1 | <sect1 id="ch-tools-gcc-pass2">
|
---|
2 | <title>Installing GCC-&gcc-version; - Pass 2</title>
|
---|
3 | <?dbhtml filename="gcc-pass2.html" dir="chapter05"?>
|
---|
4 |
|
---|
5 | <screen>&buildtime; &gcc-time-tools-pass2;
|
---|
6 | &diskspace; &gcc-compsize-tools-pass2;</screen>
|
---|
7 |
|
---|
8 | &aa-gcc-down;
|
---|
9 | &aa-gcc-dep;
|
---|
10 |
|
---|
11 | <sect2><title> </title><para> </para></sect2>
|
---|
12 |
|
---|
13 | <sect2>
|
---|
14 | <title>Re-installation of GCC</title>
|
---|
15 |
|
---|
16 | <para>The tools required to test GCC and Binutils are installed now: Tcl,
|
---|
17 | Expect and DejaGnu. Therefore we can now rebuild GCC and Binutils, linking
|
---|
18 | them against the new Glibc, and test them properly (if running the test suites
|
---|
19 | in this chapter). One thing to note, however, is that these test suites are
|
---|
20 | highly dependent on properly functioning pseudo terminals (PTYs) which are
|
---|
21 | provided by your host. These days, PTYs are most commonly implemented via the
|
---|
22 | <emphasis>devpts</emphasis> file system. You can quickly check if your host
|
---|
23 | system is set up correctly in this regard by performing a simple test:</para>
|
---|
24 |
|
---|
25 | <screen><userinput>expect -c "spawn ls"</userinput></screen>
|
---|
26 |
|
---|
27 | <para>The response might be:</para>
|
---|
28 |
|
---|
29 | <blockquote><screen>The system has no more ptys. Ask your system administrator to create more.</screen></blockquote>
|
---|
30 |
|
---|
31 | <para>If you receive the above message, your host doesn't have its PTYs set up
|
---|
32 | properly. In this case there is no point in running the test suites for GCC
|
---|
33 | and Binutils until you are able to resolve the issue. You can consult the LFS
|
---|
34 | Wiki at <ulink url="&wiki-root;"/> for more information on how to get PTYs
|
---|
35 | working.</para>
|
---|
36 |
|
---|
37 | <para>This time we will build both the C and the C++ compilers, so you'll have
|
---|
38 | to unpack both the core and the g++ tarballs (and testsuite too, if you want to
|
---|
39 | run the tests). Unpacking them in your working directory, they will all unfold
|
---|
40 | into a single <filename>&gcc-dir;/</filename> subdirectory.</para>
|
---|
41 |
|
---|
42 | <para>First correct a problem and make an essential adjustment:</para>
|
---|
43 |
|
---|
44 | <screen><userinput>patch -Np1 -i ../&gcc-nofixincludes-patch;
|
---|
45 | patch -Np1 -i ../&gcc-specs-patch;</userinput></screen>
|
---|
46 |
|
---|
47 | <para>The first patch disables the GCC "fixincludes" script. We mentioned this
|
---|
48 | briefly earlier, but a slightly more in-depth explanation of the fixincludes
|
---|
49 | process is warranted here. Under normal circumstances, the GCC fixincludes
|
---|
50 | script scans your system for header files that need to be fixed. It might find
|
---|
51 | that some Glibc header files on your host system need to be fixed, fix them and
|
---|
52 | put them in the GCC private include directory. Then, later on in
|
---|
53 | <xref linkend="chapter-building-system"/>, after we've installed the newer Glibc, this
|
---|
54 | private include directory would be searched before the system include
|
---|
55 | directory, resulting in GCC finding the fixed headers from the host system,
|
---|
56 | which would most likely not match the Glibc version actually used for the LFS
|
---|
57 | system.</para>
|
---|
58 |
|
---|
59 | <para>The second patch changes GCC's default location of the dynamic linker
|
---|
60 | (typically <filename>ld-linux.so.2</filename>). It also removes
|
---|
61 | <filename class="directory">/usr/include</filename> from GCC's include search
|
---|
62 | path. Patching now rather than adjusting the specs file after installation
|
---|
63 | ensures that our new dynamic linker gets used during the actual build of GCC.
