Simon McVittie [Thu, 11 Feb 2016 20:43:23 +0000 (20:43 +0000)]
Only compile test-bus-launch-helper, etc. if embedded tests are enabled
These source files are specific to the embedded tests and make no sense
otherwise.
Also remove a comment in the CMake build system about fixing the
build of the activation helper on Windows: the activation helper
is Unix-specific and always will be, since it relies on Unix setuid
to function.
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=94094
Simon McVittie [Mon, 15 Aug 2016 23:25:01 +0000 (00:25 +0100)]
Merge remote-tracking branch 'github/master'
Simon McVittie [Mon, 15 Aug 2016 23:21:48 +0000 (00:21 +0100)]
1.11.5
Simon McVittie [Mon, 15 Aug 2016 23:21:19 +0000 (00:21 +0100)]
Merge remote-tracking branch 'origin/master'
Simon McVittie [Mon, 15 Aug 2016 20:44:06 +0000 (21:44 +0100)]
1.11.4
Simon McVittie [Mon, 15 Aug 2016 20:38:52 +0000 (21:38 +0100)]
Merge branch 'dbus-1.10'
Simon McVittie [Mon, 15 Aug 2016 19:41:28 +0000 (20:41 +0100)]
Travis-CI: fetch a newer autoconf-archive from Debian
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Simon McVittie [Mon, 15 Aug 2016 19:12:19 +0000 (20:12 +0100)]
Travis-CI: install new autoconf-archive build-dependency
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Simon McVittie [Mon, 15 Aug 2016 19:11:49 +0000 (20:11 +0100)]
autogen.sh: fail cleanly if autoconf fails
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Simon McVittie [Mon, 15 Aug 2016 18:59:18 +0000 (19:59 +0100)]
1.10.10
Signed-off-by: Simon McVittie <smcv@debian.org>
Simon McVittie [Mon, 15 Aug 2016 14:15:13 +0000 (15:15 +0100)]
name-test, tools: add missing COVERAGE_CFLAGS
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Simon McVittie [Mon, 15 Aug 2016 13:29:56 +0000 (14:29 +0100)]
Document use of AX_CODE_COVERAGE
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Thomas Zimmermann [Fri, 12 Aug 2016 08:50:53 +0000 (10:50 +0200)]
Use AX_CODE_COVERAGE for test-coverage statistics
DBus uses custom rules in its Makefiles to implement test-coverage
statistics.
This patch implements test-coverage statistics with the autoconf macro
AX_CODE_COVERAGE. The script automatically tests for tools (e.g., gcov,
lcov), sets build variables and creates Makefile rules.
Run 'configure' with '--enable-code-coverage' to enable support for
test-coverage statistics. Run 'make check-code-coverage' to run the
tests and generate the statistics.
Signed-off-by: Thomas Zimmermann <tdz@users.sourceforge.net>
[smcv: do not alter compiler.m4; move AM_CXXFLAGS to the one place we
compile C++]
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=88922
Simon McVittie [Fri, 12 Aug 2016 17:24:10 +0000 (18:24 +0100)]
Merge branch 'dbus-1.10'
Simon McVittie [Fri, 12 Aug 2016 09:58:13 +0000 (10:58 +0100)]
Add a regression test for dbus-launch in X11
[Modified for the dbus-1.10 branch: DBUS_USE_TEST_BINARY is needed.]
Signed-off-by: Simon McVittie <smcv@debian.org>
Simon McVittie [Wed, 20 Jul 2016 12:31:50 +0000 (13:31 +0100)]
test-dbus-daemon-fork: exercise closed stdin, stdout, stderr
Signed-off-by: Simon McVittie <smcv@debian.org>
Simon McVittie [Wed, 20 Jul 2016 11:07:03 +0000 (12:07 +0100)]
test: expand dbus-launch-eval test to cover stdin being closed
Signed-off-by: Simon McVittie <smcv@debian.org>
Simon McVittie [Wed, 20 Jul 2016 12:12:45 +0000 (13:12 +0100)]
Add a simple test for dbus-daemon --fork
[Modified for the dbus-1.10 branch: DBUS_USE_TEST_BINARY is needed.]
