[bc82645e] | 1 | <sect2><title> </title><para> </para></sect2>
|
---|
| 2 |
|
---|
| 3 | <sect2>
|
---|
[ab6f1859] | 4 | <title>Re-installation of GCC</title>
|
---|
| 5 |
|
---|
[990ba6b] | 6 | <para>The tools required to test GCC and Binutils are installed now (Tcl, Expect
|
---|
| 7 | and DejaGnu). We can continue on rebuilding GCC and Binutils, link them against
|
---|
| 8 | the new Glibc, and test them properly. One thing to note, however, is that these
|
---|
| 9 | test suites are highly dependent on properly functioning pseudo terminals (PTYs)
|
---|
| 10 | which are provided by your host distribution. These days, PTYs are most commonly
|
---|
| 11 | implemented via the <emphasis>devpts</emphasis> file system. You can quickly
|
---|
| 12 | check if your host system is set up correctly in this regard by performing a
|
---|
| 13 | simple test:</para>
|
---|
| 14 |
|
---|
[21ba4e3] | 15 | <screen><userinput>expect -c "spawn ls"</userinput></screen>
|
---|
[990ba6b] | 16 |
|
---|
| 17 | <para>If you receive the message:</para>
|
---|
| 18 |
|
---|
[7e602eab] | 19 | <blockquote><screen>The system has no more ptys. Ask your system administrator to create more.</screen></blockquote>
|
---|
[990ba6b] | 20 |
|
---|
| 21 | <para>Your host distribution is not set up for proper PTY operation. In this
|
---|
| 22 | case there is no point in running the test suites for GCC and Binutils until you
|
---|
| 23 | are able to resolve the issue. You can consult the LFS Wiki at
|
---|
[35c2fa60] | 24 | <ulink url="http://wiki.linuxfromscratch.org/"/> for more information on how to
|
---|
[990ba6b] | 25 | get PTYs working.</para>
|
---|
[62d0130] | 26 |
|
---|
[aa49729] | 27 | <para>Unpack all three GCC tarballs (-core, -g++, and -testsuite) in one and the
|
---|
| 28 | same working directory. They will all unfold into a single
|
---|
| 29 | <filename>gcc-&gcc-version;/</filename> subdirectory.</para>
|
---|
[bc82645e] | 30 |
|
---|
[40b88c2] | 31 | <para>First correct one problem and make an essential adjustment:</para>
|
---|
[775a268] | 32 |
|
---|
[21ba4e3] | 33 | <screen><userinput>patch -Np1 -i ../&gcc-nofixincludes-patch;
|
---|
| 34 | patch -Np1 -i ../&gcc-specs-patch;</userinput></screen>
|
---|
[775a268] | 35 |
|
---|
[be2dd91] | 36 | <para>The first patch disables the GCC "fixincludes" script. We mentioned this
|
---|
| 37 | briefly earlier, but a slightly more in-depth explanation of the fixincludes
|
---|
| 38 | process is warranted here. Under normal circumstances, the GCC fixincludes
|
---|
| 39 | script scans your system for header files that need to be fixed. It might find
|
---|
[1cb7ced] | 40 | that some Glibc header files on your host system need to be fixed, fix them and
|
---|
[be2dd91] | 41 | put them in the GCC private include directory. Then, later on in Chapter 6,
|
---|
| 42 | after we've installed the newer Glibc, this private include directory would be
|
---|
| 43 | searched before the system include directory, resulting in GCC finding the
|
---|
| 44 | fixed headers from the host system, which would most likely not match the Glibc
|
---|
| 45 | version actually used for the LFS system.</para>
|
---|
| 46 |
|
---|
[dc19fb6] | 47 | <para>The last patch changes GCC's default location of the dynamic linker
|
---|
[76d113f] | 48 | (typically <filename>ld-linux.so.2</filename>). It also removes
|
---|
| 49 | <filename class="directory">/usr/include</filename> from GCC's include search
|
---|
| 50 | path. Patching now rather than adjusting the specs file after installation
|
---|
| 51 | ensures that our new dynamic linker gets used during the actual build of GCC.
