'. The default for action-if-false is to do nothing. And finally, the default for action-if-unknown is to abort configure and tell the installer which variable he should preset to bypass this test.
const, define const to be empty. Some C compilers that do
not define __STDC__ do support const; some compilers that
define __STDC__ do not completely support const. Programs
can simply use const as if every C compiler supported it; for
those that don't, the `Makefile' or configuration header file will
define it as empty.
Occasionally installers use a C++ compiler to compile C code, typically
because they lack a C compiler. This causes problems with const,
because C and C++ treat const differently. For example:
const int foo;
is valid in C but not in C++. These differences unfortunately cannot be
papered over by defining const to be empty.
If @command{autoconf} detects this situation, it leaves const alone,
as this generally yields better results in practice. However, using a
C++ compiler to compile C code is not recommended or supported, and
installers who run into trouble in this area should get a C compiler
like GCC to compile their C code.
volatile,
define volatile to be empty. Programs can simply use
volatile as if every C compiler supported it; for those that do
not, the `Makefile' or configuration header will define it as
empty.
If the correctness of your program depends on the semantics of
volatile, simply defining it to be empty does, in a sense, break
your code. However, given that the compiler does not support
volatile, you are at its mercy anyway. At least your
program will compile, when it wouldn't before.
In general, the volatile keyword is a feature of ANSI C, so
you might expect that volatile is available only when
__STDC__ is defined. However, Ultrix 4.3's native compiler does
support volatile, but does not defined __STDC__.
inline, do nothing.
Otherwise define inline to __inline__ or __inline
if it accepts one of those, otherwise define inline to be empty.
char is unsigned, define __CHAR_UNSIGNED__,
unless the C compiler predefines it.
long double type with more
range or precision than the double type, define
HAVE_LONG_DOUBLE.
HAVE_STRINGIZE. The stringizing operator is `#' and is
found in macros such as this:
#define x(y) #y
PROTOTYPES and __PROTOTYPES.
In the case the compiler does not handle
prototypes, you should use ansi2knr, which comes with the
Automake distribution, to unprotoize function definitions. For
function prototypes, you should first define PARAMS:
#ifndef PARAMS # if PROTOTYPES # define PARAMS(protos) protos # else /* no PROTOTYPES */ # define PARAMS(protos) () # endif /* no PROTOTYPES */ #endif
then use it this way:
size_t my_strlen PARAMS ((const char *));
This macro also defines __PROTOTYPES; this is for the benefit of
header files that cannot use macros that infringe on user name space.
CC if using the
GNU C compiler and ioctl does not work properly without
@option{-traditional}. That usually happens when the fixed header files
have not been installed on an old system. Since recent versions of the
GNU C compiler fix the header files automatically when installed,
this is becoming a less prevalent problem.
CXX or CCC (in that order) is set; if so, then set output
variable CXX to its value.
Otherwise, if the macro is invoked without an argument, then search for
a C++ compiler under the likely names (first g++ and c++
then other names). If none of those checks succeed, then as a last
resort set CXX to g++.
This macro may, however, be invoked with an optional first argument
which, if specified, must be a space separated list of C++ compilers to
search for. This just gives the user an opportunity to specify an
alternative search list for the C++ compiler. For example, if you
didn't like the default order, then you could invoke AC_PROG_CXX
like this:
AC_PROG_CXX(cl KCC CC cxx cc++ xlC aCC c++ g++ egcs gcc)
If using the GNU C++ compiler, set shell variable GXX to
`yes'. If output variable CXXFLAGS was not already set, set
it to @option{-g -O2} for the GNU C++ compiler (@option{-O2} on
systems where G++ does not accept @option{-g}), or @option{-g} for other
compilers.
CXXCPP to a command that runs the C++
preprocessor. If `$CXX -E' doesn't work, `/lib/cpp' is used.
It is only portable to run CXXCPP on files with a `.c',
`.C', or `.cc' extension.
If the current language is C++ (see section Language Choice), many of the
specific test macros use the value of CXXCPP indirectly by
calling AC_TRY_CPP, AC_CHECK_HEADER,
AC_EGREP_HEADER, or AC_EGREP_CPP.
Some preprocessors don't indicate missing include files by the error status. For such preprocessors an internal variable is set that causes other macros to check the standard error from the preprocessor and consider the test failed if any warnings have been reported. However, it is not known whether such broken preprocessors exist for C++.
F77 is not already
set in the environment, then check for g77 and f77, and
then some other names. Set the output variable F77 to the name
of the compiler found.
This macro may, however, be invoked with an optional first argument
which, if specified, must be a space separated list of Fortran 77
compilers to search for. This just gives the user an opportunity to
specify an alternative search list for the Fortran 77 compiler. For
example, if you didn't like the default order, then you could invoke
AC_PROG_F77 like this:
AC_PROG_F77(fl32 f77 fort77 xlf cf77 g77 f90 xlf90)
If using g77 (the GNU Fortran 77 compiler), then
AC_PROG_F77 will set the shell variable G77 to `yes'.
If the output variable FFLAGS was not already set in the
environment, then set it to @option{-g -02} for g77 (or @option{-O2}
where g77 does not accept @option{-g}). Otherwise, set
FFLAGS to @option{-g} for all other Fortran 77 compilers.
F77_NO_MINUS_C_MINUS_O if it
does not.
The following macros check for Fortran 77 compiler characteristics. To
check for characteristics not listed here, use AC_TRY_COMPILE
(see section Examining Syntax) or AC_TRY_RUN (see section Checking Run Time Behavior),
making sure to first set the current language to Fortran 77
AC_LANG(Fortran 77) (see section Language Choice).
FLIBS is set to these flags.
This macro is intended to be used in those situations when it is necessary to mix, e.g. C++ and Fortran 77 source code into a single program or shared library (see section `Mixing Fortran 77 With C and C++' in GNU Automake).
For example, if object files from a C++ and Fortran 77 compiler must be linked together, then the C++ compiler/linker must be used for linking (since special C++-ish things need to happen at link time like calling global constructors, instantiating templates, enabling exception support, etc.).
However, the Fortran 77 intrinsic and run-time libraries must be linked
in as well, but the C++ compiler/linker doesn't know by default how to
add these Fortran 77 libraries. Hence, the macro
AC_F77_LIBRARY_LDFLAGS was created to determine these Fortran 77
libraries.
The macro AC_F77_DUMMY_MAIN or AC_F77_MAIN will probably
also be necessary to link C/C++ with Fortran; see below.
AC_F77_LIBRARY_LDFLAGS provide their own main entry
function that initializes things like Fortran I/O, and which then calls
a user-provided entry function named e.g. MAIN__ to run the
user's program. The AC_F77_DUMMY_MAIN or AC_F77_MAIN
macro figures out how to deal with this interaction.
When using Fortran for purely numerical functions (no I/O, etcetera),
users often prefer to provide their own main and skip the Fortran
library initializations. In this case, however, one may still need to
provide a dummy MAIN__ routine in order to prevent linking errors
on some systems. AC_F77_DUMMY_MAIN detects whether any such
routine is required for linking, and what its name is; the shell
variable F77_DUMMY_MAIN holds this name, unknown when no
solution was found, and none when no such dummy main is needed.
By default, action-if-found defines F77_DUMMY_MAIN to the
name of this routine (e.g. MAIN__) if it is required.
@ovar{action-if-not-found} defaults to exiting with an error.
In order to link with Fortran routines, the user's C/C++ program should then include the following code to define the dummy main if it is needed:
#ifdef F77_DUMMY_MAIN
# ifdef __cplusplus
extern "C"
# endif
int F77_DUMMY_MAIN() { return 1; }
#endif
Note that AC_F77_DUMMY_MAIN is called automatically from
AC_F77_WRAPPERS; there is generally no need to call it explicitly
unless one wants to change the default actions.
AC_F77_DUMMY_MAIN, many Fortran libraries
allow you to provide an entry point called e.g. MAIN__ instead of
the usual main, which is then called by a main function in
the Fortran libraries that initializes things like Fortran I/O. The
AC_F77_MAIN macro detects whether it is possible to
utilize such an alternate main function, and defines F77_MAIN to
the name of the function. (If no alternate main function name is found,
F77_MAIN is simply defined to main.)
Thus, when calling Fortran routines from C that perform things like I/O,
one should use this macro and name the "main" function F77_MAIN
instead of main.
F77_FUNC(name,NAME) and
F77_FUNC_(name,NAME) to properly mangle the names of C/C++
identifiers, and identifiers with underscores, respectively, so that
they match the name-mangling scheme used by the Fortran 77 compiler.
Fortran 77 is case-insensitive, and in order to achieve this the Fortran
77 compiler converts all identifiers into a canonical case and format.
To call a Fortran 77 subroutine from C or to write a C function that is
callable from Fortran 77, the C program must explicitly use identifiers
in the format expected by the Fortran 77 compiler. In order to do this,
one simply wraps all C identifiers in one of the macros provided by
AC_F77_WRAPPERS. For example, suppose you have the following
Fortran 77 subroutine:
subroutine foobar(x,y)
double precision x, y
y = 3.14159 * x
return
end
You would then declare its prototype in C or C++ as:
#define FOOBAR_F77 F77_FUNC(foobar,FOOBAR) #ifdef __cplusplus extern "C" /* prevent C++ name mangling */ #endif void FOOBAR_F77(double *x, double *y);
Note that we pass both the lowercase and uppercase versions of the
function name to F77_FUNC so that it can select the right one.
Note also that all parameters to Fortran 77 routines are passed as
pointers (see section `Mixing Fortran 77 With C and C++' in GNU Automake).
Although Autoconf tries to be intelligent about detecting the
name-mangling scheme of the Fortran 77 compiler, there may be Fortran 77
compilers that it doesn't support yet. In this case, the above code
will generate a compile-time error, but some other behavior
(e.g. disabling Fortran-related features) can be induced by checking
whether the F77_FUNC macro is defined.
Now, to call that routine from a C program, we would do something like:
{
double x = 2.7183, y;
FOOBAR_F77(&x, &y);
}
If the Fortran 77 identifier contains an underscore
(e.g. foo_bar), you should use F77_FUNC_ instead of
F77_FUNC (with the same arguments). This is because some Fortran
77 compilers mangle names differently if they contain an underscore.
AC_F77_WRAPPERS). shellvar is
optional; if it is not supplied, the shell variable will be simply
name. The purpose of this macro is to give the caller a way to
access the name-mangling information other than through the C
preprocessor as above; for example, to call Fortran routines from some
language other than C/C++.
The following macros check for operating system services or capabilities.
xmkmf on a
trivial `Imakefile' and examining the `Makefile' that it
produces. If that fails (such as if xmkmf is not present), look
for them in several directories where they often reside. If either
method is successful, set the shell variables x_includes and
x_libraries to their locations, unless they are in directories
the compiler searches by default.
If both methods fail, or the user gave the command line option
@option{--without-x}, set the shell variable no_x to `yes';
otherwise set it to the empty string.
AC_PATH_X. It adds the C compiler flags
that X needs to output variable X_CFLAGS, and the X linker flags
to X_LIBS. Define X_DISPLAY_MISSING if X is not
available.
This macro also checks for special libraries that some systems need in
order to compile X programs. It adds any that the system needs to
output variable X_EXTRA_LIBS. And it checks for special X11R6
libraries that need to be linked with before @option{-lX11}, and adds
any found to the output variable X_PRE_LIBS.
interpval; it will be set to `yes'
if the system supports `#!', `no' if not.
CC. Define
_FILE_OFFSET_BITS and _LARGE_FILES if necessary.
Large-file support can be disabled by configuring with the @option{--disable-largefile} option.
If you use this macro, check that your program works even when
off_t is longer than long, since this is common when
large-file support is enabled. For example, it is not correct to print
an arbitrary off_t value X with printf ("%ld",
(long) X).
HAVE_LONG_FILE_NAMES.
am_cv_sys_posix_termios to
`yes'. If not, set the variable to `no'.
The following macros check for certain operating systems that need special treatment for some programs, due to exceptional oddities in their header files or libraries. These macros are warts; they will be replaced by a more systematic approach, based on the functions they make available or the environments they provide.
_ALL_SOURCE. Allows the use of some BSD
functions. Should be called before any macros that run the C compiler.
LIBS if necessary for POSIX facilities. Call this
after AC_PROG_CC and before any other macros that use POSIX
interfaces. INTERACTIVE UNIX is no longer sold, and Sun says that
they will drop support for it on 2006-07-23, so this macro is becoming
obsolescent.
_MINIX and _POSIX_SOURCE and define
_POSIX_1_SOURCE to be 2. This allows the use of POSIX
facilities. Should be called before any macros that run the C compiler.
If the existing feature tests don't do something you need, you have to write new ones. These macros are the building blocks. They provide ways for other macros to check whether various kinds of features are available and report the results.
This chapter contains some suggestions and some of the reasons why the existing tests are written the way they are. You can also learn a lot about how to write Autoconf tests by looking at the existing ones. If something goes wrong in one or more of the Autoconf tests, this information can help you understand the assumptions behind them, which might help you figure out how to best solve the problem.
These macros check the output of the C compiler system. They do not cache the results of their tests for future use (see section Caching Results), because they don't know enough about the information they are checking for to generate a cache variable name. They also do not print any messages, for the same reason. The checks for particular kinds of C features call these macros and do cache their results and print messages about what they're checking for.
When you write a feature test that could be applicable to more than one software package, the best thing to do is encapsulate it in a new macro. See section Writing Autoconf Macros, for how to do that.
The macro AC_TRY_CPP is used to check whether particular header
files exist. You can check for one at a time, or more than one if you
need several header files to all exist for some purpose.
This macro uses CPPFLAGS, but not CFLAGS, because
@option{-g}, @option{-O}, etc. are not valid options to many C
preprocessors.
Here is how to find out whether a header file contains a particular
declaration, such as a typedef, a structure, a structure member, or a
function. Use AC_EGREP_HEADER instead of running grep
directly on the header file; on some systems the symbol might be defined
in another header file that the file you are checking `#include's.
egrep regular expression
pattern, execute shell commands action-if-found, otherwise
execute action-if-not-found.
To check for C preprocessor symbols, either defined by header files or
predefined by the C preprocessor, use AC_EGREP_CPP. Here is an
example of the latter:
AC_EGREP_CPP(yes, [#ifdef _AIX yes #endif ], is_aix=yes, is_aix=no)
egrep regular expression pattern, execute shell commands
action-if-found, otherwise execute action-if-not-found.
This macro calls AC_PROG_CPP or AC_PROG_CXXCPP (depending
on which language is current, see section Language Choice), if it hasn't
been called already.
To check for a syntax feature of the C, C++ or Fortran 77 compiler, such
as whether it recognizes a certain keyword, use AC_TRY_COMPILE to
try to compile a small program that uses that feature. You can also use
it to check for structures and structure members that are not present on
all systems.
This macro double quotes both includes and function-body.
For C and C++, includes is any #include statements needed
by the code in function-body (includes will be ignored if
the currently selected language is Fortran 77). This macro also uses
CFLAGS or CXXFLAGS if either C or C++ is the currently
selected language, as well as CPPFLAGS, when compiling. If
Fortran 77 is the currently selected language then FFLAGS will be
used when compiling.
This macro does not try to link; use AC_TRY_LINK if you need to
do that (see section Examining Libraries).
To check for a library, a function, or a global variable, Autoconf
@command{configure} scripts try to compile and link a small program that
uses it. This is unlike Metaconfig, which by default uses nm
or ar on the C library to try to figure out which functions are
available. Trying to link with the function is usually a more reliable
approach because it avoids dealing with the variations in the options
and output formats of nm and ar and in the location of the
standard libraries. It also allows configuring for cross-compilation or
checking a function's runtime behavior if needed. On the other hand, it
can be slower than scanning the libraries once.
A few systems have linkers that do not return a failure exit status when
there are unresolved functions in the link. This bug makes the
configuration scripts produced by Autoconf unusable on those systems.
However, some of them can be given options that make the exit status
correct. This is a problem that Autoconf does not currently handle
automatically. If users encounter this problem, they might be able to
solve it by setting LDFLAGS in the environment to pass whatever
options the linker needs (for example, @option{-Wl,-dn} on @sc{mips
risc/os}).
AC_TRY_LINK is used to compile test programs to test for
functions and global variables. It is also used by AC_CHECK_LIB
to check for libraries (see section Library Files), by adding the library being
checked for to LIBS temporarily and trying to link a small
program.
This macro double quotes both includes and function-body.
For C and C++, includes is any #include statements needed
by the code in function-body (includes will be ignored if
the currently selected language is Fortran 77). This macro also uses
CFLAGS or CXXFLAGS if either C or C++ is the currently
selected language, as well as CPPFLAGS, when compiling. If
Fortran 77 is the currently selected language then FFLAGS will be
used when compiling. However, both LDFLAGS and LIBS will
be used during linking in all cases.
If the file compiles and links successfully, run shell commands action-if-found, otherwise run action-if-not-found.
Sometimes you need to find out how a system performs at run time, such as whether a given function has a certain capability or bug. If you can, make such checks when your program runs instead of when it is configured. You can check for things like the machine's endianness when your program initializes itself.
If you really need to test for a run-time behavior while configuring,
you can write a test program to determine the result, and compile and
run it using AC_TRY_RUN. Avoid running test programs if
possible, because this prevents people from configuring your package for
cross-compiling.
Use the following macro if you need to test run-time behavior of the system while configuring.
This macro double quotes program, the text of a program in the
current language (see section Language Choice), on which shell variable and
back quote substitutions are performed. This macro uses CFLAGS
or CXXFLAGS, CPPFLAGS, LDFLAGS, and LIBS
when compiling.
If the C compiler being used does not produce executables that run on the system where @command{configure} is being run, then the test program is not run. If the optional shell commands action-if-cross-compiling are given, they are run instead. Otherwise, @command{configure} prints an error message and exits.
In the action-if-false section, the exit status of the program is available in the shell variable `$?', but be very careful to limit yourself to positive values smaller than 127; bigger values shall be saved into a file by the program. Note also that you have simply no guarantee that this exit status is issued by the program, or by the failure of its compilation. In other words, use this feature if sadist only, it was reestablished because the Autoconf maintainers grew tired of receiving "bug reports".
Try to provide a pessimistic default value to use when cross-compiling
makes run-time tests impossible. You do this by passing the optional
last argument to AC_TRY_RUN. @command{autoconf} prints a warning
message when creating @command{configure} each time it encounters a call to
AC_TRY_RUN with no action-if-cross-compiling argument
given. You may ignore the warning, though users will not be able to
configure your package for cross-compiling. A few of the macros
distributed with Autoconf produce this warning message.
To configure for cross-compiling you can also choose a value for those parameters based on the canonical system name (see section Manual Configuration). Alternatively, set up a test results cache file with the correct values for the host system (see section Caching Results).
To provide a default for calls of AC_TRY_RUN that are embedded in
other macros, including a few of the ones that come with Autoconf, you
can call AC_PROG_CC before running them. Then, if the shell
variable cross_compiling is set to `yes', use an alternate
method to get the results instead of calling the macros.
Test programs should not write anything to the standard output. They
should return 0 if the test succeeds, nonzero otherwise, so that success
can be distinguished easily from a core dump or other failure;
segmentation violations and other failures produce a nonzero exit
status. Test programs should exit, not return, from
main, because on some systems (old Suns, at least) the argument
to return in main is ignored.
Test programs can use #if or #ifdef to check the values of
preprocessor macros defined by tests that have already run. For
example, if you call AC_HEADER_STDC, then later on in
`configure.ac' you can have a test program that includes an
ANSI C header file conditionally:
#if STDC_HEADERS # include <stdlib.h> #endif
If a test program needs to use or create a data file, give it a name that starts with `conftest', such as `conftest.data'. The @command{configure} script cleans up by running `rm -rf conftest*' after running test programs and if the script is interrupted.
Function declarations in test programs should have a prototype conditionalized for C++. In practice, though, test programs rarely need functions that take arguments.
#ifdef __cplusplus foo (int i) #else foo (i) int i; #endif
Functions that test programs declare should also be conditionalized for C++, which requires `extern "C"' prototypes. Make sure to not include any header files containing clashing prototypes.
#ifdef __cplusplus extern "C" void *malloc (size_t); #else char *malloc (); #endif
If a test program calls a function with invalid parameters (just to see
whether it exists), organize the program to ensure that it never invokes
that function. You can do this by calling it in another function that is
never invoked. You can't do it by putting it after a call to
exit, because GCC version 2 knows that exit never returns
and optimizes out any code that follows it in the same block.
If you include any header files, make sure to call the functions
relevant to them with the correct number of arguments, even if they are
just 0, to avoid compilation errors due to prototypes. GCC version 2
has internal prototypes for several functions that it automatically
inlines; for example, memcpy. To avoid errors when checking for
them, either pass them the correct number of arguments or redeclare them
with a different return type (such as char).
This section aims at presenting some systems and pointers to documentation. It may help you addressing particular problems reported by users.
Some operations are accomplished in several possible ways, depending on the UNIX variant. Checking for them essentially requires a "case statement". Autoconf does not directly provide one; however, it is easy to simulate by using a shell variable to keep track of whether a way to perform the operation has been found yet.
Here is an example that uses the shell variable fstype to keep
track of whether the remaining cases need to be checked.