Signed-off-by: Simon McVittie <smcv@debian.org>
Simon McVittie [Wed, 20 Jul 2016 11:00:34 +0000 (12:00 +0100)]
test: add a regression test for eval "$(dbus-launch --sh-syntax)"
[Modified for the dbus-1.10 branch: DBUS_USE_TEST_BINARY is needed.]
Signed-off-by: Simon McVittie <smcv@debian.org>
Simon McVittie [Fri, 12 Aug 2016 10:12:59 +0000 (11:12 +0100)]
test-dbus-launch-x11: extend to cover --exit-with-x11
Signed-off-by: Simon McVittie <smcv@debian.org>
Simon McVittie [Fri, 12 Aug 2016 09:58:13 +0000 (10:58 +0100)]
Add a regression test for dbus-launch in X11
Signed-off-by: Simon McVittie <smcv@debian.org>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97008
Simon McVittie [Wed, 20 Jul 2016 12:31:50 +0000 (13:31 +0100)]
test-dbus-daemon-fork: exercise closed stdin, stdout, stderr
Signed-off-by: Simon McVittie <smcv@debian.org>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97008
Simon McVittie [Wed, 20 Jul 2016 12:12:45 +0000 (13:12 +0100)]
Add a simple test for dbus-daemon --fork
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Colin Walters
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97008
Simon McVittie [Wed, 20 Jul 2016 11:00:34 +0000 (12:00 +0100)]
test: add a regression test for eval "$(dbus-launch --sh-syntax)"
Signed-off-by: Simon McVittie <smcv@debian.org>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97008
Simon McVittie [Thu, 21 Jul 2016 09:55:06 +0000 (10:55 +0100)]
_dbus_change_to_daemon_user (audit code path): set DBusError correctly
The other error cases here all set the error instead of just warning.
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97009
Simon McVittie [Thu, 21 Jul 2016 09:59:50 +0000 (10:59 +0100)]
Remove trailing newlines from _dbus_assert_not_reached
This was never necessary: _dbus_assert_not_reached() always added one.
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97009
Simon McVittie [Thu, 21 Jul 2016 07:24:38 +0000 (08:24 +0100)]
dbus-daemon, dbus-launch: cope with callers having closed standard fds
In Debian bug <https://bugs.debian.org/829348>, lightdm appears to
have been starting dbus-launch with at least one of the three
standard fds 0, 1, 2 (stdin, stdout, stderr) closed. This resulted
in the dbus-daemon's epoll_create1() returning a fd less than 3.
Later, _dbus_become_daemon() replaces fds 0-2 with /dev/null. As a
result, a subsequent call to _dbus_loop_add_watch() for the reload
pipe resulted in calling epoll_ctl on the non-epoll fd pointing to
/dev/null, which fails with EINVAL, resulting in the dbus-daemon
exiting unsuccessfully.
Unix programs are not normally expected to behave correctly when
launched with the standard fds not already open; but at the same time,
X11 autolaunching means that dbus-launch (and hence the dbus-daemon)
can get started from an arbitrarily precarious situation.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97008
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Thiago Macieira <thiago@kde.org>
(cherry picked from commit
c8f73a2a3a9d9d10587f596a62ebb64e8963197e)
Simon McVittie [Thu, 21 Jul 2016 07:23:12 +0000 (08:23 +0100)]
_dbus_ensure_standard_fds: new function to ensure std* fds are open
This function opens stdin, stdout, stderr pointing to /dev/null
if they aren't already open. Optionally, it can also replace
whatever is available on those fds with /dev/null.
To allow for use in contexts where only async-signal-safe functions
should be used, such as between fork() and a following exec(),
this function does not use conventional libdbus error handling
(which would require malloc). Instead, it sets errno and returns
an explanatory string.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97008
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Thiago Macieira <thiago@kde.org>
(cherry picked from commit
69123a6bd2adabbaec1f770fc4573fc3ed4ceca6)
Simon McVittie [Mon, 11 Jul 2016 10:09:26 +0000 (11:09 +0100)]
various tests: produce better diagnostics on error
Same root cause as <https://bugs.freedesktop.org/show_bug.cgi?id=96653>:
we didn't check whether the message was in fact an error reply.
