summaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* add gsm to connectable bearer for networkmanager.Lorn Potter2011-02-231-0/+5
| | | | Reviewed-by: trustme
* Removing tabs from 7388fcb83592a90aace054314e0c3e7e7a94fdae changesetCristiano di Flora2011-02-221-1/+1
| | | | (cherry picked from commit 44373d71dde16c4899377703e724d46d803ade9e)
* Fix QNetworkConfigurationManager crash due to null private pointer.Cristiano di Flora2011-02-221-12/+5
| | | | | | Reviewed-by: Aaron McCarthy Task-Number: QTBUG-17305 (cherry picked from commit 7388fcb83592a90aace054314e0c3e7e7a94fdae)
* Wheel delta is always zero in QEventTransition's eventDean Dettman2011-02-211-0/+1
| | | | | | | | When a QGraphicsSceneWheelEvent was cloned, the original event's delta property was not copied. Task-number: QTBUG-17536 Reviewed-by: Kent Hansen
* Fix warning about id maybe used when uninitialisedThiago Macieira2011-02-201-1/+1
| | | | | The code would indicate it's never used uninitialised, but gcc doesn't know it.
* Doc: setSslConfiguration also sets the CA certificatesThiago Macieira2011-02-201-2/+3
|
* Prevents crashing when ICO file has bad color table valueMauri Vehniäinen2011-02-181-0/+2
| | | | | | | | | | | When color table value is read from ICO header memory allocation is made based on this value. This case is relevant only when reading 8bit images. Therefore values over 256 will abort reading the image. Task-number: QT-4535 Merge-request: 1090 Reviewed-by: Harald Fernengel <harald.fernengel@nokia.com>
* directfb: fix crash in client/server mode of qwsDenis Oliver Kropp2011-02-172-16/+31
| | | | | | | | | | | | | | The server was crashing due to client pointer being sent as "sibling". Instead surface flags are sent and sibling code is removed. Also the window ID is transfered and used on receiver side to get a local handle to the window. Click to focus window works, but it is not raised. Merge-request: 989 Reviewed-by: Marcel Schuette <marcel.schuette@nokia.com>
* Make Qt/DirectFB less verbose about failing to load pixmaps.Anders Bakken2011-02-171-3/+2
| | | | | | | | | | Qt/DirectFB is way too verbose about not being able to use DirectFB for loading images. A lot of times DirectFB doesn't have support for e.g. pngs or gifs and currently one gets a message for every failed pixmap load, even if they will succeed when falling back to Qt. Merge-request: 2552 Reviewed-by: Marcel Schuette <marcel.schuette@nokia.com>
* Fix warnings in QtDeclarativeThiago Macieira2011-02-173-3/+3
|
* QtDBus: Make sure we can receive unknown typesThiago Macieira2011-02-171-1/+6
| | | | | | | We must make sure we advance the receiving iterator, or we end up in an infinite loop. Task-number: QTBUG-17476
* QtDBus meta object: keep methods, signals and props with unknown typesThiago Macieira2011-02-172-9/+13
| | | | | | | | | | | | | | | | | | | | | Use VoidStar as the metatype, with an unknown type called "QDBusRawType". The actual D-Bus type is saved as an hex value as a template parameter. D-Bus type Qt type comment h QDBusRawType<0x68>* Unix file descriptors ~ QDBusRawType<0x7e>* invalid type ai QDBusRawType<0x6169>* array of int32 a{i(ssy)} QDBusRawType<0x617b6928737379297d>* Note that the number in the template doesn't have to be valid. The QDBusRawType class doesn't exist anyway. I thought of just leaving the raw D-Bus type there, but who knows what kind of things can appear there, like other '>' (which may cause problems for anything trying to parse the meta object later). Task-number: QTBUG-17476
* Make the QtDBus parser not complain about unknown thingsThiago Macieira2011-02-172-30/+36
| | | | | | | Keep only the warnings about application errors. If the unknown things come from the outside, it's not our job to make noise. Task-number: QTBUG-17476
* fix-layer-getsurface-result-check-for-screen-size-determinationDenis Oliver Kropp2011-02-171-1/+1
| | | | | | | GetSurface() was checked for NOT returning DFB_OK, but it should Merge-request: 2528 Reviewed-by: Marcel Schuette <marcel.schuette@nokia.com>
* Bump QtWebKit version to 2.0.2Ademar de Souza Reis Jr2011-02-171-2/+2
| | | | | | | | | QtWebKit 2.0.1 was part of Qt-4.7.1, QtWebKit as included in 4.7.2 should be 2.0.2. Merge-request: 1095 Task-number: QTBUG-17480 Reviewed-by: Jason McDonald
* SSL: fix memory leak when loading certificates on Mac OS XMartin Petersson2011-02-161-0/+1
| | | | Reviewed-by: Markus Goetz
* Use QElapsedTimer in QFutureInterface.Thiago Macieira2011-02-162-7/+7