AC_MSG_CHECKING([how to get file system type])
fstype=no
# The order of these tests is important.
AC_TRY_CPP([#include <sys/statvfs.h>
#include <sys/fstyp.h>],
[AC_DEFINE(FSTYPE_STATVFS) fstype=SVR4])
if test $fstype = no; then
AC_TRY_CPP([#include <sys/statfs.h>
#include <sys/fstyp.h>],
[AC_DEFINE(FSTYPE_USG_STATFS) fstype=SVR3])
fi
if test $fstype = no; then
AC_TRY_CPP([#include <sys/statfs.h>
#include <sys/vmount.h>],
[AC_DEFINE(FSTYPE_AIX_STATFS) fstype=AIX])
fi
# (more cases omitted here)
AC_MSG_RESULT([$fstype])
Autoconf-generated @command{configure} scripts check for the C compiler and its features by default. Packages that use other programming languages (maybe more than one, e.g. C and C++) need to test features of the compilers for the respective languages. The following macros determine which programming language is used in the subsequent tests in `configure.ac'.
Supported languages are:
CC and CPP and use extension
`.c' for test programs.
CXX and CXXCPP and use
extension `.C' for test programs.
F77 and use extension `.f' for
test programs.
AC_LANG) on a stack, and
then select the language. Use this macro and AC_LANG_POP
in macros that need to temporarily switch to a particular language.
AC_LANG_PUSH, and remove it from the stack.
If given, language specifies the language we just quit. It is a good idea to specify it when it's known (which should be the case...), since Autoconf will detect inconsistencies.
AC_LANG_PUSH(Fortran 77) # Perform some tests on Fortran 77. # ... AC_LANG_POP(Fortran 77)
AC_REQUIRE (see section Prerequisite Macros) with an
argument of either AC_PROG_CPP or AC_PROG_CXXCPP,
depending on which language is current.
Once @command{configure} has determined whether a feature exists, what can it do to record that information? There are four sorts of things it can do: define a C preprocessor symbol, set a variable in the output files, save the result in a cache file for future @command{configure} runs, and print a message letting the user know the result of the test.
A common action to take in response to a feature test is to define a C
preprocessor symbol indicating the results of the test. That is done by
calling AC_DEFINE or AC_DEFINE_UNQUOTED.
By default, AC_OUTPUT places the symbols defined by these macros
into the output variable DEFS, which contains an option
@option{-Dsymbol=value} for each symbol defined. Unlike in
Autoconf version 1, there is no variable DEFS defined while
@command{configure} is running. To check whether Autoconf macros have
already defined a certain C preprocessor symbol, test the value of the
appropriate cache variable, as in this example:
AC_CHECK_FUNC(vprintf, [AC_DEFINE(HAVE_VPRINTF)]) if test "$ac_cv_func_vprintf" != yes; then AC_CHECK_FUNC(_doprnt, [AC_DEFINE(HAVE_DOPRNT)]) fi
If AC_CONFIG_HEADERS has been called, then instead of creating
DEFS, AC_OUTPUT creates a header file by substituting the
correct values into #define statements in a template file.
See section Configuration Header Files, for more information about this kind of
output.
AC_CONFIG_HEADERS it should not contain any `#'
characters, as make tends to eat them. To use a shell variable
(which you need to do in order to define a value containing the M4 quote
characters `[' or `]'), use AC_DEFINE_UNQUOTED instead.
description is only useful if you are using
AC_CONFIG_HEADERS. In this case, description is put into
the generated `config.h.in' as the comment before the macro define.
The following example defines the C preprocessor variable
EQUATION to be the string constant `"$a > $b"':
AC_DEFINE(EQUATION, "$a > $b")
AC_DEFINE, but three shell expansions are
performed--once--on variable and value: variable expansion
(`$'), command substitution (``'), and backslash escaping
(`\'). Single and double quote characters in the value have no
special meaning. Use this macro instead of AC_DEFINE when
variable or value is a shell variable. Examples:
AC_DEFINE_UNQUOTED(config_machfile, "$machfile") AC_DEFINE_UNQUOTED(GETGROUPS_T, $ac_cv_type_getgroups) AC_DEFINE_UNQUOTED($ac_tr_hdr)
Due to the syntactical bizarreness of the Bourne shell, do not use
semicolons to separate AC_DEFINE or AC_DEFINE_UNQUOTED
calls from other macro calls or shell code; that can cause syntax errors
in the resulting @command{configure} script. Use either spaces or
newlines. That is, do this:
AC_CHECK_HEADER(elf.h, [AC_DEFINE(SVR4) LIBS="$LIBS -lelf"])
or this:
AC_CHECK_HEADER(elf.h, [AC_DEFINE(SVR4) LIBS="$LIBS -lelf"])
instead of this:
AC_CHECK_HEADER(elf.h, [AC_DEFINE(SVR4); LIBS="$LIBS -lelf"])
Another way to record the results of tests is to set output variables, which are shell variables whose values are substituted into files that @command{configure} outputs. The two macros below create new output variables. See section Preset Output Variables, for a list of output variables that are always available.
AC_OUTPUT
substitute the variable variable into output files (typically one
or more `Makefile's). This means that AC_OUTPUT will
replace instances of `@variable@' in input files with the
value that the shell variable variable has when AC_OUTPUT
is called. This value of variable should not contain literal
newlines.
If value is given, in addition assign it to `variable'.
AC_OUTPUT insert (without substitutions) the contents of the file
named by shell variable variable into output files. This means
that AC_OUTPUT will replace instances of
`@variable@' in output files (such as `Makefile.in')
with the contents of the file that the shell variable variable
names when AC_OUTPUT is called. Set the variable to
`/dev/null' for cases that do not have a file to insert.
This macro is useful for inserting `Makefile' fragments containing
special dependencies or other make directives for particular host
or target types into `Makefile's. For example, `configure.ac'
could contain:
AC_SUBST_FILE(host_frag) host_frag=$srcdir/conf/sun4.mh
and then a `Makefile.in' could contain:
@host_frag@
Running @command{configure} in different environments can be extremely dangerous. If for instance the user runs `CC=bizarre-cc ./configure', then the cache, `config.h' and many other output files will depend upon @command{bizarre-cc} being the C compiler. If for some reason the user runs @command{/configure} again, or if it is run via `./config.status --recheck', (See section Automatic Remaking, and see section Recreating a Configuration), then the configuration can be inconsistent, composed of results depending upon two different compilers.
Such variables are named precious variables, and can be declared
as such by AC_ARG_VAR.
Being precious means that
AC_SUBST'd.
CC in `./configure
CC=bizarre-cc', it is impossible to notice it in `CC=bizarre-cc
./configure', which, unfortunately, is what most users do.
$ ./configure --silent --config-cache $ CC=cc ./configure --silent --config-cache configure: error: `CC' was not set in the previous run configure: error: changes in the environment can compromise \ the build configure: error: run `make distclean' and/or \ `rm config.cache' and start overand similarly if the variable is unset, or if its content is changed.
$ CC=/usr/bin/cc ./configure undeclared_var=raboof --silent $ ./config.status --recheck running /bin/sh ./configure undeclared_var=raboof --silent \ CC=/usr/bin/cc --no-create --no-recursion
To avoid checking for the same features repeatedly in various @command{configure} scripts (or in repeated runs of one script), @command{configure} can optionally save the results of many checks in a cache file (see section Cache Files). If a @command{configure} script runs with caching enabled and finds a cache file, it reads the results of previous runs from the cache and avoids rerunning those checks. As a result, @command{configure} can then run much faster than if it had to perform all of the checks every time.
The commands-to-set-it must have no side effects except for setting the variable cache-id, see below.
AC_CACHE_VAL that takes care of printing the
messages. This macro provides a convenient shorthand for the most
common way to use these macros. It calls AC_MSG_CHECKING for
message, then AC_CACHE_VAL with the cache-id and
commands arguments, and AC_MSG_RESULT with cache-id.
The commands-to-set-it must have no side effects except for setting the variable cache-id, see below.
It is very common to find buggy macros using AC_CACHE_VAL or
AC_CACHE_CHECK, because people are tempted to call
AC_DEFINE in the commands-to-set-it. Instead, the code that
follows the call to AC_CACHE_VAL should call
AC_DEFINE, by examining the value of the cache variable. For
instance, the following macro is broken:
AC_DEFUN([AC_SHELL_TRUE],
[AC_CACHE_CHECK([whether true(1) works], [ac_cv_shell_true_works],
[ac_cv_shell_true_works=no
true && ac_cv_shell_true_works=yes
if test $ac_cv_shell_true_works = yes; then
AC_DEFINE([TRUE_WORKS], 1
[Define if `true(1)' works properly.])
fi])
])
This fails if the cache is enabled: the second time this macro is run,
TRUE_WORKS will not be defined. The proper implementation
is:
AC_DEFUN([AC_SHELL_TRUE],
[AC_CACHE_CHECK([whether true(1) works], [ac_cv_shell_true_works],
[ac_cv_shell_true_works=no
true && ac_cv_shell_true_works=yes])
if test $ac_cv_shell_true_works = yes; then
AC_DEFINE([TRUE_WORKS], 1
[Define if `true(1)' works properly.])
fi
])
Also, commands-to-set-it should not print any messages, for
example with AC_MSG_CHECKING; do that before calling
AC_CACHE_VAL, so the messages are printed regardless of whether
the results of the check are retrieved from the cache or determined by
running the shell commands.
The names of cache variables should have the following format:
package-prefix_cv_value-type_specific-value_@ovar{additional-options}
for example, `ac_cv_header_stat_broken' or `ac_cv_prog_gcc_traditional'. The parts of the variable name are:
_cv_
The values assigned to cache variables may not contain newlines. Usually, their values will be boolean (`yes' or `no') or the names of files or functions; so this is not an important restriction.
A cache file is a shell script that caches the results of configure tests run on one system so they can be shared between configure scripts and configure runs. It is not useful on other systems. If its contents are invalid for some reason, the user may delete or edit it.
By default, @command{configure} uses no cache file (technically, it uses @option{--cache-file=/dev/null}), to avoid problems caused by accidental use of stale cache files.
To enable caching, @command{configure} accepts @option{--config-cache} (or
@option{-C}) to cache results in the file `config.cache'.
Alternatively, @option{--cache-file=file} specifies that
file be the cache file. The cache file is created if it does not
exist already. When @command{configure} calls @command{configure} scripts in
subdirectories, it uses the @option{--cache-file} argument so that they
share the same cache. See section Configuring Other Packages in Subdirectories, for information on
configuring subdirectories with the AC_CONFIG_SUBDIRS macro.
`config.status' only pays attention to the cache file if it is given the @option{--recheck} option, which makes it rerun @command{configure}.
It is wrong to try to distribute cache files for particular system types. There is too much room for error in doing that, and too much administrative overhead in maintaining them. For any features that can't be guessed automatically, use the standard method of the canonical system type and linking files (see section Manual Configuration).
The site initialization script can specify a site-wide cache file to use, instead of the usual per-program cache. In this case, the cache file will gradually accumulate information whenever someone runs a new @command{configure} script. (Running @command{configure} merges the new cache results with the existing cache file.) This may cause problems, however, if the system configuration (e.g. the installed libraries or compilers) changes and the stale cache file is not deleted.
If your configure script, or a macro called from configure.ac, happens
to abort the configure process, it may be useful to checkpoint the cache
a few times at key points using AC_CACHE_SAVE. Doing so will
reduce the amount of time it takes to re-run the configure script with
(hopefully) the error that caused the previous abort corrected.
AC_INIT.
AC_OUTPUT, but it can be quite useful to call
AC_CACHE_SAVE at key points in configure.ac.
For instance:
... AC_INIT, etc. ... # Checks for programs. AC_PROG_CC AC_PROG_GCC_TRADITIONAL ... more program checks ... AC_CACHE_SAVE # Checks for libraries. AC_CHECK_LIB(nsl, gethostbyname) AC_CHECK_LIB(socket, connect) ... more lib checks ... AC_CACHE_SAVE # Might abort... AM_PATH_GTK(1.0.2,, [AC_MSG_ERROR([GTK not in path])]) AM_PATH_GTKMM(0.9.5,, [AC_MSG_ERROR([GTK not in path])]) ... AC_OUTPUT, etc. ...
@command{configure} scripts need to give users running them several kinds of information. The following macros print messages in ways appropriate for each kind. The arguments to all of them get enclosed in shell double quotes, so the shell performs variable and back-quote substitution on them.
These macros are all wrappers around the echo shell command.
@command{configure} scripts should rarely need to run echo directly
to print messages for the user. Using these macros makes it easy to
change how and when each kind of message is printed; such changes need
only be made to the macro definitions and all of the callers will change
automatically.
To diagnose static issues, i.e., when @command{autoconf} is run, see section Reporting Messages.
AC_MSG_RESULT to print the result of the check and the
newline. The feature-description should be something like
`whether the Fortran compiler accepts C++ comments' or `for
c89'.
This macro prints nothing if @command{configure} is run with the @option{--quiet} or @option{--silent} option.
AC_MSG_CHECKING, and the result-description should be
the completion of the message printed by the call to
AC_MSG_CHECKING.
This macro prints nothing if @command{configure} is run with the @option{--quiet} or @option{--silent} option.
AC_MSG_NOTICE([checking if stack overflow is detectable])
This macro prints nothing if @command{configure} is run with the @option{--quiet} or @option{--silent} option.
The error-description should start with a lower-case letter, and "cannot" is preferred to "can't".
AC_MSG_WARN should
provide a default (back-up) behavior for the situations they warn about.
problem-description should be something like `ln -s seems to
make hard links'.
Autoconf is written on top of two layers: M4sugar, which provides convenient macros for pure M4 programming, and M4sh, which provides macros dedicated to shell script generation.
As of this version of Autoconf, these two layers are still experimental, and their interface might change in the future. As a matter of fact, anything that is not documented must not be used.
The most common brokenness of existing macros is an improper quotation. This section, which users of Autoconf can skip, but which macro writers must read, first justifies the quotation scheme that was chosen for Autoconf and then ends with a rule of thumb. Understanding the former helps one to follow the latter.
To fully understand where proper quotation is important, you first need
to know what are the special characters in Autoconf: `#' introduces
a comment inside which no macro expansion is performed, `,'
separates arguments, `[' and `]' are the quotes themselves,
and finally `(' and `)' (which m4 tries to match by
pairs).
In order to understand the delicate case of macro calls, we first have to present some obvious failures. Below they are "obvious-ified", although you find them in real life, they are usually in disguise.
Comments, introduced by a hash and running up to the newline, are opaque tokens to the top level: active characters are turned off, and there is no macro expansion:
# define([def], ine) =># define([def], ine)
Each time there can be a macro expansion, there is a quotation expansion; i.e., one level of quotes is stripped:
int tab[10]; =>int tab10; [int tab[10];] =>int tab[10];
Without this in mind, the reader will try hopelessly to use her macro
array:
define([array], [int tab[10];]) array =>int tab10; [array] =>array
How can you correctly output the intended results(2)?
Let's proceed on the interaction between active characters and macros with this small macro, which just returns its first argument:
define([car], [$1])
The two pairs of quotes above are not part of the arguments of
define; rather, they are understood by the top level when it
tries to find the arguments of define. Therefore, it is
equivalent to write:
define(car, $1)
But, while it is acceptable for a `configure.ac' to avoid unneeded quotes, it is bad practice for Autoconf macros which must both be more robust and also advocate perfect style.
At the top level, there are only two possible quotings: either you quote or you don't:
car(foo, bar, baz) =>foo [car(foo, bar, baz)] =>car(foo, bar, baz)
Let's pay attention to the special characters:
car(#) error-->EOF in argument list
The closing parenthesis is hidden in the comment; with a hypothetical quoting, the top level understood it this way:
car([#)]
Proper quotation, of course, fixes the problem:
car([#]) =>#
The reader will easily understand the following examples:
car(foo, bar) =>foo car([foo, bar]) =>foo, bar car((foo, bar)) =>(foo, bar) car([(foo], [bar)]) =>(foo car([], []) => car([[]], [[]]) =>[]
With this in mind, we can explore the cases where macros invoke macros...
The examples below use the following macros:
define([car], [$1]) define([active], [ACT, IVE]) define([array], [int tab[10]])
Each additional embedded macro call introduces other possible interesting quotations:
car(active) =>ACT car([active]) =>ACT, IVE car([[active]]) =>active
In the first case, the top level looks for the arguments of car,
and finds `active'. Because m4 evaluates its arguments
before applying the macro, `active' is expanded, which results in:
car(ACT, IVE) =>ACT
In the second case, the top level gives `active' as first and only
argument of car, which results in:
active =>ACT, IVE
i.e., the argument is evaluated after the macro that invokes it.
In the third case, car receives `[active]', which results in:
[active] =>active
exactly as we already saw above.
The example above, applied to a more realistic example, gives:
car(int tab[10];) =>int tab10; car([int tab[10];]) =>int tab10; car([[int tab[10];]]) =>int tab[10];
Huh? The first case is easily understood, but why is the second wrong,
and the third right? To understand that, you must know that after
m4 expands a macro, the resulting text is immediately subjected
to macro expansion and quote removal. This means that the quote removal
occurs twice--first before the argument is passed to the car
macro, and second after the car macro expands to the first
argument.
As the author of the Autoconf macro car, you then consider it to
be incorrect that your users have to double-quote the arguments of
car, so you "fix" your macro. Let's call it qar for
quoted car:
define([qar], [[$1]])
and check that qar is properly fixed:
qar([int tab[10];]) =>int tab[10];
Ahhh! That's much better.
But note what you've done: now that the arguments are literal strings, if the user wants to use the results of expansions as arguments, she has to use an unquoted macro call:
qar(active) =>ACT
where she wanted to reproduce what she used to do with car:
car([active]) =>ACT, IVE
Worse yet: she wants to use a macro that produces a set of cpp
macros:
define([my_includes], [#include <stdio.h>]) car([my_includes]) =>#include <stdio.h> qar(my_includes) error-->EOF in argument list
This macro, qar, because it double quotes its arguments, forces
its users to leave their macro calls unquoted, which is dangerous.
Commas and other active symbols are interpreted by m4 before
they are given to the macro, often not in the way the users expect.
Also, because qar behaves differently from the other macros,
it's an exception that should be avoided in Autoconf.
changequote is Evil
The temptation is often high to bypass proper quotation, in particular
when it's late at night. Then, many experienced Autoconf hackers
finally surrender to the dark side of the force and use the ultimate
weapon: changequote.
The M4 builtin changequote belongs to a set of primitives that
allow one to adjust the syntax of the language to adjust it to her
needs. For instance, by default M4 uses ``' and `'' as
quotes, but in the context of shell programming (and actually of most
programming languages), it's about the worst choice one can make:
because of strings and back quoted expression in shell (such as
`'this'' and ``that`'), because of literal characters in usual
programming language (as in `'0''), there are many unbalanced
``' and `''. Proper M4 quotation then becomes a nightmare, if
not impossible. In order to make M4 useful in such a context, its
designers have equipped it with changequote, which makes it
possible to chose another pair of quotes. M4sugar, M4sh, Autoconf, and
Autotest all have chosen to use `[' and `]'. Not especially
because they are unlikely characters, but because they are
characters unlikely to be unbalanced.
There are other magic primitives, such as changecom to specify
what syntactic forms are comments (it is common to see
`changecom(<!--, -->)' when M4 is used to produce HTML pages),
changeword and changesyntax to change other syntactic
details (such as the character to denote the n-th argument, `$' by
default, the parenthesis around arguments etc.).
These primitives are really meant to make M4 more useful for specific
domains: they should be considered like command line options:
@option{--quotes}, @option{--comments}, @option{--words}, and
--syntax. Nevertheless, they are implemented as M4 builtins, as
it makes M4 libraries self contained (no need for additional options).
There lies the problem...
The problem is that it is then tempting to use them in the middle of an M4 script, as opposed to its initialization. This, if not carefully thought, can lead to disastrous effects: you are changing the language in the middle of the execution. Changing and restoring the syntax is often not enough: if you happened to invoke macros in between, these macros will be lost, as the current syntax will probably not be the one they were implemented with.
When writing an autoconf macro you may occasionally need to generate special characters that are difficult to express with the standard autoconf quoting rules. For example, you may need to output the regular expression `[^[]', which matches any character other than `['. This expression contains unbalanced brackets so it cannot be put easily into an M4 macro.
You can work around this problem by using one of the following quadrigraphs:
Quadrigraphs are replaced at a late stage of the translation process, after @command{m4} is run, so they do not get in the way of M4 quoting. For example, the string `^@<:@', independently of its quotation, will appear as `^[' in the output.
The empty quadrigraph can be used:
AC_FOO is a comment, while
still being sure that @command{autom4te} will still catch unexpanded
`AC_*'. Then write `AC@&t@_FOO'.
The name `@&t@' was suggested by Paul Eggert:
I should give some credit to the `@&t@' pun. The `&' is my own invention, but the `t' came from the source code of the ALGOL68C compiler, written by Steve Bourne (of Bourne shell fame), and which used `mt' to denote the empty string. In C, it would have looked like something like:
char const mt[] = "";but of course the source code was written in Algol 68.
I don't know where he got `mt' from: it could have been his own invention, and I suppose it could have been a common pun around the Cambridge University computer lab at the time.
To conclude, the quotation rule of thumb is:
One pair of quotes per pair of parentheses.
Never over-quote, never under-quote, in particular in the definition of macros. In the few places where the macros need to use brackets (usually in C program text or regular expressions), properly quote the arguments!
