1 @c This is for making the `INSTALL' file for the distribution.
2 @c Makeinfo ignores it when processing the file from the include.
7 @node Installation, Maintenance, Library Summary, Top
8 @c %MENU% How to install the GNU C Library
9 @appendix Installing @theglibc{}
11 Before you do anything else, you should read the FAQ at
12 @url{http://sourceware.org/glibc/wiki/FAQ}. It answers common
13 questions and describes problems you may experience with compilation
16 Features can be added to @theglibc{} via @dfn{add-on} bundles. These are
17 separate tar files, which you unpack into the top level of the source
18 tree. Then you give @code{configure} the @samp{--enable-add-ons} option
19 to activate them, and they will be compiled into the library.
21 You will need recent versions of several GNU tools: definitely GCC and
22 GNU Make, and possibly others. @xref{Tools for Compilation}, below.
25 * Configuring and compiling:: How to compile and test GNU libc.
26 * Running make install:: How to install it once you've got it
28 * Tools for Compilation:: You'll need these first.
29 * Linux:: Specific advice for GNU/Linux systems.
30 * Reporting Bugs:: So they'll get fixed.
33 @node Configuring and compiling
34 @appendixsec Configuring and compiling @theglibc{}
38 @Theglibc{} cannot be compiled in the source directory. You must build
39 it in a separate build directory. For example, if you have unpacked
40 the @glibcadj{} sources in @file{/src/gnu/glibc-@var{version}},
42 @file{/src/gnu/glibc-build} to put the object files in. This allows
43 removing the whole build directory in case an error occurs, which is
44 the safest way to get a fresh start and should always be done.
46 From your object directory, run the shell script @file{configure} located
47 at the top level of the source tree. In the scenario above, you'd type
50 $ ../glibc-@var{version}/configure @var{args@dots{}}
53 Please note that even though you're building in a separate build
54 directory, the compilation may need to create or modify files and
55 directories in the source directory.
58 @code{configure} takes many options, but the only one that is usually
59 mandatory is @samp{--prefix}. This option tells @code{configure}
60 where you want @theglibc{} installed. This defaults to @file{/usr/local},
61 but the normal setting to install as the standard system library is
62 @samp{--prefix=/usr} for @gnulinuxsystems{} and @samp{--prefix=} (an
63 empty prefix) for @gnuhurdsystems{}.
65 It may also be useful to set the @var{CC} and @var{CFLAGS} variables in
66 the environment when running @code{configure}. @var{CC} selects the C
67 compiler that will be used, and @var{CFLAGS} sets optimization options
70 The following list describes all of the available options for
74 @item --prefix=@var{directory}
75 Install machine-independent data files in subdirectories of
76 @file{@var{directory}}. The default is to install in @file{/usr/local}.
78 @item --exec-prefix=@var{directory}
79 Install the library and other machine-dependent files in subdirectories
80 of @file{@var{directory}}. The default is to the @samp{--prefix}
81 directory if that option is specified, or @file{/usr/local} otherwise.
83 @item --with-headers=@var{directory}
84 Look for kernel header files in @var{directory}, not
85 @file{/usr/include}. @Theglibc{} needs information from the kernel's header
86 files describing the interface to the kernel. @Theglibc{} will normally
87 look in @file{/usr/include} for them,
88 but if you specify this option, it will look in @var{DIRECTORY} instead.
90 This option is primarily of use on a system where the headers in
91 @file{/usr/include} come from an older version of @theglibc{}. Conflicts can
92 occasionally happen in this case. You can also use this option if you want to
93 compile @theglibc{} with a newer set of kernel headers than the ones found in
96 @item --enable-add-ons[=@var{list}]
97 Specify add-on packages to include in the build. If this option is
98 specified with no list, it enables all the add-on packages it finds in
99 the main source directory; this is the default behavior. You may
100 specify an explicit list of add-ons to use in @var{list}, separated by
101 spaces or commas (if you use spaces, remember to quote them from the
102 shell). Each add-on in @var{list} can be an absolute directory name
103 or can be a directory name relative to the main source directory, or
104 relative to the build directory (that is, the current working directory).
