summaryrefslogtreecommitdiffstats
path: root/src/dbus
Commit message (Collapse)AuthorAgeFilesLines
* Doc: document what the timeout of -1 means in D-BusThiago Macieira2010-03-041-3/+14
| | | | Task-number: QTBUG-8729
* Fix an issue with the error signal in a callWithCallback not beingThiago Macieira2010-02-161-18/+9
| | | | | | | | | | delivered queued. Since we emit the signal while the locks are in place, we have to have queued delivery. Task-number: QT-760 Reviewed-by: Trust Me
* When checking to see if a signal is connected, compare the match-arguments tooThiago Macieira2010-01-261-2/+4
|
* Do the refcounting of services watched properly.Thiago Macieira2010-01-261-20/+15
| | | | Issue found while debugging KDE applications.
* fix copy-paste errorRitt Konstantin2010-01-221-1/+0
| | | | | Merge-request: 442 Reviewed-by: Thiago Macieira <thiago.macieira@nokia.com>
* minor optimizationRitt Konstantin2010-01-221-2/+3
| | | | | | | avoid needless atomic operations Merge-request: 442 Reviewed-by: Thiago Macieira <thiago.macieira@nokia.com>
* Fix D-Bus calls with QDBus::BlockWithGui.Thiago Macieira2010-01-121-1/+1
| | | | | | | | We must wait for a reply in case the reply hasn't been received yet, not if it has already been received. Simple typo. Task-number: QT-2307 Reviewed-by: Trust Me
* Update copyright year to 2010Jason McDonald2010-01-0660-60/+60
| | | | Reviewed-by: Trust Me
* Remove the need to know the service owner when adding/removing signal rules.Thiago Macieira2010-01-024-34/+21
| | | | | Task-number: QTBUG-5979 Reviewed-By: TrustMe
* Use the watched service name list in matching incoming signals too.Thiago Macieira2010-01-021-2/+8
| | | | | Task-number: QTBUG-5979 Reviewed-By: TrustMe
* Change the QtDBus watched service list to a map to its owner.Thiago Macieira2010-01-022-17/+48
| | | | | | | | | | | | | Instead of simply counting how many times we're watching a service, also map a service to its owner. This allows us to provide a faster, cached GetNameOwner too. Technically, this could mean that a service that is started in-between event loops could remain invalid for a longer period of time. I'm not sure this is an issue -- will need some testing in KDE. Task-number: QTBUG-5979 Reviewed-By: TrustMe
* Update debug output to show the connection's ptr and threadThiago Macieira2010-01-021-10/+10
|
* Merge branch '4.5' into 4.6-stagingThiago Macieira2009-12-271-1/+3
|\
| * Fix D-Bus marshalling of booleans in optimised code.Thiago Macieira2009-12-271-1/+3
| | | | | | | | | | | | | | | | | | | | | | C++ booleans are 1 byte in size, but D-Bus booleans (dbus_bool_t) are 4 bytes. That means a boolean with a zero in the LSB byte but non-zero garbage in the high bytes is a valid "false" in C++, but libdbus-1 will turn that to true when sending. Task-number: QTBUG-7041 Reviewed-By: Trust Me BT: yes
* | Merge branch '4.5' into 4.6-stagingThiago Macieira2009-12-152-1/+2
|\ \ | |/ | | | | | | Conflicts: mkspecs/macx-g++40/qplatformdefs.h
| * Fix crash in QDBusPendingReply/QDBusReply in case of unconnected calls.Thiago Macieira2009-12-092-1/+2
| | | | | | | | | | | | | | | | If we made calls on a QDBusConnection that isn't connected, the d pointer is 0. Ensure we don't crash. Task-number: QTBUG-6571 Reviewed-by: Bradley T. Hughes
| * Update license headers again.Jason McDonald2009-09-0859-236/+236
| | | | | | | | Reviewed-by: Trust Me
* | Fix compilation: private headers must be #included with ""Thiago Macieira2009-12-1012-13/+13
| | | | | | | | | | | | | | | | | | The reason is that the private headers don't exist in any of the -I lines, so they need to be found by the preprocessor relative to the source file, which means "". Task-number: QTBUG-6665 Reviewed-by: Trust Me
* | Small performance improvement: don't match the bus path.Thiago Macieira2009-11-241-3/+2
| | | | | | | | | | We trust the bus, so no need to match per path. The sender and the interface should be enough for us.