Signed-off-by: Simon McVittie <smcv@debian.org>
Simon McVittie [Fri, 12 Aug 2016 09:17:46 +0000 (10:17 +0100)]
More NEWS for 1.11
Simon McVittie [Fri, 12 Aug 2016 09:14:46 +0000 (10:14 +0100)]
Merge branch 'dbus-1.10'
Simon McVittie [Fri, 12 Aug 2016 09:12:28 +0000 (10:12 +0100)]
NEWS for 1.10
Simon McVittie [Thu, 11 Aug 2016 15:08:39 +0000 (16:08 +0100)]
Mark WaitingForOK state as unused
It should probably be used (see #97298) but the fact that it isn't
is breaking compatibility with gcc 6, so apply a quick workaround
while we look into what's wrong here.
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97282
(cherry picked from commit
21d61180819c141e779d6ecf9919e62e768b6fd9)
Simon McVittie [Mon, 11 Jul 2016 09:52:44 +0000 (10:52 +0100)]
update-activation-environment: produce better diagnostics on error
If dbus-daemon or systemd replied to our method call with an error,
we would report it as "invalid arguments" instead of the true error
name and message.
Same root cause as <https://bugs.freedesktop.org/show_bug.cgi?id=96653>.
Signed-off-by: Simon McVittie <smcv@debian.org>
Simon McVittie [Wed, 20 Jul 2016 07:55:55 +0000 (08:55 +0100)]
config-parser: remove unnecessary newline from _dbus_log()
_dbus_log() already adds a newline.
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97009
Simon McVittie [Wed, 20 Jul 2016 07:55:14 +0000 (08:55 +0100)]
_dbus_system_log: rename to _dbus_log
This is a step towards making it write to either stderr or syslog
or both, as configured globally.
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97009
Simon McVittie [Thu, 21 Jul 2016 09:29:10 +0000 (10:29 +0100)]
_dbus_init_system_log: record a syslog tag (executable name)
Instead of hard-coding "dbus", report what the executable really is.
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97009
Simon McVittie [Wed, 20 Jul 2016 09:07:12 +0000 (10:07 +0100)]
sysdeps: move _dbus_system_log() into the shared library
This is in preparation for optionally making _dbus_warn() use it.
dbus-daemon closes its stderr under some circumstances, including
when launched by dbus-launch, which makes failures in that
situation rather hard to debug.
_dbus_system_log() is the same on Unix and Windows, so move it
to dbus-sysdeps.c. _dbus_system_logv() remains platform-specific.
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97009
[smcv: move the #include for syslog.h, too]
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Simon McVittie [Wed, 20 Jul 2016 09:14:51 +0000 (10:14 +0100)]
test-bus: don't write to the syslog
We want to emulate the behaviour of the system bus, but we don't
really want to spam the system log with lots of test messages.
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97009
Thomas Zimmermann [Tue, 9 Aug 2016 16:22:04 +0000 (18:22 +0200)]
Fix misleading indentation to avoid respective compiler warning
Signed-off-by: Thomas Zimmermann <tdz@users.sourceforge.net>
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97282
Thomas Zimmermann [Tue, 9 Aug 2016 16:19:56 +0000 (18:19 +0200)]
Protect 'i' in _handle_inotify_watch by DBUS_ENABLE_VERBOSE_MODE
Signed-off-by: Thomas Zimmermann <tdz@users.sourceforge.net>
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97282
Thomas Zimmermann [Tue, 9 Aug 2016 14:48:59 +0000 (16:48 +0200)]
Initialize 'klass' in _dbus_type_reader_recurse to NULL
Initializing 'klass' in _dbus_type_reader_recurse avoids a
compile-time warning about the variable being uninitialized.
