platform/upstream/dbus.git
8 years agofix SVACE bugs 73/66573/7 accepted/tizen/common/20160422.172237 accepted/tizen/ivi/20160423.060622 accepted/tizen/mobile/20160423.055705 accepted/tizen/tv/20160423.060059 accepted/tizen/wearable/20160423.060351 submit/tizen/20160422.093504
Adrian Szyndela [Tue, 19 Apr 2016 10:37:40 +0000 (12:37 +0200)]
fix SVACE bugs

WGID 31177: strcpy/stpcpy->snprintf
WGID 31178: strcpy/stpcpy->snprintf
WGID 31179: strcpy/stpcpy->snprintf/strncpy
WGID 31180: use _dbus_strdup for string duplication
WGID 31183: sprintf->snprintf
WGID 31184: sprintf->snprintf
WGID 31185: sprintf->snprintf
WGID 31186: sprintf->snprintf
WGID 31200: added explanation and unified conditions
WGID 31210: added cast to __u64
all other sprintf->snprintf in dbus-transport-kdbus.c

Change-Id: I8a488d37ec8be66370cba7b88a464787dcfeefc3

8 years agopackaging: Integrate libdbus and dbus spec files 60/66560/5
Karol Lewandowski [Tue, 19 Apr 2016 11:23:37 +0000 (13:23 +0200)]
packaging: Integrate libdbus and dbus spec files

With dbus-1.10 depending on systemd code there is no benefit
in keeping separate spec files.

Change-Id: I3a9311e1730b51b3d4ff1dda513b564d19065aa9

8 years agopackaging: Remove dbus-x11 flavor 59/66559/4
Karol Lewandowski [Tue, 19 Apr 2016 11:00:12 +0000 (13:00 +0200)]
packaging: Remove dbus-x11 flavor

dbus-x11 is not used on tizen.org in any of available profiles.

Change-Id: I94205fbdb526a96202bd2f2165e6a5bcf2c2a080

8 years agopackaging: Remove documentation packages 58/66558/4
Karol Lewandowski [Tue, 19 Apr 2016 10:45:06 +0000 (12:45 +0200)]
packaging: Remove documentation packages

Change-Id: I3b2ddc48b26819452ffa73d25d35177997c844b1

8 years agoAdded a strategy for selecting default protocol type. 29/66529/3
Adrian Szyndela [Tue, 19 Apr 2016 09:38:07 +0000 (11:38 +0200)]
Added a strategy for selecting default protocol type.

When a message is created, it is created without context
of a bus. Thus, we have to guess what type of protocol
is correct for the message. This commit introduces
a possibility to influence guessing.

An environment variable is introduced (DBUS_DEFAULT_PROTOCOL_STRATEGY),
with following values recognized:
* first-bus - messages created for type of first opened bus;
* last-bus - messages created for type of last opened bus;
* last-message - messages created for type of last sent message;
* dbus1 - messages created with dbus1 protocol;
* gvariant - messages created with GVariant protocol.

last-bus is a default strategy.

Change-Id: I2a185761973191ad5917bc2b6f4bfb2f4a3ed547

8 years agofixed some compiler/Valgrind warnings 81/66381/2
Adrian Szyndela [Mon, 18 Apr 2016 12:27:23 +0000 (14:27 +0200)]
fixed some compiler/Valgrind warnings

Change-Id: I5c9c0f7c5e50f89e69db9e5f000b181ebc98add6

8 years agorefactoring: org.freedesktop.DBus method handling simplifications 62/66562/1
Lukasz Skalski [Tue, 19 Apr 2016 09:10:09 +0000 (11:10 +0200)]
refactoring: org.freedesktop.DBus method handling simplifications

Change-Id: I4060e9ad4a6703cdcfdccd3bce69fd2b2c958031

8 years agorefactoring: bloom filters handling moved to low-level 00/66300/1
Adrian Szyndela [Fri, 15 Apr 2016 11:24:50 +0000 (13:24 +0200)]
refactoring: bloom filters handling moved to low-level

Constructing bloom filters belongs to low level API now.

Change-Id: I95b524e91905029a1a040a95204e7008120cd89c

8 years agorefactoring: cleanup 99/66299/1
Adrian Szyndela [Fri, 15 Apr 2016 10:57:49 +0000 (12:57 +0200)]
refactoring: cleanup

String name to unique id conversion simplified.
Item names for debug strings simplified.