| | | | | | | | QElapsedTimer can also keep its on vlaid/invalid state, so we don't need to keep an additional boolean. Task-number: https://projects.maemo.org/bugzilla/show_bug.cgi?id=227660 Reviewed-by: Morten Sørvig
* Keep sending posted events after commit 7196045b78b33cf135683d5c0b4e164f95231791Bradley T. Hughes2011-02-141-4/+1
| | | | | | | | | | | | | | | | | | | | | | | | This commit put in place a timer based mechanism to ensure that posted events continue to be sent. As pointed out in commit eb1015c7bbf135af3656110a4d112377c1209db8, GetQueueStatus() can indicate that there are messages in the queue but our hook never actually receives them, hence the timer solution. However, having the message type and hwnd "guards" in the conditional for starting a timer would cause us to lose events in some situations, due to the fact that QEventDispatcherWin32::processEvents() will only allow one WM_QT_SENDPOSTEDEVENTS message per call (multiple messages will be compressed into a single one). When racing with other threads, it would be possible to end up blocking one thread that has pending posted events unintentionally. This fix removes the guards and makes sure that we start the keep alive timer when it's really needed, and only stops it when we really know that there is nothing left in the Windows message queue. Task-number: QTBUG-17014 Reviewed-by: joao Reviewed-by: Marius Storm-Olsen
* Merge branch '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public into ↵Qt Continuous Integration System2011-02-101-33/+19
|\ | | | | | | | | | | | | 4.7-integration * '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public: Fix QFileDialog Symbian native file dialog filename filtering.
| * Fix QFileDialog Symbian native file dialog filename filtering.Miikka Heikkinen2011-02-101-33/+19
| | | | | | | | | | | | | | | | | | Now QFileDialog static functions that use Symbian native file dialog should find the same files as non-native QFileDialog for any given filename filter. Task-number: QTBUG-17298 Reviewed-by: Janne Koskinen
* | Merge branch '4.7' of scm.dev.nokia.troll.no:qt/qt-qml into 4.7-integrationQt Continuous Integration System2011-02-105-16/+48
|\ \ | | | | | | | | | | | | | | | * '4.7' of scm.dev.nokia.troll.no:qt/qt-qml: PathView items were not correctly updated when rootIndex changed. Allow text to selected in a TextEdit or TextInput inside a Flickable.
| * | PathView items were not correctly updated when rootIndex changed.Martin Jones2011-02-101-16/+28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | PathView got confused because it queried the model when removing items due to the rootIndex change. This caused new items to be added since the model->count() was not zero (it was the count for the new rootIndex). Calculate the PathView modelCount similarly to the other views (rather than querying the model) and correctly handle the modelCount being set to zero. Change-Id: I9e4fb27c14aa8c77ad37ff01d95c123f909db20d Task-number: QTBUG-16366 Reviewed-by: Bea Lam
| * | Allow text to selected in a TextEdit or TextInput inside a Flickable.Andrew den Exter2011-02-104-0/+20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Set keepMouseGrabs to true when the selectByMouse property is enabled to prevent flickable from stealing the mouse grab and interrupting a selection. Change-Id: I08e11265ab9c55d239fd09aceef4fdb8e74aba9d Task-number: QTBUG-16956 Reviewed-by: Martin Jones
* | | Merge branch '4.7' of scm.dev.nokia.troll.no:qt/qt-qml into 4.7-integrationQt Continuous Integration System2011-02-103-5/+62
|\ \ \ | |/ / | | | | | | | | | | | | * '4.7' of scm.dev.nokia.troll.no:qt/qt-qml: Update modules-related tests and docs CLeanup registers after AddString
| * | Update modules-related tests and docsBea Lam2011-02-102-2/+24
| | | | | | | | | | | | | | | | | | | | | Added tests for remote imports and fixed/improved examples and text in the docs. Change-Id: I8f411a0287c4d50ec3cebe5567d803689cd5b1c7
| * | CLeanup registers after AddStringAaron Kennedy2011-02-101-3/+38
| | | | | | | | | | | | | | | Reviewed-by: Martin Jones Change-Id: I7bd2806770fe5b86fbc589644107d74981d32abe
* | | Merge branch '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1 into ↵Qt Continuous Integration System2011-02-101-0/+4
|\ \ \ | | | | | | | | | | | | | | | | | | | | | | | | 4.7-integration * '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1: Stop focus propagation of touch events when flag is set.