It is common to read Autoconf programs with snippets like:
AC_TRY_LINK( changequote(<<, >>)dnl <<#include <time.h> #ifndef tzname /* For SGI. */ extern char *tzname[]; /* RS6000 and others reject char **tzname. */ #endif>>, changequote([, ])dnl [atoi (*tzname);], ac_cv_var_tzname=yes, ac_cv_var_tzname=no)
which is incredibly useless since AC_TRY_LINK is already
double quoting, so you just need:
AC_TRY_LINK(
[#include <time.h>
#ifndef tzname /* For SGI. */
extern char *tzname[]; /* RS6000 and others reject char **tzname. */
#endif],
[atoi (*tzname);],
[ac_cv_var_tzname=yes],
[ac_cv_var_tzname=no])
The M4-fluent reader will note that these two examples are rigorously
equivalent, since m4 swallows both the `changequote(<<, >>)'
and `<<' `>>' when it collects the arguments: these
quotes are not part of the arguments!
Simplified, the example above is just doing this:
changequote(<<, >>)dnl <<[]>> changequote([, ])dnl
instead of simply:
[[]]
With macros that do not double quote their arguments (which is the rule), double-quote the (risky) literals:
AC_LINK_IFELSE([AC_LANG_PROGRAM(
[[#include <time.h>
#ifndef tzname /* For SGI. */
extern char *tzname[]; /* RS6000 and others reject char **tzname. */
#endif]],
[atoi (*tzname);])],
[ac_cv_var_tzname=yes],
[ac_cv_var_tzname=no])
See See section Quadrigraphs, for what to do if you run into a hopeless case where quoting does not suffice.
When you create a @command{configure} script using newly written macros,
examine it carefully to check whether you need to add more quotes in
your macros. If one or more words have disappeared in the m4
output, you need more quotes. When in doubt, quote.
However, it's also possible to put on too many layers of quotes. If this happens, the resulting @command{configure} script will contain unexpanded macros. The @command{autoconf} program checks for this problem by doing `grep AC_ configure'.
The Autoconf suite, including M4sugar, M4sh, and Autotest in addition to Autoconf per se, heavily rely on M4. All these different uses revealed common needs factored into a layer over @command{m4}: @command{autom4te}(3).
@command{autom4te} should basically considered as a replacement of @command{m4} itself. In particular, its handling of command line arguments is modeled after M4's:
autom4te options files
where the files are directly passed to @command{m4}. In addition to the regular expansion, it handles the replacement of the quadrigraphs (see section Quadrigraphs), and of `__oline__', the current line in the output. It supports an extended syntax for the files:
Of course, it supports the Autoconf common subset of options:
As an extension of @command{m4}, it includes the following options:
AC_DIAGNOSE, for a comprehensive list of categories. Special
values include:
WARNINGS, a comma separated list of categories, is
honored. @command{autom4te -W category} will actually
behave as if you had run:
autom4te --warnings=syntax,$WARNINGS,categoryIf you want to disable @command{autom4te}'s defaults and
WARNINGS, but (for example) enable the warnings about obsolete
constructs, you would use @option{-W none,obsolete}.
@command{autom4te} displays a back trace for errors, but not for
warnings; if you want them, just pass @option{-W error}. For instance,
on this `configure.ac':
AC_DEFUN([INNER], [AC_TRY_RUN([exit (0)])]) AC_DEFUN([OUTER], [INNER]) AC_INIT OUTERyou get:
$ autom4te -l autoconf -Wcross configure.ac:8: warning: AC_TRY_RUN called without default \ to allow cross compiling $ autom4te -l autoconf -Wcross,error configure.ac:8: error: AC_TRY_RUN called without default \ to allow cross compiling acgeneral.m4:3044: AC_TRY_RUN is expanded from... configure.ac:2: INNER is expanded from... configure.ac:5: OUTER is expanded from... configure.ac:8: the top level
file.m4f will be
replaced with file.m4. This helps tracing the macros which
are executed only when the files are frozen, typically
m4_define. For instance, running:
autom4te --melt 1.m4 2.m4f 3.m4 4.m4f input.m4is roughly equivalent to running:
m4 1.m4 2.m4 3.m4 4.m4 input.m4while
autom4te 1.m4 2.m4f 3.m4 4.m4f input.m4is equivalent to:
m4 --reload-state=4.m4f input.m4
autom4te 1.m4 2.m4 3.m4 --freeze --output=3.m4fcorresponds to
m4 1.m4 2.m4 3.m4 --freeze-state=3.m4f
As another additional feature over @command{m4}, @command{autom4te} caches its results. GNU M4 is able to produce a regular output and traces at the same time. Traces are heavily used in the GNU Build System: @command{autoheader} uses them to build `config.h.in', @command{autoreconf} to determine what GNU Build System components are used, @command{automake} to "parse" `configure.ac' etc. To save the long runs of @command{m4}, traces are cached while performing regular expansion, and conversely. This cache is (actually, the caches are) stored in the directory `autom4te.cache'. It can safely be removed at any moment (especially if for some reason @command{autom4te} considers it is trashed).
Because traces are so important to the GNU Build System, @command{autom4te} provides high level tracing features as compared to M4, and helps exploiting the cache:
automake, autoreconf etc. will
trace, so that running @command{m4} is not needed to trace them: the
cache suffices. This results in a huge speed-up.
Finally, @command{autom4te} introduces the concept of @dfn{Autom4te libraries}. They consists in a powerful yet extremely simple feature: sets of combined command line arguments:
M4sugar
M4sh
Autotest
Autoconf
As an example, if Autoconf is installed in its default location, `/usr/local', running `autom4te -l m4sugar foo.m4' is strictly equivalent to running `autom4te --include /usr/local/share/autoconf m4sugar/m4sugar.m4f --warning syntax foo.m4'. Recursive expansion applies: running `autom4te -l m4sh foo.m4', is the same as `autom4te --language M4sugar m4sugar/m4sh.m4f foo.m4', i.e., `autom4te --include /usr/local/share/autoconf m4sugar/m4sugar.m4f m4sugar/m4sh.m4f --mode 777 foo.m4'. The definition of the languages is stored in `autom4te.cfg'.
M4 by itself provides only a small, but sufficient, set of all-purpose macros. M4sugar introduces additional generic macros. Its name was coined by Lars J. Aas: "Readability And Greater Understanding Stands 4 M4sugar".
With a few exceptions, all the M4 native macros are moved in the
`m4_' pseudo-namespace, e.g., M4sugar renames define as
m4_define etc.
Some M4 macros are redefined, and are slightly incompatible with their native equivalent.
m4_undefine.
ifelse.
m4_ifdef([macro], [m4_undefine([macro])])
to recover the behavior of the builtin.
patsubst. The name m4_patsubst
is kept for future versions of M4sh, on top of GNU M4 which will
provide extended regular expression syntax via epatsubst.
m4_undefine.
regexp. The name m4_regexp
is kept for future versions of M4sh, on top of GNU M4 which will
provide extended regular expression syntax via eregexp.
m4wrap.
You are encouraged to end text with `[]', so that there are
no risks that two consecutive invocations of m4_wrap result in an
unexpected pasting of tokens, as in
m4_define([foo], [Foo]) m4_define([bar], [Bar]) m4_define([foobar], [FOOBAR]) m4_wrap([bar]) m4_wrap([foo]) =>FOOBAR
The following macros give some control over the order of the evaluation by adding or removing levels of quotes. They are meant for hard core M4 programmers.
The following example aims at emphasing the difference between (i), not
using these macros, (ii), using m4_quote, and (iii), using
m4_dquote.
$ cat example.m4 # Over quote, so that quotes are visible. m4_define([show], [$[]1 = [$1], $[]@ = [$@]]) m4_divert(0)dnl show(a, b) show(m4_quote(a, b)) show(m4_dquote(a, b)) $ autom4te -l m4sugar example.m4 $1 = a, $@ = [a],[b] $1 = a,b, $@ = [a,b] $1 = [a],[b], $@ = [[a],[b]]
M4sugar provides a means to define suspicious patterns, patterns describing tokens which should not be found in the output. For instance, if an Autoconf `configure' script includes tokens such as `AC_DEFINE', or `dnl', then most probably something went wrong (typically a macro was not evaluated because of over quotation).
M4sugar forbids all the tokens matching `^m4_' and `^dnl$'.
Of course, you might encounter exceptions to these generic rules, for instance you might have to refer to `$m4_flags'.
m4_pattern_forbid pattern.
M4sh provides portable alternatives for some common shell constructs that unfortunately are not portable in practice.
When you write a feature test that could be applicable to more than one software package, the best thing to do is encapsulate it in a new macro. Here are some instructions and guidelines for writing Autoconf macros.
Autoconf macros are defined using the AC_DEFUN macro, which is
similar to the M4 builtin m4_define macro. In addition to
defining a macro, AC_DEFUN adds to it some code that is used to
constrain the order in which macros are called (see section Prerequisite Macros).
An Autoconf macro definition looks like this:
AC_DEFUN(macro-name, macro-body)
You can refer to any arguments passed to the macro as `$1', `$2', etc. See section `How to define new macros' in GNU m4, for more complete information on writing M4 macros.
Be sure to properly quote both the macro-body and the macro-name to avoid any problems if the macro happens to have been previously defined.
Each macro should have a header comment that gives its prototype, and a brief description. When arguments have default values, display them in the prototype. For example:
# AC_MSG_ERROR(ERROR, [EXIT-STATUS = 1])
# --------------------------------------
m4_define([AC_MSG_ERROR],
[{ _AC_ECHO([configure: error: $1], 2); exit m4_default([$2], 1); }])
Comments about the macro should be left in the header comment. Most other comments will make their way into `configure', so just keep using `#' to introduce comments.
If you have some very special comments about pure M4 code, comments
that make no sense in `configure' and in the header comment, then
use the builtin dnl: it causes m4 to discard the text
through the next newline.
Keep in mind that dnl is rarely needed to introduce comments;
dnl is more useful to get rid of the newlines following macros
that produce no output, such as AC_REQUIRE.
All of the Autoconf macros have all-uppercase names starting with `AC_' to prevent them from accidentally conflicting with other text. All shell variables that they use for internal purposes have mostly-lowercase names starting with `ac_'. To ensure that your macros don't conflict with present or future Autoconf macros, you should prefix your own macro names and any shell variables they use with some other sequence. Possibilities include your initials, or an abbreviation for the name of your organization or software package.
Most of the Autoconf macros' names follow a structured naming convention that indicates the kind of feature check by the name. The macro names consist of several words, separated by underscores, going from most general to most specific. The names of their cache variables use the same convention (see section Cache Variable Names, for more information on them).
The first word of the name after `AC_' usually tells the category of feature being tested. Here are the categories used in Autoconf for specific test macros, the kind of macro that you are more likely to write. They are also used for cache variables, in all-lowercase. Use them where applicable; where they're not, invent your own categories.
C
DECL
FUNC
GROUP
HEADER
LIB
PATH
PROG
MEMBER
SYS
TYPE
VAR
After the category comes the name of the particular feature being
tested. Any further words in the macro name indicate particular aspects
of the feature. For example, AC_FUNC_UTIME_NULL checks the
behavior of the utime function when called with a NULL
pointer.
An internal macro should have a name that starts with an underscore;
Autoconf internals should therefore start with `_AC_'.
Additionally, a macro that is an internal subroutine of another macro
should have a name that starts with an underscore and the name of that
other macro, followed by one or more words saying what the internal
macro does. For example, AC_PATH_X has internal macros
_AC_PATH_X_XMKMF and _AC_PATH_X_DIRECT.
When macros statically diagnose abnormal situations, benign or fatal, they should report them using these macros. For dynamic issues, i.e., when @command{configure} is run, see section Printing Messages.
When the user runs `autoconf -W error', warnings from
AC_DIAGNOSE and AC_WARNING are reported as error, see
section Using @command{autoconf} to Create @command{configure}.
Some Autoconf macros depend on other macros having been called first in order to work correctly. Autoconf provides a way to ensure that certain macros are called if needed and a way to warn the user if macros are called in an order that might cause incorrect operation.
A macro that you write might need to use values that have previously
been computed by other macros. For example, AC_DECL_YYTEXT
examines the output of flex or lex, so it depends on
AC_PROG_LEX having been called first to set the shell variable
LEX.
Rather than forcing the user of the macros to keep track of the
dependencies between them, you can use the AC_REQUIRE macro to do
it automatically. AC_REQUIRE can ensure that a macro is only
called if it is needed, and only called once.
AC_DEFUN or else contain a call to AC_PROVIDE to indicate
that it has been called.
AC_REQUIRE must be used inside an AC_DEFUN'd macro; it
must not be called from the top level.
AC_REQUIRE is often misunderstood. It really implements
dependencies between macros in the sense that if one macro depends upon
another, the latter will be expanded before the body of the
former. In particular, `AC_REQUIRE(FOO)' is not replaced with the
body of FOO. For instance, this definition of macros:
AC_DEFUN([TRAVOLTA], [test "$body_temparature_in_celsius" -gt "38" && dance_floor=occupied]) AC_DEFUN([NEWTON_JOHN], [test "$hair_style" = "curly" && dance_floor=occupied]) AC_DEFUN([RESERVE_DANCE_FLOOR], [if date | grep '^Sat.*pm' >/dev/null 2>&1; then AC_REQUIRE([TRAVOLTA]) AC_REQUIRE([NEWTON_JOHN]) fi])
with this `configure.ac'
AC_INIT RESERVE_DANCE_FLOOR if test "$dance_floor" = occupied; then AC_MSG_ERROR([cannot pick up here, let's move]) fi
will not leave you with a better chance to meet a kindred soul at other times than Saturday night since it expands into:
test "$body_temperature_in_Celsius" -gt "38" && dance_floor=occupied test "$hair_style" = "curly" && dance_floor=occupied fi if date | grep '^Sat.*pm' >/dev/null 2>&1; then fi
This behavior was chosen on purpose: (i) it prevents messages in required macros from interrupting the messages in the requiring macros; (ii) it avoids bad surprises when shell conditionals are used, as in:
if ...; then AC_REQUIRE([SOME_CHECK]) fi ... SOME_CHECK
You are encouraged to put all AC_REQUIREs at the beginning of a
macro. You can use dnl to avoid the empty lines they leave.
Some macros should be run before another macro if both are called, but neither requires that the other be called. For example, a macro that changes the behavior of the C compiler should be called before any macros that run the C compiler. Many of these dependencies are noted in the documentation.
Autoconf provides the AC_BEFORE macro to warn users when macros
with this kind of dependency appear out of order in a
`configure.ac' file. The warning occurs when creating
@command{configure} from `configure.ac', not when running
@command{configure}.
For example, AC_PROG_CPP checks whether the C compiler
can run the C preprocessor when given the @option{-E} option. It should
therefore be called after any macros that change which C compiler is
being used, such as AC_PROG_CC. So AC_PROG_CC contains:
AC_BEFORE([$0], [AC_PROG_CPP])dnl
This warns the user if a call to AC_PROG_CPP has already occurred
when AC_PROG_CC is called.
m4 print a warning message to the standard error output if
called-macro-name has already been called. this-macro-name
should be the name of the macro that is calling AC_BEFORE. The
macro called-macro-name must have been defined using
AC_DEFUN or else contain a call to AC_PROVIDE to indicate
that it has been called.
Configuration and portability technology has evolved over the years. Often better ways of solving a particular problem are developed, or ad-hoc approaches are systematized. This process has occurred in many parts of Autoconf. One result is that some of the macros are now considered obsolete; they still work, but are no longer considered the best thing to do, hence they should be replaced with more modern macros. Ideally, @command{autoupdate} should substitute the old macro calls with their modern implementation.
Autoconf provides a simple means to obsolete a macro.
AC_DEFUN is that the user will be warned that
old-macro is now obsolete.
If she then uses @command{autoupdate}, the call to old-macro will be replaced by the modern implementation. The additional message is then printed.
The Autoconf macros follow a strict coding style. You are encouraged to follow this style, especially if you intend to distribute your macro, either by contributing it to Autoconf itself, or via other means.
The first requirement is to pay great attention to the quotation, for more details, see section The Autoconf Language, and section M4 Quotation.
Do not try to invent new interfaces. It is likely that there is a macro in Autoconf that resembles the macro you are defining: try to stick to this existing interface (order of arguments, default values, etc.). We are conscious that some of these interfaces are not perfect; nevertheless, when harmless, homogeneity should be preferred over creativity.
Be careful about clashes both between M4 symbols and between shell variables.
If you stick to the suggested M4 naming scheme (see section Macro Names),
you are unlikely to generate conflicts. Nevertheless, when you need to
set a special value, avoid using a regular macro name; rather,
use an "impossible" name. For instance, up to version 2.13, the macro
AC_SUBST used to remember what symbols were already defined
by setting AC_SUBST_symbol, which is a regular macro name.
But since there is a macro named AC_SUBST_FILE, it was just
impossible to `AC_SUBST(FILE)'! In this case,
AC_SUBST(symbol) or _AC_SUBST(symbol) should
have been used (yes, with the parentheses)...or better yet, high-level
macros such as AC_EXPAND_ONCE.
No Autoconf macro should ever enter the user-variable name space; i.e.,
except for the variables that are the actual result of running the
macro, all shell variables should start with ac_. In
addition, small macros or any macro that is likely to be embedded in
other macros should be careful not to use obvious names.
Do not use dnl to introduce comments: most of the comments you
are likely to write are either header comments which are not output
anyway, or comments that should make their way into `configure'.
There are exceptional cases where you do want to comment special M4
constructs, in which case dnl is right, but keep in mind that it
is unlikely.
M4 ignores the leading spaces before each argument, use this feature to indent in such a way that arguments are (more or less) aligned with the opening parenthesis of the macro being called. For instance, instead of
AC_CACHE_CHECK(for EMX OS/2 environment, ac_cv_emxos2, [AC_COMPILE_IFELSE([AC_LANG_PROGRAM(, [return __EMX__;])], [ac_cv_emxos2=yes], [ac_cv_emxos2=no])])
write
AC_CACHE_CHECK([for EMX OS/2 environment], [ac_cv_emxos2],
[AC_COMPILE_IFELSE([AC_LANG_PROGRAM([], [return __EMX__;])],
[ac_cv_emxos2=yes],
[ac_cv_emxos2=no])])
or even
AC_CACHE_CHECK([for EMX OS/2 environment],
[ac_cv_emxos2],
[AC_COMPILE_IFELSE([AC_LANG_PROGRAM([],
[return __EMX__;])],
[ac_cv_emxos2=yes],
[ac_cv_emxos2=no])])
When using AC_TRY_RUN or any macro that cannot work when
cross-compiling, provide a pessimistic value (typically `no').
Feel free to use various tricks to prevent auxiliary tools, such as syntax-highlighting editors, from behaving improperly. For instance, instead of:
m4_bpatsubst([$1], [$"])
use
m4_bpatsubst([$1], [$""])
so that Emacsen do not open a endless "string" at the first quote. For the same reasons, avoid:
test $[#] != 0
and use:
test $[@%:@] != 0
Otherwise, the closing bracket would be hidden inside a `#'-comment,
breaking the bracket-matching highlighting from Emacsen. Note the
preferred style to escape from M4: `$[1]', `$[@]', etc. Do
not escape when it is unneeded. Common examples of useless quotation
are `[$]$1' (write `$$1'), `[$]var' (use `$var'),
etc. If you add portability issues to the picture, you'll prefer
`${1+"$[@]"}' to `"[$]@"', and you'll prefer do something
better than hacking Autoconf :-).
When using @command{sed}, don't use @option{-e} except for indenting
purpose. With the s command, the preferred separator is `/'
unless `/' itself is used in the command, in which case you should
use `,'.
See section Macro Definitions, for details on how to define a macro. If a
macro doesn't use AC_REQUIRE and it is expected to never be the
object of an AC_REQUIRE directive, then use define. In
case of doubt, use AC_DEFUN. All the AC_REQUIRE
statements should be at the beginning of the macro, dnl'ed.
You should not rely on the number of arguments: instead of checking whether an argument is missing, test that it is not empty. It provides both a simpler and a more predictable interface to the user, and saves room for further arguments.
Unless the macro is short, try to leave the closing `])' at the beginning of a line, followed by a comment that repeats the name of the macro being defined. This introduces an additional newline in @command{configure}; normally, that is not a problem, but if you want to remove it you can use `[]dnl' on the last line. You can similarly use `[]dnl' after a macro call to remove its newline. `[]dnl' is recommended instead of `dnl' to ensure that M4 does not interpret the `dnl' as being attached to the preceding text or macro output. For example, instead of:
AC_DEFUN([AC_PATH_X], [AC_MSG_CHECKING([for X]) AC_REQUIRE_CPP() # ...omitted... AC_MSG_RESULT([libraries $x_libraries, headers $x_includes]) fi])
you would write:
AC_DEFUN([AC_PATH_X], [AC_REQUIRE_CPP()[]dnl AC_MSG_CHECKING([for X]) # ...omitted... AC_MSG_RESULT([libraries $x_libraries, headers $x_includes]) fi[]dnl ])# AC_PATH_X
If the macro is long, try to split it into logical chunks. Typically,
macros that check for a bug in a function and prepare its
AC_LIBOBJ replacement should have an auxiliary macro to perform
this setup. Do not hesitate to introduce auxiliary macros to factor
your code.
