| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
version.
|
|
|
|
|
|
|
|
|
| |
On Mac there really is no native pan gesture. Instead, this
gesture is delivered as mouse wheel events. So adding a
pan gesture to qabstractscrollarea means that it will get
both wheel and pan events. And both will try to move the
viweport. I think that the correct thing to do here is
to not apply a pan gesture by default on all scroll areas.
|
|
|
|
| |
src\3rdparty\webkit\JavaScriptCore\wtf\Assertions.h:150: badly punctuated parameter list in `#define'
|
|
|
|
|
|
| |
These classes are needed in QDirectFBPaintEngine.
Reviewed-by: Donald Carr <donald.carr@nokia.com>
|
|
|
|
|
|
|
| |
Don't try too hard to temporarily represent an image in a real directfb
surface. It leads to all sorts of rendering issues.
Reviewed-by: Donald Carr <donald.carr@nokia.com>
|
|
|
|
| |
StationQuery::query is static. No need for StationQuery instance.
|
|
|
|
| |
TimeQuery::query is static. No need for TimeQuery instance.
|
| |
|
|\ |
|
| |
| |
| |
| |
| |
| | |
This is required for QDirectFBPaintEngine to link.
Reviewed-by: Donald Carr <donald.carr@nokia.com>
|
| |
| |
| |
| |
| |
| | |
Apparently QScopedPointer requires knowledge of the class it points to.
Reviewed-by: Donald Carr <donald.carr@nokia.com>
|
| | |
|
| | |
|
| | |
|
|/
|
|
| |
Reviewed-By: Harald Fernengel
|
|
|
|
| |
Reviewed-by: Trust Me
|
|
|
|
| |
Reviewed-by: Trust Me
|
|
|
|
| |
Reviewed-by: Trust Me
|
|\ |
|
| | |
|
| |
| |
| |
| |
| |
| |
| | |
The previous commit reintroduced memcpy in detach2(). So we don't need
to do memcpy here.
Re-add the memcpy when detach3() is introduced.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Problem spotted in Designer:
==13197== Invalid read of size 4
==13197== at 0x5C9AD2A: QHash<QString, QConfFile*>::isEmpty() const (qhash.h:283)
==13197== by 0x5C98142: QHash<QString, QConfFile*>::remove(QString const&) (qhash.h:769)
==13197== by 0x5C8C9FF: QConfFile::~QConfFile() (qsettings.cpp:226)
==13197== by 0x5C9238E: QConfFileSettingsPrivate::~QConfFileSettingsPrivate() (qsettings.cpp:1228)
==13197== by 0x5C9D7AA: QScopedPointerDeleter<QObjectData>::cleanup(QObjectData*) (qscopedpointer.h:62)
==13197== by 0x5CF11A4: QScopedPointer<QObjectData, QScopedPointerDeleter<QObjectData> >::~QScopedPointer() (qscopedpointer.h:100)
==13197== by 0x5CE9B46: QObject::~QObject() (qobject.cpp:992)
==13197== by 0x5C95C90: QSettings::~QSettings() (qsettings.cpp:2754)
==13197== by 0x5CC1FDC: LibraryData::~LibraryData() (qlibrary.cpp:428)
==13197== by 0x5CC2441: QGlobalStaticDeleter<LibraryData>::~QGlobalStaticDeleter() (qglobal.h:1767)
==13197== by 0x5FE16DE: __run_exit_handlers (exit.c:78)
==13197== by 0x5FE174E: exit (exit.c:100)
==13197== by 0x5FCBA3D: (below main) (libc-start.c:252)
==13197== Address 0x0 is not stack'd, malloc'd or (recently) free'd
|
| |
| |
| |
| |
| |
| | |
Qt 4.5 did that and the node_copy function depended on it. I have some
patches from Lars that introduce detach3() and try to improve
performance a bit more.
|
| |
| |
| |
| |
| |
| |
| | |
The compatibility method QMetaObject::connect did not handle the
signal_index = -1 to connect to all signal
Reviewed-by: Thiago
|
|/
|
|
| |
Reviewed-by: Trust Me
|
|
|
|
|
|
|
|
| |
When mkdist builds qdoc, it uses a shadow build outside Qt's build. So
.qmake.cache is not present. That led qmake to try and write to /bin,
which would obviously fail.
Reviewed-by: TrustMe
|
|
|
|
|
|
|
| |
When libraryData() is 0, &libraryData()->map is 0x4. So make sure we
return 0 if libraryData() is null too.
Reviewed-by: TrustMe
|
|
|
|
|
|
| |
This is breaking the MacOS X build because -rpath is a supported
option, but it conflicts with others (namely
-mmacosx-version-min=10.4). So it's causing the Mac build to fail.
|
| |
|
|\ |
|
| |
| |
| |
| |
| |
| | |
This is a common effect that many many customers have asked for.
Reviewed-by: Samuel
|
|\ \
| |/ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This reverts commit 3ed5b5ebf5b360dedbba14c03f6ca5701b3b9290. That was
a well-intentioned commit and technically correct, but it breaks when
there are static QDBusConnection objects in something that is loaded
before QtDBus (like libraries that depend on QtDBus or applications
like qdbus)
Since QDBusConnection is ref-counted, the last instance is supposed to
disconnect. However, if we have already unloaded libdbus-1, we can't
do that. Q_GLOBAL_STATICs are destroyed too soon.
|
| | |
|
| | |
|
| |\ |
|
| | |
| | |
| | |
| | | |
Reviewed-by: axis <qt-info@nokia.com>
|
| |/
|/|
| |
| |
| |
| |
| |
| | |
examplesbase.pri sounded to generic and it was not obvious that this
was only for Symbian. Also merged the different symbian rules if they
were spread around in the .pro file.
Reviewed-by: Aleksandar Sasha Babic
|
|/
|
|
|
|
|
| |
QGLContextPrivate::bindTextureFromNativePixmap() needs to abort if it's
not allowed to return a y-inverted texture.
Reviewed-by: Trustme
|
|
|
|
|
|
|
|
|
|
|
| |
Task-number: 259998
Reviewed-by: eskil
The patch removes double drawing of SBP_LOWERTRACKVERT/HORZ
as it is redundant to draw this under the gripper as well as on
the bar itself. It also breaks for certain non-standard XP themes.
The patch was suggested by Jonathan Liu.
|
|
|
|
| |
Last change wasn't enough...
|
| |
|
|\ |
|
| |
| |
| |
| | |
Reviewed-by: TrustMe
|
|\ \
| |/
|/|
| | |
This brings support for the S60 platform into Qt.
|
| | |
|
| |\ |
|
| | | |
|
| |\ \
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Conflicts:
src/gui/graphicsview/qgraphicsitem.cpp
src/gui/kernel/qwidget.cpp
src/gui/kernel/qwidget_p.h
|
| |\ \ \
| | |_|/
| |/| |
| | | |
| | | |
| | | |
| | | | |
Conflicts:
src/corelib/io/qfsfileengine_unix.cpp
src/corelib/plugin/qlibrary.cpp
tests/auto/qfileinfo/tst_qfileinfo.cpp
|