Signed-off-by: Thomas Zimmermann <tdz@users.sourceforge.net>
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97282
Thomas Zimmermann [Tue, 9 Aug 2016 16:26:56 +0000 (18:26 +0200)]
Protect 'orig_len' in recover_unused_bytes by DBUS_ENABLE_VERBOSE_MODE
Signed-off-by: Thomas Zimmermann <tdz@users.sourceforge.net>
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97282
Thomas Zimmermann [Tue, 9 Aug 2016 16:15:25 +0000 (18:15 +0200)]
dbus-launch: Protect concat2 by DBUS_ENABLE_EMBEDDED_TESTS
Signed-off-by: Thomas Zimmermann <tdz@users.sourceforge.net>
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97282
Simon McVittie [Thu, 11 Aug 2016 15:08:39 +0000 (16:08 +0100)]
Mark WaitingForOK state as unused
It should probably be used (see #97298) but the fact that it isn't
is breaking compatibility with gcc 6, so apply a quick workaround
while we look into what's wrong here.
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97282
Simon McVittie [Fri, 29 Jul 2016 18:15:21 +0000 (19:15 +0100)]
Merge branch 'dbus-1.10'
Simon McVittie [Fri, 29 Jul 2016 18:14:18 +0000 (19:14 +0100)]
NEWS for 1.10 branch
Ioan-Adrian Ratiu [Thu, 28 Jul 2016 22:19:37 +0000 (01:19 +0300)]
configure.ac: explicitely check stdint.h
Otherwise HAVE_STDINT_H will not be defined or the var will not be
picked up from cache so builds could fail with errors like:
| ../../dbus-1.10.8/dbus/dbus-internals.h:239:8: error: ‘uintptr_t’ undeclared (first use in this function)
Signed-off-by: Ioan-Adrian Ratiu <adrian.ratiu@ni.com>
[smcv: fix Autoconf underquoting]
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Simon McVittie [Mon, 25 Jul 2016 11:11:39 +0000 (12:11 +0100)]
Merge branch 'dbus-1.10'
Simon McVittie [Mon, 25 Jul 2016 11:11:00 +0000 (12:11 +0100)]
Update NEWS for 1.11
Simon McVittie [Mon, 25 Jul 2016 10:53:13 +0000 (11:53 +0100)]
Update NEWS for merge of dbus-1.10-ci branch
Simon McVittie [Mon, 25 Jul 2016 10:43:54 +0000 (11:43 +0100)]
Merge branch 'dbus-1.10-ci' into dbus-1.10
Simon McVittie [Thu, 21 Jul 2016 07:27:36 +0000 (08:27 +0100)]
_read_subprocess_line_argv: use _dbus_ensure_standard_fds
This also gives us an opportunity to improve the error reporting.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97008
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Thiago Macieira <thiago@kde.org>
Simon McVittie [Thu, 21 Jul 2016 07:27:06 +0000 (08:27 +0100)]
dbus-launch: use _dbus_ensure_standard_fds when closing stderr
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97008
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Thiago Macieira <thiago@kde.org>
Simon McVittie [Thu, 21 Jul 2016 07:26:40 +0000 (08:26 +0100)]
dbus-launch: use _dbus_ensure_standard_fds() in the babysitter
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97008
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Thiago Macieira <thiago@kde.org>
Simon McVittie [Thu, 21 Jul 2016 07:25:21 +0000 (08:25 +0100)]
_dbus_become_daemon: use _dbus_ensure_standard_fds, and report errors
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97008
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Thiago Macieira <thiago@kde.org>
Simon McVittie [Thu, 21 Jul 2016 07:24:38 +0000 (08:24 +0100)]
dbus-daemon, dbus-launch: cope with callers having closed standard fds
In Debian bug <https://bugs.debian.org/829348>, lightdm appears to
have been starting dbus-launch with at least one of the three
standard fds 0, 1, 2 (stdin, stdout, stderr) closed. This resulted
in the dbus-daemon's epoll_create1() returning a fd less than 3.
Later, _dbus_become_daemon() replaces fds 0-2 with /dev/null. As a
result, a subsequent call to _dbus_loop_add_watch() for the reload
pipe resulted in calling epoll_ctl on the non-epoll fd pointing to
/dev/null, which fails with EINVAL, resulting in the dbus-daemon
exiting unsuccessfully.