| * | | Stop focus propagation of touch events when flag is set.Niklas Kurkisuo2011-02-091-0/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Added check to stop propagation of touch events for items that have the QGraphicsItem::ItemStopsClickFocusPropagation flag set. Merge-request: 1085 Reviewed-by: Harald Fernengel <harald.fernengel@nokia.com>
* | | | Merge branch '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public into ↵Qt Continuous Integration System2011-02-095-10/+54
|\ \ \ \ | | |_|/ | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 4.7-integration * '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public: 4.7.2 changes Change paging attributes for Symbian binaries Fix build failure Fix regression in creating mkspecs/default Changes for 4.7.2 Recreate Symbian raster backing store when WA_TranslucentBackground set
| * | | Change paging attributes for Symbian binariesIain2011-02-092-6/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Removed UNPAGED workaround from QtCore and QtSql on OS releases where data paging is supported. Depends on new macro being defined in the OS builds (SYMBIAN_DLL_DATA_EXPORTS_SUPPORTED) to allow us to detect this. Data exports are supported in all S^3 builds after early 2009, however, this macro is only added recently (and thus isn't defined in, eg. Symbian Foundation PDKs (v3.0.4), or current releases (v0.9) of the S^3 SDK from Nokia). Thus building Qt against these older S^3 builds will still result in QtCore and QtSql being unpaged. Also, remove PAGED keyword from all Qt-based software. This changes the code paging field in the Symbian (E32Image) header from "paged" to "default". Thus it is left to the configuration of the particular device whether paging is used for the binary or not. All devices that support code paging should have it turned on (the value is stored in the HAL, so can be checked using eg. fshell); data paging may or may not be turned on depending on the device characteristics. Leaving both code and data paging as "default" means that any limitations in the device (eg. around flash wear) can be controlled by that device by disabling the appropriate types of paging. This change also leaves the way open to deprecated PAGED as a keyword in the Symbian build systems, so you can only opt out of paging. This makes more sense, as the default for code paging will be on, the default for data paging will be on if the device can handle it, and binaries should only be opting out of paging if they have some specific real-time requirements that paging inhibits. Task-number: QT-3503 Task-number: QTBUG-13931 Reviewed-by: Shane Kearns
| * | | Fix build failureGareth Stockwell2011-02-092-6/+6
| | | | | | | | | | | | | | | | | | | | Task-number: QTBUG-17288 Reviewed-by: trustme
| * | | Recreate Symbian raster backing store when WA_TranslucentBackground setGareth Stockwell2011-02-093-4/+42
| | | | | | | | | | | | | | | | | | | | Task-number: QTBUG-17288 Reviewed-by: Jani Hautakangas
* | | | Merge branch '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-2 into ↵Qt Continuous Integration System2011-02-091-0/+6
|\ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 4.7-integration * '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-2: Fixed tst_qpixmap::toImageDeepCopy autotest for GL graphicssystem.
