Commit 0c03b505 was meant to clear all the fds indexed by j in
[0, n_fds), which socket_disconnect() can't be allowed to close
(because on failure the caller remains responsible for closing them);
but instead it closed the one we failed to add to the main loop
(fd i), repeatedly.
Similarly, it was meant to invalidate all the watches indexed by j
in [i, n_fds) (the one we failed to add to the main loop and the ones
we didn't try to add to the main loop yet), which socket_disconnect()
can't be allowed to see (because it would fail to remove them from
the main loop and hit an assertion failure); but instead it invalidated
fd i, repeatedly.
These happen to be the same thing if you only have one fd, resulting
in the test-case passing on an IPv4-only system, but failing on a
system with both IPv4 and IPv6.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=89104
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
(cherry picked from commit c9aa00ce73)
This also covers _dbus_server_new_for_socket(), which is one of the
worse places in terms of complexity of the error-unwinding path
(3 labels).
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=89104
(cherry picked from commit ac6c4d384d)
dbus_realloc() doesn't guarantee to set errno (if it did, the
only reasonable thing it could set it to would be ENOMEM). In
particular, faking OOM conditions doesn't set it. This can cause an
assertion failure when OOM tests assert that the only error that can
validly occur is DBUS_ERROR_NO_MEMORY.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=89104
(cherry picked from commit 9ded6907e6)
_dbus_server_finalize_base() asserts that the socket has been
disconnected, but in some OOM code paths we would call it without
officially disconnecting. Do so.
This means we need to be a bit more careful about what is
socket_disconnect()'s responsibility to clean up, what is
_dbus_server_new_for_socket()'s responsibility, and what is the caller's
responsibility.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=89104
(cherry picked from commit 0c03b505a9)
We assert that every watch is invalidated before it is freed, but
in some OOM code paths this didn't happen.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=89104
(cherry picked from commit 1ce34beef8)
While reviewing fd.o#101354, Philip Withnall pointed out that if we
rejected a connection in the new code there, we didn't log why. It
turns out we didn't log that in the more normal code path either.
Redo the error handling so that failure to set up a connection
is logged.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103592
(cherry picked from commit 56847ae818)
In the debug build, enable features that are off by default. In the
reduced build, explicitly disable features, some of which are
on by default. In the legacy build, check that we can compile the
default feature-set without inotify, dnotify, systemd, etc.
Reviewed-by: Philip Withnall <withnall@endlessm.com>
[smcv: Rebase onto 1.13.x branch, fix minor conflicts]
Signed-off-by: Simon McVittie <smcv@collabora.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=101354
(cherry picked from commit 3c031ef5aa)
This was always leaked, both on success and on error.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103597
(cherry picked from commit 37d5af203c)
We could free all of these without having ever successfully
initialized them.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103597
(cherry picked from commit 6d08f5c04e)
If _dbus_string_init() fails, it doesn't guarantee that the string
is initialized to anything in particular. Worse, if
_dbus_string_init (&buffer) fails, p would never have been initialized
at all, due to the use of the short-circuiting || operator.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103597
(cherry picked from commit 0ea0e4b0fd)
`echo -e` is a GNU extension, and in particular not available on
Darwin.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103493
[smcv: Added commit message]
Reviewed-by: Simon McVittie <smcv@collabora.com>
(cherry picked from commit 216ae1adfc)
We're not going to replace deprecated functions here, similar to commit
88e0ccb2 in the dbus-1.10 branch.
Signed-off-by: Simon McVittie <smcv@collabora.com>
These were in git but not distributed in source tarballs, and in fact
not hooked up to the Autotools build system at all.
test/data/valid-introspection-files was mentioned in the CMake build
system (copied from the source directory to the build directory), but
according to `git grep` is not used for anything.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103420
Our official source releases are Autotools "make dist" tarballs, but
there's no reason why CMake users can't use those too, and we already
include the CMake build files.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103420
If we distribute the entire directory in "make dist" tarballs, then
we include the generated files cmake/DBus1Config.cmake and
cmake/DBus1ConfigVersion.cmake, which we should not.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103420
test-bus exercises the parser by trying to parse every file in these
directories. A couple of files were accidentally left out, meaning
those parsing code paths are tested when we build from git, but not
when we build from a tarball release.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103420
cmake/DBus1Config.cmake, cmake/DBus1ConfigVersion.cmake and dbus-1.pc
are all generated by AC_CONFIG_FILES, so they are automatically listed
in $(CONFIG_CLEAN_FILES) and cleaned in "make distclean" without further
help from us.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103420
It's generated by configure, so we should not distribute it.
Because it's generated by configure, it is automatically cleaned up
by "make distclean" via $(CONFIG_CLEAN_FILES).
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103420
I'm hoping for 1.11.24 to act as 1.12rc1, with a remarkably similar
1.12.0 release shortly after if all goes well.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Previous cmake versions seemed to be more tolerant.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103015
Signed-off-by: Ralf Habacker <ralf.habacker@freenet.de>
Reviewed-by: Simon McVittie <smcv@collabora.com>
- include <windows.h> to be able to use constants
- let versioninfo be visible in explorer by adding a "Translation" value
- change FILEOS from VOS_NT_WINDOWS32, which was intended for Windows NT,
to VOS__WINDOWS32
- stop setting FILEFLAGS 0x20 (VS_FF_SPECIALBUILD), which is not
appropriate here because we build the normal version, not a special
version
- use constants
- fix strings
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103015
Signed-off-by: Ralf Habacker <ralf.habacker@freenet.de>
Reviewed-by: Simon McVittie <smcv@collabora.com>
The recent implementation generates a timestamp containing eol on
linux hosts, which generates unparseable versioninfo.rc.
This commit raises the minimal required cmake version to 3.0.2.
Signed-off-by: Ralf Habacker <ralf.habacker@freenet.de>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103015
Reviewed-by: Simon McVittie <smcv@collabora.com>
That's what is checked for by LT_LANG([Windows Resource]) further
up, and is what we now use during the build. Its value is typically
i686-w64-mingw32-windres.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103015
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
We have two variables that both expand to i686-w64-mingw32-windres,
namely WINDRES and RC, and we might as well use the same one as
in dbus/ here. However, it seems we can't wrap windres in libtool
when producing an executable: if we use .rc.lo, my Automake 1.15.1
doesn't realise that it needs to include disable-uac.lo in the
list of objects, whereas if we use .rc.o, Ralf's libtool 2.4.2 and
Automake 1.13.4 disagree on where the output should go
(.libs/disable-uac.o vs. disable-uac.o) and the link fails.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103015
If __LINE__ doesn't work in MSVC's resource compiler, then removing
the #line directive altogether seems a simpler fix than redefining
__LINE__ to the wrong value.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103015
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
The resource file used to #include this, but it was unnecessary,
and Ralf removed it in commit e3a14eb.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103015
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
libtool has built-in support for Windows resources, and we even
enable it in configure.ac. What it doesn't have is a built-in rule
for generating Libtool objects using that built-in support, but
we can add one.
We have to generate Libtool pseudo-objects (.lo) rather than native
object files (.o) so that we get both a PIC object for the shared
library and a non-PIC object for the static library.
This mimics the libtool invocations used for compiling C and C++.
Note that $(RC) is typically i686-w64-mingw32-windres, the same as
our project-specific variable $(WINDRES) which was previously used here.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=103015
Reviewed-by: Philip Withnall <withnall@endlessm.com>
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
