<refentry id="glib-cross-compiling" revision="8 Apr 2003">
<refmeta>
<refentrytitle>Cross-compiling the GLib package</refentrytitle>
<manvolnum>3</manvolnum>
<refmiscinfo>GLib Library</refmiscinfo>
</refmeta>
<refnamediv>
<refname>Cross-compiling the GLib Package</refname>
<refpurpose>
How to cross-compile GLib
</refpurpose>
</refnamediv>
<refsect1 id="cross">
<title>Building the Library for a different architecture</title>
<para>
Cross-compilation is the process of compiling a program or
library on a different architecture or operating system then
it will be run upon. GLib is slightly more difficult to
cross-compile than many packages because much of GLib is
about hiding differences between different systems.
</para>
<para>
These notes cover things specific to cross-compiling GLib;
for general information about cross-compilation, see the
<application>autoconf</application> info pages.
</para>
<para>
GLib tries to detect as much information as possible about
the target system by compiling and linking programs without
actually running anything; however, some information GLib
needs is not available this way. This information needs
to be provided to the configure script via a "cache file"
or by setting the cache variables in your environment.
</para>
<para>
As an example of using a cache file, to cross compile for
the "MingW32" Win32 runtine environment on a Linux system,
create a file 'win32.cache' with the following contents:
</para>
<programlisting>
glib_cv_long_long_format=ll
glib_cv_stack_grows=no
</programlisting>
<para>
Then execute the following commands:
</para>
<programlisting>
PATH=/path/to/mingw32-compiler/bin:$PATH
chmod a-w win32.cache # prevent configure from changing it
./configure --cache-file=win32.cache --host=mingw32
</programlisting>
<para>
The complete list of cache file variables follows. Most
of these won't need to be set in most cases.
</para>
</refsect1>
<refsect1 id="cache-file-variables">
<title>Cache file variables</title>
<formalpara>
<title>glib_cv_long_long_format=[ll/q/I64]</title>
<para>
Format used by <function>printf()</function> and
<function>scanf()</function> for 64 bit integers. "ll" is
the C99 standard, and what is used by the 'trio' library
that GLib builds if your <function>printf()</function> is
insufficiently capable.
Doesn't need to be set if you are compiling using trio.
</para>
</formalpara>
<formalpara>
<title>glib_cv_stack_grows=[yes/no]</title>
<para>
Whether the stack grows up or down. Most places will want "no",
A few architectures, such as PA-RISC need "yes".
</para>
</formalpara>
<formalpara>
<title>glib_cv_working_bcopy=[yes/no]</title>
<para>
Whether your <function>bcopy()</function> can handle overlapping
copies. Only needs to be set if you don't have
<function>memmove()</function>. (Very unlikely)
</para>
</formalpara>
<formalpara>
<title>glib_cv_sane_realloc=[yes/np]</title>
<para>
Whether your <function>realloc()</function> conforms to ANSI C
and can handle <literal>NULL</literal> as the first argument.
Defaults to "yes" and probably doesn't need to be set.
</para>
</formalpara>
<formalpara>
<title>glib_cv_have_strlcpy=[yes/no]</title>
<para>
Whether you have <function>strlcpy()</function> that matches
OpenBSD. Defaults to "no", which is safe, since GLib uses a
built-in version in that case.
</para>
</formalpara>
<formalpara>
<title>glib_cv_va_val_copy=[yes/no]</title>
<para>
Whether <type>va_list</type> can be copied as a pointer. If set
to "no", then <function>memcopy()</function> will be used. Only
matters if you don't have <function>va_copy()</function> or
<function>__va_copy()</function>. (So, doesn't matter for GCC.)
Defaults to "yes" which is slightly more common than "no".
</para>
</formalpara>
<formalpara>
<title>glib_cv_rtldglobal_broken=[yes/no]</title>
<para>
Whether you have a bug found in OSF/1 v5.0. Defaults to "no".
</para>
</formalpara>
<formalpara>
<title>glib_cv_uscore=[yes/no]</title>
<para>
Whether an underscore needs to be prepended to symbols when
looking them up via <function>dlsym()</function>. Only needs to
be set if your system uses
<function>dlopen()</function>/<function>dlsym()</function>.
</para>
</formalpara>
<formalpara>
<title>ac_cv_func_posix_getpwuid_r=[yes/no]</title>
<para>
Whether you have a getpwuid_r function (in your C library,
not your thread library) that conforms to the POSIX spec.
(Takes a 'struct passwd **' as the final argument)
</para>
</formalpara>
<formalpara>
<title>ac_cv_func_nonposix_getpwuid_r=[yes/no]</title>
<para>
Whether you have some variant of <function>getpwuid_r()</function>
that doesn't conform to to the POSIX spec, but GLib might be able
to use (or might segfault.) Only needs to be set if
<literal>ac_cv_func_posix_getpwuid_r</literal> is not set. It's
safest to set this to "no".
</para>
</formalpara>
<formalpara>
<title>glib_cv_use_pid_surrogate=[yes/no]</title>
<para>
Whether to use a <function>setpriority()</function> on the PID of
the thread as a method for setting the priority of threads. This
only needs to be set when using POSIX threads.
</para>
</formalpara>
<formalpara>
<title>ac_cv_func_printf_unix98=[yes/no]</title>
<para>
Whether your <function>printf()</function> family supports Unix98
style <literal>%N$</literal> positional parameters. Defaults to
"no".
</para>
</formalpara>
<formalpara>
<title>ac_cv_func_vsnprintf_c99=[yes/no]</title>
<para>
Whether you have a <function>vsnprintf()</function> with C99
semantics. (C99 semantics means returning the number of bytes
that would have been written had the output buffer had enough
space.) Defaults to "no".
</para>
</formalpara>
</refsect1>
</refentry>