| Age | Commit message (Collapse) | Author |
|---|
|
Instructing gcc (or clang) to prepare for capturing coverage data is
easy: Just build with: CXXFLAGS=--coverage
The hard part is that our buildsystem uses relative paths and so
confuses the hell out of lcov as it assumes this way that all our *.cc
files are in the same directory… by changing to absolute paths in the
compile rules we solve this problem.
Still not perfect as it refers to build/include files for most headers
and our forking/threading code isn't properly captured, but good enough
to see red reports for now:
CXXFLAGS=--coverage make
make test
./test/integration/run-tests -q
lcov --no-external --directory . --capture --output-file apt.info
genhtml --output-directory ./coverage/ apt.info
Git-Dch: Ignore
|
|
We have xz/lzma support for a while, but only via an external binary
provided by xz-utils. Now that the Debian archive provides xz by default
and dpkg pre-depends on the library provided by liblzma-dev we can switch
now to use this library as well to avoid requiring an external binary.
For now the binary is in a prio:required package, but this might change
in the future.
API wise it is quiet similar to bz2 code expect that it doesn't provide
file I/O methods, so we piece this together on our own.
|
|
Now that the last few commits resolved the issues we can finally enable
a bunch of compiler warnings by default.
Git-Dch: ignore
|
|
Beside being a bit cleaner it hopefully also resolves oddball problems
I have with high levels of parallel jobs.
Git-Dch: Ignore
Reported-By: iwyu (include-what-you-use)
|
|
also adds namespaced attributes for good usage
Git-Dch: Ignore
|
|
|
|
manpages sometimes refer to distro-specific things like the name of the
package providing the achive-keyring. Having a central place to
configure this helps in having it consistent in the manpages and allows
to load this info from other places in the buildsystem as well later.
|
|
The upgrade releated code is moved into upgrade.{cc,h} and
all pkg*Upgrade* prototypes are included in algorihms.h to
avoid breaking API (unless build with APT_9_CLEANER_HEADERS).
|
|
It works for a while now in manual tests, now lets see how it will
perform if enabled for all by default automatically.
|
|
Git-Dch: Ignore
|
|
The buildsystem tried to build l10n for test applications which never
produced the output it expected causing it to try building it all the
time.
|
|
fixing parallel build in the handcrafted buildsystem is a pain,
so its not enabled by default, but its works for me – sometimes
Git-Dch: Ignore
|
|
Building manpages becames more consistent this way and
it is simpler to ignore build artefacts, too.
|
|
- fix build failure when building without NLS (closes: #671587)
|
|
- demote debiandoc-sgml to Build-Depends-Indep
* doc/makefile:
- separate translation building of debiandoc from manpages
so that we don't need to build debiandoc for binary packages
|
|
- print a message detailing how to get config.guess and config.sub
in case they are not in /usr/share/misc (Closes: #677312)
|
|
- remove inttypes.h compatibility as providing such a c99 types
compatibility conflicts with the usage of c99 type long long
|
|
(it creates changelog from cvs)
|
|
|
|
- ensure that all sources end up in the srclist so that we don't
forget to extract half of the translation strings
|
|
manpage building instead of doing it at setup time, so we can properly
depend on it
|
|
- build manpages with the correct l10n.gentext.default.language setting
to get the correct section titles provided by docbook
|
|
- make apt and apt-utils packages depend on manpages instead of full doc
|
|
- reshuffle dependencies so that parallel building seems to work
|
|
- remove the libz-dev alternative from zlib1g-dev build-dependency
- do the same for bz2 builtin if available
* apt-pkg/contrib/fileutl.cc:
- use libz2 library for (de)compression instead of the bzip2 binary as
the first is a dependency of dpkg and the later just priority:optional
so we gain 'easier' access to bz2-compressed Translation files this way
|
|
compatible with users accessing it directly, but this prepares for a drop
of this strict requirement in the future
|
|
to the more standard PACKAGE_VERSION and make it work in every file
|
|
causes timestamp changes for the mo files which therefore can't
be refcounted by dpkg for your M-A: same packages
(Closes: #659333, LP: #924628)
The commit also enables a top-level 'make update-po' and does
all the needed changes to let this work now that update-po might
be called in a freshly checkout tree
|
|
(and be it if it tries to announce that…)
|
|
|
|
size are pretty unlikely for now, but we need it for deb
packages which could become bigger than 4GB now (LP: #815895)
|
|
|
|
all (sub)directories are build which are needed (e.g. obj/ and co)
|
|
- move some strings into apt-verbatim.ent to avoid showing them in
apt-doc.pot as they are untranslatable anyway (e.g. manpage references)
|
|
- ensure that the build fails if documentation building fails
|
|
(This is the revert of the glibc-abi-compatibility-hack)
|
|
- add ngettext support with P_()
|
|
|
|
|
|
This commit sets up our faked library extension to trick the build
system into building the packages with libc6.9-6 while we are actually
already at libc6.10-6. Oh dear...
|
|
|
|
|
|
information any longer
|
|
|
|
|
|
by the new added build-dependency autotools-dev instead
|
|
so it is not needed to change x-files to just increase a version number
(because this is far to easy to forget one of the files, which will not
result in failures or is bad in general, but is inconsitent.)
|
|
Backported from lp:~mvo/apt/debian-experimental
Based on patch by Eugene V. Lyubimkin
|
|
[cmdline/apt-get.cc] check for availability ofstatfs.f_type
Patch from Robert Millan, thanks! (Closes: #509313)
|
|
flag again and a manual mv to the correct name is no longer needed.
(used in the buildprocess of po4a manpages)
|