Unix programs are not normally expected to behave correctly when
launched with the standard fds not already open; but at the same time,
X11 autolaunching means that dbus-launch (and hence the dbus-daemon)
can get started from an arbitrarily precarious situation.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97008
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Thiago Macieira <thiago@kde.org>
Simon McVittie [Thu, 21 Jul 2016 07:23:12 +0000 (08:23 +0100)]
_dbus_ensure_standard_fds: new function to ensure std* fds are open
This function opens stdin, stdout, stderr pointing to /dev/null
if they aren't already open. Optionally, it can also replace
whatever is available on those fds with /dev/null.
To allow for use in contexts where only async-signal-safe functions
should be used, such as between fork() and a following exec(),
this function does not use conventional libdbus error handling
(which would require malloc). Instead, it sets errno and returns
an explanatory string.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=97008
Signed-off-by: Simon McVittie <smcv@debian.org>
Reviewed-by: Thiago Macieira <thiago@kde.org>
Ralf Habacker [Wed, 25 May 2016 08:54:32 +0000 (10:54 +0200)]
dbus: do not include unused afxres.h on windows.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=96181
Reviewed-by: simon.mcvittie@collabora.co.uk
Simon McVittie [Mon, 11 Jul 2016 11:10:39 +0000 (12:10 +0100)]
dbus-launch: add --exit-with-x11 option
This is more suitable for distributions' Xsession scripts: it verifies
that X is already available, and so never results in an attempt to poll
stdin.
We read the machine UUID because it is needed to set the X atoms.
x11_init() assumes that the machine UUID (global variable) has been
set, either via read_machine_uuid_if_needed() or save_machine_uuid().
This is pretty tangled, but to make The Right Thing happen
automatically, we'd need to redo dbus-launch in terms of DBusError.
Reviewed-by: Will Thompson
Reviewed-by: Thiago Macieira
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=39197
Signed-off-by: Simon McVittie <smcv@debian.org>
Deepika Aggarwal [Thu, 3 Dec 2015 05:52:56 +0000 (11:22 +0530)]
Be explicit about enum comparison in bus_config_load()
XML_Parse() can potentially return values other than OK (1) or ERROR (0),
and they aren't errors.
Signed-off-by: Deepika Aggarwal <deepika.a@samsung.com>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=93205
[smcv: improve commit message, fix whitespace]
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Deepika Aggarwal [Mon, 7 Dec 2015 11:26:06 +0000 (16:56 +0530)]
bus: reassure static analysis tool that server slot allocation can't fail
The NULL-dereference is not actually possible in this case, because we
know that the allocation and setup were done previously.
Signed-off-by: Deepika Aggarwal <deepika.a@samsung.com>
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=93210
Nick Lewycky [Thu, 28 Jan 2016 20:58:43 +0000 (12:58 -0800)]
Add an invalid state in DBusAuthState to use for invalid lines in auth scripts.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=93909
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Simon McVittie [Thu, 30 Jun 2016 13:47:21 +0000 (14:47 +0100)]
NEWS for 1.11.x
Simon McVittie [Thu, 30 Jun 2016 13:41:23 +0000 (14:41 +0100)]
Merge branch 'dbus-1.10'
Simon McVittie [Thu, 30 Jun 2016 13:40:24 +0000 (14:40 +0100)]
Update NEWS
Philip Withnall [Tue, 15 Dec 2015 09:25:36 +0000 (09:25 +0000)]
spec: Clarify the marshaling format in a few minor ways
This is an attempt to make that section a little clearer. I don’t think
any factual inaccuracies have been fixed (because I couldn’t find any).
Including some wording and an example by Simon McVittie.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=93382
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Simon McVittie [Thu, 30 Jun 2016 12:57:29 +0000 (13:57 +0100)]
Merge branch 'dbus-1.10' into dbus-1.10-ci
Simon McVittie [Thu, 30 Jun 2016 12:56:44 +0000 (13:56 +0100)]
Merge branch 'dbus-1.10'
WaLyong Cho [Tue, 7 Jun 2016 14:26:42 +0000 (23:26 +0900)]
activation: set children oom_score_adj to 0
If dbus is running as systemd service, dbus daemon is running with
oom_score_adj -900 by OOMScoreAdjust=-900. And children will also have
same value with dbus daemon.
