lfs-book/chapter05/gcc-pass2-inst.xml

<sect2><title>&nbsp;</title><para>&nbsp;</para></sect2>

<sect2>
<title>Re-installation of GCC</title>

<para>Unpack all three GCC tarballs in one and the same working directory.
They will all unfold into a single <filename>gcc-&gcc-version;/</filename>
subdir.</para>

<para>First correct one problem and make an essential adjustment:</para>

<para><screen><userinput>patch -Np1 -i ../gcc-&gcc-version;-no_fixincludes-2.patch
patch -Np1 -i ../gcc-&gcc-specs-version;.patch</userinput></screen></para>

<para>The first patch disables the GCC "fixincludes" script.  We mentioned this
briefly earlier, but a slightly more in-depth explanation of the fixincludes 
process is warranted here.  Under normal circumstances, the GCC fixincludes
script scans your system for header files that need to be fixed.  It might find
that the Glibc header files on your host system need to be fixed, fix them and
put them in the GCC private include directory.  Then, later on in Chapter 6,
after we've installed the newer Glibc, this private include directory would be
searched before the system include directory, resulting in GCC finding the
fixed headers from the host system, which would most likely not match the Glibc
version actually used for the LFS system.</para>

<para>The last patch changes GCC's default location of the dynamic linker,
a simple substitution of "/lib/ld-linux.so.2" with "/tools/lib/ld-linux.so.2"
in <filename>config/i386/linux.h</filename>. Patching now rather than adjusting
the specs file after installation ensures that our new dynamic linker gets
used during the actual build of GCC. That is, all the final (and temporary)
binaries created during the build will link against the new Glibc.</para>

<para>Create a separate build directory again:</para>

<para><screen><userinput>mkdir ../gcc-build
cd ../gcc-build</userinput></screen></para>

<para>Before starting to build GCC, remember to unset any environment
variables that override the default optimization flags.</para>

<para>Now prepare GCC to be compiled:</para>

<para><screen><userinput>../gcc-&gcc-version;/configure --prefix=/tools \
&nbsp;&nbsp;&nbsp;&nbsp;--with-local-prefix=/tools \
&nbsp;&nbsp;&nbsp;&nbsp;--enable-clocale=gnu --enable-shared \
&nbsp;&nbsp;&nbsp;&nbsp;--enable-threads=posix --enable-__cxa_atexit \
&nbsp;&nbsp;&nbsp;&nbsp;--enable-languages=c,c++</userinput></screen></para>

<para>Compile the package:</para>

<para><screen><userinput>make</userinput></screen></para>
<para>There is no need to use the <userinput>bootstrap</userinput> target now,
as the compiler we're using to compile this GCC has been built from the exact
same sources.</para>

<para>Test the results:</para>

<para><screen><userinput>make -k check</userinput></screen></para>

<para>The <userinput>-k</userinput> flag is used to make the test suite run
through to completion and not stop at the first failure. The GCC test suite is
very comprehensive and is almost guaranteed to generate a few failures. To get
a summary of the test suite results, run this:</para>
 
<para><screen><userinput>../gcc-3*/contrib/test_summary | less</userinput></screen></para>

<para>You can compare your results to those posted to the gcc-testresults
mailing list for similar configurations to your own. For an example of how
current GCC-3.3.1 should look on i686-pc-linux-gnu, see
<ulink url="http://gcc.gnu.org/ml/gcc-testresults/2003-08/msg01612.html"/>.</para>

<para>Note that the results contain:</para>

<screen>* 1 XPASS (unexpected pass) for g++
* 26 XPASS's for libstdc++
* 1 FAIL for libstdc++</screen>
 
<para>The unexpected pass for g++ is due to the use of
<userinput>--enable-__cxa_atexit</userinput>. Apparently not all platforms
supported by GCC have support for "__cxa_atexit" in their C libraries, so this
test is not always expected to pass.</para>

<para>The 26 unexpected passes for libstdc++ are due to the use of
<userinput>--enable-clocale=gnu</userinput>, which is the correct choice on
Glibc-based systems of versions 2.2.5 and above. The underlying locale support
in the GNU C library is superior to that of the otherwise selected "generic"
model (which may be applicable if for instance you were using Newlibc, Sun-libc
or whatever libc). The libstdc++ test suite is apparently expecting the
"generic" model, hence those tests are not always expected to pass.</para>

<para>The failure for libstdc++ is in
<filename>26_numerics/c99_classification_macros_c.cc</filename> and is a
long-standing known failure (since at least January 2002) that the developers
are apparently unable to easily fix.</para>

<para>And finally install the package:</para>                                              

<para><screen><userinput>make install</userinput></screen></para>

</sect2>