Change-Id: I934c6e3f64708db777dfdff2ffb4b4f5e7fa41a2

8 years agorefactoring: using local types where available 98/66298/1
Adrian Szyndela [Fri, 15 Apr 2016 10:51:58 +0000 (12:51 +0200)]
refactoring: using local types where available

Change-Id: I91cb72a2c52d1d55fdfd8afb1e6599483d637ea1

8 years agorefactoring: low-level API types changed to low-level 97/66297/1
Adrian Szyndela [Fri, 15 Apr 2016 09:41:18 +0000 (11:41 +0200)]
refactoring: low-level API types changed to low-level

Change-Id: I934800df6da5ad5fa34366683541782eb88daa9f

8 years agorefactoring: coding style 96/66296/1
Adrian Szyndela [Fri, 15 Apr 2016 09:24:25 +0000 (11:24 +0200)]
refactoring: coding style

Change-Id: I2a39f9b86a5ebbde11ee496ecc40a6cc08e6a1a1

8 years agofix for memleaks 16/66016/2
Adrian Szyndela [Thu, 14 Apr 2016 12:35:57 +0000 (14:35 +0200)]
fix for memleaks

Freeing non-converted message after conversion
Freeing string objects
Freeing unique name in kdbus transport at end-of-life
Freeing matchmaker in kdbus transport at end-of-life

Change-Id: Iae4e231c52b78af3efd5a8a366fc01e50b03feeb

8 years agochecking policy: don't check 'ins' if not a method call 84/65584/1
Adrian Szyndela [Mon, 11 Apr 2016 12:31:59 +0000 (14:31 +0200)]
checking policy: don't check 'ins' if not a method call

Change-Id: I666bc6474475906aa0ca2de96e47bfaa2c402ce2

8 years agoUpdate to dbus-1.10.6 15/64815/1 accepted/tizen/common/20160406.143833 accepted/tizen/ivi/20160406.014807 accepted/tizen/mobile/20160406.014711 accepted/tizen/tv/20160406.014732 accepted/tizen/wearable/20160406.014752 submit/tizen/20160405.144419
Adrian Szyndela [Thu, 18 Feb 2016 14:17:01 +0000 (15:17 +0100)]
Update to dbus-1.10.6

Note: from version 1.10 it's required to have systemd support to be enabled not
only in dbus-daemon code but also in libdbus.  This is because maintainers removed
built-in systemd support code (dbus/sd-daemon.c) and now depend on libsystemd-
provided functionality.

Additionally, update pkgconfig name to libsystemd (as required by configure.ac)

Change-Id: I2056086a9281543695643ac31fd3ab8648a4d205

8 years agoMOVe License patch (LGPL2.1+, BSD2.0 license files)
Hyotaek, Shim [Fri, 18 Mar 2016 06:21:43 +0000 (15:21 +0900)]
MOVe License patch (LGPL2.1+, BSD2.0 license files)

Signed-off-by: Hyotaek, Shim <hyotaek.shim@samsung.com>
Change-Id: Ie0e4364412bdf83cde0dd33faddef7a22b856f58

8 years agoadd dbuspolicy support
Adrian Szyndela [Tue, 15 Dec 2015 13:25:50 +0000 (14:25 +0100)]
add dbuspolicy support

Change-Id: I1bef17d3930a2ca626d3c002eaa10cf6f16c8aac

8 years agodbus:modified to default allow own & send method_call for system bus
sanghyeok.oh [Thu, 4 Feb 2016 06:36:53 +0000 (15:36 +0900)]
dbus:modified to default allow own & send method_call for system bus

Change-Id: I439cfcbf9c8e44217f6a906836c579dd952b9ec7
Signed-off-by: sanghyeok.oh <sanghyeok.oh@samsung.com>
8 years agoFixed remarshalling and getting signature
Adrian Szyndela [Tue, 2 Feb 2016 13:27:50 +0000 (14:27 +0100)]
Fixed remarshalling and getting signature

A few fixes:
- remarshalling to GVariant lacked locking of messages. It caused
  missing signature and offsets.
- dbus_message_get_signature returned signature with outer parentheses.
- removed unused param in _dbus_header_load_gvariant().
- editorial corrections in kdbus_decode_dbus_message().

Change-Id: I29ca21cef6769b725be0c98580f54313bda842c8

8 years agoRemembering last offset and position at the message level
Adrian Szyndela [Mon, 25 Jan 2016 08:38:10 +0000 (09:38 +0100)]
Remembering last offset and position at the message level

In GVariant, offsets in structs are stored for every end
of variable-sized elements except the last one.
We never know if an element we just added is the last one.
Therefore, the offset may be added when next element is added.
However, some users use multiple iterators. This change fixes
it at the root level - last offset value and position is kept
in the message.

Change-Id: I954657424d9da075edb10e5630b27df5bca8f917

8 years agoDisable GVARIANT feature temporally
INSUN PYO [Mon, 18 Jan 2016 10:10:51 +0000 (19:10 +0900)]
Disable GVARIANT feature temporally

Signed-off-by: INSUN PYO <insun.pyo@samsung.com>
Change-Id: I84cf169d64f59b6b2086ade14a827b252ab44b97

8 years agoAdd kdbus transport
Adrian Szyndela [Thu, 7 Jan 2016 13:48:48 +0000 (14:48 +0100)]
Add kdbus transport

This reverts commit 197db0df78e0a9f82df4162885854f99ff749505,
which reverts commit 771f5155c9a393a242329988d56661a51fc04e6c.