105 For example, @samp{--enable-add-ons=nptl,../glibc-libidn-@var{version}}.
107 @item --enable-kernel=@var{version}
108 This option is currently only useful on @gnulinuxsystems{}. The
109 @var{version} parameter should have the form X.Y.Z and describes the
110 smallest version of the Linux kernel the generated library is expected
111 to support. The higher the @var{version} number is, the less
112 compatibility code is added, and the faster the code gets.
114 @item --with-binutils=@var{directory}
115 Use the binutils (assembler and linker) in @file{@var{directory}}, not
116 the ones the C compiler would default to. You can use this option if
117 the default binutils on your system cannot deal with all the constructs
118 in @theglibc{}. In that case, @code{configure} will detect the
119 problem and suppress these constructs, so that the library will still be
120 usable, but functionality may be lost---for example, you can't build a
121 shared libc with old binutils.
124 Use this option if your computer lacks hardware floating-point support
125 and your operating system does not emulate an FPU.
127 @c disable static doesn't work currently
128 @c @item --disable-static
129 @c Don't build static libraries. Static libraries aren't that useful these
130 @c days, but we recommend you build them in case you need them.
132 @item --disable-shared
133 Don't build shared libraries even if it is possible. Not all systems
134 support shared libraries; you need ELF support and (currently) the GNU
137 @item --disable-profile
138 Don't build libraries with profiling information. You may want to use
139 this option if you don't plan to do profiling.
141 @item --disable-versioning
142 Don't compile the shared libraries with symbol version information.
143 Doing this will make the resulting library incompatible with old
144 binaries, so it's not recommended.
146 @item --enable-static-nss
147 Compile static versions of the NSS (Name Service Switch) libraries.
148 This is not recommended because it defeats the purpose of NSS; a program
149 linked statically with the NSS libraries cannot be dynamically
150 reconfigured to use a different name database.
153 By default the C library is built with support for thread-local storage
154 if the used tools support it. By using @samp{--without-tls} this can be
155 prevented though there generally is no reason since it creates
156 compatibility problems.
158 @item --enable-hardcoded-path-in-tests
159 By default, dynamic tests are linked to run with the installed C library.
160 This option hardcodes the newly built C library path in dynamic tests
161 so that they can be invoked directly.
163 @item --build=@var{build-system}
164 @itemx --host=@var{host-system}
165 These options are for cross-compiling. If you specify both options and
166 @var{build-system} is different from @var{host-system}, @code{configure}
167 will prepare to cross-compile @theglibc{} from @var{build-system} to be used
168 on @var{host-system}. You'll probably need the @samp{--with-headers}
169 option too, and you may have to override @var{configure}'s selection of
170 the compiler and/or binutils.
172 If you only specify @samp{--host}, @code{configure} will prepare for a
173 native compile but use what you specify instead of guessing what your
174 system is. This is most useful to change the CPU submodel. For example,
175 if @code{configure} guesses your machine as @code{i686-pc-linux-gnu} but
176 you want to compile a library for 586es, give
177 @samp{--host=i586-pc-linux-gnu} or just @samp{--host=i586-linux} and add
178 the appropriate compiler flags (@samp{-mcpu=i586} will do the trick) to
181 If you specify just @samp{--build}, @code{configure} will get confused.
183 @item --with-pkgversion=@var{version}
184 Specify a description, possibly including a build number or build
185 date, of the binaries being built, to be included in
186 @option{--version} output from programs installed with @theglibc{}.
187 For example, @option{--with-pkgversion='FooBar GNU/Linux glibc build
188 123'}. The default value is @samp{GNU libc}.
190 @item --with-bugurl=@var{url}
191 Specify the URL that users should visit if they wish to report a bug,
192 to be included in @option{--help} output from programs installed with
193 @theglibc{}. The default value refers to the main bug-reporting
194 information for @theglibc{}.
197 To build the library and related programs, type @code{make}. This will
198 produce a lot of output, some of which may look like errors from
199 @code{make} but isn't. Look for error messages from @code{make}
200 containing @samp{***}. Those indicate that something is seriously wrong.
202 The compilation process can take a long time, depending on the
203 configuration and the speed of your machine. Some complex modules may
204 take a very long time to compile, as much as several minutes on slower
205 machines. Do not panic if the compiler appears to hang.
