Simon McVittie [Thu, 12 Jul 2018 16:38:04 +0000 (17:38 +0100)]
embedded tests: Make it easier to run a single test-case
When running tests under "make check" or similar to take advantage
of facilities like AM_TESTS_ENVIRONMENT and AX_VALGRIND_CHECK, it's
more straightforward to set an environment variable than to pass a
command-line option.
Signed-off-by: Simon McVittie <smcv@collabora.com> Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://gitlab.freedesktop.org/dbus/dbus/issues/218
Simon McVittie [Thu, 15 Nov 2018 18:55:57 +0000 (18:55 +0000)]
internals: Assume compiler supports a subset of ISO varargs syntax
We have considerable anecdotal evidence that every relevant compiler
supports at least the small part of ISO varargs syntax that we need
here, because tools/tool-common.h has contained
#define VERBOSE(...) do {} while (0)
since dbus 1.9.2 (2014) and nobody has complained yet. With that in
mind, let's simplify.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Ralf Habacker [Thu, 25 Oct 2018 18:14:24 +0000 (20:14 +0200)]
Refactor cmake checks for DBUS_VA_COPY and DBUS_VA_COPY_ARRAY
For test case execution, CheckCSourceCompiles is now used instead
of try_compile and the determination of DBUS_VA_AS_ARRAY is
performed with a separate test instead of evaluating the result
of HAVE_VA_COPY and HAVE___VA_COPY.
The tests are performed for all supported compilers. Since older
MSVC compilers (< 2013) do not support va_copy(), the macro
_DBUS_VA_ASSIGN(a1,a2) with the implementation { a1 = a2; } is used
as a fallback.
Use SELINUX_CB_POLICYLOAD instead of AVC_CALLBACK_RESET callback
Use SELINUX_CB_POLICYLOAD instead of AVC_CALLBACK_RESET callback as this
only seems necessary on policy reload and not if the enforcing mode is
changing.
See discussion at https://marc.info/?l=selinux&m=152173501930182&w=2
Stop using avc_init() and use avc_open() instead. With this commit
dbus-daemon will stop using a thread to monitor the avc netlink and will
poll it instead.
Ralf Habacker [Fri, 26 Oct 2018 13:12:29 +0000 (15:12 +0200)]
Avoid double slashes in paths created by pkg-config
If in a .pc variable a path is created from another
variable, such as exec_prefix=${prefix}/lib, prefix
must not contain a trailing slash to avoid double
slashes in the generated path.
Simon McVittie [Wed, 31 Oct 2018 15:09:24 +0000 (15:09 +0000)]
tests: Build installable helpers whenever any tests are enabled
We might need these to run tests at build-time, so we should build them
whenever either modular or embedded tests are enabled, even if
installed-tests aren't.
We haven't noticed this bug until now because $(installable_helpers)
only contained test-apparmor-activation, which isn't normally needed at
build-time because the AppArmor test can only work when run as root.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Simon McVittie [Mon, 22 Oct 2018 09:20:37 +0000 (10:20 +0100)]
cmake: Install d-d-launch-helper to ${CMAKE_INSTALL_LIBEXECDIR}
Instead of hard-coding the lib subdirectory of the installation prefix,
use the ${CMAKE_INSTALL_LIBEXECDIR} provided by the GNUInstallDirs
module, which defaults to the libexec subdirectory of the installation
prefix. This is consistent with the Autotools build system, which
installs to ${exec_prefix}/libexec unless modified with --libexecdir.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Ralf Habacker [Tue, 23 Oct 2018 19:56:25 +0000 (21:56 +0200)]
Cleanup of the cmake definitions of source directories
Since the cmake build system is located below the cmake/
subdirectory, references to the source files used such as
../../bus or ${CMAKE_SOURCE_DIR}/../dbus etc. are required.
To standardize and simplify this, a cmake variable is now
defined in each of the listed directories, which contains
the corresponding path.
Ralf Habacker [Sat, 20 Oct 2018 20:34:33 +0000 (22:34 +0200)]
Refactor cmake checks for DBUS_VA_COPY and DBUS_VA_COPY_ARRAY
For test case execution, CheckCSourceCompiles is now used instead
of try_compile and the determination of DBUS_VA_AS_ARRAY is
performed with a separate test instead of evaluating the result
of HAVE_VA_COPY and HAVE___VA_COPY.