Change-Id: I4133020c585a6f9b42f487ffae4d2005b25a1d08

8 years agodbus:MOVe:modify license ID
sanghyeok.oh [Wed, 30 Dec 2015 07:12:18 +0000 (16:12 +0900)]
dbus:MOVe:modify license ID

Change-Id: Iea6db9311c918f99059bb98e8bd396b36387d902
Signed-off-by: sanghyeok.oh <sanghyeok.oh@samsung.com>
8 years agoRevert "Add kdbus transport"
Hyotaek, Shim [Wed, 16 Dec 2015 06:27:54 +0000 (15:27 +0900)]
Revert "Add kdbus transport"

This reverts commit 771f5155c9a393a242329988d56661a51fc04e6c.

Change-Id: I73b6adc6b76b63e972f7228c88c18dabda81b006

8 years agoAdd kdbus transport
Adrian Szyndela [Fri, 30 Oct 2015 10:19:33 +0000 (11:19 +0100)]
Add kdbus transport

This commit introduces ABI break due to size change of DBusMessageIter
structure. Consequently, all packages depending on libdbus need to be rebuilt.

Other authors:
Paweł Szewczyk <p.szewczyk@samsung.com>
Karol Lewandowski <k.lewandowsk@samsung.com>
and possibly others

Change-Id: Ie04b34295c38e5aaac63982996fa9eddc97dd696

8 years agoRevert "Perform Cynara runtime policy checks by default"
Lukasz Skalski [Wed, 23 Sep 2015 09:24:29 +0000 (11:24 +0200)]
Revert "Perform Cynara runtime policy checks by default"

This reverts commit e8610297cf7031e94eb314a2e8c11246f4405403.

Change-Id: Ifb60464c705f5b4e92f02f9e809d23e982d31c3d
Signed-off-by: Lukasz Skalski <l.skalski@samsung.com>
8 years agoPerform Cynara runtime policy checks by default
Jacek Bukarewicz [Tue, 23 Jun 2015 09:08:48 +0000 (11:08 +0200)]
Perform Cynara runtime policy checks by default

This change introduces http://tizen.org/privilege/internal/dbus privilege
which is supposed to be available only to trusted system resources.
Checks for this privilege are used in place of certain allow rules to
make security policy more strict.

For system bus sending and receiving signals now requires
http://tizen.org/privilege/internal/dbus privilege. Requesting name
ownership and sending methods is still denied by default.

For session bus http://tizen.org/privilege/internal/dbus privilege
is now required for requesting name, calling methods, sending and receiving
signals.

Services are supposed to override these default settings to implement their
own security policy.

Change-Id: Ifb4a160bf6e0638404e0295a2e4fa3077efd881c
Signed-off-by: Jacek Bukarewicz <j.bukarewicz@samsung.com>
8 years agopackaging: remove character class negation for sed
José Bollo [Wed, 29 Apr 2015 13:58:13 +0000 (15:58 +0200)]
packaging: remove character class negation for sed

When running the command sed provided by toybox, the expression
[^[:cntrl:]] is not understood, what forbids to detect DBUS and
cause further errors.

This patch removes this expression.

Change-Id: If30543fadddf8b6811e14b548a747be99612894b
Signed-off-by: José Bollo <jose.bollo@open.eurogiciel.org>
8 years agoDo not package libdbus-1.la file
Jacek Bukarewicz [Fri, 20 Mar 2015 11:16:55 +0000 (12:16 +0100)]
Do not package libdbus-1.la file

Change-Id: I8de1d91e6ab3c891de9549bfb94ae637d7ff664a

8 years agoAdd missing security manifests
Jacek Bukarewicz [Fri, 14 Nov 2014 09:47:29 +0000 (10:47 +0100)]
Add missing security manifests

Change-Id: I765c1f189a4496620e5edc8c02b3055db5b45c09

8 years agoDo not rely on Cynara cache when processing check rules
Jacek Bukarewicz [Tue, 3 Mar 2015 16:37:39 +0000 (17:37 +0100)]
Do not rely on Cynara cache when processing check rules

Cynara cache was required when processing messages that have been
blocked at the sender's message queue. Reliance on cache turned out to be
unacceptable due to the fact that some policies are not cacheable. For example
responses provided by Cynara askuser plugin might be single-use and thus cannot
be cached.

The solution is to attach deferred message to the message object when policy result
is unavailable. Upon next bus_check_privilege call use response from Cynara which
is saved in deferred message object.

Change-Id: I17152343540d7b8d13ad3540c25c043d57aa5949

8 years agoFix several BusResult/dbus_bool_t mismatches
Jacek Bukarewicz [Mon, 9 Feb 2015 15:25:31 +0000 (16:25 +0100)]
Fix several BusResult/dbus_bool_t mismatches

They were found by temporarily redefining BusResult in the following way:

 typedef enum { BUS_RESULT_TRUE_E, BUS_RESULT_FALSE_E, BUS_RESULT_LATER_E } bus_result_t;
 typedef struct { bus_result_t result; } BusResult;
 #define BUS_RESULT_TRUE ((BusResult){BUS_RESULT_TRUE_E})
 #define BUS_RESULT_FALSE ((BusResult){BUS_RESULT_FALSE_E})
 #define BUS_RESULT_LATER ((BusResult){BUS_RESULT_LATER_E})

It doesn't compile because equality operator is not defined for structs.
Also, structs are not allowed in switch statement. However, some errors
indicated type mismatches which are now fixed.