| * | | | Fixed tst_qpixmap::toImageDeepCopy autotest for GL graphicssystem.Samuel Rødal2011-02-091-0/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We need to copy the image if painting is active. Reviewed-by: Jani Hautakangas
* | | | | Merge branch '4.7' of scm.dev.nokia.troll.no:qt/qt-qml into 4.7-integrationQt Continuous Integration System2011-02-098-73/+59
|\ \ \ \ \ | | |_|_|/ | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * '4.7' of scm.dev.nokia.troll.no:qt/qt-qml: Fix auto test failure. PathView regression: dragging the path didn't update currentIndex Fix docs for variant and list basic types Add test for importing plugins that contain QML files Call onAdd() for first items added to ListView and GridView Use same values for Text.lineHeightMode and QTextBlock::lineHeightMode from master. Don't clear pre-edit text when a graphics items loses focus.
| * | | | Merge branch '4.7' of scm.dev.nokia.troll.no:qt/qt-qml into 4.7Bea Lam2011-02-0910-40/+110
| |\ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Conflicts: tests/auto/declarative/qdeclarativemoduleplugin/qdeclarativemoduleplugin.pro Change-Id: I8da521f05ebc266d91caa582852155b571d5cc63
| | * | | | PathView regression: dragging the path didn't update currentIndexMartin Jones2011-02-091-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Some refectoring while fixing QTBUG-13687 resulted in the private setOffset() method being called rather than the public. The public version was responsible for updating currentIndex. Change-Id: Iac9c7a19d6fa64550b9498e77b8983512e199370 Task-number: QTBUG-17319 Reviewed-by: Michael Brasser
| | * | | | Use same values for Text.lineHeightMode and QTextBlock::lineHeightMode from ↵Yann Bodson2011-02-094-10/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | master. - MultiplyHeight becomes ProportionalHeight - PixelHeight becomes FixedHeight Change-Id: I2a1ebc6ff9db7e62f513919f19773f985b08f8d7 Reviewed-by: Michael Brasser
| | * | | | Don't clear pre-edit text when a graphics items loses focus.Andrew den Exter2011-02-091-12/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Don't forceably reset the input method state when a graphics item loses focus, reset the input context and let it handle sending the appropriate events itself. That way behavior is consistent with QWidget and the input context can opt to commit its current pre-edit text instead of discarding it. And don't change the focus until after the input context has been reset or the input method events won't be delivered. Change-Id: Ic545bccab6bf671709c1d06d499217baf614e6f9 Task-number: QTBUG-16997 Reviewed-by: axis
| * | | | | Call onAdd() for first items added to ListView and GridViewBea Lam2011-02-092-50/+35
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The bug was caused by the fact that the itemsInserted() implementation in ListView and GridView returned early for cases where (model->count() <= 1) and so did not call onAdd() unlike for other cases. This change also fixes some incorrect values in the header() and footer() unit tests for ListView and GridView. Change-Id: I24f5c86afcd62c26e17b0932f257f4767a287b8e Task-number: QTBUG-15642 Reviewed-by: Martin Jones
* | | | | | Merge branch '4.7' of scm.dev.nokia.troll.no:qt/qt-qml into 4.7-integrationQt Continuous Integration System2011-02-081-12/+64
|\ \ \ \ \ \ | | |/ / / / | |/| | | | | | | | | | | | | | | | * '4.7' of scm.dev.nokia.troll.no:qt/qt-qml: Support seperate versions of installed modules
| * | | | | Support seperate versions of installed modulesmae2011-02-081-12/+64
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | QML supports versioned types in modules. There's a version major and a version minor. This makes it possible to have a module com.organisation.fancycomponents with version 1.0, and later you could ship a new module com.organisation.fancycomponents which contains a more recent version 1.1 or 2.0 AND also the old versions to keep old code running. This is good. The problem is that this is difficult with certain QA procedures. It's hard to verify that a new module is indeed 100% compatible with the previous versions. The change extends the import mechanism by adding optional versioning to the component patch. With the patch, you can add a new module com.organisation.fancycomponents.2.0 which will be loaded when the QML file specifies "import com.organisation.fancycomponents 2.0". The patch works as follows: if you try to load com.organisation.fancycomponents in version 2.0, the engine first looks for com/organisation/fancycomponents.2.0, then for com/organisation/fancycomponents.2 then for com.organisation/fancycomponents. Reviewed-by: Aaron Kennedy Task-number: QTBUG-16455
* | | | | | Merge branch '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1 into ↵Qt Continuous Integration System2011-02-0813-83/+86
|\ \ \ \ \ \ | | |_|_|_|/ | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 4.7-integration * '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1: (21 commits) Fix compilation of QMutableSetIterator::value() with QT_STRICT_ITERATORS Fix the compilation error that the previous fix didn't fix Use the thread-default glib context for the DBus connection in the ICD bearer manager. Check that _POSIX_THREAD_SAFE_FUNCTIONS is larger than zero. Use an increasing size for the getpwuid_r buffer. Make sure we use at least 1024 bytes when calling getpwuid_r. Fix compilation error: symbol is namespaced. Don't crash when creating backtrace for built-in JS function (2nd try) Revert "Improved performance of mapFromGlobal/mapToGlobal on X11" QtDBus: do not wait for reply for AddMatch and RemoveMatch Make qdbus show all types, even those that it doesn't know about Doc: Fix the docs saying what the locale codec is used for Autotest: simple improvements Autotest: Use QElapsedTimer for timings directfb: Pixmap creation always premultiplied the alpha even when it is already premultiplied in the image. Don't crash when BMP color table is broken Designer: Fix a crash in the Signal-Slot-Editor. complain about unmatched quotes/parens use const ref move multiple inclusion check for feature files to correct location ...