The tests are performed for all supported compilers. Since older
MSVC compilers (< 2013) do not support va_copy(), the macro
_DBUS_VA_ASSIGN(a1,a2) with the implementation { a1 = a2; } is used
as a fallback.
Reviewed-by: Simon McVittie <smcv@collabora.com>
Bug: https://gitlab.freedesktop.org/dbus/dbus/merge_requests/18
Simon McVittie [Tue, 23 Oct 2018 10:43:23 +0000 (11:43 +0100)]
dbus-daemon test: Allow much longer for pending fd timeout
The timeout we're using here is 0.5s (500ms), but the actual time taken
is unbounded, because the OS scheduler might not schedule our process
for an arbitrary length of time after we become runnable.
We previously allowed up to 1 second, but in the CI jobs for dbus!9
and dbus!18 we've seen this take up to 3.4 seconds (presumably
because other tests, or other jobs running on the same shared
infrastructure, starved this process). Allow up to 10 seconds to guard
against spurious failures.
The timeout used in the production system.conf is 150 seconds (2½
minutes), and we're only using the shorter 500ms timeout here to make
the test complete more quickly, so ±10 seconds is relatively
insignificant: the main thing is that it's finite.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Simon McVittie [Mon, 22 Oct 2018 10:45:45 +0000 (11:45 +0100)]
build: Never use poll() on Darwin family (macOS, etc.) or Interix
Doing a runtime check in configure.ac (AC_RUN_IFELSE) has several
disadvantages:
* It doesn't work when cross-compiling. For example, if we build macOS
binaries on a Linux system, we'd assume that poll() works, but in
fact it won't.
* It checks the build system capabilities, but that is not necessarily
appropriate if (for example) a macOS 10.10 user builds binaries that
could be used by macOS 10.12 or macOS 10.9 users.
* It checks for one specific failure mode, but macOS seems to have a
history of various implementation issues in poll().
* If we want it to work in CMake, we have to duplicate it in the CMake
build system.
None of these is a showstopper on its own, but the combination of all
of them makes the current approach to avoiding the broken poll() on
macOS look unreliable. libcurl, a widely-portable library making
extensive use of sockets, specifically doesn't use poll() on Darwin
(macOS, iOS, etc.) or on Interix; let's follow their example here.
See also https://bugzilla.gnome.org/show_bug.cgi?id=302672 and
https://daniel.haxx.se/blog/2016/10/11/poll-on-mac-10-12-is-broken/
for some relevant history.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Resolves: https://gitlab.freedesktop.org/dbus/dbus/issues/232
Ralf Habacker [Mon, 22 Oct 2018 07:27:30 +0000 (09:27 +0200)]
cmake: Fix incorrect header list separator in configure checks
CMake expects a semicolon-separated list of headers, not a
space-separated list. In particular, this meant we failed to detect
getpwnam_r() on Linux, and fell back to getpwnam().
Ralf Habacker [Tue, 16 Oct 2018 10:21:55 +0000 (12:21 +0200)]
Keep macro values in config.h.cmake synchronized with Autotools
Autotools sets the value of HAVE_xxx macros in config.h to 1 if
the corresponding value exists. This has not been used consistently
in config.h.cmake so far.
Bug: https://gitlab.freedesktop.org/dbus/dbus/issues/117 Reviewed-by: Simon McVittie <smcv@collabora.com>
Simon McVittie [Thu, 18 Oct 2018 16:36:37 +0000 (17:36 +0100)]
desktop-file: Treat backslashes in section names as literal
The Desktop Entry Specification doesn't give any special meaning to
backslashes in section names: a line "[\n]" starts a section whose
name is the two characters (backslash, n), not a section whose name
is a newline. GKeyFile in GLib matches this interpretation.
In practice, the only section used by dbus-daemon is "D-BUS Service",
only way this could make a difference is if someone had written it
as "D-BUS\sService". According to
https://codesearch.debian.net/search?q=%5C%5BD-BUS%5C%5CsService%5C%5D
there is no instance of that pattern in Debian.
Signed-off-by: Simon McVittie <smcv@collabora.com>
David King [Fri, 12 Oct 2018 12:58:43 +0000 (13:58 +0100)]
is_valid_section_name: Fix logical expression
Group names in desktop files may contain all ASCII characters, except
control characters and '[' and ']'. Rather than accepting all values,
thanks to a logical operator confusion found by GCC warning
-Wlogical-op, instead explicitly reject the invalid values.