Change-Id: I0eb5368359f342e0f4239a2ad95d34b9a8e10a23
Signed-off-by: Jacek Bukarewicz <j.bukarewicz@samsung.com>
8 years agoAdd <check own="..." > support
Jacek Bukarewicz [Thu, 27 Nov 2014 10:26:21 +0000 (11:26 +0100)]
Add <check own="..." > support

Policy result unavailability is handled like send rules - dispatching
messages from the sender is blocked and resumed when result becomes
available.

Handler of "RequestName" method needs to return BUS_RESULT_LATER when
policy result is not known therefore its return type is modified.
Since bus message handlers are put into function pointer array other
message handler function singatures are also affected.

Change-Id: I4c2cbd4585e41fccd8a30f825a8f0d342ab56755

8 years agoHandle receive rule result unavailability and message broadcasts
Jacek Bukarewicz [Fri, 28 Nov 2014 11:39:33 +0000 (12:39 +0100)]
Handle receive rule result unavailability and message broadcasts

When message is sent to the addressed recipient and receive rule
result is unavailable we don't want to block the sender
as it most likely will be the privileged service, so instead we queue
it at the recipient. Any further messages sent to it will be queued to
maintain message order. Once the answer from Cynara arrives messages are
dispatched from the recipient queue. In such case full dispatch is
performed - messages are sent to addressed recipient and other
interested connections.
Messages sent to non-addressed recipients (eavesdroppers or broadcast
message recipients) are handled in a similar way. The difference is
that it is not full dispatch meaning message is sent to a single recipient.

Change-Id: Iecd5395f75a4c7811fa97247a37d8fc4d42e8814

8 years agoDisable message dispatching when send rule result is not known
Jacek Bukarewicz [Fri, 28 Nov 2014 11:07:39 +0000 (12:07 +0100)]
Disable message dispatching when send rule result is not known

When unicast message to addressed recipient is sent and policy result
is not available message dispatch from the sender is disabled.
This also means that any further messages from the given connection are
put into the incoming queue. If response is received message dispatching
is resumed. This time answer is expected to be in cache so the message is
processed synchronously.
Receive rule result unavailability is not yet handled - such messages are
rejected. Also, if message is sent to non-addressed recipient message
is silently dropped.

Change-Id: Ia45905baf667ca42f386c1def108eca190d615bb

8 years agoIntegration of asynchronous security checks
Jacek Bukarewicz [Thu, 27 Nov 2014 17:11:05 +0000 (18:11 +0100)]
Integration of asynchronous security checks

This commit introduces basic framework for asynchronous policy
checks and Cynara integration code. Functions for checking security
policy can now return third value - BUS_RESULT_LATER denoting check
result unavailability. Whenever policy checker cannot decide on the
result of the check it is supposed to allocate DeferredMessage structure
that will be passed to the upper layers which can decide what should be
done in such situation.
Proper handling of such case will be implemented in subsequent commits.
Currently such return value results in message denial.

Change-Id: I324b6ab68442e493853d8fe219c7a37fbd831872

8 years agopolicy: add <check> element
Patrick Ohly [Wed, 30 Jul 2014 08:00:59 +0000 (10:00 +0200)]
policy: add <check> element

The new <check> element is almost the same as <allow> and <deny>. The
difference is that it has an additional "privilege" parameter which
will be tested at runtime. Depending on the outcome of the check, the
rule turns into an allow or deny rule.

Executing these checks will be implemented separately. At the moment,
a <check> is basically the same as <deny>.

The representation of a rule grows by one additional pointer and needs
one additional bit to represent <check> in addition to <allow>/<deny>.
Reordering elements might mitigate this effect.

Change-Id: I25baa802fdf41413a78200273c3a0b17ae7f1cfa

8 years agoGetConnectionCredentials - add smack support
Patrick Ohly [Fri, 20 Jun 2014 14:55:00 +0000 (16:55 +0200)]
GetConnectionCredentials - add smack support

A process should never change its Smack label while connected to
D-Bus. If it did, we would end up with race conditions around
permission checking. Therefore we can retrieve the Smack label once,
when the process connects, and use that label whenever it is needed.

A new public libdbus API also gets added: dbus_connection_get_smack_label()
This is primarily for dbus-daemon, but may also be useful for other applications
creating direct connections.

Change-Id: I16ec50a031809aab879a543ec2d7effd56768bf1

8 years agoDrop capabilities in user dbus session
Rafal Krypa [Wed, 29 Oct 2014 10:39:37 +0000 (11:39 +0100)]
Drop capabilities in user dbus session

In the systemd user service file, drop all process capabilities. The
capabilities will be inherited from parent process (systemd --user). They
are meant to be inherited by launcher process, but are of no use for
dbus. Since they would give unneeded privileges to dbus, they should be
dropped.

