| 1 |
@c Copyright (C) 2001-2022 Free Software Foundation, Inc. |
| 2 |
@c This is part of the GCC manual. |
| 3 |
@c For copying conditions, see the file gcc.texi. |
| 4 |
|
| 5 |
@node Makefile |
| 6 |
@subsection Makefile Targets |
| 7 |
@cindex makefile targets |
| 8 |
@cindex targets, makefile |
| 9 |
|
| 10 |
These targets are available from the @samp{gcc} directory: |
| 11 |
|
| 12 |
@table @code |
| 13 |
@item all |
| 14 |
This is the default target. Depending on what your build/host/target |
| 15 |
configuration is, it coordinates all the things that need to be built. |
| 16 |
|
| 17 |
@item doc |
| 18 |
Produce info-formatted documentation and man pages. Essentially it |
| 19 |
calls @samp{make man} and @samp{make info}. |
| 20 |
|
| 21 |
@item dvi |
| 22 |
Produce DVI-formatted documentation. |
| 23 |
|
| 24 |
@item pdf |
| 25 |
Produce PDF-formatted documentation. |
| 26 |
|
| 27 |
@item html |
| 28 |
Produce HTML-formatted documentation. |
| 29 |
|
| 30 |
@item man |
| 31 |
Generate man pages. |
| 32 |
|
| 33 |
@item info |
| 34 |
Generate info-formatted pages. |
| 35 |
|
| 36 |
@item mostlyclean |
| 37 |
Delete the files made while building the compiler. |
| 38 |
|
| 39 |
@item clean |
| 40 |
That, and all the other files built by @samp{make all}. |
| 41 |
|
| 42 |
@item distclean |
| 43 |
That, and all the files created by @command{configure}. |
| 44 |
|
| 45 |
@item maintainer-clean |
| 46 |
Distclean plus any file that can be generated from other files. Note |
| 47 |
that additional tools may be required beyond what is normally needed to |
| 48 |
build GCC. |
| 49 |
|
| 50 |
@item srcextra |
| 51 |
Generates files in the source directory that are not version-controlled but |
| 52 |
should go into a release tarball. |
| 53 |
|
| 54 |
@item srcinfo |
| 55 |
@itemx srcman |
| 56 |
Copies the info-formatted and manpage documentation into the source |
| 57 |
directory usually for the purpose of generating a release tarball. |
| 58 |
|
| 59 |
@item install |
| 60 |
Installs GCC. |
| 61 |
|
| 62 |
@item uninstall |
| 63 |
Deletes installed files, though this is not supported. |
| 64 |
|
| 65 |
@item check |
| 66 |
Run the testsuite. This creates a @file{testsuite} subdirectory that |
| 67 |
has various @file{.sum} and @file{.log} files containing the results of |
| 68 |
the testing. You can run subsets with, for example, @samp{make check-gcc}. |
| 69 |
You can specify specific tests by setting @env{RUNTESTFLAGS} to be the name |
| 70 |
of the @file{.exp} file, optionally followed by (for some tests) an equals |
| 71 |
and a file wildcard, like: |
| 72 |
|
| 73 |
@smallexample |
| 74 |
make check-gcc RUNTESTFLAGS="execute.exp=19980413-*" |
| 75 |
@end smallexample |
| 76 |
|
| 77 |
Note that running the testsuite may require additional tools be |
| 78 |
installed, such as Tcl or DejaGnu. |
| 79 |
@end table |
| 80 |
|
| 81 |
The toplevel tree from which you start GCC compilation is not |
| 82 |
the GCC directory, but rather a complex Makefile that coordinates |
| 83 |
the various steps of the build, including bootstrapping the compiler |
| 84 |
and using the new compiler to build target libraries. |
| 85 |
|
| 86 |
When GCC is configured for a native configuration, the default action |
| 87 |
for @command{make} is to do a full three-stage bootstrap. This means |
| 88 |
that GCC is built three times---once with the native compiler, once with |
| 89 |
the native-built compiler it just built, and once with the compiler it |
| 90 |
built the second time. In theory, the last two should produce the same |
| 91 |
results, which @samp{make compare} can check. Each stage is configured |
| 92 |
separately and compiled into a separate directory, to minimize problems |
| 93 |
due to ABI incompatibilities between the native compiler and GCC. |
| 94 |
|
| 95 |
If you do a change, rebuilding will also start from the first stage |
| 96 |
and ``bubble'' up the change through the three stages. Each stage |
| 97 |
is taken from its build directory (if it had been built previously), |
| 98 |
rebuilt, and copied to its subdirectory. This will allow you to, for |
| 99 |
example, continue a bootstrap after fixing a bug which causes the |
| 100 |
stage2 build to crash. It does not provide as good coverage of the |
| 101 |
compiler as bootstrapping from scratch, but it ensures that the new |
| 102 |
code is syntactically correct (e.g., that you did not use GCC extensions |
| 103 |