207 If you want to run a parallel make, simply pass the @samp{-j} option
208 with an appropriate numeric parameter to @code{make}. You need a recent
209 GNU @code{make} version, though.
211 To build and run test programs which exercise some of the library
212 facilities, type @code{make check}. If it does not complete
213 successfully, do not use the built library, and report a bug after
214 verifying that the problem is not already known. @xref{Reporting Bugs},
215 for instructions on reporting bugs. Note that some of the tests assume
216 they are not being run by @code{root}. We recommend you compile and
217 test @theglibc{} as an unprivileged user.
219 Before reporting bugs make sure there is no problem with your system.
220 The tests (and later installation) use some pre-existing files of the
221 system such as @file{/etc/passwd}, @file{/etc/nsswitch.conf} and others.
222 These files must all contain correct and sensible content.
224 To format the @cite{GNU C Library Reference Manual} for printing, type
225 @w{@code{make dvi}}. You need a working @TeX{} installation to do
226 this. The distribution builds the on-line formatted version of the
227 manual, as Info files, as part of the build process. You can build
228 them manually with @w{@code{make info}}.
230 The library has a number of special-purpose configuration parameters
231 which you can find in @file{Makeconfig}. These can be overwritten with
232 the file @file{configparms}. To change them, create a
233 @file{configparms} in your build directory and add values as appropriate
234 for your system. The file is included and parsed by @code{make} and has
235 to follow the conventions for makefiles.
237 It is easy to configure @theglibc{} for cross-compilation by
238 setting a few variables in @file{configparms}. Set @code{CC} to the
239 cross-compiler for the target you configured the library for; it is
240 important to use this same @code{CC} value when running
241 @code{configure}, like this: @samp{CC=@var{target}-gcc configure
242 @var{target}}. Set @code{BUILD_CC} to the compiler to use for programs
243 run on the build system as part of compiling the library. You may need to
244 set @code{AR} to cross-compiling versions of @code{ar}
245 if the native tools are not configured to work with
246 object files for the target you configured for. When cross-compiling
247 @theglibc{}, it may be tested using @samp{make check
248 test-wrapper="@var{srcdir}/scripts/cross-test-ssh.sh @var{hostname}"},
249 where @var{srcdir} is the absolute directory name for the main source
250 directory and @var{hostname} is the host name of a system that can run
251 the newly built binaries of @theglibc{}. The source and build
252 directories must be visible at the same locations on both the build
253 system and @var{hostname}.
255 In general, when testing @theglibc{}, @samp{test-wrapper} may be set
256 to the name and arguments of any program to run newly built binaries.
257 This program must preserve the arguments to the binary being run, its
258 working directory, all environment variables set as part of testing
259 and the standard input, output and error file descriptors. If
260 @samp{@var{test-wrapper} env} will not work to run a program with
261 environment variables set, then @samp{test-wrapper-env} must be set to
262 a program that runs a newly built program with environment variable
263 assignments in effect, those assignments being specified as
264 @samp{@var{var}=@var{value}} before the name of the program to be run.
267 @node Running make install
268 @appendixsec Installing the C Library
271 To install the library and its header files, and the Info files of the
272 manual, type @code{env LANGUAGE=C LC_ALL=C make install}. This will
273 build things, if necessary, before installing them; however, you should
274 still compile everything first. If you are installing @theglibc{} as your
275 primary C library, we recommend that you shut the system down to
276 single-user mode first, and reboot afterward. This minimizes the risk
277 of breaking things when the library changes out from underneath.
279 @samp{make install} will do the entire job of upgrading from a
280 previous installation of @theglibc{} version 2.x. There may sometimes
282 left behind from the previous installation, but those are generally
283 harmless. If you want to avoid leaving headers behind you can do
284 things in the following order.
286 You must first build the library (@samp{make}), optionally check it
287 (@samp{make check}), switch the include directories and then install
288 (@samp{make install}). The steps must be done in this order. Not moving
289 the directory before install will result in an unusable mixture of header
290 files from both libraries, but configuring, building, and checking the
291 library requires the ability to compile and run programs against the old
292 library. The new @file{/usr/include}, after switching the include
293 directories and before installing the library should contain the Linux
294 headers, but nothing else. If you do this, you will need to restore
295 any headers from libraries other than @theglibc{} yourself after installing the
298 You can install @theglibc{} somewhere other than where you configured it to go
299 by setting the @code{install_root} variable on the command line for
300 @samp{make install}. The value of this variable is prepended to all the
301 paths for installation. This is useful when setting up a chroot
302 environment or preparing a binary distribution. The directory should be
303 specified with an absolute file name.