|
---|
64 | That is, all the final (and temporary) binaries created during the build will
|
---|
65 | link against the new Glibc.</para>
|
---|
66 |
|
---|
67 | <important><para>The above patches are <emphasis>critical</emphasis> in ensuring
|
---|
68 | a successful overall build. Do not forget to apply them.</para></important>
|
---|
69 |
|
---|
70 | <para>Create a separate build directory again:</para>
|
---|
71 |
|
---|
72 | <screen><userinput>mkdir ../gcc-build
|
---|
73 | cd ../gcc-build</userinput></screen>
|
---|
74 |
|
---|
75 | <para>Before starting to build GCC, remember to unset any environment
|
---|
76 | variables that override the default optimization flags.</para>
|
---|
77 |
|
---|
78 | <para>Now prepare GCC for compilation:</para>
|
---|
79 |
|
---|
80 | <screen><userinput>../&gcc-dir;/configure --prefix=/tools \
|
---|
81 | --with-local-prefix=/tools \
|
---|
82 | --enable-clocale=gnu --enable-shared \
|
---|
83 | --enable-threads=posix --enable-__cxa_atexit \
|
---|
84 | --enable-languages=c,c++</userinput></screen>
|
---|
85 |
|
---|
86 | <para>The meaning of the new configure options:</para>
|
---|
87 |
|
---|
88 | <itemizedlist>
|
---|
89 | <listitem><para><userinput>--enable-clocale=gnu</userinput>: This option
|
---|
90 | ensures the correct locale model is selected for the C++ libraries under all
|
---|
91 | circumstances. If the configure script finds the <emphasis>de_DE</emphasis>
|
---|
92 | locale installed, it will select the correct <emphasis>gnu</emphasis> locale
|
---|
93 | model. However, people who don't install the <emphasis>de_DE</emphasis> locale
|
---|
94 | would run the risk of building ABI incompatible C++ libraries due to the wrong
|
---|
95 | <emphasis>generic</emphasis> locale model being selected.</para></listitem>
|
---|
96 |
|
---|
97 | <listitem><para><userinput>--enable-threads=posix</userinput>: This enables
|
---|
98 | C++ exception handling for multi-threaded code.</para></listitem>
|
---|
99 |
|
---|
100 | <listitem><para><userinput>--enable-__cxa_atexit</userinput>: This option
|
---|
101 | allows use of __cxa_atexit, rather than atexit, to register C++ destructors for
|
---|
102 | local statics and global objects and is essential for fully standards-compliant
|
---|
103 | handling of destructors. It also affects the C++ ABI and therefore results in
|
---|
104 | C++ shared libraries and C++ programs that are interoperable with other Linux
|
---|
105 | distributions.</para></listitem>
|
---|
106 |
|
---|
107 | <listitem><para><userinput>--enable-languages=c,c++</userinput>: This option
|
---|
108 | ensures that both the C and C++ compilers are built.</para></listitem>
|
---|
109 | </itemizedlist>
|
---|
110 |
|
---|
111 | <para>Compile the package:</para>
|
---|
112 |
|
---|
113 | <screen><userinput>make</userinput></screen>
|
---|
114 |
|
---|
115 | <para>There is no need to use the <emphasis>bootstrap</emphasis> target now,
|
---|
116 | as the compiler we're using to compile this GCC was built from the exact same
|
---|
117 | version of the GCC sources we used earlier.</para>
|
---|
118 |
|
---|
119 | <para>Compilation is now complete. As mentioned earlier, we don't recommend
|
---|
120 | running the test suites for the temporary tools here in this chapter. If you
|
---|
121 | still want to run the GCC test suite anyway, the following command will do
|
---|
122 | so:</para>
|
---|
123 |
|
---|