To avoid this, set the child itself values after fork () to 0.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=32851
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Руслан Ижбулатов [Thu, 23 Jun 2016 12:26:45 +0000 (12:26 +0000)]
Use dbus_set_error_from_message() to check for an error
Do not rely on dbus_message_get_args() to turn an ERROR message
into DBusError.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=96653
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Thiago Macieira [Tue, 29 Mar 2016 19:15:15 +0000 (12:15 -0700)]
Quote the DBUS_SESSION_BUS_ADDRESS variable in the shell file
Some D-Bus daemon versions set multiple addresses separated by semi-colon,
which breaks sourcing of the file.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=94746
Signed-off-by: Thiago Macieira <thiago@kde.org>
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Thiago Macieira [Tue, 29 Mar 2016 19:15:15 +0000 (12:15 -0700)]
Quote the DBUS_SESSION_BUS_ADDRESS variable in the shell file
Some D-Bus daemon versions set multiple addresses separated by semi-colon,
which breaks sourcing of the file.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=94746
Signed-off-by: Thiago Macieira <thiago@kde.org>
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Ralf Habacker [Mon, 16 May 2016 13:35:01 +0000 (15:35 +0200)]
Migrate platform specific spawn tests into cross platform tests.
Also enable segfault checks on windows because the reason why it
has been disabled has been fixed with bug #95155.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=95191
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Simon McVittie [Mon, 16 May 2016 12:48:37 +0000 (13:48 +0100)]
travis: compile with mingw in both production and debug flavours
This should detect anything that is only a build failure when embedded
tests or verbose mode are enabled, such as the missing include fixed
in commit 4858faf. I'm not enabling embedded tests and verbose mode
orthogonally because we don't want CI builds to take too long.
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
(cherry picked from commit
8f6fe1451ead14b82475177f40bbc2c256cba876)
Simon McVittie [Wed, 18 May 2016 17:14:16 +0000 (18:14 +0100)]
Merge branch 'dbus-1.10' into dbus-1.10-ci
Simon McVittie [Mon, 16 May 2016 13:10:36 +0000 (14:10 +0100)]
test-segfault: add missing include of <stdio.h> on Windows
Needed for fprintf (stderr, ...).
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
(cherry picked from commit
07b7dcd7178f927cd0b3a3282396b7f99c0b1d29)
Simon McVittie [Mon, 16 May 2016 12:48:37 +0000 (13:48 +0100)]
travis: compile with mingw in both production and debug flavours
This should detect anything that is only a build failure when embedded
tests or verbose mode are enabled, such as the missing include fixed
in commit 4858faf. I'm not enabling embedded tests and verbose mode
orthogonally because we don't want CI builds to take too long.
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
Simon McVittie [Mon, 16 May 2016 13:10:36 +0000 (14:10 +0100)]
test-segfault: add missing include of <stdio.h> on Windows
Needed for fprintf (stderr, ...).
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
Ralf Habacker [Sat, 14 May 2016 09:42:04 +0000 (11:42 +0200)]
Fix memory leak issue in spawn code on Windows.
In _dbus_babysitter_block_for_child_exit () use spawning
thread handle to have a reliable way to detect spawning
thread termination.
See https://bugs.freedesktop.org/show_bug.cgi?id=95191#c33
for more informations.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=95191
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Ralf Habacker [Sat, 14 May 2016 08:00:20 +0000 (10:00 +0200)]
Fix compiling on windows/mingw with automake build system.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=95191
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Ralf Habacker [Mon, 16 May 2016 10:57:04 +0000 (12:57 +0200)]
Merge branch 'dbus-1.10'
Ralf Habacker [Mon, 16 May 2016 10:52:25 +0000 (12:52 +0200)]
Fix ambiguous setup of DBusBabySitter struct member child_handle on Windows.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=95191
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Simon McVittie [Fri, 13 May 2016 13:54:33 +0000 (14:54 +0100)]
dbus-spawn-win: Do not attempt to call child_setup (again)
This reinstates part of commit
420f3474, fixing a mis-merge in
commit
8c0d5980.