| * | | | | Fix compilation of QMutableSetIterator::value() with QT_STRICT_ITERATORSDavid Faure2011-02-081-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Merge-request: 1078 Reviewed-by: Thiago Macieira <thiago.macieira@nokia.com>
| * | | | | Use the thread-default glib context for the DBus connection in the ICD ↵Jocelyn Turcotte2011-02-081-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | bearer manager. This allows the bearer management plugin to be ran on a thread different than the main one. Task-number: QTBUG-17199 Reviewed-by: Kranthi <kranthi.kumar-kuntala@nokia.com>
| * | | | | Check that _POSIX_THREAD_SAFE_FUNCTIONS is larger than zero.Thiago Macieira2011-02-081-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | POSIX allows this symbol to be defined to -1 to indicate that the interfaces aren't present. On FreeBSD 8, it's defined to -1, but the functions are present (supposedly, they don't work -- just stubs). Guessing that OpenBSD is the same. Also removing the QT_NO_THREAD, since Qt can't be built without threading support. It's only used by our bootstrapped tools, which in turn use QCoreApplication only (no QApplication). Reviewed-By: Bradley T. Hughes
| * | | | | Use an increasing size for the getpwuid_r buffer.Thiago Macieira2011-02-081-1/+12
| | | | | | | | | | | | | | | | | | | | | | | | Reviewed-by: Bradley T. Hughes
| * | | | | Make sure we use at least 1024 bytes when calling getpwuid_r.Thiago Macieira2011-02-081-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | On some systems, sysconf may return -1 to indicate that there's no such limit (limitless), so don't create a buffer of size -1. Reviewed-by: Bradley T. Hughes
| * | | | | Don't crash when creating backtrace for built-in JS function (2nd try)Kent Hansen2011-02-084-4/+30
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Commit 147df10403ba280b3f04c1e3d6c4b1cf386abe5d did not quite fix the issue; other places need the same checks. When the JIT is enabled, frames for built-in JS host calls (such as Array.prototype.forEach) are not fully initialized. In particular, the CodeBlock register of such frames is not set (see comment in JITCall.cpp). We need to check if the codeBlock is actually valid before we start using it. This fixes the crash(es) but not the problem of actually getting the arguments for such frames through the API. There's also a related problem when a QtScript function (newFunction()) is called as a callback of a built-in JS host function (QTBUG-17287). These problems will go away once JavaScriptCore is updated to a more recent version (4.8 at the earliest), since the native-vs-script frame handling has been unified. Task-number: QTBUG-17137 Reviewed-by: Olivier Goffart
| * | | | | Revert "Improved performance of mapFromGlobal/mapToGlobal on X11"Denis Dzyubenko2011-02-071-28/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change introduced regressions with some window managers (for example with metacity when you maximize the window) This reverts commit cdd776a91e65bf5c30cea1bab9823134a3f797d0.
generated by cgit v0.12 at 2024-11-12 15:42:24 (GMT)