Change-Id: I89a1a1b21d07380f68c9933aab272ebe2b08a889
Signed-off-by: Rafal Krypa <r.krypa@samsung.com>
8 years agoUpdate packaging to 1.8.2
Patrick Ohly [Fri, 20 Jun 2014 13:29:53 +0000 (15:29 +0200)]
Update packaging to 1.8.2

Change-Id: I8c0bdcaa11802e992c43c629d3f413d21bd9d159

8 years agoadd a small script in /etc/profile.d/dbus.sh to export DBUS_SESSION_BUS_ADDRESS
Stephane Desneux [Thu, 23 Oct 2014 11:21:39 +0000 (13:21 +0200)]
add a small script in /etc/profile.d/dbus.sh to export DBUS_SESSION_BUS_ADDRESS

If the dbus session address is not defined, this small script will try to
pick the address from the environment of the systemd --user process for the current
user.

Typically, this allows to log on a target through ssh, then run su - <user> and
have the same environment as the shells inside the user session.

Change-Id: Id6133077bf9943c4203f7d993b8942dc1455bef5
Signed-off-by: Stephane Desneux <stephane.desneux@open.eurogiciel.org>
8 years agoMake dbus-1 dependency provided by dbus
Jacek Bukarewicz [Thu, 4 Sep 2014 08:50:59 +0000 (10:50 +0200)]
Make dbus-1 dependency provided by dbus

dbus-1 was originally provided by dbus package. Commit 8eeae5fd70
fixed circular dependency between dbus and systemd. It also moved
dbus-1 provision from dbus to dbus-devel probably by mistake.
Certain packages like gumd or polkit have explicit dependency on
dbus-1 which makes them install dbus-devel unnecessarily.

Change-Id: I77a6561eff6f8c6ce84d849df936dda2c290a4d2
Signed-off-by: Jacek Bukarewicz <j.bukarewicz@samsung.com>
8 years agodbus-x11 should not try to build into a pure wayland platform.
Ronan Le Martret [Wed, 29 Jan 2014 11:06:31 +0000 (12:06 +0100)]
dbus-x11 should not try to build into a pure wayland platform.

This package is designed for an X server platform.

In a system released, based exclusively on Wayland platform, it should be in a excluded status.

Tizen devel  | Tizen release
_____________|______________
succeeded    | succeeded
excluded     | excluded
broken       |
failed       |
unresolvable |
blocked      |
locked       |
note:
    - The disabled status is only
      allowed for OBS administration.

Change-Id: I1495b66a2e32b580223229ead6bc66426ff680a6
Signed-off-by: Ronan Le Martret <ronan@fridu.net>
8 years agolibdbus:build-require pkg-config
Adrian Negreanu [Mon, 25 Nov 2013 12:11:14 +0000 (14:11 +0200)]
libdbus:build-require pkg-config

otherwise, configure fails to find libsmack, which
happens to be the first lib searched with pkg-config

  checking for _NSGetEnviron... no
  checking for LIBSMACK... configure: error: libsmack is
  required to enable smack support
  error: Bad exit status from /var/tmp/rpm-tmp.VUiY9N

Change-Id: Ibf01a52ae6274503e2490f835fbd686e985807e8
Signed-off-by: Adrian Negreanu <adrian.m.negreanu@intel.com>
8 years agopackaging: enable Smack support
Patrick McCarty [Mon, 7 Oct 2013 19:39:15 +0000 (12:39 -0700)]
packaging: enable Smack support

Change-Id: Ib400feec8193bb5c98347b778e0aa92a30d954c3
Signed-off-by: Patrick McCarty <patrick.mccarty@linux.intel.com>
8 years agoInstall dbus directories
Chengwei Yang [Tue, 10 Sep 2013 10:26:16 +0000 (18:26 +0800)]
Install dbus directories

At previous, these directories owned by libdbus. However, the previous
commit 8eeae5f split libdbus into a separate .spec and doesn't install
these directoris.

These directoris are critical to dbus, for example, without
/etc/dbus-1/{system.d,session.d}, the dbus-daemon system bus and session
bus can not launch.

Change-Id: Ia98bd56171492c90e4a078b39adf08cc802cb955

8 years agofix systemd->dbus->systemd circular dependency
Adrian Negreanu [Tue, 6 Aug 2013 12:02:04 +0000 (15:02 +0300)]
fix systemd->dbus->systemd circular dependency

extract libdbus and dbus-devel as separate packages

Change-Id: Ia097a2b3fee2911ee89000dd2d5762a112bf17f6
Signed-off-by: Adrian Negreanu <adrian.m.negreanu@intel.com>
8 years agoupdating changelog
Michael Leibowitz [Mon, 22 Jul 2013 12:56:06 +0000 (05:56 -0700)]
updating changelog

8 years agoEnforce smack policy from conf file
Brian McGillion [Mon, 6 Feb 2012 16:48:30 +0000 (18:48 +0200)]
Enforce smack policy from conf file

8 years agoEnable checking of smack context from DBus interface
Brian McGillion [Mon, 6 Feb 2012 16:46:05 +0000 (18:46 +0200)]
Enable checking of smack context from DBus interface

Conflicts:
bus/driver.c
cmake/CMakeLists.txt

Change-Id: Ibc9d1ccb86c3b28d8df3a4becf33ba30234832d8

8 years agoresetting manifest requested domain to floor
Alexandru Cornea [Fri, 28 Jun 2013 20:42:49 +0000 (23:42 +0300)]
resetting manifest requested domain to floor