In order to highlight the recommended coding style, here is a macro written the old way:
dnl Check for EMX on OS/2. dnl _AC_EMXOS2 AC_DEFUN(_AC_EMXOS2, [AC_CACHE_CHECK(for EMX OS/2 environment, ac_cv_emxos2, [AC_COMPILE_IFELSE([AC_LANG_PROGRAM(, return __EMX__;)], ac_cv_emxos2=yes, ac_cv_emxos2=no)]) test "$ac_cv_emxos2" = yes && EMXOS2=yes])
and the new way:
# _AC_EMXOS2
# ----------
# Check for EMX on OS/2.
define([_AC_EMXOS2],
[AC_CACHE_CHECK([for EMX OS/2 environment], [ac_cv_emxos2],
[AC_COMPILE_IFELSE([AC_LANG_PROGRAM([], [return __EMX__;])],
[ac_cv_emxos2=yes],
[ac_cv_emxos2=no])])
test "$ac_cv_emxos2" = yes && EMXOS2=yes[]dnl
])# _AC_EMXOS2
When writing your own checks, there are some shell-script programming
techniques you should avoid in order to make your code portable. The
Bourne shell and upward-compatible shells like the Korn shell and Bash
have evolved over the years, but to prevent trouble, do not take
advantage of features that were added after UNIX version 7, circa
1977. You should not use shell functions, aliases, negated character
classes, or other features that are not found in all Bourne-compatible
shells; restrict yourself to the lowest common denominator. Even
unset is not supported by all shells! Also, include a space
after the exclamation point in interpreter specifications, like this:
#! /usr/bin/perl
If you omit the space before the path, then 4.2BSD based systems (such as Sequent DYNIX) will ignore the line, because they interpret `#! /' as a 4-byte magic number. Some old systems have quite small limits on the length of the `#!' line too, for instance 32 bytes (not including the newline) on SunOS 4.
The set of external programs you should run in a @command{configure} script is fairly small. See section `Utilities in Makefiles' in GNU Coding Standards, for the list. This restriction allows users to start out with a fairly small set of programs and build the rest, avoiding too many interdependencies between packages.
Some of these external utilities have a portable subset of features; see section Limitations of Usual Tools.
There are several families of shells, most prominently the Bourne family and the C shell family which are deeply incompatible. If you want to write portable shell scripts, avoid members of the C shell family.
Below we describe some of the members of the Bourne shell family.
foo= false $foo echo "Don't use it: $?"
cat ${FOO=`bar`}
BASH_VERSION is set. To disable its extensions and require
POSIX compatibility, run `set -o posix'. See section `Bash POSIX Mode' in The GNU Bash Reference Manual, for
details.
ZSH_VERSION is set. By default @command{zsh} is not
compatible with the Bourne shell: you have to run `emulate sh' and
set NULLCMD to `:'. See section `Compatibility' in The Z Shell Manual, for details.
Zsh 3.0.8 is the native @command{/bin/sh} on Mac OS X 10.0.3.
The following discussion between Russ Allbery and Robert Lipe is worth reading:
Russ Allbery:
The GNU assumption that @command{/bin/sh} is the one and only shell leads to a permanent deadlock. Vendors don't want to break user's existant shell scripts, and there are some corner cases in the Bourne shell that are not completely compatible with a POSIX shell. Thus, vendors who have taken this route will never (OK..."never say never") replace the Bourne shell (as @command{/bin/sh}) with a POSIX shell.
Robert Lipe:
This is exactly the problem. While most (at least most System V's) do have a Bourne shell that accepts shell functions most vendor @command{/bin/sh} programs are not the POSIX shell.
So while most modern systems do have a shell _somewhere_ that meets the POSIX standard, the challenge is to find it.
Don't rely on `\' being preserved just because it has no special meaning together with the next symbol. in the native @command{/bin/sh} on OpenBSD 2.7 `\"' expands to `"' in here-documents with unquoted delimiter. As a general rule, if `\\' expands to `\' use `\\' to get `\'.
With OpenBSD 2.7's @command{/bin/sh}
$ cat <<EOF > \" \\ > EOF " \
and with Bash:
bash-2.04$ cat <<EOF > \" \\ > EOF \" \
Many older shells (including the Bourne shell) implement here-documents inefficiently. Users can generally speed things up by using a faster shell, e.g., by using the command `bash ./configure' rather than plain `./configure'.
Some shells can be extremely inefficient when there are a lot of here-documents inside a single statement. For instance if your `configure.ac' includes something like:
if <cross_compiling>; then assume this and that else check this check that check something else ... on and on forever ... fi
A shell parses the whole if/fi construct, creating
temporary files for each here document in it. Some shells create links
for such here-documents on every fork, so that the clean-up code
they had installed correctly removes them. It is creating the links
that the shell can take forever.
Moving the tests out of the if/fi, or creating multiple
if/fi constructs, would improve the performance
significantly. Anyway, this kind of construct is not exactly the
typical use of Autoconf. In fact, it's even not recommended, because M4
macros can't look into shell conditionals, so we may fail to expand a
macro when it was expanded before in a conditional path, and the
condition turned out to be false at run-time, and we end up not
executing the macro at all.
Some file descriptors shall not be used, since some systems, admittedly arcane, use them for special purpose:
3 --- some systems may open it to `/dev/tty'. 4 --- used on the Kubota Titan.
Don't redirect several times the same file descriptor, as you are doomed to failure under Ultrix.
ULTRIX V4.4 (Rev. 69) System #31: Thu Aug 10 19:42:23 GMT 1995 UWS V4.4 (Rev. 11) $ eval 'echo matter >fullness' >void illegal io $ eval '(echo matter >fullness)' >void illegal io $ (eval '(echo matter >fullness)') >void Ambiguous output redirect.
In each case the expected result is of course `fullness' containing `matter' and `void' being empty.
Don't try to redirect the standard error of a command substitution: it must be done inside the command substitution: when running `: `cd /zorglub` 2>/dev/null' expect the error message to escape, while `: `cd /zorglub 2>/dev/null`' works properly.
It is worth noting that Zsh (but not Ash nor Bash) makes it possible in assignments though: `foo=`cd /zorglub` 2>/dev/null'.
Most shells, if not all (including Bash, Zsh, Ash), output traces on stderr, even for sub-shells. This might result in undesired content if you meant to capture the standard-error output of the inner command:
$ ash -x -c '(eval "echo foo >&2") 2>stderr' $ cat stderr + eval echo foo >&2 + echo foo foo $ bash -x -c '(eval "echo foo >&2") 2>stderr' $ cat stderr + eval 'echo foo >&2' ++ echo foo foo $ zsh -x -c '(eval "echo foo >&2") 2>stderr' # Traces on startup files deleted here. $ cat stderr +zsh:1> eval echo foo >&2 +zsh:1> echo foo foo
You'll appreciate the various levels of detail...
One workaround is to grep out uninteresting lines, hoping not to remove good ones...
Don't try to move/delete open files, such as in `exec >foo; mv foo bar', see See section Limitations of Shell Builtins, @command{mv} for more details.
While @command{autoconf} and friends will usually be run on some Unix variety, it can and will be used on other systems, most notably DOS variants. This impacts several assumptions regarding file and path names.
For example, the following code:
case $foo_dir in
/*) # Absolute
;;
*)
foo_dir=$dots$foo_dir ;;
esac
will fail to properly detect absolute paths on those systems, because they can use a drivespec, and will usually use a backslash as directory separator. The canonical way to check for absolute paths is:
case $foo_dir in
[\\/]* | ?:[\\/]* ) # Absolute
;;
*)
foo_dir=$dots$foo_dir ;;
esac
Make sure you quote the brackets if appropriate and keep the backslash as first character (see section Limitations of Shell Builtins).
Also, because the colon is used as part of a drivespec, these systems don't
use it as path separator. When creating or accessing paths, use the
PATH_SEPARATOR output variable instead. @command{configure} sets this
to the appropriate value (`:' or `;') when it starts up.
File names need extra care as well. While DOS-based environments that are Unixy enough to run @command{autoconf} (such as DJGPP) will usually be able to handle long file names properly, there are still limitations that can seriously break packages. Several of these issues can be easily detected by the @href{ftp://ftp.gnu.org/gnu/non-gnu/doschk/doschk-1.1.tar.gz, doschk} package.
A short overview follows; problems are marked with SFN/LFN to indicate where they apply: SFN means the issues are only relevant to plain DOS, not to DOS boxes under Windows, while LFN identifies problems that exist even under Windows.
AC_CONFIG_HEADER(config.h) AC_CONFIG_FILES([source.c foo.bar]) AC_OUTPUTbut it causes problems on DOS, as it requires `config.h.in', `source.c.in' and `foo.bar.in'. To make your package more portable to DOS-based environments, you should use this instead:
AC_CONFIG_HEADER(config.h:config.hin) AC_CONFIG_FILES([source.c:source.cin foo.bar:foobar.in]) AC_OUTPUT
Contrary to a persistent urban legend, the Bourne shell does not
systematically split variables and backquoted expressions, in particular
on the right-hand side of assignments and in the argument of case.
For instance, the following code:
case "$given_srcdir" in .) top_srcdir="`echo "$dots" | sed 's,/$,,'`" *) top_srcdir="$dots$given_srcdir" ;; esac
is more readable when written as:
case $given_srcdir in .) top_srcdir=`echo "$dots" | sed 's,/$,,'` *) top_srcdir=$dots$given_srcdir ;; esac
and in fact it is even more portable: in the first case of the
first attempt, the computation of top_srcdir is not portable,
since not all shells properly understand "`..."..."...`".
Worse yet, not all shells understand "`...\"...\"...`"
the same way. There is just no portable way to use double-quoted
strings inside double-quoted backquoted expressions (pfew!).
$@
${var:-value}
sh, don't accept the
colon for any shell substitution, and complain and die.
${var=literal}
: ${var='Some words'}
otherwise some shells, such as on Digital Unix V 5.0, will die because
of a "bad substitution".
Solaris' @command{/bin/sh} has a frightening bug in its interpretation
of this. Imagine you need set a variable to a string containing
`}'. This `}' character confuses Solaris' @command{/bin/sh}
when the affected variable was already set. This bug can be exercised
by running:
$ unset foo
$ foo=${foo='}'}
$ echo $foo
}
$ foo=${foo='}' # no error; this hints to what the bug is
$ echo $foo
}
$ foo=${foo='}'}
$ echo $foo
}}
^ ugh!
It seems that `}' is interpreted as matching `${', even
though it is enclosed in single quotes. The problem doesn't happen
using double quotes.
${var=expanded-value}
default="yu,yaa"
: ${var="$default"}
will set var to `M-yM-uM-,M-yM-aM-a', i.e., the 8th bit of
each char will be set. You won't observe the phenomenon using a simple
`echo $var' since apparently the shell resets the 8th bit when it
expands $var. Here are two means to make this shell confess its sins:
$ cat -v <<EOF $var EOFand
$ set | grep '^var=' | cat -vOne classic incarnation of this bug is:
default="a b c"
: ${list="$default"}
for c in $list; do
echo $c
done
You'll get `a b c' on a single line. Why? Because there are no
spaces in `$list': there are `M- ', i.e., spaces with the 8th
bit set, hence no IFS splitting is performed!!!
One piece of good news is that Ultrix works fine with `:
${list=$default}'; i.e., if you don't quote. The bad news is
then that QNX 4.25 then sets list to the last item of
default!
The portable way out consists in using a double assignment, to switch
the 8th bit twice on Ultrix:
list=${list="$default"}
...but beware of the `}' bug from Solaris (see above). For safety,
use:
test "${var+set}" = set || var={value}
`commands`
$ pwd /tmp $ test -n "`cd /`" && pwd /The result of `foo=`exit 1`' is left as an exercise to the reader.
$(commands)
$ showrev -c /bin/sh | grep version
Command version: SunOS 5.8 Generic 109324-02 February 2001
$ echo $(echo blah)
syntax error: `(' unexpected
nor does IRIX 6.5's Bourne shell:
$ uname -a IRIX firebird-image 6.5 07151432 IP22 $ echo $(echo blah) $(echo blah)
When setting several variables in a row, be aware that the order of the evaluation is undefined. For instance `foo=1 foo=2; echo $foo' gives `1' with sh on Solaris, but `2' with Bash. You must use `;' to enforce the order: `foo=1; foo=2; echo $foo'.
Don't rely on the exit status of an assignment: Ash 0.2 does not change the status and propagates that of the last statement:
$ false || foo=bar; echo $? 1 $ false || foo=`:`; echo $? 0
and to make things even worse, QNX 4.25 just sets the exit status to 0 in any case:
$ foo=`exit 1`; echo $? 0
To assign default values, follow this algorithm:
: ${var='my literal'}
: ${var="$default"}
var=${var="$default"}
test "${var+set}" = set || var='${indirection}'
In most cases `var=${var="$default"}' is fine, but in case of doubt, just use the latter. See section Shell Substitutions, items `${var:-value}' and `${var=value}' for the rationale.
Some shell variables should not be used, since they can have a deep influence on the behavior of the shell. In order to recover a sane behavior from the shell, some variables should be unset, but @command{unset} is not portable (see section Limitations of Shell Builtins) and a fallback value is needed. We list these values below.
CDPATH
cd is verbose, so idioms such as
`abs=`cd $rel && pwd`' break because abs receives the path
twice.
Setting CDPATH to the empty value is not enough for most shells.
A simple path separator is enough except for zsh, which prefers a
leading dot:
zsh-3.1.6$ mkdir foo && (CDPATH=: cd foo) /tmp/foo zsh-3.1.6$ (CDPATH=:. cd foo) /tmp/foo zsh-3.1.6$ (CDPATH=.: cd foo) zsh-3.1.6$(of course we could just @command{unset}
CDPATH, since it also
behaves properly if set to the empty string).
Life wouldn't be so much fun if @command{bash} and @command{zsh} had the
same behavior:
bash-2.02$ mkdir foo && (CDPATH=: cd foo) bash-2.02$ (CDPATH=:. cd foo) bash-2.02$ (CDPATH=.: cd foo) /tmp/fooOf course, even better style would be to use
PATH_SEPARATOR instead
of a `:'.
Therefore, a portable solution to neutralize CDPATH is
CDPATH=${ZSH_VERSION+.}$PATH_SEPARATOR
Note that since @command{zsh} supports @command{unset}, you may unset
CDPATH using PATH_SEPARATOR as a fallback, see
section Limitations of Shell Builtins.
IFS
IFS to backslash. Indeed,
Bourne shells use the first character (backslash) when joining the
components in `"$@"' and some shells then re-interpret (!) the
backslash escapes, so you can end up with backspace and other strange
characters.
LANG
LC_ALL
LC_COLLATE
LC_CTYPE
LC_MESSAGES
LC_NUMERIC
LC_TIME
LANGUAGE
LINENO
LINENO.
Its value is the line number of the beginning of the current command.
Autoconf attempts to execute @command{configure} with a modern shell.
If no such shell is available, it attempts to implement LINENO
with a Sed prepass that replaces the each instance of the string
$LINENO (not followed by an alphanumeric character) with the
line's number.
You should not rely on LINENO within @command{eval}, as the
behavior differs in practice. Also, the possibility of the Sed
prepass means that you should not rely on $LINENO when quoted,
when in here-documents, or when in long commands that cross line
boundaries. Subshells should be OK, though. In the following
example, lines 1, 6, and 9 are portable, but the other instances of
LINENO are not:
$ cat lineno echo 1. $LINENO cat <<EOF 3. $LINENO 4. $LINENO EOF ( echo 6. $LINENO ) eval 'echo 7. $LINENO' echo 8. '$LINENO' echo 9. $LINENO ' 10.' $LINENO $ bash-2.05 lineno 1. 1 3. 2 4. 2 6. 6 7. 1 8. $LINENO 9. 9 10. 9 $ zsh-3.0.6 lineno 1. 1 3. 2 4. 2 6. 6 7. 7 8. $LINENO 9. 9 10. 9 $ pdksh-5.2.14 lineno 1. 1 3. 2 4. 2 6. 6 7. 0 8. $LINENO 9. 9 10. 9 $ sed '=' <lineno | > sed ' > N > s,$,-, > : loop > s,^\([0-9]*\)\(.*\)[$]LINENO\([^a-zA-Z0-9_]\),\1\2\1\3, > t loop > s,-$,, > s,^[0-9]*\n,, > ' | > sh 1. 1 3. 3 4. 4 6. 6 7. 7 8. 8 9. 9 10. 10
NULLCMD
NULLCMD is
`:', while @command{zsh}, even in Bourne shell compatibility mode,
sets NULLCMD to `cat'. If you forgot to set NULLCMD,
your script might be suspended waiting for data on its standard input.
status
zsh (at least 3.1.6),
hence read-only. Do not use it.
PATH_SEPARATOR
PATH_SEPARATOR output
variable accordingly.
On DJGPP systems, the PATH_SEPARATOR environment variable can be
set to either `:' or `;' to control the path separator
@command{bash} uses to set up certain environment variables (such as
PATH). Since this only works inside @command{bash}, you want
@command{configure} to detect the regular DOS path separator
(`;'), so it can be safely substituted in files that may not support
`;' as path separator. So it is recommended to either unset this
variable or set it to `;'.
RANDOM
RANDOM, a variable that returns a different
integer when used. Most of the time, its value does not change when it
is not used, but on IRIX 6.5 the value changes all the time. This
can be observed by using @command{set}.
No, no, we are serious: some shells do have limitations! :)
You should always keep in mind that any built-in or command may support
options, and therefore have a very different behavior with arguments
starting with a dash. For instance, the innocent `echo "$word"'
can give unexpected results when word starts with a dash. It is
often possible to avoid this problem using `echo "x$word"', taking
the `x' into account later in the pipe.
fnmatch, @command{bash} fails to properly
handle backslashes in character classes:
bash-2.02$ case /tmp in [/\\]*) echo OK;; esac bash-2.02$This is extremely unfortunate, since you are likely to use this code to handle UNIX or MS-DOS absolute paths. To work around this bug, always put the backslash first:
bash-2.02$ case '\TMP' in [\\/]*) echo OK;; esac OK bash-2.02$ case /tmp in [\\/]*) echo OK;; esac OKSome shells, such as Ash 0.3.8, are confused by empty
case/esac:
ash-0.3.8 $ case foo in esac; error-->Syntax error: ";" unexpected (expecting ")")Many shells still do not support parenthesized cases, which is a pity for those of us using tools that rely on balanced parentheses. For instance, Solaris 2.8's Bourne shell:
$ case foo in (foo) echo foo;; esac
error-->syntax error: `(' unexpected
echo is probably the most surprising source of
portability troubles. It is not possible to use `echo' portably
unless both options and escape sequences are omitted. New applications
which are not aiming at portability should use `printf' instead of
`echo'.
Don't expect any option. See section Preset Output Variables, ECHO_N
etc. for a means to simulate @option{-c}.
Do not use backslashes in the arguments, as there is no consensus on
their handling. On `echo '\n' | wc -l', the @command{sh} of
Digital Unix 4.0, MIPS RISC/OS 4.52, answer 2, but the Solaris'
@command{sh}, Bash and Zsh (in @command{sh} emulation mode) report 1.
Please note that the problem is truly @command{echo}: all the shells
understand `'\n'' as the string composed of a backslash and an
`n'.
Because of these problems, do not pass a string containing arbitrary
characters to @command{echo}. For example, `echo "$foo"' is safe
if you know that foo's value cannot contain backslashes and cannot
start with `-', but otherwise you should use a here-document like
this:
cat <<EOF $foo EOF
$?;
unfortunately, some shells, such as the DJGPP port of Bash 2.04, just
perform `exit 0'.
bash-2.04$ foo=`exit 1` || echo fail fail bash-2.04$ foo=`(exit 1)` || echo fail fail bash-2.04$ foo=`(exit 1); exit` || echo fail bash-2.04$Using `exit $?' restores the expected behavior. Some shell scripts, such as those generated by @command{autoconf}, use a trap to clean up before exiting. If the last shell command exited with nonzero status, the trap also exits with nonzero status so that the invoker can tell that an error occurred. Unfortunately, in some shells, such as Solaris 8 @command{sh}, an exit trap ignores the
exit command's status. In these shells, a trap
cannot determine whether it was invoked by plain exit or by
exit 1. Instead of calling exit directly, use the
AC_MSG_ERROR macro that has a workaround for this problem.
#! /bin/sh echo $FOO FOO=bar echo $FOO exec /bin/sh $0when run with `FOO=foo' in the environment, these shells will print alternately `foo' and `bar', although it should only print `foo' and then a sequence of `bar's. Therefore you should @command{export} again each environment variable that you update.
for arg do echo "$arg" doneYou may not leave the
do on the same line as for,
since some shells improperly grok:
for arg; do echo "$arg" doneIf you want to explicitly refer to the positional arguments, given the `$@' bug (see section Shell Substitutions), use:
for arg in ${1+"$@"}; do
echo "$arg"
done
if ! cmp -s file file.new; then mv file.new file fiuse:
if cmp -s file file.new; then :; else mv file.new file fiThere are shells that do not reset the exit status from an @command{if}:
$ if (exit 42); then true; fi; echo $? 42whereas a proper shell should have printed `0'. This is especially bad in Makefiles since it produces false failures. This is why properly written Makefiles, such as Automake's, have such hairy constructs:
if test -f "$file"; then install "$file" "$dest" else : fi
set x $my_list; shift
test program is the way to perform many file and string
tests. It is often invoked by the alternate name `[', but using
that name in Autoconf code is asking for trouble since it is an M4 quote
character.