|
---|
| 52 | That is, all the final (and temporary) binaries created during the build will
|
---|
| 53 | link against the new Glibc.</para>
|
---|
| 54 |
|
---|
| 55 | <important><para>These patches are <emphasis>critical</emphasis> in ensuring a
|
---|
| 56 | successful overall build. Do not forget to apply them.</para></important>
|
---|
[775a268] | 57 |
|
---|
| 58 | <para>Create a separate build directory again:</para>
|
---|
[bc82645e] | 59 |
|
---|
[21ba4e3] | 60 | <screen><userinput>mkdir ../gcc-build
|
---|
| 61 | cd ../gcc-build</userinput></screen>
|
---|
[bc82645e] | 62 |
|
---|
[ab6f1859] | 63 | <para>Before starting to build GCC, remember to unset any environment
|
---|
| 64 | variables that override the default optimization flags.</para>
|
---|
| 65 |
|
---|
[775a268] | 66 | <para>Now prepare GCC to be compiled:</para>
|
---|
| 67 |
|
---|
[21ba4e3] | 68 | <screen><userinput>../gcc-&gcc-version;/configure --prefix=/tools \
|
---|
[148bb04] | 69 | --with-local-prefix=/tools \
|
---|
[6b2f69c] | 70 | --enable-clocale=gnu --enable-shared \
|
---|
[0939e573] | 71 | --enable-threads=posix --enable-__cxa_atexit \
|
---|
[21ba4e3] | 72 | --enable-languages=c,c++</userinput></screen>
|
---|
[775a268] | 73 |
|
---|
[aa49729] | 74 | <para>The meaning of the new configure options:</para>
|
---|
| 75 |
|
---|
| 76 | <itemizedlist>
|
---|
| 77 | <listitem><para><userinput>--enable-threads=posix</userinput>: This enables
|
---|
| 78 | C++ exception handling for multi-threaded code.</para></listitem>
|
---|
| 79 |
|
---|
| 80 | <listitem><para><userinput>--enable-__cxa_atexit</userinput>: This option
|
---|
| 81 | allows use of __cxa_atexit, rather than atexit, to register C++ destructors for
|
---|
| 82 | local statics and global objects and is essential for fully standards-compliant
|
---|
| 83 | handling of destructors. It also affects the C++ ABI and therefore results in
|
---|
| 84 | C++ shared libraries and C++ programs that are interoperable with other Linux
|
---|
| 85 | distributions.</para></listitem>
|
---|
| 86 |
|
---|
| 87 | <listitem><para><userinput>--enable-clocale=gnu</userinput>: This option ensures
|
---|
| 88 | the correct locale model is selected for the C++ libraries under all
|
---|
| 89 | circumstances. If the configure script finds the <emphasis>de_DE</emphasis>
|
---|
| 90 | locale installed, it will select the correct model of <emphasis>gnu</emphasis>.
|
---|
| 91 | However, people who don't install the <emphasis>de_DE</emphasis> locale, run the
|
---|
| 92 | risk of building ABI incompatible C++ libraries due to the wrong locale model of
|
---|
| 93 | <emphasis>generic</emphasis> being selected.</para></listitem>
|
---|
| 94 |
|
---|
| 95 | <listitem><para><userinput>--enable-languages=c,c++</userinput>: This option is
|
---|
| 96 | needed to ensure that both C and C++ compilers are built.</para></listitem>
|
---|
| 97 | </itemizedlist>
|
---|
| 98 |
|
---|
[775a268] | 99 | <para>Compile the package:</para>
|
---|
| 100 |
|
---|
[21ba4e3] | 101 | <screen><userinput>make</userinput></screen>
|
---|
[ccaabde] | 102 |
|
---|
[775a268] | 103 | <para>There is no need to use the <userinput>bootstrap</userinput> target now,
|
---|
[ccaabde] | 104 | as the compiler we're using to compile this GCC was built from the exact same
|
---|
| 105 | version of the GCC sources we used earlier.</para>
|
---|
| 106 |
|
---|
[1668d8e] | 107 | <note><para>It's worth pointing out that running the GCC test suite here
|
---|