8 years agoUpdate to dbus 1.6.12
Anas Nashif [Wed, 19 Jun 2013 10:16:37 +0000 (06:16 -0400)]
Update to dbus 1.6.12

Signed-off-by: Anas Nashif <anas.nashif@intel.com>
8 years agoupdate changelog
Simon McVittie [Wed, 19 Jun 2013 14:31:03 +0000 (15:31 +0100)]
update changelog

- TZPC-3044, CVE-2013-2168: fix local denial of service (backport from 1.6.12)
- TZPC-1971: make libdbus thread-safe by default (backport from 1.7.4/1.7.6)

8 years agopackaging: ensure ownership for /var/lib/dbus
Patrick McCarty [Fri, 7 Jun 2013 22:53:20 +0000 (15:53 -0700)]
packaging: ensure ownership for /var/lib/dbus

For x86_64 builds, %{_localstatedir}%{_libdir}/dbus expands to
/var/usr/lib64/dbus, which is incorrect.

This commit fixes the path in the spec to make sure the intended
directory, /var/lib/dbus, is created and owned by this package.

Change-Id: I60947c4e2c221ad2f83d01bd87dc76fc8f5e4433

8 years agoUpdate to 1.6.10
Anas Nashif [Mon, 27 May 2013 03:43:02 +0000 (23:43 -0400)]
Update to 1.6.10

8 years agoupdate to 1.6.10
Anas Nashif [Sun, 26 May 2013 18:52:43 +0000 (14:52 -0400)]
update to 1.6.10

8 years agoAdapt for rebase onto upstream dbus-1.6.8 git tag
Simon McVittie [Tue, 16 Apr 2013 17:34:16 +0000 (18:34 +0100)]
Adapt for rebase onto upstream dbus-1.6.8 git tag

When building from git we need a BuildRequires on xmlto, since we can
no longer rely on the pre-generated HTML documentation from the tarball.

That pulls in libxslt-tools, which means dbus.devhelp is generated, so
the build fails because we weren't accounting for that; so explicitly
require libxslt-tools too (dbus uses it both via xmlto, and directly),
and package dbus.devhelp.

Change-Id: I7505fc6fd1e7c4fc23a53f349a4fdd18d88237ec

8 years agoCleanup lib name in spec
Anas Nashif [Thu, 28 Mar 2013 13:15:58 +0000 (06:15 -0700)]
Cleanup lib name in spec

8 years agouse libname directly
Anas Nashif [Thu, 28 Mar 2013 13:15:12 +0000 (06:15 -0700)]
use libname directly

8 years agoFixed package groups
Anas Nashif [Fri, 22 Mar 2013 18:36:11 +0000 (11:36 -0700)]
Fixed package groups

8 years agoNo OOMScoreAdjust for user session
Anas Nashif [Wed, 16 Jan 2013 03:13:39 +0000 (19:13 -0800)]
No OOMScoreAdjust for user session

8 years agoadd user session units
Anas Nashif [Mon, 31 Dec 2012 15:56:22 +0000 (07:56 -0800)]
add user session units

8 years agoadd user session units
Anas Nashif [Mon, 31 Dec 2012 15:50:37 +0000 (07:50 -0800)]
add user session units

8 years agoinitial changelog
Simon McVittie [Fri, 30 Nov 2012 16:52:54 +0000 (16:52 +0000)]
initial changelog

8 years agoregenerate packaging/dbus-x11.spec with ( cd packaging && sh pre_checkin.sh )
Simon McVittie [Thu, 29 Nov 2012 11:10:36 +0000 (11:10 +0000)]
regenerate packaging/dbus-x11.spec with ( cd packaging && sh pre_checkin.sh )

Otherwise, local builds with 'gbs build' try to apply the nonexistent
patch 0001-fix-sba-for-dbus-activation.patch, and fail.

8 years agofixed splitting
Anas Nashif [Fri, 16 Nov 2012 18:45:18 +0000 (10:45 -0800)]
fixed splitting

8 years agofixed splitting
Anas Nashif [Fri, 16 Nov 2012 18:25:40 +0000 (10:25 -0800)]
fixed splitting

8 years agofixed splitting
Anas Nashif [Fri, 16 Nov 2012 18:24:32 +0000 (10:24 -0800)]
fixed splitting

8 years agoFixed package splitting
Anas Nashif [Fri, 16 Nov 2012 17:26:26 +0000 (09:26 -0800)]
Fixed package splitting

split dbus-x11 correctly

8 years agospec cleanup and removed patches
Anas Nashif [Fri, 16 Nov 2012 01:50:16 +0000 (17:50 -0800)]
spec cleanup and removed patches

8 years agoadd packaging
Anas Nashif [Tue, 30 Oct 2012 23:28:05 +0000 (16:28 -0700)]
add packaging

8 years ago1.10.6 dbus-1.10.6
Simon McVittie [Tue, 1 Dec 2015 18:19:17 +0000 (18:19 +0000)]
1.10.6

8 years agoNEWS for 1.10 branch
Simon McVittie [Mon, 23 Nov 2015 21:10:09 +0000 (21:10 +0000)]
NEWS for 1.10 branch

8 years agouid-permissions test: don't assert that root can UpdateActivationEnvironment
Simon McVittie [Fri, 20 Nov 2015 17:19:51 +0000 (17:19 +0000)]
uid-permissions test: don't assert that root can UpdateActivationEnvironment

Since 1.10.4 this is hard-coded to be disallowed when an activation
helper is used. That would be a security flaw waiting to happen,
and makes little sense anyway, because the activation helper sanitises
its environment.