If you need to make multiple checks using test, combine them with
the shell operators `&&' and `||' instead of using the
test operators @option{-a} and @option{-o}. On System V, the
precedence of @option{-a} and @option{-o} is wrong relative to the unary
operators; consequently, POSIX does not specify them, so using them
is nonportable. If you combine `&&' and `||' in the same
statement, keep in mind that they have equal precedence.
You may use `!' with @command{test}, but not with @command{if}:
`test ! -r foo || exit 1'.
test might interpret its argument as an
option (e.g., `string = "-n"').
Contrary to a common belief, `test -n string' and `test
-z string' are portable, nevertheless many shells (such
as Solaris 2.5, AIX 3.2, UNICOS 10.0.0.6, Digital Unix 4 etc.) have
bizarre precedence and may be confused if string looks like an
operator:
$ test -n = test: argument expectedIf there are risks, use `test "xstring" = x' or `test "xstring" != x' instead. It is frequent to find variations of the following idiom:
test -n "`echo $ac_feature | sed 's/[-a-zA-Z0-9_]//g'`" && actionto take an action when a token matches a given pattern. Such constructs should always be avoided by using:
echo "$ac_feature" | grep '[^-a-zA-Z0-9_]' >/dev/null 2>&1 && actionUse
case where possible since it is faster, being a shell builtin:
case $ac_feature in *[!-a-zA-Z0-9_]*) action;; esacAlas, negated character classes are probably not portable, although no shell is known to not support the POSIX.2 syntax `[!...]' (when in interactive mode, @command{zsh} is confused by the `[!...]' syntax and looks for an event in its history because of `!'). Many shells do not support the alternative syntax `[^...]' (Solaris, Digital Unix, etc.). One solution can be:
expr "$ac_feature" : '.*[^-a-zA-Z0-9_]' >/dev/null && actionor better yet
expr "x$ac_feature" : '.*[^-a-zA-Z0-9_]' >/dev/null && action`expr "Xfoo" : "Xbar"' is more robust than `echo "Xfoo" | grep "^Xbar"', because it avoids problems when `foo' contains backslashes.
$ cat trap.sh trap 'echo $?' 0 (exit 42); exit 0 $ zsh trap.sh 42 $ bash trap.sh 0The portable solution is then simple: when you want to `exit 42', run `(exit 42); exit 42', the first @command{exit} being used to set the exit status to 42 for Zsh, and the second to trigger the trap and pass 42 as exit status for Bash. The shell in FreeBSD 4.0 has the following bug: `$?' is reset to 0 by empty lines if the code is inside @command{trap}.
$ trap 'false echo $?' 0 $ exit 0Fortunately, this bug only affects @command{trap}.
In a sense, yes, because if it doesn't exist, the shell will produce an exit status of failure, which is correct for @command{false}, but not for @command{true}.
CDPATH, you can test for its existence and use
it provided you give a neutralizing value when @command{unset} is
not supported:
if (unset FOO) >/dev/null 2>&1; then unset=unset else unset=false fi $unset CDPATH || CDPATH=:See section Special Shell Variables, for some neutralizing values. Also, see section Limitations of Shell Builtins, documentation of @command{export}, for the case of environment variables.
The small set of tools you can expect to find on any machine can still include some limitations you should be aware of.
$ gawk 'function die () { print "Aaaaarg!" }
BEGIN { die () }'
gawk: cmd. line:2: BEGIN { die () }
gawk: cmd. line:2: ^ parse error
$ gawk 'function die () { print "Aaaaarg!" }
BEGIN { die() }'
Aaaaarg!
If you want your program to be deterministic, don't depend on for
on arrays:
$ cat for.awk
END {
arr["foo"] = 1
arr["bar"] = 1
for (i in arr)
print i
}
$ gawk -f for.awk </dev/null
foo
bar
$ nawk -f for.awk </dev/null
bar
foo
Some AWK, such as HPUX 11.0's native one, have regex engines fragile to
inner anchors:
$ echo xfoo | $AWK '/foo|^bar/ { print }'
$ echo bar | $AWK '/foo|^bar/ { print }'
bar
$ echo xfoo | $AWK '/^bar|foo/ { print }'
xfoo
$ echo bar | $AWK '/^bar|foo/ { print }'
bar
Either do not depend on such patterns (i.e., use `/^(.*foo|bar)/',
or use a simple test to reject such AWK.
link (or, in
newer systems, rename).
$ uname -a OSF1 medusa.sis.pasteur.fr V5.1 732 alpha $ date "+%s" %s
AS_DIRNAME (see section Programming in M4sh). For example:
dir=`dirname "$file"` # This is not portable. dir=`AS_DIRNAME(["$file"])` # This is more portable.This handles a few subtleties in the standard way required by POSIX. For example, under UN*X, should `dirname //1' give `/'? Paul Eggert answers:
No, under some older flavors of Unix, leading `//' is a special path name: it refers to a "super-root" and is used to access other machines' files. Leading `///', `////', etc. are equivalent to `/'; but leading `//' is special. I think this tradition started with Apollo Domain/OS, an OS that is still in use on some older hosts.
POSIX allows but does not require the special treatment for `//'. It says that the behavior of dirname on path names of the form `//([^/]+/*)?' is implementation defined. In these cases, GNU @command{dirname} returns `/', but it's more portable to return `//' as this works even on those older flavors of Unix.
> printf "foo\n|foo\n" | egrep '^(|foo|bar)$' |foo > printf "bar\nbar|\n" | egrep '^(foo|bar|)$' bar| > printf "foo\nfoo|\n|bar\nbar\n" | egrep '^(foo||bar)$' foo |bar@command{egrep} also suffers the limitations of @command{grep}.
length, substr, match and index.
expr '' \| ''GNU/Linux and POSIX.2-1992 return the empty string for this case, but traditional Unix returns `0' (Solaris is one such example). In the latest POSIX draft, the specification has been changed to match traditional Unix's behavior (which is bizarre, but it's too late to fix this). Please note that the same problem does arise when the empty string results from a computation, as in:
expr bar : foo \| foo : barAvoid this portability problem by avoiding the empty string.
expr a : b \| ''unfortunately this behaves exactly as the original expression, see the `@command{expr' (`:')} entry for more information. Older @command{expr} implementations (e.g. SunOS 4 @command{expr} and Solaris 8 @command{/usr/ucb/expr}) have a silly length limit that causes @command{expr} to fail if the matched substring is longer than 120 bytes. In this case, you might want to fall back on `echo|sed' if @command{expr} fails. Don't leave, there is some more! The QNX 4.25 @command{expr}, in addition of preferring `0' to the empty string, has a funny behavior in its exit status: it's always 1 when parentheses are used!
$ val=`expr 'a' : 'a'`; echo "$?: $val" 0: 1 $ val=`expr 'a' : 'b'`; echo "$?: $val" 1: 0 $ val=`expr 'a' : '\(a\)'`; echo "?: $val" 1: a $ val=`expr 'a' : '\(b\)'`; echo "?: $val" 1: 0In practice this can be a big problem if you are ready to catch failures of @command{expr} programs with some other method (such as using @command{sed}), since you may get twice the result. For instance
$ expr 'a' : '\(a\)' || echo 'a' | sed 's/^\(a\)$/\1/'will output `a' on most hosts, but `aa' on QNX 4.25. A simple work around consists in testing @command{expr} and use a variable set to @command{expr} or to @command{false} according to the result.
$ touch foo
$ find . -name foo -exec echo "{}-{}" \;
{}-{}
while GNU @command{find} reports `./foo-./foo'.
grep to `/dev/null'. Check the exit
status of grep to determine whether it found a match.
Don't use multiple regexps with @option{-e}, as some grep will only
honor the last pattern (eg., IRIX 6.5 and Solaris 2.5.1). Anyway,
Stardent Vistra SVR4 grep lacks @option{-e}... Instead, use
alternation and egrep.
exec > foo mv foo barnor can
exec > foo rm -f foo
$ echo a | sed 's/x/x/;;s/x/x/' sed: 1: "s/x/x/;;s/x/x/": invalid command code ;Input should have reasonably long lines, since some @command{sed} have an input buffer limited to 4000 bytes. Alternation, `\|', is common but POSIX.2 does not require its support, so it should be avoided in portable scripts. Solaris 8 @command{sed} does not support alternation; e.g. `sed '/a\|b/d'' deletes only lines that contain the literal string `a|b'. Anchors (`^' and `$') inside groups are not portable. Nested parenthesization in patterns (e.g., `\(\(a*\)b*)\)') is quite portable to modern hosts, but is not supported by some older @command{sed} implementations like SVR3. Of course the option @option{-e} is portable, but it is not needed. No valid Sed program can start with a dash, so it does not help disambiguating. Its sole usefulness is helping enforcing indenting as in:
sed -e instruction-1 \
-e instruction-2
as opposed to
sed instruction-1;instruction-2Contrary to yet another urban legend, you may portably use `&' in the replacement part of the
s command to mean "what was
matched". All descendents of Bell Lab's V7 @command{sed} (at least; we
don't have first hand experience with older @command{sed}s) have
supported it.
s/keep me/kept/g # a t end # b s/.*/deleted/g # c : end # don
delete me # 1 delete me # 2 keep me # 3 delete me # 4you get
deleted delete me kept deletedinstead of
deleted deleted kept deletedWhy? When processing 1, a matches, therefore sets the t flag, b jumps to d, and the output is produced. When processing line 2, the t flag is still set (this is the bug). Line a fails to match, but @command{sed} is not supposed to clear the t flag when a substitution fails. Line b sees that the flag is set, therefore it clears it, and jumps to d, hence you get `delete me' instead of `deleted'. When processing 3 t is clear, a matches, so the flag is set, hence b clears the flags and jumps. Finally, since the flag is clear, 4 is processed properly. There are two things one should remind about `t' in @command{sed}. Firstly, always remember that `t' jumps if some substitution succeeded, not only the immediately preceding substitution, therefore, always use a fake `t clear; : clear' to reset the t flag where indeed. Secondly, you cannot rely on @command{sed} to clear the flag at each new cycle. One portable implementation of the script above is:
t clear : clear s/keep me/kept/g t end s/.*/deleted/g : end
echo as a workaround.
GNU @command{touch} 3.16r (and presumably all before that) fails to work
on SunOS 4.1.3 when the empty file is on an NFS-mounted 4.2 volume.
Make itself suffers a great number of limitations, only a few of which being listed here. First of all, remember that since commands are executed by the shell, all its weaknesses are inherited...
$<
$ cat Makefile _am_include = # _am_quote = all:; @echo this is test $ make Make: Must be a separator on rules line 2. Stop. $ cat Makefile2 am_include = # am_quote = all:; @echo this is test $ make -f Makefile2 this is test
VPATH
VPATH causes Sun
@command{make} to only execute the first set of double-colon rules.
A few kinds of features can't be guessed automatically by running test
programs. For example, the details of the object-file format, or
special options that need to be passed to the compiler or linker. You
can check for such features using ad-hoc means, such as having
@command{configure} check the output of the uname program, or
looking for libraries that are unique to particular systems. However,
Autoconf provides a uniform method for handling unguessable features.
Like other GNU @command{configure} scripts, Autoconf-generated @command{configure} scripts can make decisions based on a canonical name for the system type, which has the form: `cpu-vendor-os', where os can be `system' or `kernel-system'
@command{configure} can usually guess the canonical name for the type of
system it's running on. To do so it runs a script called
@command{config.guess}, which infers the name using the uname
command or symbols predefined by the C preprocessor.
Alternately, the user can specify the system type with command line arguments to @command{configure}. Doing so is necessary when cross-compiling. In the most complex case of cross-compiling, three system types are involved. The options to specify them are:
If you mean to override the result of @command{config.guess}, use
@option{--build}, not @option{--host}, since the latter enables
cross-compilation. For historical reasons, passing @option{--host} also
changes the build type. Therefore, whenever you specify --host,
be sure to specify --build too. This will be fixed in the
future.
./configure --build=i686-pc-linux-gnu --host=m68k-coff
will enter cross-compilation mode, but @command{configure} will fail if it can't run the code generated by the specified compiler if you configure as follows:
./configure CC=m68k-coff-gcc
@command{configure} recognizes short aliases for many system types; for example, `decstation' can be used instead of `mips-dec-ultrix4.2'. @command{configure} runs a script called @command{config.sub} to canonicalize system type aliases.
This section deliberately omits the description of the obsolete interface, see section Hosts and Cross-Compilation.
The following macros make the system type available to @command{configure} scripts.
The variables `build_alias', `host_alias', and
`target_alias' are always exactly the arguments of @option{--build},
@option{--host}, and @option{--target}; in particular, they are left empty
if the user did not use them, even if the corresponding
AC_CANONICAL macro was run. Any configure script may use these
variables anywhere. These are the variables that should be used when in
interaction with the user.
If you need to recognize some special environments based on their system type, run the following macros to get canonical system names. These variables are not set before the macro call.
If you use these macros, you must distribute @command{config.guess} and
@command{config.sub} along with your source code. See section Outputting Files, for
information about the AC_CONFIG_AUX_DIR macro which you can use
to control in which directory @command{configure} looks for those scripts.
build, and its
three individual parts build_cpu, build_vendor, and
build_os.
If @option{--build} was specified, then build is the
canonicalization of build_alias by @command{config.sub},
otherwise it is determined by the shell script @command{config.guess}.
host, and its
three individual parts host_cpu, host_vendor, and
host_os.
If @option{--host} was specified, then host is the
canonicalization of host_alias by @command{config.sub},
otherwise it defaults to build.
target, and its
three individual parts target_cpu, target_vendor, and
target_os.
If @option{--target} was specified, then target is the
canonicalization of target_alias by @command{config.sub},
otherwise it defaults to host.
Note that there can be artifacts due to the backward compatibility code. See section Hosts and Cross-Compilation, for more.
How do you use a canonical system type? Usually, you use it in one or
more case statements in `configure.ac' to select
system-specific C files. Then, using AC_CONFIG_LINKS, link those
files which have names based on the system name, to generic names, such
as `host.h' or `target.c' (see section Creating Configuration Links). The
case statement patterns can use shell wild cards to group several
cases together, like in this fragment:
case $target in
i386-*-mach* | i386-*-gnu*)
obj_format=aout emulation=mach bfd_gas=yes ;;
i960-*-bout) obj_format=bout ;;
esac
and later in `configure.ac', use:
AC_CONFIG_LINKS(host.h:config/$machine.h
object.h:config/$obj_format.h)
Note that the above example uses $target because it's taken from
a tool which can be built on some architecture ($build), run on
another ($host), but yet handle data for a third architecture
($target). Such tools are usually part of a compiler suite, they
generate code for a specific $target.
However $target should be meaningless for most packages. If you
want to base a decision on the system where your program will be run,
make sure you use the $host variable, as in the following
excerpt:
case $host in
*-*-msdos* | *-*-go32* | *-*-mingw32* | *-*-cygwin* | *-*-windows*)
MUMBLE_INIT="mumble.ini"
;;
*)
MUMBLE_INIT=".mumbleinit"
;;
esac
AC_SUBST([MUMBLE_INIT])
You can also use the host system type to find cross-compilation tools.
See section Generic Program and File Checks, for information about the AC_CHECK_TOOL
macro which does that.
@command{configure} scripts support several kinds of local configuration decisions. There are ways for users to specify where external software packages are, include or exclude optional features, install programs under modified names, and set default values for @command{configure} options.
Some packages require, or can optionally use, other software packages that are already installed. The user can give @command{configure} command line options to specify which such external software to use. The options have one of these forms:
--with-package[=arg] --without-package
For example, @option{--with-gnu-ld} means work with the GNU linker instead of some other linker. @option{--with-x} means work with The X Window System.
The user can give an argument by following the package name with `=' and the argument. Giving an argument of `no' is for packages that are used by default; it says to not use the package. An argument that is neither `yes' nor `no' could include a name or number of a version of the other package, to specify more precisely which other package this program is supposed to work with. If no argument is given, it defaults to `yes'. @option{--without-package} is equivalent to @option{--with-package=no}.
@command{configure} scripts do not complain about @option{--with-package} options that they do not support. This behavior permits configuring a source tree containing multiple packages with a top-level @command{configure} script when the packages support different options, without spurious error messages about options that some of the packages support. An unfortunate side effect is that option spelling errors are not diagnosed. No better approach to this problem has been suggested so far.
For each external software package that may be used, `configure.ac'
should call AC_ARG_WITH to detect whether the @command{configure}
user asked to use it. Whether each package is used or not by default,
and which arguments are valid, is up to you.
The option's argument is available to the shell commands
action-if-given in the shell variable withval, which is
actually just the value of the shell variable with_package,
with any @option{-} characters changed into `_'. You may use that
variable instead, if you wish.
The argument help-string is a description of the option that looks like this:
--with-readline support fancy command line editing
help-string may be more than one line long, if more detail is needed. Just make sure the columns line up in `configure --help'. Avoid tabs in the help string. You'll need to enclose it in `[' and `]' in order to produce the leading spaces.
You should format your help-string with the macro
AC_HELP_STRING (see section Making Your Help Strings Look Pretty).
AC_ARG_WITH that does not
support providing a help string.
If a software package has optional compile-time features, the user can give @command{configure} command line options to specify whether to compile them. The options have one of these forms:
--enable-feature[=arg] --disable-feature
These options allow users to choose which optional features to build and install. @option{--enable-feature} options should never make a feature behave differently or cause one feature to replace another. They should only cause parts of the program to be built rather than left out.
The user can give an argument by following the feature name with `=' and the argument. Giving an argument of `no' requests that the feature not be made available. A feature with an argument looks like @option{--enable-debug=stabs}. If no argument is given, it defaults to `yes'. @option{--disable-feature} is equivalent to @option{--enable-feature=no}.
@command{configure} scripts do not complain about @option{--enable-feature} options that they do not support. This behavior permits configuring a source tree containing multiple packages with a top-level @command{configure} script when the packages support different options, without spurious error messages about options that some of the packages support. An unfortunate side effect is that option spelling errors are not diagnosed. No better approach to this problem has been suggested so far.
For each optional feature, `configure.ac' should call
AC_ARG_ENABLE to detect whether the @command{configure} user asked
to include it. Whether each feature is included or not by default, and
which arguments are valid, is up to you.
The option's argument is available to the shell commands
action-if-given in the shell variable enableval, which is
actually just the value of the shell variable
enable_feature, with any @option{-} characters changed into
`_'. You may use that variable instead, if you wish. The
help-string argument is like that of AC_ARG_WITH
(see section Working With External Software).
You should format your help-string with the macro
AC_HELP_STRING (see section Making Your Help Strings Look Pretty).
AC_ARG_ENABLE that does not
support providing a help string.
Properly formatting the `help strings' which are used in
AC_ARG_WITH (see section Working With External Software) and AC_ARG_ENABLE
(see section Choosing Package Options) can be challenging. Specifically, you want
your own `help strings' to line up in the appropriate columns of
`configure --help' just like the standard Autoconf `help
strings' do. This is the purpose of the AC_HELP_STRING macro.
Expands into an help string that looks pretty when the user executes
`configure --help'. It is typically used in AC_ARG_WITH
(see section Working With External Software) or AC_ARG_ENABLE (see section Choosing Package Options). The following example will make this clearer.
AC_DEFUN(TEST_MACRO,
[AC_ARG_WITH(foo,
AC_HELP_STRING([--with-foo],
[use foo (default is NO)]),
ac_cv_use_foo=$withval, ac_cv_use_foo=no),
AC_CACHE_CHECK(whether to use foo,
ac_cv_use_foo, ac_cv_use_foo=no)])
Please note that the call to AC_HELP_STRING is unquoted.
Then the last few lines of `configure --help' will appear like
this:
--enable and --with options recognized: --with-foo use foo (default is NO)
The AC_HELP_STRING macro is particularly helpful when the
left-hand-side and/or right-hand-side are composed of macro
arguments, as shown in the following example.
AC_DEFUN(MY_ARG_WITH,
[AC_ARG_WITH([$1],
AC_HELP_STRING([--with-$1], [use $1 (default is $2)]),
ac_cv_use_$1=$withval, ac_cv_use_$1=no),
AC_CACHE_CHECK(whether to use $1, ac_cv_use_$1, ac_cv_use_$1=$2)])
Some software packages require complex site-specific information. Some examples are host names to use for certain services, company names, and email addresses to contact. Since some configuration scripts generated by Metaconfig ask for such information interactively, people sometimes wonder how to get that information in Autoconf-generated configuration scripts, which aren't interactive.
Such site configuration information should be put in a file that is
edited only by users, not by programs. The location of the file
can either be based on the prefix variable, or be a standard
location such as the user's home directory. It could even be specified
by an environment variable. The programs should examine that file at
run time, rather than at compile time. Run time configuration is more
convenient for users and makes the configuration process simpler than
getting the information while configuring. See section `Variables for Installation Directories' in GNU Coding Standards, for more information on where to put data files.
Autoconf supports changing the names of programs when installing them.
In order to use these transformations, `configure.ac' must call the
macro AC_ARG_PROGRAM.
program_transform_name a sequence of
sed commands for changing the names of installed programs.
If any of the options described below are given to @command{configure},
program names are transformed accordingly. Otherwise, if
AC_CANONICAL_TARGET has been called and a @option{--target} value
is given, the target type followed by a dash is used as a prefix.
Otherwise, no program name transformation is done.
