Difference between revisions of "Configure"
Vollstrecker (Talk | contribs) m (Removed unnecessary Repeats) |
(→More information) |
||
| Line 352: | Line 352: | ||
Of course, you can always take a look at the [http://www.airs.com/ian/configure GNU configure page]. | Of course, you can always take a look at the [http://www.airs.com/ian/configure GNU configure page]. | ||
| − | |||
| − | |||
Revision as of 11:56, 12 December 2008
Introduction
configure configures aMule to adapt to many kinds of systems.
Its purpose is to make it easy to port it to many platforms and to compile it and use it in all platforms in the very same way.
You can set lots of parameters in it to adapt it to your system and to tweak its features.
NOTE: This article is focused in aMule's configure. Other configure scripts work in a similar way, but their options will differ. This article might be useful for them as well, but have in mind that they might have options not listed here and lack some listed here too.
Main usage
Its usage is ./configure [OPTION]... [VARIABLE=VALUE]...
So, as you can see, you can define its behaviour both through options and through variables
Options
You can define one or more options. The available options in aMule's configure are listed below. To better understand their meaning, they have been grouped into several tables with options which affect the behaviour of configure in a similar way.
Please note that configure will (in most cases) not complain if you enter an invalid option or variable. So make sure you enter the options correctly (no typos) or they will have no effect and configure won't behave in the way you expected.
Main options
These options should exist in any configure script around the world. They won't run the configure script to configure your application's compilation, but only output information you might need.
| -h or --help | This will display a help message summing up all configure options |
| --help=short | This will display a help message summing up all configure options specific to this application (doesn't display generic configure options) |
| -V or --version | This will display the version information |
Non-compilation common options
These options should also exist on any configure script around the world. They will set some non-compilation related preferences.
| -q or --quiet or --silent | Do not print the `checking...' messages, only the final summed up ouput |
| --cache-file=FILE | Store the checkings' ouputs in FILE in a script format, so you can reuse without the need to rerun configure (read the cached file for more information) |
| -C or --config-cache | Same as --cache-file=config.cache |
| -n or --no-create | Just run configure, but do not create any files |
| --srcdir=DIR | Look for the sources in DIR. By default, the configure dir or its parent (../) |
Common options
These options should also exist on any configure script around the world. This time, these ones will affect the compilation of the application, so be careful when using them. You should take a close look at the application's README document (if available) if you want to use any of them (and even if you don't, still take a look at it, since you might have to use some).
| --prefix=PREFIX | Set where to install architecture-independent files (data) in PREFIX. By dafault, this is set to /usr/local |
| --exec-prefix=EPREFIX | Set where to install architecture-dependent files (mainly binaries) in EPREFIX. By default, this is set to the same value as PREFIX. |
| --bindir=DIR | Where to store user executables. By default, EPREFIX/bin |
| --sbindir=DIR | Where to store system administrator executables. By default, EPREFIX/sbin |
| --libexecdir=DIR | Where to store program executable. By default EPREFIX/libexec |
| --datadir=DIR | Where to store read-only architecture independent data. By default, EPREFIX/share |
| --sysconfdir=DIR | Where to install read-only single-machine data. By default, PREFIX/etc |
| --sharedstatedir=DIR | Where to store modifiable architacture-independent data. By default PREFIX/com |
| --localstatedir=DIR | Where to store modifiable single-machine data. By default, PREFIX/var |
| --libdir=DIR | Where to search libs object code libraries. By default, EPREFIX/lib |
| --includedir=DIR | Directory containing the C header files. By default, PREFIX/include |
| --oldincludedir=DIR | Directory containing C header files for non-GCC. By default, /usr/include |
| --infodir=DIR | Where to store info and documentation files. By default, EPREFIX/info |
| --mandir=DIR | Where to store man pages. By default, EPREFIX/man |
| --program-prefix=PREFIX | Prepend PREFIX to installed program names |
| --program-suffix=SUFFIX | Append SUFFIX to installed program names |
| --program-transform-name=SUBST | Run sed SUBST on installed program names |
| --disable-FEATURE | Do not include FEATURE. Equivalent to --enable-FEATURE=no (see below) |
| --enable-FEATURE | Include FEATURE. Equivalent to --enable-FEATURE=yes (see below) |
| --enable-FEATURE=VALUE | Include FEATURE and specify it's value to VALUE. The meaning of the value should be explained in the configure --help output of the script |
| --without-PACKAGE | Do not use PACKAGE. Equivalent to --with-PACKAGE=no (see below) |
| --with-PACKAGE | use PACKAGE. Equivalent to --with-PACKAGE=yes (see below) |
| --with-PACKAGE=VALUE | Use PACKAGE and set its value to VALUE |
| --build=BUILD | Configure for building on BUILD platform. By default, it guesses the platform it is being run on |
| --host=HOST | Cross-compile to build programs to run on HOST platforms. By default, same as BUILD |
| --target=TARGET | Configure for building compilers for TARGET platfroms. By default, same as HOST |
Very-common options
In most configure scripts you will also be able to use these options.