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Ralf Habacker [Fri, 13 May 2016 13:32:44 +0000 (15:32 +0200)]
Merge branch 'dbus-1.10'
Ralf Habacker [Wed, 11 May 2016 06:06:04 +0000 (08:06 +0200)]
Prefix debug messages displayed with _dbus_verbose() with a timestamp.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=95191
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Ralf Habacker [Tue, 10 May 2016 14:51:25 +0000 (16:51 +0200)]
Display thread id in _dbus_verbose to be able to see threading issues.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=95191
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Ralf Habacker [Tue, 10 May 2016 14:53:57 +0000 (16:53 +0200)]
On Windows make access to member 'refcount' of struct DBusBabysitter thread safe.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=95191
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Ralf Habacker [Thu, 12 May 2016 22:56:42 +0000 (00:56 +0200)]
Eliminates a race condition accessing DBusBabysitter instance at startup of babysitter() on Windows.
Ensure that the babysitter thread already owns its one reference to the
babysitter when it starts up, and eliminates the race condition.
This patch requires that DBusBabysitter refcounting is thread-safe
and is based on an analysis and proposal of Simon Mc Vittie.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=95191
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Ralf Habacker [Mon, 9 May 2016 21:38:21 +0000 (23:38 +0200)]
Fix memory leak in _dbus_replace_install_prefix() on success case on Windows.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=95191
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Ralf Habacker [Mon, 9 May 2016 19:50:56 +0000 (21:50 +0200)]
Merge branch 'dbus-1.10'
Ralf Habacker [Mon, 9 May 2016 15:31:17 +0000 (17:31 +0200)]
Suppress Windows popups and jit debugger when app crashes with exception.
Based on a patch from Yiyang Fei.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=95155
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Lennart Poettering [Wed, 4 May 2016 16:06:17 +0000 (18:06 +0200)]
sysdeps: increase listen() backlog of AF_UNIX sockets to SOMAXCONN
Previously, the listen() backlog was set to an arbitrary 30. This means
that if dbus-daemon is overloaded only 30 more connections may be queued
by the kernel, before connect() fails with EAGAIN. (Note that EAGAIN !=
EINPROGRESS -- the latter is what is returned if a connection is queued
and being processed for asynchronous sockets; EAGAIN in this case is
really an error, that cannot be recovered from).
Most software simply sets SOMAXCONN as backlog for AF_UNIX sockets, to
allow queuing of as many connections as the kernel allows. SOMAXCONN is
128 on Linux, which is not particularly high, but at least higher than
30.
This patch changes dbus-daemon to do the same.
I noticed this when flooding dbus-daemon with a lot of connections,
where it pretty quickly ceased to respond, much earlier than it really
should.
Note that the backlog has nothing to do with the number of concurrent
connections allowed, it simply controls how many queued, but not
accept()ed connections there may be on the listening socket.
https://bugs.freedesktop.org/show_bug.cgi?id=95264
Ralf Habacker [Mon, 2 May 2016 05:19:41 +0000 (07:19 +0200)]
Merge branch 'dbus-1.10'
Yiyang Fei [Wed, 27 Apr 2016 15:10:06 +0000 (08:10 -0700)]
Fix crash in test-spawn unit test app on Windows.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=95160
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Ralf Habacker [Fri, 29 Apr 2016 07:47:51 +0000 (09:47 +0200)]
Fix assert in test-spawn caused by missing initialization of DBusError instance on gcc builds.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=95160
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
(cherry picked from commit
9323a621e868d6a5b628b89696b1efe0300ff939)
Ralf Habacker [Fri, 29 Apr 2016 07:47:51 +0000 (09:47 +0200)]
Fix assert in test-spawn caused by missing initialization of DBusError instance on gcc builds.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=95160
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Ralf Habacker [Thu, 14 Apr 2016 11:58:13 +0000 (13:58 +0200)]
dbus_internal_do_not_use_try_message_data: Calling _dbus_message_loader_get_max_message_size(loader) is only useful for its return value, which is ignored (CID 54725).
Reported by Coverity: CID 54725: Useless call (USELESS_CALL)
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=90021
Simon McVittie [Mon, 7 Mar 2016 20:54:05 +0000 (20:54 +0000)]
1.11.3
Simon McVittie [Mon, 7 Mar 2016 19:45:15 +0000 (19:45 +0000)]
Merge branch 'dbus-1.10' into dbus-1.10-ci