Signed-off-by: David King <dking@redhat.com> Fixes: https://gitlab.freedesktop.org/dbus/dbus/issues/208
Simon McVittie [Tue, 16 Oct 2018 14:44:59 +0000 (15:44 +0100)]
CONTRIBUTING.md: Update and rewrite
This file hadn't kept up with reality, and needs updating for Gitlab.
Take the opportunity to rewrite it.
Much of the text, particularly about commit messages, was taken from
Wayland's contributing guide (thanks to Ander Conselvan de Oliveira,
Bryce Harrington, Eric Engestrom, Pekka Paalanen and Daniel Stone).
Signed-off-by: Simon McVittie <smcv@collabora.com>
Ralf Habacker [Mon, 31 Oct 2011 14:53:25 +0000 (15:53 +0100)]
Use separate bus configuration files for executing test cases with a temporary session bus
Instead of the previous adaptation of the existing template
for the session bus, a separate template is now used, which
can be more easily adapted to the requirements of the test
applications.
EAGAIN and EWOULDBLOCK are documented to possibly be numerically equal,
for instance in errno(3), and a simple logical OR check will trigger the
-Wlogical-op warning of GCC. The GCC developers consider the warning to
work as-designed in this case:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69602
Avoid such a warning by explicitly checking if the values are identical.
Fixes: https://gitlab.freedesktop.org/dbus/dbus/issues/225 Signed-off-by: David King <dking@redhat.com> Reviewed-by: Simon McVittie <smcv@collabora.com>
Ralf Habacker [Mon, 11 Dec 2017 08:14:49 +0000 (09:14 +0100)]
Build dbus-run-session on non Windows platforms with cmake
dbus-run-session is the preferred way to run a temporary D-Bus
session scoped to the lifetime of a process, for example
dbus-run-session -- make check.
Simon McVittie [Tue, 4 Sep 2018 14:38:12 +0000 (07:38 -0700)]
cmake: Define HAVE_PRCTL, HAVE_RAISE according to existing checks
Otherwise test-segfault will not be able to disable core dumps, making
it extremely slow and noisy to run the tests under cmake.
I added the missing checks in commit be55374f, but didn't add the
corresponding symbols to config.h.cmake.
Fixes: be55374f "cmake: check for the necessary symbols for test-segfault.c" Closes: https://gitlab.freedesktop.org/dbus/dbus/issues/227 Signed-off-by: Simon McVittie <smcv@collabora.com>
Simon McVittie [Thu, 4 Oct 2018 16:26:42 +0000 (17:26 +0100)]
ci: Mark many Gitlab jobs to be run manually
freedesktop.org Gitlab doesn't currently have enough test runners
available to run all of this every time. For higher-risk changes
(for example those that change the build system) we can run the
complete set through the web UI.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Simon McVittie [Wed, 3 Oct 2018 16:51:49 +0000 (17:51 +0100)]
ci: Teach ci-install.sh to install wine on Debian 9 'stretch'
Signed-off-by: Simon McVittie <smcv@collabora.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=108177 Acked-by: Philip Withnall <withnall@endlessm.com>
Simon McVittie [Wed, 3 Oct 2018 16:51:35 +0000 (17:51 +0100)]
ci: Explicitly install cmake
Travis-CI workers have cmake preinstalled, but Gitlab-CI Docker images
typically don't.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=108177 Acked-by: Philip Withnall <withnall@endlessm.com>
Simon McVittie [Wed, 3 Oct 2018 16:25:43 +0000 (17:25 +0100)]
ci: Add Gitlab-CI configuration
This uses the same shell scripts as Travis-CI, with slightly different
settings. We use Docker containers for all our Gitlab-CI runs, so take
the opportunity to use Debian 9 'stretch' as our baseline, and
relegate Ubuntu 14.04 'trusty' to to a secondary build.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=108177 Acked-by: Philip Withnall <withnall@endlessm.com>
Previously we didn't accept this as relocatable, but actually it's
fine: ${prefix} is still equivalent to ${libdir}/pkgconfig/../..,
so our relocation setup can work. Accept the result of expanding
"${prefix}" as an acceptable value for --exec-prefix, and accept the
results of expanding "${exec_prefix}/lib" etc. as acceptable values
for --libdir.
Note the use of single vs. double quotes here. A case statement that
matches '${prefix}' tests for the literal string «${prefix}»,
whereas a case that matches "${prefix}" tests for the string that is
the value of the variable named «prefix» that is set by the
--prefix command-line argument.