You can specify name transformations by giving @command{configure} these command line options:
sed substitution expression on the names.
These transformations are useful with programs that can be part of a cross-compilation development environment. For example, a cross-assembler running on a Sun 4 configured with @option{--target=i960-vxworks} is normally installed as `i960-vxworks-as', rather than `as', which could be confused with a native Sun 4 assembler.
You can force a program name to begin with `g', if you don't want
GNU programs installed on your system to shadow other programs with
the same name. For example, if you configure GNU diff with
@option{--program-prefix=g}, then when you run `make install' it is
installed as `/usr/local/bin/gdiff'.
As a more sophisticated example, you could use
--program-transform-name='s/^/g/; s/^gg/g/; s/^gless/less/'
to prepend `g' to most of the program names in a source tree,
excepting those like gdb that already have one and those like
less and lesskey that aren't GNU programs. (That is
assuming that you have a source tree containing those programs that is
set up to use this feature.)
One way to install multiple versions of some programs simultaneously is to append a version number to the name of one or both. For example, if you want to keep Autoconf version 1 around for awhile, you can configure Autoconf version 2 using @option{--program-suffix=2} to install the programs as `/usr/local/bin/autoconf2', `/usr/local/bin/autoheader2', etc. Nevertheless, pay attention that only the binaries are renamed, therefore you'd have problems with the library files which might overlap.
Here is how to use the variable program_transform_name in a
`Makefile.in':
PROGRAMS = cp ls rm
transform = @program_transform_name@
install:
for p in $(PROGRAMS); do \
$(INSTALL_PROGRAM) $$p $(DESTDIR)$(bindir)/`echo $$p | \
sed '$(transform)'`; \
done
uninstall:
for p in $(PROGRAMS); do \
rm -f $(DESTDIR)$(bindir)/`echo $$p | sed '$(transform)'`; \
done
It is guaranteed that program_transform_name is never empty, and
that there are no useless separators. Therefore you may safely embed
program_transform_name within a sed program using `;':
transform = @program_transform_name@ transform_exe = s/$(EXEEXT)$$//;$(transform);s/$$/$(EXEEXT)/
Whether to do the transformations on documentation files (Texinfo or
man) is a tricky question; there seems to be no perfect answer,
due to the several reasons for name transforming. Documentation is not
usually particular to a specific architecture, and Texinfo files do not
conflict with system documentation. But they might conflict with
earlier versions of the same files, and man pages sometimes do
conflict with system documentation. As a compromise, it is probably
best to do name transformations on man pages but not on Texinfo
manuals.
Autoconf-generated @command{configure} scripts allow your site to provide default values for some configuration values. You do this by creating site- and system-wide initialization files.
If the environment variable @command{CONFIG_SITE} is set, @command{configure} uses its value as the name of a shell script to read. Otherwise, it reads the shell script `prefix/share/config.site' if it exists, then `prefix/etc/config.site' if it exists. Thus, settings in machine-specific files override those in machine-independent ones in case of conflict.
Site files can be arbitrary shell scripts, but only certain kinds of
code are really appropriate to be in them. Because @command{configure}
reads any cache file after it has read any site files, a site file can
define a default cache file to be shared between all Autoconf-generated
@command{configure} scripts run on that system (see section Cache Files). If
you set a default cache file in a site file, it is a good idea to also
set the output variable CC in that site file, because the cache
file is only valid for a particular compiler, but many systems have
several available.
You can examine or override the value set by a command line option to
@command{configure} in a site file; options set shell variables that have
the same names as the options, with any dashes turned into underscores.
The exceptions are that @option{--without-} and @option{--disable-} options
are like giving the corresponding @option{--with-} or @option{--enable-}
option and the value `no'. Thus, @option{--cache-file=localcache}
sets the variable cache_file to the value `localcache';
@option{--enable-warnings=no} or @option{--disable-warnings} sets the variable
enable_warnings to the value `no'; @option{--prefix=/usr} sets the
variable prefix to the value `/usr'; etc.
Site files are also good places to set default values for other output
variables, such as CFLAGS, if you need to give them non-default
values: anything you would normally do, repetitively, on the command
line. If you use non-default values for prefix or
exec_prefix (wherever you locate the site file), you can set them
in the site file if you specify it with the @command{CONFIG_SITE}
environment variable.
You can set some cache values in the site file itself. Doing this is useful if you are cross-compiling, so it is impossible to check features that require running a test program. You could "prime the cache" by setting those values correctly for that system in `prefix/etc/config.site'. To find out the names of the cache variables you need to set, look for shell variables with `_cv_' in their names in the affected @command{configure} scripts, or in the Autoconf M4 source code for those macros.
The cache file is careful to not override any variables set in the site
files. Similarly, you should not override command-line options in the
site files. Your code should check that variables such as prefix
and cache_file have their default values (as set near the top of
@command{configure}) before changing them.
Here is a sample file `/usr/share/local/gnu/share/config.site'. The command `configure --prefix=/usr/share/local/gnu' would read this file (if @command{CONFIG_SITE} is not set to a different file).
# config.site for configure # # Change some defaults. test "$prefix" = NONE && prefix=/usr/share/local/gnu test "$exec_prefix" = NONE && exec_prefix=/usr/local/gnu test "$sharedstatedir" = '$prefix/com' && sharedstatedir=/var test "$localstatedir" = '$prefix/var' && localstatedir=/var # Give Autoconf 2.x generated configure scripts a shared default # cache file for feature test results, architecture-specific. if test "$cache_file" = /dev/null; then cache_file="$prefix/var/config.cache" # A cache file is only valid for one C compiler. CC=gcc fi
Below are instructions on how to configure a package that uses a @command{configure} script, suitable for inclusion as an `INSTALL' file in the package. A plain-text version of `INSTALL' which you may use comes with Autoconf.
These are generic installation instructions.
The @command{configure} shell script attempts to guess correct values for various system-dependent variables used during compilation. It uses those values to create a `Makefile' in each directory of the package. It may also create one or more `.h' files containing system-dependent definitions. Finally, it creates a shell script `config.status' that you can run in the future to recreate the current configuration, and a file `config.log' containing compiler output (useful mainly for debugging @command{configure}).
It can also use an optional file (typically called `config.cache' and enabled with @option{--cache-file=config.cache} or simply @option{-C}) that saves the results of its tests to speed up reconfiguring. (Caching is disabled by default to prevent problems with accidental use of stale cache files.)
If you need to do unusual things to compile the package, please try to figure out how @command{configure} could check whether to do them, and mail diffs or instructions to the address given in the `README' so they can be considered for the next release. If you are using the cache, and at some point `config.cache' contains results you don't want to keep, you may remove or edit it.
The file `configure.ac' (or `configure.in') is used to create
`configure' by a program called autoconf. You only need
`configure.ac' if you want to change it or regenerate
`configure' using a newer version of autoconf.
The simplest way to compile this package is:
cd to the directory containing the package's source code and type
`./configure' to configure the package for your system. If you're
using csh on an old version of System V, you might need to type
`sh ./configure' instead to prevent csh from trying to
execute @command{configure} itself.
Running @command{configure} takes awhile. While running, it prints some
messages telling which features it is checking for.
Some systems require unusual options for compilation or linking that the @command{configure} script does not know about. Run @samp{./configure --help} for details on some of the pertinent environment variables.
You can give @command{configure} initial values for variables by setting them in the environment. You can do that on the command line like this:
./configure CC=c89 CFLAGS=-O2 LIBS=-lposix
See section Defining Variables, for more details.
You can compile the package for more than one kind of computer at the
same time, by placing the object files for each architecture in their
own directory. To do this, you must use a version of @command{make}
that supports the VPATH variable, such as GNU @command{make}.
@command{cd} to the directory where you want the object files and
executables to go and run the @command{configure} script.
@command{configure} automatically checks for the source code in the
directory that @command{configure} is in and in `..'.
If you have to use a @command{make} that does not support the
VPATH variable, you have to compile the package for one
architecture at a time in the source code directory. After you have
installed the package for one architecture, use `make distclean'
before reconfiguring for another architecture.
By default, `make install' will install the package's files in `/usr/local/bin', `/usr/local/man', etc. You can specify an installation prefix other than `/usr/local' by giving @command{configure} the option @option{--prefix=path}.
You can specify separate installation prefixes for architecture-specific files and architecture-independent files. If you give @command{configure} the option @option{--exec-prefix=path}, the package will use path as the prefix for installing programs and libraries. Documentation and other data files will still use the regular prefix.
In addition, if you use an unusual directory layout you can give options like @option{--bindir=path} to specify different values for particular kinds of files. Run `configure --help' for a list of the directories you can set and what kinds of files go in them.
If the package supports it, you can cause programs to be installed with an extra prefix or suffix on their names by giving @command{configure} the option @option{--program-prefix=PREFIX} or @option{--program-suffix=SUFFIX}.
Some packages pay attention to @option{--enable-feature} options to @command{configure}, where feature indicates an optional part of the package. They may also pay attention to @option{--with-package} options, where package is something like `gnu-as' or `x' (for the X Window System). The `README' should mention any @option{--enable-} and @option{--with-} options that the package recognizes.
For packages that use the X Window System, @command{configure} can usually find the X include and library files automatically, but if it doesn't, you can use the @command{configure} options @option{--x-includes=dir} and @option{--x-libraries=dir} to specify their locations.
There may be some features @command{configure} cannot figure out automatically, but needs to determine by the type of machine the package will run on. Usually, assuming the package is built to be run on the same architectures, @command{configure} can figure that out, but if it prints a message saying it cannot guess the machine type, give it the @option{--build=type} option. type can either be a short name for the system type, such as `sun4', or a canonical name which has the form:
cpu-company-system
where system can have one of these forms:
os kernel-os
See the file `config.sub' for the possible values of each field. If `config.sub' isn't included in this package, then this package doesn't need to know the machine type.
If you are building compiler tools for cross-compiling, you should use the @option{--target=type} option to select the type of system they will produce code for.
If you want to use a cross compiler, that generates code for a platform different from the build platform, you should specify the host platform (i.e., that on which the generated programs will eventually be run) with @option{--host=type}.
If you want to set default values for @command{configure} scripts to
share, you can create a site shell script called `config.site' that
gives default values for variables like CC, cache_file,
and prefix. @command{configure} looks for
`prefix/share/config.site' if it exists, then
`prefix/etc/config.site' if it exists. Or, you can set the
CONFIG_SITE environment variable to the location of the site
script. A warning: not all @command{configure} scripts look for a site
script.
Variables not defined in a site shell script can be set in the environment passed to @command{configure}. However, some packages may run configure again during the build, and the customized values of these variables may be lost. In order to avoid this problem, you should set them in the @command{configure} command line, using `VAR=value'. For example:
./configure CC=/usr/local2/bin/gcc
will cause the specified gcc to be used as the C compiler (unless it is overridden in the site shell script).
@command{configure} recognizes the following options to control how it operates.
@command{configure} also accepts some other, not widely useful, options. Run `configure --help' for more details.
The @command{configure} script creates a file named `config.status', which actually configures, instantiates, the template files. It also records the configuration options that were specified when the package was last configured in case reconfiguring is needed.
Synopsis:
./config.status option... [file...]
It configures the files, if none are specified, all the templates are instantiated. The files must be specified without their dependencies, as in
./config.status foobar
not
./config.status foobar:foo.in:bar.in
The supported options are:
`config.status' checks several optional environment variables that can alter its behavior:
You can use `./config.status' in your Makefiles. For example, in the dependencies given above (see section Automatic Remaking), `config.status' is run twice when `configure.ac' has changed. If that bothers you, you can make each run only regenerate the files for that rule:
config.h: stamp-h
stamp-h: config.h.in config.status
./config.status config.h
echo > stamp-h
Makefile: Makefile.in config.status
./config.status Makefile
The calling convention of `config.status' has changed, see section Obsolete `config.status' Invocation, for details.
Autoconf changes, and throughout the years some constructs are obsoleted. Most of the changes involve the macros, but the tools themselves, or even some concepts, are now considered obsolete.
You may completely skip this chapter if you are new to Autoconf, its intention is mainly to help maintainers updating their packages by understanding how to move to more modern constructs.
`config.status' now supports arguments to specify the files to instantiate, see section Recreating a Configuration, for more details. Before, environment variables had to be used.
AC_OUTPUT and AC_CONFIG_COMMANDS in
`configure.ac'.
AC_OUTPUT and
AC_CONFIG_FILES in `configure.ac'.
#define statements. The
default is the arguments given to AC_CONFIG_HEADERS; if that
macro was not called, `config.status' ignores this variable.
AC_CONFIG_LINKS; if that macro was not called,
`config.status' ignores this variable.
In section Recreating a Configuration, using this old interface, the example would be:
config.h: stamp-h
stamp-h: config.h.in config.status
CONFIG_COMMANDS= CONFIG_LINKS= CONFIG_FILES= \
CONFIG_HEADERS=config.h ./config.status
echo > stamp-h
Makefile: Makefile.in config.status
CONFIG_COMMANDS= CONFIG_LINKS= CONFIG_HEADERS= \
CONFIG_FILES=Makefile ./config.status
(If `configure.ac' does not call AC_CONFIG_HEADERS, there is
no need to set @command{CONFIG_HEADERS} in the make rules, equally
for @command{CONFIG_COMMANDS} etc.)
In order to produce `config.h.in', @command{autoheader} needs to
build or to find templates for each symbol. Modern releases of Autoconf
use AH_VERBATIM and AH_TEMPLATE (see section Autoheader Macros), but in older releases a file, `acconfig.h', contained the
list of needed templates. @command{autoheader} copies comments and
#define and #undef statements from `acconfig.h' in
the current directory, if present. This file used to be mandatory if
you AC_DEFINE any additional symbols.
Modern releases of Autoconf also provide AH_TOP and
AH_BOTTOM if you need to prepend/append some information to
`config.h.in'. Ancient versions of Autoconf had a similar feature:
if `./acconfig.h' contains the string `@TOP@',
@command{autoheader} copies the lines before the line containing
`@TOP@' into the top of the file that it generates. Similarly,
if `./acconfig.h' contains the string `@BOTTOM@',
@command{autoheader} copies the lines after that line to the end of the
file it generates. Either or both of those strings may be omitted. An
even older alternate way to produce the same effect in jurasik versions
of Autoconf is to create the files `file.top' (typically
`config.h.top') and/or `file.bot' in the current
directory. If they exist, @command{autoheader} copies them to the
beginning and end, respectively, of its output.
In former versions of Autoconf, the files used in preparing a software package for distribution were:
configure.ac --. .------> autoconf* -----> configure
+---+
[aclocal.m4] --+ `---.
[acsite.m4] ---' |
+--> [autoheader*] -> [config.h.in]
[acconfig.h] ----. |
+-----'
[config.h.top] --+
[config.h.bot] --'
Use only the AH_ macros, `configure.ac' should be
self-contained, and should not depend upon `acconfig.h' etc.
The @command{autoupdate} program updates a `configure.ac' file that calls Autoconf macros by their old names to use the current macro names. In version 2 of Autoconf, most of the macros were renamed to use a more uniform and descriptive naming scheme. See section Macro Names, for a description of the new scheme. Although the old names still work (see section Obsolete Macros, for a list of the old macros and the corresponding new names), you can make your `configure.ac' files more readable and make it easier to use the current Autoconf documentation if you update them to use the new macro names.
If given no arguments, @command{autoupdate} updates `configure.ac',
backing up the original version with the suffix `~' (or the value
of the environment variable SIMPLE_BACKUP_SUFFIX, if that is
set). If you give @command{autoupdate} an argument, it reads that file
instead of `configure.ac' and writes the updated file to the
standard output.
@command{autoupdate} accepts the following options:
Several macros are obsoleted in Autoconf, for various reasons (typically they failed to quote properly, couldn't be extended for more recent issues etc.). They are still supported, but deprecated: their use should be avoided.
During the jump from Autoconf version 1 to version 2, most of the macros were renamed to use a more uniform and descriptive naming scheme, but their signature did not change. See section Macro Names, for a description of the new naming scheme. Below, there is just the mapping from old names to new names for these macros, the reader is invited to refer to the definition of the new macro for the signature and the description.
The user is encouraged to use either AC_CANONICAL_BUILD, or
AC_CANONICAL_HOST, or AC_CANONICAL_TARGET, depending on
the needs. Using AC_CANONICAL_TARGET is enough to run the two
other macros.
AC_CHECK_TYPE, deprecated because of its flaws. Firstly, although
it is a member of the CHECK clan, singular sub-family, it does
more than just checking. Second, missing types are not
typedef'd, they are #define'd, which can lead to
incompatible code in the case of pointer types.
This use of AC_CHECK_TYPE is obsolete and discouraged, see
section Generic Type Checks, for the description of the current macro.
If the type type is not defined, define it to be the C (or C++) builtin type default; e.g., `short' or `unsigned'.
This macro is equivalent to:
AC_CHECK_TYPE([type],
[AC_DEFINE([type], [default],
[Define to `default' if <sys/types.h>
does not define.])])
In order to keep backward compatibility, the two versions of
AC_CHECK_TYPE are implemented, selected by a simple heuristics:
You are encouraged either to use a valid builtin type, or to use the
equivalent modern code (see above), or better yet, to use
AC_CHECK_TYPES together with
#if !HAVE_LOFF_T typedef loff_t off_t; #endif
AC_TRY_LINK (see section Examining Libraries), with the addition that it prints `checking for
echo-text' to the standard output first, if echo-text is
non-empty. Use AC_MSG_CHECKING and AC_MSG_RESULT instead
to print messages (see section Printing Messages).
CYGWIN is set to `yes'. Don't use this macro, the dignified
means to check the nature of the host is using
AC_CANONICAL_HOST. As a matter of fact this macro is defined as:
AC_REQUIRE([AC_CANONICAL_HOST])[]dnl
case $host_os in
*cygwin* ) CYGWIN=yes;;
* ) CYGWIN=no;;
esac
Beware that the variable CYGWIN has a very special meaning when
running CygWin32, and should not be changed. That's yet another reason
not to use this macro.
DIRENT HAVE_DIRENT_H
SYSNDIR HAVE_SYS_NDIR_H
SYSDIR HAVE_SYS_DIR_H
NDIR HAVE_NDIR_H
LIBS. This macro used to be defined as
AC_CHECK_LIB(seq, getmntent, LIBS="-lseq $LIBS")now it is just
AC_FUNC_GETMNTENT.
EXEEXT based on the output of the
compiler, which is now done automatically. Typically set to empty
string if Unix and `.exe' if Win32 or OS/2.
wait3 is found and fills in the contents of its third argument
(a `struct rusage *'), which HP-UX does not do, define
HAVE_WAIT3.
These days portable programs should use waitpid, not
wait3, as wait3 is being removed from the Open Group
standards, and will not appear in the next revision of POSIX.
AC_CHECK_LIB with a
function argument of main. In addition, library can
be written as any of `foo', @option{-lfoo}, or `libfoo.a'. In
all of those cases, the compiler is passed @option{-lfoo}. However,
library cannot be a shell variable; it must be a literal name.
AC_INIT used to have a single argument, and was
equivalent to:
AC_INIT AC_CONFIG_SRCDIR(unique-file-in-source-dir)
int is 16 bits wide, define INT_16_BITS.
Use `AC_CHECK_SIZEOF(int)' instead.
LIBS. If you were using it to get getmntent, use
AC_FUNC_GETMNTENT instead. If you used it for the NIS versions
of the password and group functions, use `AC_CHECK_LIB(sun,
getpwnam)'. Up to Autoconf 2.13, it used to be
AC_CHECK_LIB(sun, getmntent, LIBS="-lsun $LIBS")now it is defined as
AC_FUNC_GETMNTENT AC_CHECK_LIB(sun, getpwnam)
AC_LANG_SAVE, remove it from the stack, and call
AC_LANG(language).
AC_LANG) on a stack.
The current language does not change. AC_LANG_PUSH is preferred.
AC_CONFIG_LINKS. An updated
version of:
AC_LINK_FILES(config/$machine.h config/$obj_format.h,
host.h object.h)
is:
AC_CONFIG_LINKS(host.h:config/$machine.h
object.h:config/$obj_format.h)
LONG_64_BITS if the C type long int is 64 bits wide.
Use the generic macro `AC_CHECK_SIZEOF([long int])' instead.
NEED_MEMORY_H if the mem functions were
defined in `memory.h'. Today it is equivalent to
`AC_CHECK_HEADERS(memory.h)'. Adjust your code to depend upon
HAVE_MEMORY_H, not NEED_MEMORY_H, see See section Standard Symbols.
AC_CYGWIN but checks for the MingW32 compiler
environment and sets MINGW32.
OBJEXT based on the output of the
compiler, after .c files have been excluded. Typically set to `o'
if Unix, `obj' if Win32. Now the compiler checking macros handle
this automatically.
m4 print a message to the standard error output warning that
this-macro-name is obsolete, and giving the file and line number
where it was called. this-macro-name should be the name of the
macro that is calling AC_OBSOLETE. If suggestion is given,
it is printed at the end of the warning message; for example, it can be
a suggestion for what to use instead of this-macro-name.
For instance
AC_OBSOLETE([$0], [; use AC_CHECK_HEADERS(unistd.h) instead])dnlYou are encouraged to use
AU_DEFUN instead, since it gives better
services to the user.
AC_OUTPUT with argument is deprecated, this obsoleted
interface is equivalent to:
AC_CONFIG_FILES(file...)