* | Doc: Update the documentation for the serviceOwnerChanged signal.Thiago Macieira2009-11-241-0/+6
| | | | | | | | | | Add the note that this signal will cause many activations and that QDBusServiceWatcher is better.
* | Fix the detection of when this process gets names on the bus.Thiago Macieira2009-11-242-12/+11
| | | | | | | | | | | | | | | | | | | | | | | | Previously, we were relying on NameOwnerChanged to notify us of our own names. This worked because we got all NameOwnerChanged that happened on the bus. Now, we only get those we're interested in. Instead of watching for newOwner==baseService, let's just use the NameAcquired and NameLost signals, that the D-Bus server sends to us anyway. Task-number: QTBUG-5979
* | Revert "Disable the deprecation warning for this signal since we're using it ↵Thiago Macieira2009-11-241-0/+5
| | | | | | | | | | | | now" This reverts commit f01c631f84f9373b06d0572bfed7a4e8213a46a7.
* | Revert "Revert "Stop using the NameOwnerChanged signal without arg0 in ↵Thiago Macieira2009-11-241-3/+0
| | | | | | | | | | | | | | | | QtDBus itself"" This reverts commit db1cff73726ff12d50139769dc8071f9e9173261. Task-number: QTBUG-5979
* | Merge branch '4.6' of scm.dev.nokia.troll.no:qt/qt into 4.6Simon Hausmann2009-11-191-2/+2
|\ \
| * | Ran the script utils/normalizeOlivier Goffart2009-11-181-2/+2
| | | | | | | | | | | | Over src/ tools/ examples/ and demos/
* | | Remember to match the actual arguments too.Thiago Macieira2009-11-192-0/+20
| | |
* | | Disable the deprecation warning for this signal since we're using it nowThiago Macieira2009-11-191-5/+0
| | | | | | | | | | | | Task-number: QTBUG-5972
* | | Revert "Stop using the NameOwnerChanged signal without arg0 in QtDBus itself"Thiago Macieira2009-11-191-0/+3
|/ / | | | | | | | | | | This partially reverts commit 1176ecf0b533279e5a1c97f183e5c5f1c57fb188. Task-number: QTBUG-5972
* | Merge branch '4.6' into core-4.6Thiago Macieira2009-11-092-6/+10
|\ \
| * \ Merge remote branch 'mainline/4.6' into 4.6Oswald Buddenhagen2009-11-062-6/+6
| |\ \
| | * \ Merge commit 'origin/4.6' into 4.6Olivier Goffart2009-11-051-5/+5
| | |\ \
| | | * | Doc: Fixed qdoc warnings.David Boddie2009-11-041-5/+5
| | | | | | | | | | | | | | | | | | | | Reviewed-by: Trust Me
| | * | | Merge commit 'origin/4.6' into 4.6Olivier Goffart2009-11-049-166/+803
| | |\ \ \ | | | |/ /
| | * | | One signal/slot connection is enough, when exporting N signals to DBusDavid Faure2009-10-301-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When exporting an object to DBus, and the object has N signals, this code was connecting the QDBusConnectionPrivate multiple times to the destroyed signal of the object. Merge-request: 1961 Reviewed-by: Olivier Goffart <ogoffart@trolltech.com>
| * | | | compile fix for QDBusServiceWatcher with namespaced Qthjk2009-11-061-0/+4
| | |/ / | |/| |
* | | | Let QDBusInterface work even if the introspection failed.Thiago Macieira2009-11-091-2/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If the introspection failed, we won't have a meta object, meaning you cannot connect to signals or invoke methods via the slot mechanism. But you can still make calls. The isValid() functionality is preserved because there is no owner for this object. Reviewed-By: Harald Fernengel
* | | | Fix annotations added for complex types in signals.Thiago Macieira2009-11-081-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Signals are badly mishandled in QtDBus. In the spec, they are "in" arguments, but we generate "Out" annotations for them. What's worse, we were using the wrong count. QTBUG-914 is about fixing the directionality. While that task isn't fixed, at least use the right count. Task-number: QTBUG-5563 Reviewed-by: TrustMe
* | | | Fix enabling of level 2 debugging in QtDBusThiago Macieira2009-11-081-1/+2
|/ / /
* | | Doc: add documentation for QDBusServiceWatcher class.Thiago Macieira2009-10-281-0/+163
| | |
* | | Add QDBusServiceWatcher class.Thiago Macieira2009-10-283-69/+393
| | |
* | | Add a warning to user's connecting to serviceOwnerChanged directlyThiago Macieira2009-10-281-1/+7
| | | | | | | | | | | | We want people to not use this signal directly.