| --with-gnu-ld | Use the GNU ld as linker, without testing if it is the default linker in the system |
| --with-x | Use the X Window System |
| --x-includes=DIR | DIR contains X's include files |
| --x-libraries=DIR | DIR contains X's libraries |
aMule options
The following options are specific to aMule's configure script. They will allow you to enable and disable them, so you should really take a look at them or you might sadly find out you lack things you would like to have in your aMule system ;-)
| --enable-maintainer-mode | Activate some extra instructions only useful for developers (and totaly non-recommended to normal users) |
| --disable-dependency-tracking | Makes the application compile faster, but causes problems if you change the source code and try to compile again. Dependency tracking allows the make process to know which things to rebuild when something changes. If you do not plan to compile the application again with those same sources, you can safely use this option |
| --enable-dependency-tracking | Keep track of the dependencies so that on next build, only changed files will be recompiled. It slows down the compilation a little. This is the default |
| --enable-ccache | Enable ccache. This will make following compilations faster since only the modified bits of code will be recompiled. If you don't plan to compile aMule very often, don't enable this, since it takes longer to compile on the first compilation and it requires lots of disk space |
| --enable-profile | Enable code profiling (which allows to see how much time and resources are spent on each part of the code, though it will make aMule run much slower, so, in most cases, don't use it). The output file is gmon.out and can be viewed with gprof. For more information, man gprof and man gcc (in the man pages of gcc, search for the command line option -pg) |
| --disable-debug | Disables outputing information of the application execution. This information is useful to track bugs, so you might not want to use this option or you will not be able to report bugs properly |
| --enable-optimize | Optimizes the code to run (a little) faster. Using this option will not allow you to create useful bug reports so you might not want to enable it |
| --disable-rpath | Do not write the libraries' path into the binary. Useful only if you plan to move the binary to other systems or if you move your libraries often. Only use it if you know what you're doing, since it will make aMule run slower |
| --enable-static | Creates statically linked executables, which means that they will need no libraries since any needed library will be included in the executable file. The executable might run faster, and will be easily portable between systems using the same platform, but will be incredibly bigger in size |
| --disable-embedded_crypto | Do not use aMule's cryptography code, instead, use external libraries' code. This option is not recommended and it could possibly break the compilation |
| --disable-nls | Do not use Native Language Support. This way gettext won't be required, but translations will be disabled |
| --disable-monolithic | disable building of the monolithic aMule app. This is normally not recommended, unless you really know what you are doing (e.g. to compile only the daemon) |
| --enable-amule-daemon | Compile aMule daemon |
| --enable-amulecmd | Compile the aMule command line client |
| --enable-webserver | Compile aMule's WebServer |
| --enable-amule-gui | Compile aMule remote GUI |
| --enable-cas | Compile C aMule Statistics |
| --enable-wxcas | Compile aMule GUI Statistics |
| --disable-ed2k | Don't compile aMule's ed2k links handler |
| --enable-alc | Compile aMuleLinkCreator GUI version |
| --enable-alcc | Compile aMuleLinkCreator for console |
| --disable-gtktest | Do not try to compile a test GTK (either GTK1 or GTK2 depending on your system's configuration) program to see if your system has a proper GTK configuration. This will make the configure script run a (very) little faster, but you should still not use it unless you know what you are doing, since the configure script could miss some tests. |
aMule compilation settings
This options are specific to aMule's configure script, but they are not directly related to aMule. They just tell were some libraries or headers are located in your system, so that aMule can find them when compiling. Use them if you have those files in non-standard paths or when you know what you are doing.