AC_CONFIG_COMMANDS([default],
extra-cmds, init-cmds)
AC_OUTPUT
AC_CONFIG_COMMANDS.
Here is an unrealistic example:
fubar=27
AC_OUTPUT_COMMANDS([echo this is extra $fubar, and so on.],
[fubar=$fubar])
AC_OUTPUT_COMMANDS([echo this is another, extra, bit],
[echo init bit])
Aside from the fact that AC_CONFIG_COMMANDS requires an
additional key, an important difference is that
AC_OUTPUT_COMMANDS is quoting its arguments twice, while
AC_CONFIG_COMMANDS. This means that AC_CONFIG_COMMANDS
can safely be given macro calls as arguments:
AC_CONFIG_COMMANDS(foo, [my_FOO()])conversely, where one level of quoting was enough for literal strings with
AC_OUTPUT_COMMANDS, you need two with
AC_CONFIG_COMMANDS. The following lines are equivalent:
AC_OUTPUT_COMMANDS([echo "Square brackets: []"]) AC_CONFIG_COMMANDS([default], [[echo "Square brackets: []"]])
LIBS. This
macro used to
AC_CHECK_LIB(intl, strftime, LIBS="-lintl $LIBS")now it just calls
AC_FUNC_STRFTIME instead.
HAVE_RESTARTABLE_SYSCALLS. This macro does
not check if system calls are restarted in general--it tests whether a
signal handler installed with signal (but not sigaction)
causes system calls to be restarted. It does not test if system calls
can be restarted when interrupted by signals that have no handler.
These days portable programs should use sigaction with
SA_RESTART if they want restartable system calls. They should
not rely on HAVE_RESTARTABLE_SYSCALLS, since nowadays whether a
system call is restartable is a dynamic issue, not a configuration-time
issue.
USG if the BSD string functions are defined in
`strings.h'. You should no longer depend upon USG, but on
HAVE_STRING_H, see See section Standard Symbols.
LIBS if on
Xenix. Also, if `dirent.h' is being checked for, added
@option{-ldir} to LIBS. Now it is merely an alias of
AC_HEADER_DIRENT instead, plus some code to detect whether
running XENIX on which you should not depend:
AC_MSG_CHECKING([for Xenix])
AC_EGREP_CPP(yes,
[#if defined M_XENIX && !defined M_UNIX
yes
#endif],
[AC_MSG_RESULT([yes]); XENIX=yes],
[AC_MSG_RESULT([no]); XENIX=])
Autoconf version 2 is mostly backward compatible with version 1. However, it introduces better ways to do some things, and doesn't support some of the ugly things in version 1. So, depending on how sophisticated your `configure.ac' files are, you might have to do some manual work in order to upgrade to version 2. This chapter points out some problems to watch for when upgrading. Also, perhaps your @command{configure} scripts could benefit from some of the new features in version 2; the changes are summarized in the file `NEWS' in the Autoconf distribution.
If you have an `aclocal.m4' installed with Autoconf (as opposed to in a particular package's source directory), you must rename it to `acsite.m4'. See section Using @command{autoconf} to Create @command{configure}.
If you distribute `install.sh' with your package, rename it to
`install-sh' so make builtin rules won't inadvertently
create a file called `install' from it. AC_PROG_INSTALL
looks for the script under both names, but it is best to use the new name.
If you were using `config.h.top', `config.h.bot', or
`acconfig.h', you still can, but you will have less clutter if you
use the AH_ macros. See section Autoheader Macros.
Add `@CFLAGS@', `@CPPFLAGS@', and `@LDFLAGS@' in your `Makefile.in' files, so they can take advantage of the values of those variables in the environment when @command{configure} is run. Doing this isn't necessary, but it's a convenience for users.
Also add `@configure_input@' in a comment to each input file for
AC_OUTPUT, so that the output files will contain a comment saying
they were produced by @command{configure}. Automatically selecting the
right comment syntax for all the kinds of files that people call
AC_OUTPUT on became too much work.
Add `config.log' and `config.cache' to the list of files you
remove in distclean targets.
If you have the following in `Makefile.in':
prefix = /usr/local exec_prefix = $(prefix)
you must change it to:
prefix = @prefix@ exec_prefix = @exec_prefix@
The old behavior of replacing those variables without `@' characters around them has been removed.
Many of the macros were renamed in Autoconf version 2. You can still use the old names, but the new ones are clearer, and it's easier to find the documentation for them. See section Obsolete Macros, for a table showing the new names for the old macros. Use the @command{autoupdate} program to convert your `configure.ac' to using the new macro names. See section Using @command{autoupdate} to Modernize @file{configure.ac}.
Some macros have been superseded by similar ones that do the job better,
but are not call-compatible. If you get warnings about calling obsolete
macros while running @command{autoconf}, you may safely ignore them, but
your @command{configure} script will generally work better if you follow
the advice it prints about what to replace the obsolete macros with. In
particular, the mechanism for reporting the results of tests has
changed. If you were using echo or AC_VERBOSE (perhaps
via AC_COMPILE_CHECK), your @command{configure} script's output will
look better if you switch to AC_MSG_CHECKING and
AC_MSG_RESULT. See section Printing Messages. Those macros work best
in conjunction with cache variables. See section Caching Results.
If you were checking the results of previous tests by examining the
shell variable DEFS, you need to switch to checking the values of
the cache variables for those tests. DEFS no longer exists while
@command{configure} is running; it is only created when generating output
files. This difference from version 1 is because properly quoting the
contents of that variable turned out to be too cumbersome and
inefficient to do every time AC_DEFINE is called. See section Cache Variable Names.
For example, here is a `configure.ac' fragment written for Autoconf version 1:
AC_HAVE_FUNCS(syslog)
case "$DEFS" in
*-DHAVE_SYSLOG*) ;;
*) # syslog is not in the default libraries. See if it's in some other.
saved_LIBS="$LIBS"
for lib in bsd socket inet; do
AC_CHECKING(for syslog in -l$lib)
LIBS="$saved_LIBS -l$lib"
AC_HAVE_FUNCS(syslog)
case "$DEFS" in
*-DHAVE_SYSLOG*) break ;;
*) ;;
esac
LIBS="$saved_LIBS"
done ;;
esac
Here is a way to write it for version 2:
AC_CHECK_FUNCS(syslog)
if test $ac_cv_func_syslog = no; then
# syslog is not in the default libraries. See if it's in some other.
for lib in bsd socket inet; do
AC_CHECK_LIB($lib, syslog, [AC_DEFINE(HAVE_SYSLOG)
LIBS="$LIBS -l$lib"; break])
done
fi
If you were working around bugs in AC_DEFINE_UNQUOTED by adding
backslashes before quotes, you need to remove them. It now works
predictably, and does not treat quotes (except back quotes) specially.
See section Setting Output Variables.
All of the boolean shell variables set by Autoconf macros now use `yes' for the true value. Most of them use `no' for false, though for backward compatibility some use the empty string instead. If you were relying on a shell variable being set to something like 1 or `t' for true, you need to change your tests.
When defining your own macros, you should now use AC_DEFUN
instead of define. AC_DEFUN automatically calls
AC_PROVIDE and ensures that macros called via AC_REQUIRE
do not interrupt other macros, to prevent nested `checking...'
messages on the screen. There's no actual harm in continuing to use the
older way, but it's less convenient and attractive. See section Macro Definitions.
You probably looked at the macros that came with Autoconf as a guide for how to do things. It would be a good idea to take a look at the new versions of them, as the style is somewhat improved and they take advantage of some new features.
If you were doing tricky things with undocumented Autoconf internals (macros, variables, diversions), check whether you need to change anything to account for changes that have been made. Perhaps you can even use an officially supported technique in version 2 instead of kludging. Or perhaps not.
To speed up your locally written feature tests, add caching to them. See whether any of your tests are of general enough usefulness to encapsulate into macros that you can share.
The introduction of the previous section (see section Upgrading From Version 1) perfectly suits this section...
Autoconf version 2.50 is mostly backward compatible with version 2.13. However, it introduces better ways to do some things, and doesn't support some of the ugly things in version 2.13. So, depending on how sophisticated your `configure.ac' files are, you might have to do some manual work in order to upgrade to version 2.50. This chapter points out some problems to watch for when upgrading. Also, perhaps your @command{configure} scripts could benefit from some of the new features in version 2.50; the changes are summarized in the file `NEWS' in the Autoconf distribution.
The most important changes are invisible to you: the implementation of most macros have completely changed. This allowed more factorization of the code, better error messages, a higher uniformity of the user's interface etc. Unfortunately, as a side effect, some construct which used to (miraculously) work might break starting with Autoconf 2.50. The most common culprit is bad quotation.
For instance, in the following example, the message is not properly quoted:
AC_INIT AC_CHECK_HEADERS(foo.h,, AC_MSG_ERROR(cannot find foo.h, bailing out)) AC_OUTPUT
Autoconf 2.13 simply ignores it:
$ autoconf-2.13; ./configure --silent creating cache ./config.cache configure: error: cannot find foo.h $
while Autoconf 2.50 will produce a broken `configure':
$ autoconf-2.50; ./configure --silent configure: error: cannot find foo.h ./configure: exit: bad non-numeric arg `bailing' ./configure: exit: bad non-numeric arg `bailing' $
The message needs to be quoted, and the AC_MSG_ERROR invocation
too!
AC_INIT
AC_CHECK_HEADERS(foo.h,,
[AC_MSG_ERROR([cannot find foo.h, bailing out])])
AC_OUTPUT
Many many (and many more) Autoconf macros were lacking proper quotation,
including no less than... AC_DEFUN itself!
$ cat configure.in AC_DEFUN([AC_PROG_INSTALL], [# My own much better version ]) AC_INIT AC_PROG_INSTALL AC_OUTPUT $ autoconf-2.13 autoconf: Undefined macros: ***BUG in Autoconf--please report*** AC_FD_MSG ***BUG in Autoconf--please report*** AC_EPI configure.in:1:AC_DEFUN([AC_PROG_INSTALL], configure.in:5:AC_PROG_INSTALL $ autoconf-2.50 $
Because Autoconf has been dormant for years, Automake provided
Autoconf-like macros for a while. Autoconf 2.50 now provides better
versions of these macros, integrated in the AC_ namespace,
instead of AM_. But in order to ease the upgrading via
@command{autoupdate}, bindings to such AM_ macros are provided.
Unfortunately Automake did not quote the name of these macros!
Therefore, when @command{m4} finds something like
`AC_DEFUN(AM_TYPE_PTRDIFF_T, ...)' in `aclocal.m4',
AM_TYPE_PTRDIFF_T is
expanded, replaced with its Autoconf definition.
Fortunately Autoconf catches pre-AC_INIT expansions, and will
complain, in its own words:
$ cat configure.in AC_INIT AM_TYPE_PTRDIFF_T $ aclocal-1.4 $ autoconf ./aclocal.m4:17: error: m4_defn: undefined macro: _m4_divert_diversion actypes.m4:289: AM_TYPE_PTRDIFF_T is expanded from... ./aclocal.m4:17: the top level $
Future versions of Automake will simply no longer define most of these macros, and will properly quote the names of the remaining macros. But you don't have to wait for it to happen to do the right thing right now: do not depend upon macros from Automake as it is simply not its job to provide macros (but the one it requires by itself):
$ cat configure.in AC_INIT AM_TYPE_PTRDIFF_T $ rm aclocal.m4 $ autoupdate autoupdate: `configure.in' is updated $ cat configure.in AC_INIT AC_CHECK_TYPES([ptrdiff_t]) $ aclocal-1.4 $ autoconf $
Based on the experience of compiler writers, and after long public debates, many aspects of the cross-compilation chain have changed:
The relationship between build, host, and target have been cleaned up: the chain of default is now simply: target defaults to host, host to build, and build to the result of @command{config.guess}. Nevertheless, in order to ease the transition from 2.13 to 2.50, the following transition scheme is implemented. Do not rely on it, as it will be completely disabled in a couple of releases (we cannot keep it, as it proves to cause more problems than to cure).
They all default to the result of running @command{config.guess}, unless you specify either @option{--build} or @option{--host}. In this case, the default becomes the system type you specified. If you specify both, and they're different, @command{configure} will enter cross compilation mode, so it won't run any tests that require execution.
Hint: if you mean to override the result of @command{config.guess},
prefer @option{--build} over @option{--host}. In the future,
@option{--host} will not override the name of the build system type.
Whenever you specify --host, be sure to specify --build
too.
For backward compatibility, @command{configure} will accept a system type as an option by itself. Such an option will override the defaults for build, host and target system types. The following configure statement will configure a cross toolchain that will run on NetBSD/alpha but generate code for GNU Hurd/sparc, which is also the build platform.
./configure --host=alpha-netbsd sparc-gnu
In Autoconf, the variables build, host, and target
had a different semantics before and after the invocation of
AC_CANONICAL_BUILD etc. Now, the argument of @option{--build} is
strictly copied into build_alias, and is left empty otherwise.
After the AC_CANONICAL_BUILD, build is set to the
canonicalized build type. To ease the transition, before, its contents
is the same as that of build_alias. Do not rely on this
broken feature.
For consistency with the backward compatibility scheme exposed above, when @option{--host} is specified by @option{--build} isn't, the build system will be assumed to be the same as @option{--host}, and `build_alias' will be set to that value. Eventually, this historically incorrect behavior will go away.
The former scheme to enable cross-compilation proved to cause more harm than good, in particular, it used to be triggered too easily, leaving regular end users puzzled in front of cryptic error messages. @command{configure} could even enter cross-compilation mode, only because the compiler was not functional. This is mainly because @command{configure} used to try to detect cross-compilation, instead of waiting for an explicit flag from the user.
Now, @command{configure} enters cross-compilation mode iff @option{--host} is passed.
That's the short documentation. To ease the transition between 2.13 and its successors, a more complicated scheme is implemented. Do not rely on the following, as it will be removed in a near future.
If you specify @option{--host}, but not @option{--build}, when
@command{configure} performs the first compiler test it will try to run
an executable produced by the compiler. If the execution fails, it will
enter cross-compilation mode. This is fragile. Moreover, by the time
the compiler test is performed, it may be too late to modify the
build-system type: other tests may have already been performed.
Therefore, whenever you specify --host, be sure to specify
--build too.
./configure --build=i686-pc-linux-gnu --host=m68k-coff
will enter cross-compilation mode. The former interface, which consisted in setting the compiler to a cross-compiler without informing @command{configure} is obsolete. For instance, @command{configure} will fail if it can't run the code generated by the specified compiler if you configure as follows:
./configure CC=m68k-coff-gcc
AC_LIBOBJ vs. LIBOBJS
Up to Autoconf 2.13, the replacement of functions was triggered via the
variable LIBOBJS. Since Autoconf 2.50, the macro
AC_LIBOBJ should be used instead (see section Generic Function Checks).
Starting at Autoconf 2.53, the use of LIBOBJS is an error.
This change is mandated by the unification of the GNU Build System
components. In particular, the various fragile techniques used to parse
a `configure.ac' are all replaced with the use of traces. As a
consequence, any action must be traceable, which obsoletes critical
variable assignments. Fortunately, LIBOBJS was the only problem.
At the time this documentation is written, Automake does not rely on
traces yet, but this is planed for a near future. Nevertheless, to
ease the transition, and to guarantee this future Automake release will
be able to use Autoconf 2.53, using LIBOBJS directly will make
@command{autoconf} fail. But note that the output, @command{configure},
is correct and fully functional: you have some delay to adjust your
source.
There are two typical uses of LIBOBJS: asking for a replacement
function, and adjusting LIBOBJS for Automake and/or Libtool.
As for function replacement, the fix is immediate: use
AC_LIBOBJ. For instance:
LIBOBJS="$LIBOBJS fnmatch.o" LIBOBJS="$LIBOBJS malloc.$ac_objext"
should be replaced with:
AC_LIBOBJ([fnmatch]) AC_LIBOBJ([malloc])
When asked for automatic de-ANSI-fication, Automake needs
LIBOBJS'ed filenames to have `$U' appended to the
base names. Libtool requires the definition of LTLIBOBJS, which
suffixes are mapped to `.lo'. Although Autoconf provides them with
means to free the user to do that by herself, by the time of this
writing, none do. Therefore, it is common to see `configure.ac'
end with:
# This is necessary so that .o files in LIBOBJS are also built via # the ANSI2KNR-filtering rules. LIBOBJS=`echo "$LIBOBJS" | sed 's/\.o /\$U.o /g;s/\.o$/\$U.o/'` LTLIBOBJS=`echo "$LIBOBJS" | sed 's/\.o/\.lo/g'` AC_SUBST(LTLIBOBJS)
First, note that this code is wrong, because `.o' is not the
only possible extension(4)! Because the token LIBOBJS is now
forbidden, you will have to replace this snippet with:
# This is necessary so that .o files in LIBOBJS are also built via
# the ANSI2KNR-filtering rules.
LIB@&t@OBJS=`echo "$LIB@&t@OBJS" |
sed 's,\.[[^.]]* ,$U&,g;s,\.[[^.]]*$,$U&,'`
LTLIBOBJS=`echo "$LIB@&t@OBJS" |
sed 's,\.[[^.]]* ,.lo ,g;s,\.[[^.]]*$,.lo,'`
AC_SUBST(LTLIBOBJS)
Unfortunately, @command{autoupdate} cannot help here, since... this is not a macro! Of course, first make sure your release of Automake and/or Libtool still requires these.
Note: This section describes an experimental feature which will be part of Autoconf in a forthcoming release. Although we believe Autotest is stabilizing, this documentation describes an interface which might change in the future: do not depend upon Autotest without subscribing to the Autoconf mailing lists.
It is paradoxical that portable projects depend on nonportable tools to run their test suite. Autoconf by itself is the paragon of this problem: although it aims at perfectly portability, up to 2.13, its test suite was using DejaGNU, a rich and complex testing framework, but which is far from being standard on Unix systems. Worse yet, it was likely to be missing on the most fragile platforms, the very platforms that are most likely to torture Autoconf and exhibit deficiencies.
To circumvent this problem many package maintainers have developed their own testing framework, based on simple shell scripts whose sole output are their exit status: the test succeeded, or failed. In addition, most of these tests share some common patterns, what results in lots of duplicated code, tedious maintenance etc.
Following exactly the same reasoning that yielded to the inception of Autoconf, Autotest provides a test suite generation frame work, based on M4 macros, building a portable shell script. The suite itself is equipped with automatic logging and tracing facilities which greatly diminish the interaction with bug reporters, and simple timing reports.
Autoconf itself has been using Autotest for years, and we do attest that it has considerably improved the strength of the test suite, and the quality of bug reports. Other projects are known to use some generation of Autotest, such as Bison, Free Recode, Free Wdiff, GNU Tar, each of them having different needs, what slowly polishes Autotest as a general testing framework.
Nonetheless, compared to DejaGNU, Autotest is inadequate for interactive tool testing, which is probably its main limitation.
Generating testing or validation suites using Autotest is rather easy.
The whole validation suite is held in a file to be processed through
@command{autom4te}, itself using GNU m4 under the scene, to
produce a stand-alone Bourne shell script which then gets distributed.
Neither @command{autom4te} nor GNU m4 are not needed anymore at
the installer end.
Each test of the validation suite should be part of some test group. A test group is a sequence of interwoven tests that ought to be executed together, usually because one test in the group creates data files than a later test in the same group needs to read. Complex test groups make later debugging more tedious. It is much better keeping keep only a few tests per test group, and if you can put only one test per test group, this is just ideal.
For all but the simplest packages, some file such as `testsuite.at' does not fully hold all test sources, as these are often easier to maintain in separate files. Each of these separate files holds a single test group, or a sequence of test groups all addressing some common functionality in the package. In such cases, file `testsuite.at' only initializes the whole validation suite, and sometimes do elementary health checking, before listing include statements for all other test files. The special file `package.m4', containing the identification of the package, is automatically included if found.
The validation scripts that Autotest produces are by convention called @command{testsuite}. When run, @command{testsuite} executes each test group in turn, producing only one summary line per test to say if that particular test succeeded or failed. At end of all tests, summarizing counters get printed. If any test failed, one debugging script gets automatically generated for each test group which failed. These debugging scripts are named `testsuite.nn', where nn is the sequence number of the test group. In the ideal situation, none of the tests fail, and consequently, no debugging script is generated out of validation.
The automatic generation of debugging scripts for failed test has the purpose of easing the chase for bugs.
It often happens in practice that individual tests in the validation
suite need to get information coming out of the configuration process.
Some of this information, common for all validation suites, is provided
through the file `atconfig', automatically created by
AC_CONFIG_TESTDIR. For configuration informations which your
testing environment specifically needs, you might prepare an optional
file named `atlocal.in', instantiated by AC_CONFIG_FILES.
The configuration process produces `atconfig' and `atlocal'
out of these two input files, and these two produced files are
automatically read by the `testsuite' script.
Here is a diagram showing the relationship between files.
Files used in preparing a software package for distribution:
subfile-1.at ->.
... \
subfile-i.at ---->-- testsuite.at -->.
... / \
subfile-n.at ->' >-- autom4te* -->testsuite
/
[package.m4] ->'
Files used in configuring a software package:
.--> atconfig
/
[atlocal.in] --> config.status* --<
\
`--> [atlocal]
Files created during the test suite execution:
atconfig -->. .--> testsuite.log
\ /
>-- testsuite* --<
/ \
[atlocal] ->' `--> [testsuite.nn*]
When run, the test suite creates a log file named after itself, e.g., a test suite named @command{testsuite} creates `testsuite.log'. It contains a lot of information, usually more than maintainers actually need, but therefore most of the time it contains all that is needed:
CC for subsequent runs(5). Autoconf faced exactly the same problem, and solved it by asking
users to pass the variable definitions as command line arguments.