Signed-off-by: Simon McVittie <smcv@collabora.com> Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=107662
Simon McVittie [Wed, 29 Aug 2018 13:01:02 +0000 (14:01 +0100)]
build: Give better error messages if relocation is impossible
There are two reasons why we might reject relocation: the exec_prefix
differing from the prefix, or the libdir not being a first-level
subdirectory named "lib" or "lib64" of the prefix. Make it clearer
which one failed and why.
Signed-off-by: Simon McVittie <smcv@collabora.com> Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=107662
Simon McVittie [Tue, 3 Jul 2018 19:26:33 +0000 (20:26 +0100)]
containers: Share code for what happens when we lose a connection
This improves test coverage, because
bus_container_instance_lost_connection() was previously only called
when we failed to set up a connection with a server due to OOM, but
it is now also called (instead of being duplicated) when we are told
to clean up a connection because it has disconnected.
To make sure that connections from containers can never cheat their
way into being treated as uncontained, do not set their
contained_data_slot to NULL.
Signed-off-by: Simon McVittie <smcv@collabora.com> Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=107739
Simon McVittie [Fri, 17 Aug 2018 16:02:45 +0000 (17:02 +0100)]
Add more test coverage for config file parsing
minimal.conf is a valid config file added to make it obvious why
the new invalid config files are invalid.
Signed-off-by: Simon McVittie <smcv@collabora.com> Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=107739
Simon McVittie [Wed, 29 Aug 2018 17:33:01 +0000 (18:33 +0100)]
bus_config_parser_check_doctype: Remove, unused
We have never checked the <!DOCTYPE> of busconfig XML since the libxml
parser was removed in 2013, and the libxml parser was broken before
that anyway. The recommended Expat parser (our only parser since 2013)
does not appear to have ever validated this, so now does not seem like
the time to start. Just ignore the <!DOCTYPE> if there is one.
(We never validated this particularly strictly anyway;
<!DOCTYPE busconfig SYSTEM "http://example.com/bees"> would have been
treated as perfectly valid.)
Signed-off-by: Simon McVittie <smcv@collabora.com> Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=107739
Simon McVittie [Fri, 3 Aug 2018 14:33:08 +0000 (15:33 +0100)]
containers test: Exercise GetConnectionInstance() on dbus-daemon itself
This is an easy bit of missing test coverage detected by running the
test suite with gcov.
Signed-off-by: Simon McVittie <smcv@collabora.com> Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=107739
Simon McVittie [Fri, 17 Aug 2018 15:35:37 +0000 (16:35 +0100)]
BusConfigParser test: Check that all limits are equal, not just one
Signed-off-by: Simon McVittie <smcv@collabora.com> Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=107739
Simon McVittie [Fri, 17 Aug 2018 15:33:06 +0000 (16:33 +0100)]
bus_context_get_policy: Remove, unused
Signed-off-by: Simon McVittie <smcv@collabora.com> Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=107739
Simon McVittie [Fri, 17 Aug 2018 15:31:00 +0000 (16:31 +0100)]
bus_connections_get_context: Remove, unused
Signed-off-by: Simon McVittie <smcv@collabora.com> Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=107739
These do not appear in code coverage statistics, and `git grep`
reveals that they are unused.
Signed-off-by: Simon McVittie <smcv@collabora.com> Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=107739
Simon McVittie [Fri, 3 Aug 2018 14:21:53 +0000 (15:21 +0100)]
dbus-cleanup-sockets: Mark functions noreturn as suggested by clang
Signed-off-by: Simon McVittie <smcv@collabora.com> Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=107739
Simon McVittie [Thu, 23 Aug 2018 08:01:03 +0000 (09:01 +0100)]
Do not apply __attribute__((__malloc__)) to dbus_realloc()
As noted in GLib commit c879f50f, gcc's interpretation of the malloc
attribute has become more strict over time, which could result in
miscompilation. The new definition is that in addition to assuming
that the returned memory block is newly-allocated, gcc now assumes
that it does not contain any valid pointers. This is OK for
uninitialized or zero-initialized memory returned by dbus_malloc()
or dbus_malloc0(), but not valid for dbus_realloc(), which might be
used for a dynamically-sized array of (structures containing)
valid pointers.
See https://gitlab.gnome.org/GNOME/glib/issues/1465
Signed-off-by: Simon McVittie <smcv@collabora.com> Reviewed-by: Philip Withnall <withnall@endlessm.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=107741
projects / thirdparty / dbus.git / log
