While writing the wording to deprecate XML policy installed by packages
into ${sysconfdir}, I realised we didn't give a typical example of
what packages *should* do.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Now that dbus 1.10 has become widely available, we should start to
treat ${sysconfdir} as reserved for the sysadmin, and encourage
third-party software packages to install their integration files into
${datadir}.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Multi-configuration generators (VS, Xcode) append a per-configuration
subdirectory to the specified CMAKE_xxx_OUTPUT_DIRECTORY} directory.
To use the real output paths in test applications, this subdirectory
must be added to the corresponding cmake variables.
This patch introduces a new cmake macro add_session_test_executable,
which uses dbus-run-session to start a dbus-daemon process with a
temporary session bus in the background and the desired client file.
add_session_test_executable requires additional environment variables
defined in the top level CMakeLists.txt.
Bug: https://gitlab.freedesktop.org/dbus/dbus/issues/135
Signed-off-by: Ralf Habacker <ralf.habacker@freenet.de>
This patch introduces a new cmake macro add_session_test_executable,
which uses dbus-run-session to start a dbus-daemon process with a
temporary session bus in the background and the desired client file.
add_session_test_executable requires additional environment variables
defined in the top level CMakeLists.txt.
Bug: https://gitlab.freedesktop.org/dbus/dbus/issues/135
Signed-off-by: Ralf Habacker <ralf.habacker@freenet.de>
Instead of the variables EXECUTABLE_OUTPUT_PATH and
LIBRARY_OUTPUT_PATH, which have become obsolete since
cmake 3.x, CMAKE_xxx_OUTPUT_PATH is now used to define
output paths in the build directory.
It is not safe to assume that the result of getenv() is not invalidated
by a subsequent call to setenv(), and dbus#240 demonstrates that this
can be a problem in practice when running Windows code under Wine.
Closes: https://gitlab.freedesktop.org/dbus/dbus/issues/240
Signed-off-by: Simon McVittie <smcv@collabora.com>
This all seems to have been written by Red Hat or by Collabora, so we
have permission to relicense it under the MIT/X11 license, and we
might as well do so.
Signed-off-by: Simon McVittie <smcv@collabora.com>
This results in one less special case in test-main, which will be
significant when we want to make the tests more data-driven.
Signed-off-by: Simon McVittie <smcv@collabora.com>
This will make it possible to unify the wrapper code that runs them.
I'm using a plain C string rather than a DBusString to make it
more straightforward to carve out tests into their own executables.
Signed-off-by: Simon McVittie <smcv@collabora.com>
We don't usually mass-remove trailing whitespace from the actual source
code because it would complicate cherry-picking bug fixes to older
branches, but that reasoning doesn't really apply to the comments
containing copyright and licensing notices.
Removing trailing whitespace makes it much easier to move code around:
we have a commit hook that rejects commits containing trailing
whitespace, but that commit hook counts moving a file as a delete + add
pair, so it objects to moving code that contains trailing whitespace.
Signed-off-by: Simon McVittie <smcv@collabora.com>
libdbus has been thread-safe by default since 1.7.6 (2013):
dbus_threads_init_default() is called on a just-in-time basis
whenever needed, and _dbus_thread_init_debug() is equivalent to that.
Signed-off-by: Simon McVittie <smcv@collabora.com>
If lists are in a completely arbitrary order, sorting them consistently
means that there is only one correct place to insert a new entry, avoiding
the merge conflicts that would occur if we always append new entries.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Patterns in the top-level .gitignore match in all subdirectories, so
there's no need to repeat ourselves quite so much for generic
C, Autotools and gcov patterns.
Signed-off-by: Simon McVittie <smcv@collabora.com>
If lists are in a completely arbitrary order, sorting them consistently
means that there is only one correct place to insert a new entry, avoiding
the merge conflicts that would occur if we always append new entries.
Signed-off-by: Simon McVittie <smcv@collabora.com>
The anchored patterns match only in the top directory; the unanchored
patterns match in any subdirectory. While I'm moving the unanchored
patterns around anyway, sort them in lexicographical (LC_ALL=C sort)
order.
Signed-off-by: Simon McVittie <smcv@collabora.com>
There shouldn't be an aclocal.m4 in any subdirectory, for instance,
so there's no need to ignore it.
Signed-off-by: Simon McVittie <smcv@collabora.com>
We put these in build-aux/ since 15e6b4f5, and we ignore all of
build-aux/, so we don't need to ignore them separately.
Signed-off-by: Simon McVittie <smcv@collabora.com>
If the environment variable DBUS_TEST_SLOW is not set or is 0
fewer iterations are performed in test-refs enough to demonstrate
that it works, rather than seriously trying to reproduce race
conditions.
This solves the long runtime that Wine 3.x (4 times longer than
native Windows 7) or Windows 7 as a VirtualBox 5.22 guest (10
times longer) is to be seen.
Fixes#244
This means we don't need to distinguish between DBUS_NAME_TEST_EXEC and
DBUS_TEST_EXEC any more, because all test helper executables are in the
same place, both during build and when installed (we don't install
test-privserver since no installed test requires it yet, but in
principle we could).
Signed-off-by: Simon McVittie <smcv@collabora.com>
I've assumed that trivial changes from Scott James Remnant (correcting
some bus names) and Marcus Brinkmann (adding a <config.h> include
tree-wide) are not significant here. All the other potential copyright
holders from the git log are listed, and all have given permission to
relicense their dbus contributions under the MIT/X11 license.
Signed-off-by: Simon McVittie <smcv@collabora.com>
We don't want to set these globally via the normal CFLAGS, because if
we did, AddressSanitizer would catch test-segfault deliberately
segfaulting, and "helpfully" turn it into exit status 1, which in turn
makes our test fail because it asserts that the segfault is reported
as a segfault.
A typical use with gcc as compiler, on a reasonably recent Debian,
would be:
./configure SANITIZE_CFLAGS="-fsanitize=address -fsanitize=undefined -fPIE -pie"
Signed-off-by: Simon McVittie <smcv@collabora.com>
It's only used in test code. We have to put it in its own translation
unit with no non-libc dependencies so that we can compile a copy of it
without AddressSanitizer support, because in a subsequent commit we will
special-case test-segfault to be compiled without using AddressSanitizer,
which would make linking to an AddressSanitizer-instrumented libdbus fail.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Running tests under instrumentation (libasan) or emulation (qemu)
is not fast.
Inspired by the --timeout-factor option in
<https://salsa.debian.org/ci-team/autopkgtest>.
Signed-off-by: Simon McVittie <smcv@collabora.com>
If we cast a user-supplied pointer to DBusBasicValue *, that's like
promising that the pointer is to an object at least as strictly aligned
as a DBusBasicValue. In practice, it often isn't: for example, when
we marshal a dbus_uint32_t, a pointer to it is only guaranteed to be
32-bit-aligned, whereas DBusBasicValue typically requires 64-bit
alignment.
Signed-off-by: Simon McVittie <smcv@collabora.com>
This results in (harmless) leak reports when running under the
AddressSanitizer, which could make real leaks harder to find.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Autotools creates executable applications in the respective
subdirectory of the build directory, while cmake creates
them in <build-root>/bin.
This leads to different paths in the file created
from org.freedesktop.DBus.TestSuite.PrivServer.service.in,
which are fixed by the new variable.
Bug: https://gitlab.freedesktop.org/dbus/dbus/issues/135
Signed-off-by: Ralf Habacker <ralf.habacker@freenet.de>
Reviewed-by: Simon McVittie <smcv@collabora.com>