To better understand these options, take a look at the common options chapter in this article.
| --with-zlib=PREFIX | Look for zLib files in PREFIX |
| --with-curl-config=CONFIG | CONFIG is the path of curl-config |
| --with-gdlib-prefix=PREFIX | Look for LibGD files in PREFIX |
| --with-gdlib-exec-prefix=PREFIX | Look for GDLib libraries in PREFIX. |
| --with-gdlib-config=CONFIG | CONFIG is the path of gdlib-config |
| --with-libpng-prefix=PREFIX | Look for LibPNG files in PREFIX |
| --with-libpng-exec-prefix=PREFIX | Look for LibPNG libraries in PREFIX |
| --with-libpng-config=CONFIG | CONFIG is the path of libpng-config |
| --with-wx-prefix=PREFIX | Look for wxWidgets files in PREFIX |
| --with-wx-exec-prefix=PREFIX | Look for wxWidgets libraries in PREFIX |
| --with-wx-config=CONFIG | CONFIG is the path of wx-config |
| --with-wxbase-prefix=PREFIX | Look for wxBase files in PREFIX |
| --with-wxbase-exec-prefix=PREFIX | Look for wxBase libraries in PREFIX |
| --with-wxbase-config=CONFIG | CONFIG is the path of wxbase-config |
| --with-gtk-prefix=PREFIX | Look for GTK (not GTK2) files in PREFIX |
| --with-gtk-exec-prefix=PREFIX | Look for GTK (not GTK2) libraries in PREFIX |
| --with-crypto-prefix=PREFIX | Look for Crypto++ files in PREFIX |
| --with-libiconv-prefix=PREFIX | Look for LibIConv files in PREFIX/include and PREFIX/lib |
| --without-libiconv-prefix | Don't search for LibIConv in includedir and libdir |
| --with-included-gettext | Use the GNU gettext library included in the intl/ directory instead of the system's. This is only for strange systems where gettext, for some reason, will not handle aMule's catalog files |
| --with-libintl-prefix=PREFIX | Look for LibIntl files in PREFIX/include and PREFIX/lib' |
| --without-libintl-prefix | Don't search for LibIntl in includedir and libdir |
| --with-ccache-prefix=PREFIX | Look for ccache files in PREFIX |
Variables
Before you take a look at the variables and what they mean, you must take a look at the two variable usage types available.
Usage types
You can either use environment variables or you can enter variables directly into the configure command.
Environment variables
If you plan to use environment variables, you will affect the behaviour of configure and any other program you run in that same shell until you end your session. You must set them before running configure. To set them, run
export VAR=VALUE'
where VAR is the name of the environmental variable and VALUE is it's value.
Configure variables
If you don't want to use environment variables, you can pass them into the configure command, this way:
configure VAR=VALUE
where VAR stands for the variable name and VALUE is its value.
Variables and what they mean
Do NOT touch these variables if you don't know what you are doing!
These are some of the most commonly used variables for configure:
| CC | C compiler command to use |
| CXX | C++ compiler command to use (e.g. g++) |
| CPP | C preprocessor command to use |
| CFLAGS | C compiler flags |
| CXXCPP | C++ preprocessor command to use |
| CXXFLAGS | C++ compiler flags |
| LDFLAGS | Linker flags. Usually used to specify additional library directories (e.g. LDFLAGS="-L/usr/lib") |
| CPPFLAGS | C/C++ preprocessor flags (e.g. CPPFLAGS="-I/usr/include") |
More information
This article is mainly an explanation (dummy-proof) of the output of configure --help.
Anyway, it could (rarely) be out of date. So, if there's something you want to do with configure that is not listed here, run configure --help and see if it is listed there.
Of course, you can always take a look at the GNU configure page.