305 @Theglibc{} includes a daemon called @code{nscd}, which you
306 may or may not want to run. @code{nscd} caches name service lookups; it
307 can dramatically improve performance with NIS+, and may help with DNS as
310 One auxiliary program, @file{/usr/libexec/pt_chown}, is installed setuid
311 @code{root}. This program is invoked by the @code{grantpt} function; it
312 sets the permissions on a pseudoterminal so it can be used by the
313 calling process. This means programs like @code{xterm} and
314 @code{screen} do not have to be setuid to get a pty. (There may be
315 other reasons why they need privileges.) If you are using a
316 Linux kernel with the @code{devptsfs} or @code{devfs} filesystems
317 providing pty slaves, you don't need this program; otherwise you do.
318 The source for @file{pt_chown} is in @file{login/programs/pt_chown.c}.
320 After installation you might want to configure the timezone and locale
321 installation of your system. @Theglibc{} comes with a locale
322 database which gets configured with @code{localedef}. For example, to
323 set up a German locale with name @code{de_DE}, simply issue the command
324 @samp{localedef -i de_DE -f ISO-8859-1 de_DE}. To configure all locales
325 that are supported by @theglibc{}, you can issue from your build directory the
326 command @samp{make localedata/install-locales}.
328 To configure the locally used timezone, set the @code{TZ} environment
329 variable. The script @code{tzselect} helps you to select the right value.
330 As an example, for Germany, @code{tzselect} would tell you to use
331 @samp{TZ='Europe/Berlin'}. For a system wide installation (the given
332 paths are for an installation with @samp{--prefix=/usr}), link the
333 timezone file which is in @file{/usr/share/zoneinfo} to the file
334 @file{/etc/localtime}. For Germany, you might execute @samp{ln -s
335 /usr/share/zoneinfo/Europe/Berlin /etc/localtime}.
337 @node Tools for Compilation
338 @appendixsec Recommended Tools for Compilation
339 @cindex installation tools
340 @cindex tools, for installing library
342 We recommend installing the following GNU tools before attempting to
347 GNU @code{make} 3.79 or newer
349 You need the latest version of GNU @code{make}. Modifying @theglibc{}
350 to work with other @code{make} programs would be so difficult that
351 we recommend you port GNU @code{make} instead. @strong{Really.} We
352 recommend GNU @code{make} version 3.79. All earlier versions have severe
353 bugs or lack features.
356 GCC 4.3 or newer, GCC 4.6 recommended
358 GCC 4.3 or higher is required; as of this writing, GCC 4.6 is the
359 compiler we advise to use to build @theglibc{}.
361 You can use whatever compiler you like to compile programs that use
364 Check the FAQ for any special compiler issues on particular platforms.
367 GNU @code{binutils} 2.20 or later
369 You must use GNU @code{binutils} (as and ld) to build @theglibc{}.
370 No other assembler or linker has the necessary functionality at the
374 GNU @code{texinfo} 4.5 or later
376 To correctly translate and install the Texinfo documentation you need
377 this version of the @code{texinfo} package. Earlier versions do not
378 understand all the tags used in the document, and the installation
379 mechanism for the info files is not present or works differently.
382 GNU @code{awk} 3.0, or higher
384 @code{Awk} is used in several places to generate files.
385 @code{gawk} 3.0 is known to work.
390 Perl is not required, but it is used if present to test the
391 installation. We may decide to use it elsewhere in the future.
394 GNU @code{sed} 3.02 or newer
396 @code{Sed} is used in several places to generate files. Most scripts work
397 with any version of @code{sed}. The known exception is the script
398 @code{po2test.sed} in the @code{intl} subdirectory which is used to
399 generate @code{msgs.h} for the test suite. This script works correctly
400 only with GNU @code{sed} 3.02. If you like to run the test suite, you
401 should definitely upgrade @code{sed}.
406 If you change any of the @file{configure.in} files you will also need
410 GNU @code{autoconf} 2.53 or higher
414 and if you change any of the message translation files you will need
418 GNU @code{gettext} 0.10.36 or later
422 You may also need these packages if you upgrade your source tree using
423 patches, although we try to avoid this.