Use BecomeMonitor() instead, as our way to assert that root and
messagebus are privileged.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=93036
Reviewed-by: Iain Lane
8 years agoname-test: don't run these tests with autotools if targeting Windows.
Simon McVittie [Tue, 17 Nov 2015 16:42:01 +0000 (16:42 +0000)]
name-test: don't run these tests with autotools if targeting Windows.

The wrapper shell script that sets up their environment is nowhere
near being portable. In particular, it uses dbus-run-session,
which is Unix-specific.

[rh: Add autotools scope in commit title]
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92899
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Rewieved-by: Ralf Habacker <ralf.habacker@freenet.de>
8 years agoWindows _dbus_get_autolaunch_address: don't leak shm_name
Simon McVittie [Wed, 18 Nov 2015 22:38:20 +0000 (23:38 +0100)]
Windows _dbus_get_autolaunch_address: don't leak shm_name

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92899
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
8 years ago1.10.5
Simon McVittie [Tue, 17 Nov 2015 23:46:34 +0000 (23:46 +0000)]
1.10.5

8 years agodbus 1.10.4
Simon McVittie [Tue, 17 Nov 2015 21:40:36 +0000 (21:40 +0000)]
dbus 1.10.4

8 years agoNEWS
Simon McVittie [Tue, 17 Nov 2015 18:22:37 +0000 (18:22 +0000)]
NEWS

8 years agobus-driver: Support returning org.freedesktop.DBus UID and PID
Jan Alexander Steffens (heftig) [Sat, 14 Nov 2015 15:32:43 +0000 (16:32 +0100)]
bus-driver: Support returning org.freedesktop.DBus UID and PID

Attempting to call SetEnvironment on systemd causes it to inquire
about the caller's connection UID and PID. If this check fails,
the call is rejected.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92857
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
[smcv: go back to DBUS_ERROR_UNIX_PROCESS_ID_UNKNOWN as the error code
for failure to determine the pid]

8 years agotest/sd-activation: Test UpdateActivationEnvironment forwarding
Jan Alexander Steffens (heftig) [Sat, 14 Nov 2015 15:32:42 +0000 (16:32 +0100)]
test/sd-activation: Test UpdateActivationEnvironment forwarding

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92857
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
8 years agobus_driver_handle_update_activation_environment: Forward to systemd
Jan Alexander Steffens (heftig) [Sun, 15 Nov 2015 15:00:33 +0000 (16:00 +0100)]
bus_driver_handle_update_activation_environment: Forward to systemd

If we use systemd activation, forward all UpdateActivationEnvironment
requests to org.freedesktop.systemd1.Manager.SetEnvironment, in order
to ensure variables needed by D-Bus services are available when these
services are launched by systemd.

Since UpdateActivationEnvironment is not available on the system bus,
this only applies to user buses.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92857
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
8 years agobus_driver_handle_update_activation_environment: Error on system buses
Jan Alexander Steffens (heftig) [Sat, 14 Nov 2015 15:32:37 +0000 (16:32 +0100)]
bus_driver_handle_update_activation_environment: Error on system buses

The default policy already disallows calls on system buses. Since any
bus with a service helper cleans the environment anyway, there's no
point in allowing this to be called.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92857
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
8 years agofdpass test: assert that TCP connections don't say they can pass fds
Simon McVittie [Tue, 17 Nov 2015 13:23:30 +0000 (13:23 +0000)]
fdpass test: assert that TCP connections don't say they can pass fds

As well as slightly increasing our test coverage, this ensures
that at least one test-case in this test is not skipped, working
around CMake's lack of support for the Automake-style exit code
produced by GTest executables (they exit 77 if everything was skipped).

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92887
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>
8 years agoFix recursive loop in backtrace generator on windows.
Ralf Habacker [Sun, 15 Nov 2015 18:54:03 +0000 (19:54 +0100)]
Fix recursive loop in backtrace generator on windows.

Backtrace generator called _dbus_warn(), which calls
backtrace generator recursively with DBUS_FATAL_WARNINGS=1.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92721
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
8 years agoDo not fail with fatal message skipping GetConnectionUnixProcessID test-bus test...
Ralf Habacker [Fri, 13 Nov 2015 18:48:16 +0000 (19:48 +0100)]
Do not fail with fatal message skipping GetConnectionUnixProcessID test-bus test on windows.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92721
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
8 years agoFix test-bus test for GetConnectionUnixUser driver method on windows.
Ralf Habacker [Fri, 13 Nov 2015 18:41:43 +0000 (19:41 +0100)]
Fix test-bus test for GetConnectionUnixUser driver method on windows.