* | | Stop using the NameOwnerChanged signal without arg0 in QtDBus itselfThiago Macieira2009-10-283-4/+37
| | | | | | | | | | | | | | | | | | | | | | | | We were using this signal to update the signal hooks when the remote service changed. That meant each Qt app received every single service creation, change or destruction. Now we only watch the services we're really interested in.
* | | Move the bulk of the signal connecting/disconnecting code to ↵Thiago Macieira2009-10-283-60/+82
| | | | | | | | | | | | | | | | | | | | | QDBusConnectionPrivate I'll need to recurse into the signal connection mechanism in the next commit.
* | | Fix the use of the owner when connecting the service "watcher".Thiago Macieira2009-10-282-3/+10
| | | | | | | | | | | | | | | | | | | | | Use null services to indicate we're not interested in the owner, but empty-but-not-null to indicate we don't know what the owner is. Since empty service names are not valid, this will mean that this rule won't match.
* | | Use the new argument-based rule-matching in QDBusAbstractInterface.Thiago Macieira2009-10-281-6/+22
| | | | | | | | | | | | | | | This allows us to listen only to the activations we're really interested in.
* | | Add new public API to QDBusConnection for connecting with stringThiago Macieira2009-10-282-22/+78
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | matching. The bus allows us to match string arguments when receiving messages. This is very useful for the NameOwnerChanged signal, whose first argument is usually what we're interested in. By using these new functions, you can restrict receiving of signals to those that you truly want, instead of receiving NameOwnerChanged for all services registered/unregistered on the bus.
* | | Add the ability for the match-rule builder to add argument matching.Thiago Macieira2009-10-283-6/+16
|/ / | | | | | | I'll use this feature to match the NameOwnerChanged signal from the bus.
* | Fixing some GCC warningsJoão Abecasis2009-10-071-0/+1
| | | | | | | | Reviewed-by: Markus Goetz
* | Use dbus_threads_init_default instead of QMutex wrappersThiago Macieira2009-10-074-151/+4
| | | | | | | | | | | | The minimum version for the recursive mutexes is D-Bus 0.93. That's also the same version that introduced the default thread functions. So we don't need to provide ours anymore, just use the default.
* | Use the recursive mutex functions for D-Bus only.Thiago Macieira2009-10-071-27/+5
| | | | | | | | | | | | | | | | | | | | | | | | At least in D-Bus 1.2.16, there's a deadlock caused by locking an already-locked mutex. When Qt is linked to libdbus-1 instead of dynamically loading it, it maintains legacy compatibility by only providing the non-recursive mutexes, which triggers this deadlock. The recursive functions have been present since 0.93, so I guess the non-recursive variants aren't tested. Report: https://bugzilla.novell.com/show_bug.cgi?id=482749#c30