[1fb160c] | 108 | is considered not as important as running it in Chapter 6.</para></note>
|
---|
[1668d8e] | 109 |
|
---|
[ab6f1859] | 110 | <para>Test the results:</para>
|
---|
[775a268] | 111 |
|
---|
[21ba4e3] | 112 | <screen><userinput>make -k check</userinput></screen>
|
---|
[775a268] | 113 |
|
---|
| 114 | <para>The <userinput>-k</userinput> flag is used to make the test suite run
|
---|
| 115 | through to completion and not stop at the first failure. The GCC test suite is
|
---|
| 116 | very comprehensive and is almost guaranteed to generate a few failures. To get
|
---|
| 117 | a summary of the test suite results, run this:</para>
|
---|
| 118 |
|
---|
[21ba4e3] | 119 | <screen><userinput>../gcc-&gcc-version;/contrib/test_summary | less</userinput></screen>
|
---|
[775a268] | 120 |
|
---|
| 121 | <para>You can compare your results to those posted to the gcc-testresults
|
---|
| 122 | mailing list for similar configurations to your own. For an example of how
|
---|
[bef49bf] | 123 | current GCC-3.3.1 should look on i686-pc-linux-gnu, see
|
---|
| 124 | <ulink url="http://gcc.gnu.org/ml/gcc-testresults/2003-08/msg01612.html"/>.</para>
|
---|
[775a268] | 125 |
|
---|
| 126 | <para>Note that the results contain:</para>
|
---|
| 127 |
|
---|
| 128 | <screen>* 1 XPASS (unexpected pass) for g++
|
---|
[bd10d48] | 129 | * 1 FAIL (unexpected failure) for g++
|
---|
[9e24a7d] | 130 | * 2 FAIL for gcc
|
---|
| 131 | * 26 XPASS's for libstdc++</screen>
|
---|
[775a268] | 132 |
|
---|
| 133 | <para>The unexpected pass for g++ is due to the use of
|
---|
| 134 | <userinput>--enable-__cxa_atexit</userinput>. Apparently not all platforms
|
---|
| 135 | supported by GCC have support for "__cxa_atexit" in their C libraries, so this
|
---|
| 136 | test is not always expected to pass.</para>
|
---|
| 137 |
|
---|
| 138 | <para>The 26 unexpected passes for libstdc++ are due to the use of
|
---|
| 139 | <userinput>--enable-clocale=gnu</userinput>, which is the correct choice on
|
---|
| 140 | Glibc-based systems of versions 2.2.5 and above. The underlying locale support
|
---|
| 141 | in the GNU C library is superior to that of the otherwise selected "generic"
|
---|
| 142 | model (which may be applicable if for instance you were using Newlibc, Sun-libc
|
---|
| 143 | or whatever libc). The libstdc++ test suite is apparently expecting the
|
---|
| 144 | "generic" model, hence those tests are not always expected to pass.</para>
|
---|
| 145 |
|
---|
[bd10d48] | 146 | <para>Unexpected failures often cannot be avoided. The GCC developers are
|
---|
| 147 | usually aware of them but haven't yet gotten around to fixing them. In short,
|
---|
| 148 | unless your results are vastly different from those at the above URL, it is safe
|
---|
| 149 | to continue on.</para>
|
---|
| 150 |
|
---|
[775a268] | 151 | <para>And finally install the package:</para>
|
---|
| 152 |
|
---|
[21ba4e3] | 153 | <screen><userinput>make install</userinput></screen>
|
---|
[bc82645e] | 154 |
|
---|
[076ddfe] | 155 | <note><para>At this point it is strongly recommended to repeat the sanity check
|
---|
[f861982] | 156 | we performed earlier in the chapter. Refer back to
|
---|
| 157 | <xref linkend="ch05-locking-glibc"/> and repeat the check. If the results are
|
---|
[1668d8e] | 158 | wrong, then most likely you forgot to apply the above mentioned GCC Specs
|
---|
[f861982] | 159 | patch.</para></note>
|
---|
[076ddfe] | 160 |
|
---|
[bc82645e] | 161 | </sect2>
|
---|
| 162 |
|
---|