by mistake), and avoids spurious bootstrap comparison |
| 104 |
failures@footnote{Except if the compiler was buggy and miscompiled |
| 105 |
some of the files that were not modified. In this case, it's best |
| 106 |
to use @command{make restrap}.}. |
| 107 |
|
| 108 |
Other targets available from the top level include: |
| 109 |
|
| 110 |
@table @code |
| 111 |
@item bootstrap-lean |
| 112 |
Like @code{bootstrap}, except that the various stages are removed once |
| 113 |
they're no longer needed. This saves disk space. |
| 114 |
|
| 115 |
@item bootstrap2 |
| 116 |
@itemx bootstrap2-lean |
| 117 |
Performs only the first two stages of bootstrap. Unlike a three-stage |
| 118 |
bootstrap, this does not perform a comparison to test that the compiler |
| 119 |
is running properly. Note that the disk space required by a ``lean'' |
| 120 |
bootstrap is approximately independent of the number of stages. |
| 121 |
|
| 122 |
@item stage@var{N}-bubble (@var{N} = 1@dots{}4, profile, feedback) |
| 123 |
Rebuild all the stages up to @var{N}, with the appropriate flags, |
| 124 |
``bubbling'' the changes as described above. |
| 125 |
|
| 126 |
@item all-stage@var{N} (@var{N} = 1@dots{}4, profile, feedback) |
| 127 |
Assuming that stage @var{N} has already been built, rebuild it with the |
| 128 |
appropriate flags. This is rarely needed. |
| 129 |
|
| 130 |
@item cleanstrap |
| 131 |
Remove everything (@samp{make clean}) and rebuilds (@samp{make bootstrap}). |
| 132 |
|
| 133 |
@item compare |
| 134 |
Compares the results of stages 2 and 3. This ensures that the compiler |
| 135 |
is running properly, since it should produce the same object files |
| 136 |
regardless of how it itself was compiled. |
| 137 |
|
| 138 |
@item profiledbootstrap |
| 139 |
Builds a compiler with profiling feedback information. In this case, |
| 140 |
the second and third stages are named @samp{profile} and @samp{feedback}, |
| 141 |
respectively. For more information, see the installation instructions. |
| 142 |
|
| 143 |
@item restrap |
| 144 |
Restart a bootstrap, so that everything that was not built with |
| 145 |
the system compiler is rebuilt. |
| 146 |
|
| 147 |
@item stage@var{N}-start (@var{N} = 1@dots{}4, profile, feedback) |
| 148 |
For each package that is bootstrapped, rename directories so that, |
| 149 |
for example, @file{gcc} points to the stage@var{N} GCC, compiled |
| 150 |
with the stage@var{N-1} GCC@footnote{Customarily, the system compiler |
| 151 |
is also termed the @file{stage0} GCC.}. |
| 152 |
|
| 153 |
You will invoke this target if you need to test or debug the |
| 154 |
stage@var{N} GCC@. If you only need to execute GCC (but you need |
| 155 |
not run @samp{make} either to rebuild it or to run test suites), |
| 156 |
you should be able to work directly in the @file{stage@var{N}-gcc} |
| 157 |
directory. This makes it easier to debug multiple stages in |
| 158 |
parallel. |
| 159 |
|
| 160 |
@item stage |
| 161 |
For each package that is bootstrapped, relocate its build directory |
| 162 |
to indicate its stage. For example, if the @file{gcc} directory |
| 163 |
points to the stage2 GCC, after invoking this target it will be |
| 164 |
renamed to @file{stage2-gcc}. |
| 165 |
|
| 166 |
@end table |
| 167 |
|
| 168 |
If you wish to use non-default GCC flags when compiling the stage2 and |
| 169 |
stage3 compilers, set @code{BOOT_CFLAGS} on the command line when doing |
| 170 |
@samp{make}. |
| 171 |
|
| 172 |
Usually, the first stage only builds the languages that the compiler |
| 173 |
is written in: typically, C and maybe Ada. If you are debugging a |
| 174 |
miscompilation of a different stage2 front-end (for example, of the |
| 175 |
Fortran front-end), you may want to have front-ends for other languages |
| 176 |
in the first stage as well. To do so, set @code{STAGE1_LANGUAGES} |
| 177 |
on the command line when doing @samp{make}. |
| 178 |
|
| 179 |
For example, in the aforementioned scenario of debugging a Fortran |
| 180 |
front-end miscompilation caused by the stage1 compiler, you may need a |
| 181 |
command like |
| 182 |
|
| 183 |
@example |
| 184 |
make stage2-bubble STAGE1_LANGUAGES=c,fortran |
| 185 |
@end example |
| 186 |
|
| 187 |
Alternatively, you can use per-language targets to build and test |
| 188 |
languages that are not enabled by default in stage1. For example, |
| 189 |
@command{make f951} will build a Fortran compiler even in the stage1 |
| 190 |
build directory. |
| 191 |
|