Because GetConnectionUnixUser is not supported on windows it fails with
DBUS_ERROR_FAIL.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92721
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
8 years agoSkip launch helper activation tests on windows silently.
Ralf Habacker [Thu, 12 Nov 2015 11:49:46 +0000 (12:49 +0100)]
Skip launch helper activation tests on windows silently.

The system bus is unsupported there and Windows does not
have any concept of setuid binaries, so it can't ever
actually work.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92721
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
8 years agoFix memory leak in _dbus_win_set_error_from_win_error().
Ralf Habacker [Wed, 11 Nov 2015 14:02:44 +0000 (15:02 +0100)]
Fix memory leak in _dbus_win_set_error_from_win_error().

There is no need to make a local (leaked) copy of the message,
because dbus_set_error() already makes a copy of its parameters.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92721
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
8 years agoFix memory leaks in bus_activation_service_reload_test() in case of errors.
Ralf Habacker [Wed, 11 Nov 2015 14:02:00 +0000 (15:02 +0100)]
Fix memory leaks in bus_activation_service_reload_test() in case of errors.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92721
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
8 years agoRevert "Fix memory leaks in bus_activation_service_reload_test() in case of errors."
Ralf Habacker [Wed, 11 Nov 2015 14:01:11 +0000 (15:01 +0100)]
Revert "Fix memory leaks in bus_activation_service_reload_test() in case of errors."

This reverts commit d8569ac647da4d62c7a251d8e809f2d545a77837.

8 years agoFix memory leaks in bus_activation_service_reload_test() in case of errors.
Ralf Habacker [Mon, 9 Nov 2015 12:38:59 +0000 (13:38 +0100)]
Fix memory leaks in bus_activation_service_reload_test() in case of errors.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92721
Reviewed-by: https://bugs.freedesktop.org/show_bug.cgi?id=92721
8 years agoFix test-bus segfault_service_no_auto_start test on windows.
Ralf Habacker [Mon, 9 Nov 2015 19:38:03 +0000 (20:38 +0100)]
Fix test-bus segfault_service_no_auto_start test on windows.

Windows returns unhandled exceptions from a running child
by specific exit codes and not by signals as on UNIX.

Therefore we use DBUS_ERROR_SPAWN_CHILD_EXITED for propagating
unhandled exceptions to the parent too.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92721
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
8 years agoupdate NEWS
Simon McVittie [Fri, 6 Nov 2015 17:26:40 +0000 (18:26 +0100)]
update NEWS

8 years agoFix test cases running client and server dispatch design issue.
Ralf Habacker [Fri, 6 Nov 2015 13:03:23 +0000 (14:03 +0100)]
Fix test cases running client and server dispatch design issue.

DBus test cases running the server *and* client loop in the same
process assumed that all messages send from the server has to be
received in one client dispatch, which is not the case in all
environments.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92721
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
8 years agoFix bug unrefing connection too early in check_hello_message().
Ralf Habacker [Fri, 6 Nov 2015 13:02:29 +0000 (14:02 +0100)]
Fix bug unrefing connection too early in check_hello_message().

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92721
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
8 years agoCompile fix on Windows.
Ralf Habacker [Wed, 4 Nov 2015 13:21:20 +0000 (14:21 +0100)]
Compile fix on Windows.

This commit fixes a regression introduced with commit 04b8a7a772cfa9ae8ea6ce452d1fb7f23e25fd3f.

8 years agoFix warning: variable 'ret' set but not used [-Wunused-but-set-variable].
Ralf Habacker [Mon, 2 Nov 2015 20:56:32 +0000 (21:56 +0100)]
Fix warning: variable 'ret' set but not used [-Wunused-but-set-variable].

Only set valid child status in case exit code has been gotten from
child process, otherwise signal failure through thread return value.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92721
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
8 years agoTest system bus config files on Unix only
Ralf Habacker [Mon, 2 Nov 2015 23:23:56 +0000 (00:23 +0100)]
Test system bus config files on Unix only

Previously, we didn't consistently test parsing of every file in
valid-config-files-system/ everywhere that we tested valid-config-files/.
We now test it on Unix.

The system bus is not supported on Windows, so we do not test
valid-config-files-system/ there.

valid-config-files/many-rules.conf contains <user> and <group> rules
which are not applicable to Windows. Copy the original many-rules.conf
to valid-config-files-system/ so that it will be tested on Unix, and
remove the non-portable rules from valid-config-files/many-rules.conf.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92721
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
[rh:base patch came from Simon]

8 years agoDon't use _dbus_warn() for intentionally-skipped tests
Simon McVittie [Mon, 2 Nov 2015 18:19:02 +0000 (18:19 +0000)]
Don't use _dbus_warn() for intentionally-skipped tests

The tests are run with _dbus_warn() fatal, so if a particular test is
not applicable on the current platform, we shouldn't call it.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=92721
Reviewed-by: Ralf Habacker <ralf.habacker@freenet.de>