426 @appendixsec Specific advice for @gnulinuxsystems{}
427 @cindex kernel header files
429 If you are installing @theglibc{} on @gnulinuxsystems{}, you need to have
430 the header files from a 2.6.19.1 or newer kernel around for reference.
431 These headers must be installed using @samp{make headers_install}; the
432 headers present in the kernel source directory are not suitable for
433 direct use by @theglibc{}. You do not need to use that kernel, just have
434 its headers installed where @theglibc{} can access them, referred to here as
435 @var{install-directory}. The easiest way to do this is to unpack it
436 in a directory such as @file{/usr/src/linux-@var{version}}. In that
437 directory, run @samp{make headers_install
438 INSTALL_HDR_PATH=@var{install-directory}}. Finally, configure @theglibc{}
439 with the option @samp{--with-headers=@var{install-directory}/include}.
440 Use the most recent kernel you can get your hands on. (If you are
441 cross-compiling @theglibc{}, you need to specify
442 @samp{ARCH=@var{architecture}} in the @samp{make headers_install}
443 command, where @var{architecture} is the architecture name used by the
444 Linux kernel, such as @samp{x86} or @samp{powerpc}.)
446 After installing @theglibc{}, you may need to remove or rename
447 directories such as @file{/usr/include/linux} and
448 @file{/usr/include/asm}, and replace them with copies of directories
449 such as @file{linux} and @file{asm} from
450 @file{@var{install-directory}/include}. All directories present in
451 @file{@var{install-directory}/include} should be copied, except that
452 @theglibc{} provides its own version of @file{/usr/include/scsi}; the
453 files provided by the kernel should be copied without replacing those
454 provided by @theglibc{}. The @file{linux}, @file{asm} and
455 @file{asm-generic} directories are required to compile programs using
456 @theglibc{}; the other directories describe interfaces to the kernel but
457 are not required if not compiling programs using those interfaces.
458 You do not need to copy kernel headers if you did not specify an
459 alternate kernel header source using @samp{--with-headers}.
461 The Filesystem Hierarchy Standard for @gnulinuxsystems{} expects some
462 components of the @glibcadj{} installation to be in
463 @file{/lib} and some in @file{/usr/lib}. This is handled automatically
464 if you configure @theglibc{} with @samp{--prefix=/usr}. If you set some other
465 prefix or allow it to default to @file{/usr/local}, then all the
466 components are installed there.
469 @appendixsec Reporting Bugs
470 @cindex reporting bugs
471 @cindex bugs, reporting
473 There are probably bugs in @theglibc{}. There are certainly
474 errors and omissions in this manual. If you report them, they will get
475 fixed. If you don't, no one will ever know about them and they will
476 remain unfixed for all eternity, if not longer.
478 It is a good idea to verify that the problem has not already been
479 reported. Bugs are documented in two places: The file @file{BUGS}
480 describes a number of well known bugs and the central @glibcadj{}
481 bug tracking system has a
483 @url{http://sourceware.org/bugzilla/}. The WWW
484 interface gives you access to open and closed reports. A closed report
485 normally includes a patch or a hint on solving the problem.
487 To report a bug, first you must find it. With any luck, this will be the
488 hard part. Once you've found a bug, make sure it's really a bug. A
489 good way to do this is to see if @theglibc{} behaves the same way
490 some other C library does. If so, probably you are wrong and the
491 libraries are right (but not necessarily). If not, one of the libraries
492 is probably wrong. It might not be @theglibc{}. Many historical
493 Unix C libraries permit things that we don't, such as closing a file
496 If you think you have found some way in which @theglibc{} does not
497 conform to the ISO and POSIX standards (@pxref{Standards and
498 Portability}), that is definitely a bug. Report it!
500 Once you're sure you've found a bug, try to narrow it down to the
501 smallest test case that reproduces the problem. In the case of a C
502 library, you really only need to narrow it down to one library
503 function call, if possible. This should not be too difficult.
505 The final step when you have a simple test case is to report the bug.
506 Do this at @value{REPORT_BUGS_TO}.
508 If you are not sure how a function should behave, and this manual
509 doesn't tell you, that's a bug in the manual. Report that too! If the
510 function's behavior disagrees with the manual, then either the library
511 or the manual has a bug, so report the disagreement. If you find any
512 errors or omissions in this manual, please report them to the
513 bug database. If you refer to specific
514 sections of the manual, please include the section names for easier