| Commit message (Collapse) | Author | Age | Files | Lines |
|\
| |
| |
| |
| |
| |
| | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-2:
Improve performance of clipping to a scaled QRectF
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Clipping to a scaled QRectF was so far hitting the slow
code path in the raster paintengine that generates clip spans
for every line. This is not needed as we also clip translated
QRectF's to integer rects.
This patch does the same for scaled rectangles and ensures the
clipping really happens at the closest pixel boundary.
Reviewed-by: Samuel
|
|\ \
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
Suppress a warning about killing a timer that isn't active.
ICC: Don't print warning 1259 since it's too annoying.
Autotest: Add some comments about this obscure test.
Add a warning about trying to release a timer ID that isn't active
Don't need to set FD_CLOEXEC since qt_safe_* will have done that.
Check for null-pointer to avoid a crash in textedit demo.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The timer was active, but it got killed by ~QObject. The QObject
d-pointer is deleted after the timers (as is expected), which means
that the QBasicTimer destructors in QAbstractItemViewPrivate are run
after the timers are already gone.
Reviewed-by: Trust Me
|
| | |
| | |
| | |
| | |
| | | |
ICC 12 keeps outputting this when you copy a type of larger size to
a smaller, or from int to float.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
More information, see commit d9bf386d917c64ad5d8f11f9daadf82b2be9d531
(old qt-history) and old TT task 90183.
Reviewed-By: Trust Me
|
| | |
| | |
| | |
| | | |
Reviewed-By: Trust Me
|
| | |
| | |
| | |
| | | |
Reviewed-By: Trust Me
|
| | |
| | |
| | |
| | |
| | | |
Task-number: QTBUG-16125
Reviewed-by: trustme
|
|\ \ \
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public:
Resize event for QDesktopWidget was sent too early
|
| |\ \ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public:
Resize event for QDesktopWidget was sent too early
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Some recent changes had caused a change in events sent from avkon
to QSymbianControl, causing QDesktopWidget to be get resize event
before the client area of application was correct.
Moved the resize event sending from HandleStatusPaneSizeChange
to HandleResourceChange (case KEikDynamicLayoutVariantSwitch).
Task-number: QTBUG-16095
Reviewed-by: Sami Merila
|
|\ \ \ \ \
| |/ / / /
|/| | | /
| | |_|/
| |/| |
| | | |
| | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-2:
Prevent crash in drawhelper code when the cpu has MMXEXT but no SSE.
|
|/ / /
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
We can't use the qdrawhelper_sse.cpp or qdrawhelper_sse3dnow.cpp when
SSE is not run-time detected, as those are compiled with -msse and
MMXEXT is just a subset of SSE. The compiler might choose to use an
instruction not in the subset, causing a crash at run-time.
Task-number: QTBUG-15693
Reviewed-by: Thiago Macieira
|
|\ \ \
| |/ /
|/| /
| |/
| |
| |
| |
| | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
Add extra type of embedded license.
Fix possible null pointer dereference.
|
| |
| |
| |
| |
| |
| | |
Marketing needs this for commercial customers.
Reviewed-by: Trust Me
|
| |
| |
| |
| |
| |
| |
| | |
When parsing a new connection an access point with the same SSID may not
have been previously seen.
Task-number: QTBUG-15276
|
|\ \
| |/
| |
| |
| |
| |
| |
| | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
Ensure that DBus is connected before all uses.
Fix invalid configurations being added to bearermonitor list.
|
| |\
|/ / |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
In early system startup applications may try to use ICD before it is
contactable. Ensure that the connection to ICD is established before all
calls are made.
If ICD is still not contactable QDBusServiceWatcher is used to monitor
registration of the com.nokia.icd2 address and reconnection is attempted
once ICD is started.
Task-number: Maemo 199755
|
| | |
|
|\ \
| |/
|/|
| |
| |
| |
| | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-2:
Don't delete FBO when resetting glyph cache
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
When the glyph cache is cleared because the max texture size has been
exceeded, we shouldn't delete the fbo as this is only created in
setContext() and will thus not be recreated. All subsequent resizing
of the cache will fail.
Task-number: QT-3971
Reviewed-by: Samuel
|
|\ \
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
Trivial: Fix coding style
Prevent ::flush from being called on QGLWindowSurface if no painting happened.
|
| | |
| | |
| | |
| | | |
Fix coding style of merge request
|
| | |
| | |
| | |
| | |
| | |
| | | |
Merge-request: 2527
Reviewed-by: Harald Fernengel <harald.fernengel@nokia.com>
Reviewed-by: Samuel Rødal
|
|\ \ \
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
* '4.7' of scm.dev.nokia.troll.no:qt/qt-qml:
Docs: implications of creating network access managers in other threads
Clarify that XmlListModel is read only
Revert a fix made for bug QTBUG-15341
|
| | | |
| | | |
| | | |
| | | | |
Task-number: QTBUG-16032
|
| | | | |
|
| | | |
| | | |
| | | |
| | | | |
Reverted a commit 2eee49127b80b5b56c605f76ccea004b03d89577 "Remove active selection when TextEdit loses focus". Contrary to TextInput, by default TextEdit keeps the selection visible when the editor loses active focus. If this is not wanted, a property called persistentSelection can be set false to make selection dependant on the focus.
|
|\ \ \ \
| |_|_|/
|/| | |
| | | |
| | | |
| | | |
| | | |
| | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public:
Fixed a bug in the input methods on S60 with QGraphicsWebView.