Autotest requires this rule too, but has no means to enforce it; the log
then contains a trace of the variables the user changed.
AT_TESTED).
The `testsuite.at' is a Bourne shell script making use of special
Autotest M4 macros. It often contains a call to AT_INIT nears
its beginning followed by one call to m4_include per source file
for tests. Each such included file, or the remainder of
`testsuite.at' if include files are not used, contain a sequence of
test groups. Each test group begins with one call to AT_SETUP,
it contains an arbitrary number of shell commands or calls to
AT_CHECK, and it completes with one call to AT_CLEANUP.
Autotest test suites rely on the PATH to find the tested program.
This saves from generating the absolute paths to the various tools, and
makes it possible to test installed programs. Therefore, knowing what
programs are being exercised is crucial to understand some problems in
the test suite itself, or its occasional misuses. It is a good idea to
also subscribe foreign programs you depend upon, to ease incompatibility
diagnostics.
AT_KEYWORDS.
Several invocations within a test group accumulate new keywords. In other words, don't fear registering several times the same keyword in a test group.
m4
expansion. The contents ought to end with an end of line.
The commands must not redirect the standard output, nor the standard error.
If status, or stdout, or stderr is `ignore', then the corresponding value is not checked.
The special value `expout' for stdout means the expected output of the commands is the content of the file `expout'. If stdout is `stdout', then the standard output of the commands is available for further tests in the file `stdout'. Similarly for stderr with `expout' and `stderr'.
Autotest test suites support the following arguments:
clean Makefile targets.
By default all the tests are performed (or described with @option{--list}) in the default environment first silently, then verbosely, but the environment, set of tests, and verbosity level can be tuned:
AUTOTEST_PATH specifies the testing path to prepend
to PATH. It handles specially relative paths (not starting with
`/'): they are considered to be relative to the top level of the
package being built. All the directories are made absolute, first
starting from the top level build tree, then from the
source tree. For instance `./testsuite
AUTOTEST_PATH=tests:bin' for a `/src/foo-1.0' source package built
in `/tmp/foo' results in `/tmp/foo/tests:/tmp/foo/bin' and
then `/src/foo-1.0/tests:/src/foo-1.0/bin' being prepended to
PATH.
AT_SETUP or AT_KEYWORDS) match all the keywords
of the comma separated list keywords.
Running `./testsuite -k autoupdate,FUNC' will select all the tests
tagged with `autoupdate' and `FUNC' (as in
`AC_CHECK_FUNC', `AC_FUNC_FNMATCH' etc.) while
`./testsuite -k autoupdate -k FUNC' runs all the tests tagged with
`autoupdate' or `FUNC'.
For putting Autotest into movement, you need some configuration and Makefile machinery. We recommend, at least if your package uses deep or shallow hierarchies, that you use `tests/' as the name of the directory holding all your tests and their `Makefile'. Here is a check list of things to do.
AT_PACKAGE_STRING, the
full signature of the package, and AT_PACKAGE_BUGREPORT, the
address to which bug reports should be sent. For sake of completeness,
we suggest that you also define AT_PACKAGE_NAME,
AT_PACKAGE_TARNAME, and AT_PACKAGE_VERSION.
See section Initializing @command{configure}, for a description of these variables. We
suggest the following Makefile excerpt:
$(srcdir)/package.m4: $(top_srcdir)/configure.ac
{ \
echo '# Signature of the current package.'; \
echo 'm4_define([AT_PACKAGE_NAME], [@PACKAGE_NAME@])'; \
echo 'm4_define([AT_PACKAGE_TARNAME], [@PACKAGE_TARNAME@])'; \
echo 'm4_define([AT_PACKAGE_VERSION], [@PACKAGE_VERSION@])'; \
echo 'm4_define([AT_PACKAGE_STRING], [@PACKAGE_STRING@])'; \
echo 'm4_define([AT_PACKAGE_BUGREPORT], [@PACKAGE_BUGREPORT@])'; \
} >$(srcdir)/package.m4
Be sure to distribute `package.m4' and to put it into the source
hierarchy: the test suite ought to be shipped!
AT_CONFIG macro from within file `configure.ac'.
This macro accepts one argument, which is the directory, relative to the
test directory, where the executables are prepared.
AC_CONFIG_FILES command includes substitution for
`tests/atconfig' and also, as appropriate, `tests/atlocal'.
With Automake, here is a minimal example about how to link `make check' with a validation suite.
EXTRA_DIST = testsuite.at testsuite
TESTSUITE = $(srcdir)/testsuite
check-local: atconfig atlocal $(TESTSUITE)
$(SHELL) $(TESTSUITE)
AUTOTEST = $(AUTOM4TE) --language=autotest
$(TESTSUITE): $(srcdir)/testsuite.at
$(AUTOTEST) -I $(srcdir) $.at -o $.tmp
mv $.tmp $
You might want to list explicitly the dependencies, i.e., the list of the files `testsuite.at' includes.
With strict Autoconf, you might need to add lines inspired from the following:
subdir = tests
atconfig: $(top_builddir)/config.status
cd $(top_builddir) && \
$(SHELL) ./config.status $(subdir)/$
atlocal: $(srcdir)/atlocal.in $(top_builddir)/config.status
cd $(top_builddir) && \
$(SHELL) ./config.status $(subdir)/$
and manage to have `atconfig.in' and $(EXTRA_DIST)
distributed.
Several questions about Autoconf come up occasionally. Here some of them are addressed.
What are the restrictions on distributing @command{configure}
scripts that Autoconf generates? How does that affect my
programs that use them?
There are no restrictions on how the configuration scripts that Autoconf produces may be distributed or used. In Autoconf version 1, they were covered by the GNU General Public License. We still encourage software authors to distribute their work under terms like those of the GPL, but doing so is not required to use Autoconf.
Of the other files that might be used with @command{configure}, `config.h.in' is under whatever copyright you use for your `configure.ac'. `config.sub' and `config.guess' have an exception to the GPL when they are used with an Autoconf-generated @command{configure} script, which permits you to distribute them under the same terms as the rest of your package. `install-sh' is from the X Consortium and is not copyrighted.
Why does Autoconf require GNU M4?
Many M4 implementations have hard-coded limitations on the size and number of macros that Autoconf exceeds. They also lack several builtin macros that it would be difficult to get along without in a sophisticated application like Autoconf, including:
m4_builtin m4_indir m4_bpatsubst __file__ __line__
Autoconf requires version 1.4 or above of GNU M4 because it uses frozen state files.
Since only software maintainers need to use Autoconf, and since GNU M4 is simple to configure and install, it seems reasonable to require GNU M4 to be installed also. Many maintainers of GNU and other free software already have most of the GNU utilities installed, since they prefer them.
If Autoconf requires GNU M4 and GNU M4 has an Autoconf
@command{configure} script, how do I bootstrap? It seems like a chicken
and egg problem!
This is a misunderstanding. Although GNU M4 does come with a @command{configure} script produced by Autoconf, Autoconf is not required in order to run the script and install GNU M4. Autoconf is only required if you want to change the M4 @command{configure} script, which few people have to do (mainly its maintainer).
Why not use Imake instead of @command{configure} scripts?
Several people have written addressing this question, so I include adaptations of their explanations here.
The following answer is based on one written by Richard Pixley:
Autoconf generated scripts frequently work on machines that it has never been set up to handle before. That is, it does a good job of inferring a configuration for a new system. Imake cannot do this.
Imake uses a common database of host specific data. For X11, this makes sense because the distribution is made as a collection of tools, by one central authority who has control over the database.
GNU tools are not released this way. Each GNU tool has a maintainer; these maintainers are scattered across the world. Using a common database would be a maintenance nightmare. Autoconf may appear to be this kind of database, but in fact it is not. Instead of listing host dependencies, it lists program requirements.
If you view the GNU suite as a collection of native tools, then the problems are similar. But the GNU development tools can be configured as cross tools in almost any host+target permutation. All of these configurations can be installed concurrently. They can even be configured to share host independent files across hosts. Imake doesn't address these issues.
Imake templates are a form of standardization. The GNU coding standards address the same issues without necessarily imposing the same restrictions.
Here is some further explanation, written by Per Bothner:
One of the advantages of Imake is that it easy to generate large Makefiles using
cpp's `#include' and macro mechanisms. However,cppis not programmable: it has limited conditional facilities, and no looping. Andcppcannot inspect its environment.All of these problems are solved by using
shinstead ofcpp. The shell is fully programmable, has macro substitution, can execute (or source) other shell scripts, and can inspect its environment.
Paul Eggert elaborates more:
With Autoconf, installers need not assume that Imake itself is already installed and working well. This may not seem like much of an advantage to people who are accustomed to Imake. But on many hosts Imake is not installed or the default installation is not working well, and requiring Imake to install a package hinders the acceptance of that package on those hosts. For example, the Imake template and configuration files might not be installed properly on a host, or the Imake build procedure might wrongly assume that all source files are in one big directory tree, or the Imake configuration might assume one compiler whereas the package or the installer needs to use another, or there might be a version mismatch between the Imake expected by the package and the Imake supported by the host. These problems are much rarer with Autoconf, where each package comes with its own independent configuration processor.
Also, Imake often suffers from unexpected interactions between
makeand the installer's C preprocessor. The fundamental problem here is that the C preprocessor was designed to preprocess C programs, not `Makefile's. This is much less of a problem with Autoconf, which uses the general-purpose preprocessorm4, and where the package's author (rather than the installer) does the preprocessing in a standard way.
Finally, Mark Eichin notes:
Imake isn't all that extensible, either. In order to add new features to Imake, you need to provide your own project template, and duplicate most of the features of the existing one. This means that for a sophisticated project, using the vendor-provided Imake templates fails to provide any leverage--since they don't cover anything that your own project needs (unless it is an X11 program).
On the other side, though:
The one advantage that Imake has over @command{configure}: `Imakefile's tend to be much shorter (likewise, less redundant) than `Makefile.in's. There is a fix to this, however--at least for the Kerberos V5 tree, we've modified things to call in common `post.in' and `pre.in' `Makefile' fragments for the entire tree. This means that a lot of common things don't have to be duplicated, even though they normally are in @command{configure} setups.
You may be wondering, Why was Autoconf originally written? How did it get into its present form? (Why does it look like gorilla spit?) If you're not wondering, then this chapter contains no information useful to you, and you might as well skip it. If you are wondering, then let there be light...
In June 1991 I was maintaining many of the GNU utilities for the Free Software Foundation. As they were ported to more platforms and more programs were added, the number of @option{-D} options that users had to select in the `Makefile' (around 20) became burdensome. Especially for me--I had to test each new release on a bunch of different systems. So I wrote a little shell script to guess some of the correct settings for the fileutils package, and released it as part of fileutils 2.0. That @command{configure} script worked well enough that the next month I adapted it (by hand) to create similar @command{configure} scripts for several other GNU utilities packages. Brian Berliner also adapted one of my scripts for his CVS revision control system.
Later that summer, I learned that Richard Stallman and Richard Pixley were developing similar scripts to use in the GNU compiler tools; so I adapted my @command{configure} scripts to support their evolving interface: using the file name `Makefile.in' as the templates; adding `+srcdir', the first option (of many); and creating `config.status' files.
As I got feedback from users, I incorporated many improvements, using Emacs to search and replace, cut and paste, similar changes in each of the scripts. As I adapted more GNU utilities packages to use @command{configure} scripts, updating them all by hand became impractical. Rich Murphey, the maintainer of the GNU graphics utilities, sent me mail saying that the @command{configure} scripts were great, and asking if I had a tool for generating them that I could send him. No, I thought, but I should! So I started to work out how to generate them. And the journey from the slavery of hand-written @command{configure} scripts to the abundance and ease of Autoconf began.
Cygnus @command{configure}, which was being developed at around that time, is table driven; it is meant to deal mainly with a discrete number of system types with a small number of mainly unguessable features (such as details of the object file format). The automatic configuration system that Brian Fox had developed for Bash takes a similar approach. For general use, it seems to me a hopeless cause to try to maintain an up-to-date database of which features each variant of each operating system has. It's easier and more reliable to check for most features on the fly--especially on hybrid systems that people have hacked on locally or that have patches from vendors installed.
I considered using an architecture similar to that of Cygnus @command{configure}, where there is a single @command{configure} script that reads pieces of `configure.in' when run. But I didn't want to have to distribute all of the feature tests with every package, so I settled on having a different @command{configure} made from each `configure.in' by a preprocessor. That approach also offered more control and flexibility.
I looked briefly into using the Metaconfig package, by Larry Wall, Harlan Stenn, and Raphael Manfredi, but I decided not to for several reasons. The @command{Configure} scripts it produces are interactive, which I find quite inconvenient; I didn't like the ways it checked for some features (such as library functions); I didn't know that it was still being maintained, and the @command{Configure} scripts I had seen didn't work on many modern systems (such as System V R4 and NeXT); it wasn't very flexible in what it could do in response to a feature's presence or absence; I found it confusing to learn; and it was too big and complex for my needs (I didn't realize then how much Autoconf would eventually have to grow).
I considered using Perl to generate my style of @command{configure}
scripts, but decided that M4 was better suited to the job of simple
textual substitutions: it gets in the way less, because output is
implicit. Plus, everyone already has it. (Initially I didn't rely on
the GNU extensions to M4.) Also, some of my friends at the
University of Maryland had recently been putting M4 front ends on
several programs, including tvtwm, and I was interested in trying
out a new language.
Since my @command{configure} scripts determine the system's capabilities automatically, with no interactive user intervention, I decided to call the program that generates them Autoconfig. But with a version number tacked on, that name would be too long for old UNIX file systems, so I shortened it to Autoconf.
In the fall of 1991 I called together a group of fellow questers after
the Holy Grail of portability (er, that is, alpha testers) to give me
feedback as I encapsulated pieces of my handwritten scripts in M4 macros
and continued to add features and improve the techniques used in the
checks. Prominent among the testers were Fran@,cois Pinard, who came up
with the idea of making an `autoconf' shell script to run m4
and check for unresolved macro calls; Richard Pixley, who suggested
running the compiler instead of searching the file system to find
include files and symbols, for more accurate results; Karl Berry, who
got Autoconf to configure TeX and added the macro index to the
documentation; and Ian Lance Taylor, who added support for creating a C
header file as an alternative to putting @option{-D} options in a
`Makefile', so he could use Autoconf for his UUCP package.
The alpha testers cheerfully adjusted their files again and again as the
names and calling conventions of the Autoconf macros changed from
release to release. They all contributed many specific checks, great
ideas, and bug fixes.
In July 1992, after months of alpha testing, I released Autoconf 1.0, and converted many GNU packages to use it. I was surprised by how positive the reaction to it was. More people started using it than I could keep track of, including people working on software that wasn't part of the GNU Project (such as TCL, FSP, and Kerberos V5). Autoconf continued to improve rapidly, as many people using the @command{configure} scripts reported problems they encountered.
Autoconf turned out to be a good torture test for M4 implementations.
UNIX m4 started to dump core because of the length of the
macros that Autoconf defined, and several bugs showed up in GNU
m4 as well. Eventually, we realized that we needed to use some
features that only GNU M4 has. 4.3BSD m4, in
particular, has an impoverished set of builtin macros; the System V
version is better, but still doesn't provide everything we need.
More development occurred as people put Autoconf under more stresses
(and to uses I hadn't anticipated). Karl Berry added checks for X11.
david zuhn contributed C++ support. Fran@,cois Pinard made it diagnose
invalid arguments. Jim Blandy bravely coerced it into configuring
GNU Emacs, laying the groundwork for several later improvements.
Roland McGrath got it to configure the GNU C Library, wrote the
@command{autoheader} script to automate the creation of C header file
templates, and added a @option{--verbose} option to @command{configure}.
Noah Friedman added the @option{--autoconf-dir} option and
AC_MACRODIR environment variable. (He also coined the term
autoconfiscate to mean "adapt a software package to use
Autoconf".) Roland and Noah improved the quoting protection in
AC_DEFINE and fixed many bugs, especially when I got sick of
dealing with portability problems from February through June, 1993.
A long wish list for major features had accumulated, and the effect of
several years of patching by various people had left some residual
cruft. In April 1994, while working for Cygnus Support, I began a major
revision of Autoconf. I added most of the features of the Cygnus
@command{configure} that Autoconf had lacked, largely by adapting the
relevant parts of Cygnus @command{configure} with the help of david zuhn
and Ken Raeburn. These features include support for using
`config.sub', `config.guess', @option{--host}, and
@option{--target}; making links to files; and running @command{configure}
scripts in subdirectories. Adding these features enabled Ken to convert
GNU as, and Rob Savoye to convert DejaGNU, to using
Autoconf.
I added more features in response to other peoples' requests. Many
people had asked for @command{configure} scripts to share the results of
the checks between runs, because (particularly when configuring a large
source tree, like Cygnus does) they were frustratingly slow. Mike
Haertel suggested adding site-specific initialization scripts. People
distributing software that had to unpack on MS-DOS asked for a way to
override the `.in' extension on the file names, which produced file
names like `config.h.in' containing two dots. Jim Avera did an
extensive examination of the problems with quoting in AC_DEFINE
and AC_SUBST; his insights led to significant improvements.
Richard Stallman asked that compiler output be sent to `config.log'
instead of `/dev/null', to help people debug the Emacs
@command{configure} script.
I made some other changes because of my dissatisfaction with the quality
of the program. I made the messages showing results of the checks less
ambiguous, always printing a result. I regularized the names of the
macros and cleaned up coding style inconsistencies. I added some
auxiliary utilities that I had developed to help convert source code
packages to use Autoconf. With the help of Fran@,cois Pinard, I made
the macros not interrupt each others' messages. (That feature revealed
some performance bottlenecks in GNU m4, which he hastily
corrected!) I reorganized the documentation around problems people want
to solve. And I began a test suite, because experience had shown that
Autoconf has a pronounced tendency to regress when we change it.
Again, several alpha testers gave invaluable feedback, especially Fran@,cois Pinard, Jim Meyering, Karl Berry, Rob Savoye, Ken Raeburn, and Mark Eichin.
Finally, version 2.0 was ready. And there was much rejoicing. (And I have free time again. I think. Yeah, right.)
Copyright (C) 2000 Free Software Foundation, Inc. 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed.
To use this License in a document you have written, include a copy of the License in the document and put the following copyright and license notices just after the title page:
Copyright (C) year your name. Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.1 or any later version published by the Free Software Foundation; with the Invariant Sections being list their titles, with the Front-Cover Texts being list, and with the Back-Cover Texts being list. A copy of the license is included in the section entitled ``GNU Free Documentation License''.
If you have no Invariant Sections, write "with no Invariant Sections" instead of saying which ones are invariant. If you have no Front-Cover Texts, write "no Front-Cover Texts" instead of "Front-Cover Texts being list"; likewise for Back-Cover Texts.
If your document contains nontrivial examples of program code, we recommend releasing these examples in parallel under your choice of free software license, such as the GNU General Public License, to permit their use in free software.
This is an alphabetical list of the environment variables that Autoconf checks.
Jump to: c - i - l - n - p - r - s - w
This is an alphabetical list of the variables that Autoconf can substitute into files that it creates, typically one or more `Makefile's. See section Setting Output Variables, for more information on how this is done.
Jump to: a - b - c - d - e - f - g - h - i - k - l - m - n - o - p - r - s - t - x - y
This is an alphabetical list of the C preprocessor symbols that the
Autoconf macros define. To work with Autoconf, C source code needs to
use these names in #if directives.
Jump to: _ - c - d - f - g - h - i - l - m - n - o - p - r - s - t - u - v - w - x - y
This is an alphabetical list of the Autoconf macros. To make the list easier to use, the macros are listed without their preceding `AC_'.
Jump to: a - b - c - d - e - f - g - h - i - l - m - o - p - r - s - t - u - v - w - x - y
This is an alphabetical list of the M4, M4sugar, and M4sh macros. To make the list easier to use, the macros are listed without their preceding `m4_' or `AS_'.
Jump to: b - d - m - p - q - u
This is an alphabetical list of the Autotest macros. To make the list easier to use, the macros are listed without their preceding `AT_'.
Jump to: c - d - i - k - s - t
This is an alphabetical list of the programs and functions which portability is discussed in this document.
Jump to: @ - a - c - e - f - g - l - m - s - u - v
alloca
chown
closedir
error_at_line
fnmatch
fork
fseeko
getgroups
getloadavg
getmntent
getpgid
getpgrp
lstat, lstat
malloc
memcmp
mktime
mmap
select
setpgrp
setvbuf
snprintf
sprintf
sscanf
stat
strcoll
strerror_r
strftime
strnlen, strnlen
strtod
unlink
utime
va_copy
va_list
vfork
vprintf
vsnprintf
vsprintf
This is an alphabetical list of the files, tools, and concepts introduced in this document.
Jump to: " - $ - @ - ` - a - b - c - d - e - f - h - i - l - m - o - p - q - r - s - t - u - v - z
AUTOTEST_PATH
dnl, dnl
undefined macro: _m4_divert_diversion
VPATH
This document was generated on 10 July 2002 using texi2html 1.56k.