124 | <screen><userinput>make -k check</userinput></screen>
|
---|
125 |
|
---|
126 | <para>The <emphasis>-k</emphasis> flag is used to make the test suite run
|
---|
127 | through to completion and not stop at the first failure. The GCC test suite is
|
---|
128 | very comprehensive and is almost guaranteed to generate a few failures. To get
|
---|
129 | a summary of the test suite results, run this:</para>
|
---|
130 |
|
---|
131 | <screen><userinput>../&gcc-dir;/contrib/test_summary</userinput></screen>
|
---|
132 |
|
---|
133 | <para>(For just the summaries, pipe the output through
|
---|
134 | <userinput>grep -A7 Summ</userinput>.)</para>
|
---|
135 |
|
---|
136 | <para>You can compare your results to those posted to the gcc-testresults
|
---|
137 | mailing list for similar configurations to your own. For an example of how
|
---|
138 | current GCC-&gcc-version; should look on i686-pc-linux-gnu, see
|
---|
139 | <ulink url="http://gcc.gnu.org/ml/gcc-testresults/2004-01/msg00826.html"/>.</para>
|
---|
140 |
|
---|
141 | <para>Note that the results contain:</para>
|
---|
142 |
|
---|
143 | <screen>* 1 XPASS (unexpected pass) for g++
|
---|
144 | * 1 FAIL (unexpected failure) for gcc
|
---|
145 | * 24 XPASS's for libstdc++</screen>
|
---|
146 |
|
---|
147 | <para>The unexpected pass for g++ is due to the use of
|
---|
148 | <emphasis>--enable-__cxa_atexit</emphasis>. Apparently not all platforms
|
---|
149 | supported by GCC have support for "__cxa_atexit" in their C libraries, so this
|
---|
150 | test is not always expected to pass.</para>
|
---|
151 |
|
---|
152 | <para>The 24 unexpected passes for libstdc++ are due to the use of
|
---|
153 | <emphasis>--enable-clocale=gnu</emphasis>. This option, which is the correct
|
---|
154 | choice on Glibc-based systems of versions 2.2.5 and above, enables in the GNU C
|
---|
155 | library a locale support that is superior to the otherwise selected
|
---|
156 | <emphasis>generic</emphasis> model (which may be applicable if for instance you
|
---|
157 | were using Newlibc, Sun-libc or whatever other libc). The libstdc++ test suite
|
---|
158 | is apparently expecting the <emphasis>generic</emphasis> model, hence those
|
---|
159 | tests are not always expected to pass.</para>
|
---|
160 |
|
---|
161 | <para>Having a few unexpected failures often cannot be avoided. The GCC
|
---|
162 | developers are usually aware of these, but haven't yet gotten around to fixing
|
---|
163 | them. In short, unless your results are vastly different from those at the
|
---|
164 | above URL, it is safe to continue.</para>
|
---|
165 |
|
---|
166 | <para>And finally install the package:</para>
|
---|
167 |
|
---|
168 | <screen><userinput>make install</userinput></screen>
|
---|
169 |
|
---|
170 | <note><para>At this point it is strongly recommended to repeat the sanity check
|
---|
171 | we performed earlier in this chapter. Refer back to
|
---|
172 | <xref linkend="ch-tools-adjusting"/> and repeat the little test compilation. If
|
---|
173 | the result is wrong, then most likely you forgot to apply the above mentioned
|
---|
174 | GCC Specs patch.</para></note>
|
---|
175 |
|
---|
176 | </sect2>
|
---|
177 |
|
---|
178 | <sect2><title> </title><para> </para>
|
---|
179 | <para>The details on this package are found in <xref linkend="contents-gcc"/>.</para>
|
---|
180 | <para> </para></sect2>
|
---|
181 |
|
---|
182 | </sect1>
|
---|
183 |
|
---|