Fixed several compile and deployment issues in the mmf phonon plugin.
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
The bug happened because we did not expect to receive a call to
CancelFepInlineEdit() when the widget had not previously been given
any preedit text by S60. However, this can happen, and if it happens
at the same time that there is a temporarily visible character in the
widget (because it will be obscured as a password a second later),
then the temporary character would be discarded.
Fixed by checking in the beginning of CancelFepInlineEdit() whether
we currently have any temporary preedit text or not, and if we do, we
don't clear the contents, since it is the input context itself that
generated the preedit text, not S60.
Also made sure that the temporary preedit text is committed if we
receive a call to UpdateFepInlineTextL(), which may sometimes happen
without a call to StartFepInlineEditL().
Task: QT-4314
AutoTest: Included
RevBy: Shane Kearns
|
| | | |
| | | |
| | | |
| | | | |
RevBy: Trust me
|
|\ \ \ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-2:
QApplication: fix crash exit whith native windows
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Regression introduced in 8dd8db250d92521fda619bdcf3e1c859b37b2da0
Patch comes from the task
Task-number: QTBUG-15774
|
|\ \ \ \ \
| |_|_|_|/
|/| | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-water-staging:
Cocoa: popup hides behind window
Cocoa: combobox does not hightlight when mouse re-hovers the first item
Cocoa: Fix addChildWindow bug where we connect a grandparent to a child
Cocoa: cannot use staysOnTop flag for native file dialogs
Cocoa: make sure stays on top child windows are not levelled down
Cocoa: native child filedialogs sometimes shows non-native on screen
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Reason: when cocoa receives a mouse press/release in a window, it
finds the correct view inside that window and sends the mouse event
to it. But NSWindow also does some other stuff just before sending the
event, like raise and lower windows. So when we override sendEvent, and
more over, do not call [super sendEvent], we stop this raise/lower etc
from working. So, to make this work again, I partially revert change
0b2eab87ad3bd73a0744469a45c29ca098649c9b
Task-number: QTBUG-15638, QTBUG-1517
Reviewed-by: Fabien Freling
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The reason is that on Mac, the highlight is supposed to switch off when
the mouse moves out of the drop down menu. On X11, it stays. So there is
a separate code path for this in qcombobox.cpp. But it fails to clear the
index set in the view when the mouse leaves, which stops the item from
re-highligh when the mouse re-enters.
Reviewed-by: ogoffart
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
A plain bug where we ask for a list of widgets, but forget that
qFindChildren is recursive, which is not what we want.
Reviewed-by: jbache
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
We cannot mix staysOnTop and native file dialogs, since Cocoa will
anyway set it back to NSModalPanelWindowLevel when running it
app modal. There are ways to work around this issue, but the file
dialog also has a button for showing a "create directory" modal
panel, and this we cannot control.
Reviewed-by: cduclos
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
[NSWindow addChild] levels the child to the level of the parent. In Qt
we do not want this. So we do a check after setting up the parent-child
relationship for this.
Reviewed-by: cduclos
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The reason is that we make all child dialogs of a modal dialogs
non-modaly shaddowed, as documented. But we forgot to check if
a window is actually visible before raising it to front.
Also adding harmless auto release pool
Rev-By: prasanth
|
|\ \ \ \ \
| |_|_|_|/
|/| | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
Abort working replies with error when network session fails.
qrgb565: Use the trivial qt_memconvert() and qt_rectconvert()
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Otherwise the connections will stay in a zombie state until the TCP
keep alive timer times out in a couple of hours.
Task-number: Maemo 201619
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Use memcpy() instead of the generic conversion routines where both template
arguments are qrgb565. Yields a massive performance gain for tiling opaque
(background) images in WebKit on 16-bit displays.
This was previously restricted to Q_WS_QWS for some unknown reason.
Reviewed-by: Benjamin Poulain
|
|\ \ \ \ \
| |_|_|_|/
|/| | | |
| | | | |
| | | | | |
* '4.7' of scm.dev.nokia.troll.no:qt/qt-qml:
Remove active selection when TextEdit loses focus
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Task-number: QTBUG-15341
Reviewed-by: Yann Bodson
|
|\ \ \ \ \
| |_|_|_|/
|/| | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public:
Non-hacky fix for qwidget autotest
Fix qapplication autotest deployment
Minor optimization
Fix crash when creating more than one QApplication in single test case
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Turns out there is a nice way for getting current CBA,
CEikButtonGroupContainer::Current(), so using that instead of the
previous hack.
Task-number: QTBUG-15915
Reviewed-by: Janne Koskinen
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Necessary default deployment was getting removed from qapplication test
Task-number: QTBUG-15915
Reviewed-by: Janne Koskinen
|