| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
| |
We keep around a engine-global index array which we only grow when we
need to. This should hopefully speed up text drawing on some devices.
|
| |
|
|
|
|
|
|
| |
glDrawElements needs to be handled properly, using triangle strips
rather than triangle lists, so the code has been removed for now.
The VBO caching did not gain us any particular performance. We rather
cache the client-side buffers for the same speed-up but with less
clean-up complexity.
|
| |\ |
|
| | |\
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
4.6-integration
* '4.6' of scm.dev.nokia.troll.no:qt/oslo-staging-2: (49 commits)
Fixed qt_x11_wait_for_window_manager
Revert change 34f1758 on non-Synbian platforms
Fixed QGifHandler::loopCount().
Fix tst_QAccessiblity failure.
fix compile error when linuxinput keyboard driver is compiled as plugin
Fixed the context menu test case in the qgraphicsscene autotest.
fix regression from Qt 4.5 wrt missing text pixels in QTabBar
Cleanup after "Changed qgraphicswidget autotest to use qWaitForWindowShown"
Cleanup after "QGraphicsWidget is painted twice on the first show"
Changed qgraphicswidget autotest to use qWaitForWindowShown.
Improved QTest::qWaitForWindowShown on X11.
Stabilize tst_QGraphicsWidget::initialShow2 (new test)
Make sure we define S_IFDIR on Windows
Fix copy/pasto
Crash when closing any top-level widget on Symbian.
Cache the sizes of the images in an animated GIF.
Fix the QAbstractSlider autotest.
Use OpenVG scissor on 90/180/270 rotations and simple clips.
Optimize single-rect IntersectClip in OpenVG using the scissor
QGraphicsWidget is painted twice on the first show.
...
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
When we wait for the window to be shown by looking for ReparentNotify,
MapNotify, etc events in the event queue, we should check if those events come
for the right window, otherwise we might exit too early if there are event for
an other window in the queue.
Reviewed-by: Leonardo Sobral Cunha
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The change introduces behavior changes (including crashing on QWS).
It is only critical for Symbian. To reduce risk, we only apply it
on the Symbian platform for now.
Task-number: Autotest regression
Reviewed-by: Jesper
|
| | | |
| | |
| | |
| | |
| | | |
Task-number: QTBUG-7037
Reviewed-by: Kim
|
| | | |\
| | | |
| | | |
| | | |
| | | | |
Conflicts:
dist/changes-4.6.2
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This started to fail after 34f1758428282a327c12b0d8040061c1f67ecc7f.
Or actually, the test crashes on my machine. Reason is that the the test
first creates a top-level line edit (which then gets its own backing
store). The line edit is then reparented into another top-level. When
the line edit is destroyed extra->topextra->backingStore is true and we
delete the backing store it first got when created as a top-level.
However, the line edit was reparented so this backing store is not the
"active" one. We should still delete topextra->backingstore, but we must
also remove any pointer references to the line edit in the "active" backing
store.
Reviewed-by: jbarron
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
When creating the driver instance two parameters are given to the
driver. But it accepts only the name of the device to be used.
Reviewed-by: Paul
Merge-request: 2288
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
The line spacing fix showed a bug in QCommonStylePrivate::tabLayout.
Since QFontMetrics::height() now usually returns one pixel less than
in Qt 4.5, the tab bar is one pixel smaller.
Squeezing the tab rect vertically can result in missing pixels.
This depends on anti-aliasing settings and font size.
The new behaviour in tabLayout is now: If we have to shift the tab
rect, then we move its position instead of changing its height.
Task-number: QTBUG-7137
Reviewed-by: jbache
|
| | | |\ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.6-integration
* '4.6' of scm.dev.nokia.troll.no:qt/oslo-staging-2: (42 commits)
Cleanup after "Changed qgraphicswidget autotest to use qWaitForWindowShown"
Cleanup after "QGraphicsWidget is painted twice on the first show"
Changed qgraphicswidget autotest to use qWaitForWindowShown.
Improved QTest::qWaitForWindowShown on X11.
Stabilize tst_QGraphicsWidget::initialShow2 (new test)
Make sure we define S_IFDIR on Windows
Fix copy/pasto
Crash when closing any top-level widget on Symbian.
Cache the sizes of the images in an animated GIF.
Fix the QAbstractSlider autotest.
Use OpenVG scissor on 90/180/270 rotations and simple clips.
Optimize single-rect IntersectClip in OpenVG using the scissor
QGraphicsWidget is painted twice on the first show.
Compiler warning in QAbstractSlider.
Optimize QPathClipper::pathToRect.
QGraphicsScene: Use QPainter::setClipRect instead of setClipPath if possible.
QStyleSheetStyle: Fix combinaison of border-image and border-radius
Revert "QAbstractScrollArea: Wheel over a scrollarea that has only one horizontal scrollbar"
Proper Fav icon is not shown, for all the links default fav icon shown
Update changes-4.6.2.
...
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Commit: dda8a57c085216db609f822837c50bae38006b4e
We don't want to reset 'updateAll' at that point, for the same reason as
mentioned in the above commit. More details in the task.
Task-number: QTBUG-6956
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The function is supposed to wait until the window has been managed by the
window manager on X11 - i.e. it has been reparented to a frame, mapped and
received at least one Expose event after that.
Reviewed-by: Olivier Goffart
|
| | | | | |
| | | | |
| | | | |
| | | | | |
Reviewed-by: Thierry Bastian
|
| | | | | |
| | | | |
| | | | |
| | | | | |
Reviewed-by: trustme
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The window surface must be deleted while the window is valid because on
some graphics systems a notifcation is sent to the window when the
surface is released.
This fix is also an optmization as we no longer process any backing
store requests while deleting the top-level. Previously it would
handle requests from the window itself and all its children.
Task-number: QT-2513
Reviewed-by: jbarron
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Rework the previous commit a bit and include caching of image sizes.
Task-number: QTBUG-6696
Reviewed-by: Kim
|
| | |\ \ \ \
| | |_|_|/
| |/| | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.6-integration
* '4.6' of scm.dev.nokia.troll.no:qt/berlin-staging-1:
don't use QKeySequence::mnemonic() after all
don't use stylesheet for just making labels bold
remove the rounded extra frame around the main message editor
add const
|
| | | | | |
| | | | |
| | | | |
| | | | | |
Reviewed-by: hjk
|
| | |\ \ \ \
| | |/ / /
| |/| | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.6-integration
* '4.6' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
Compile.
Another fix for the non unified title+toolbar regarding text under icons
My 4.6.2 changes
Added my changes to changes file.
Doc: Correcting image
My changes for 4.6.2
My changelog for 4.6.2
My Changelog entries for 4.6.2
Add the QMAKE_FILE_EXT variable to extra compilers generation.
Doc: Removed promisse to fix a problem
Add some entries to the change log
Change log updated
Improve raster graphics system performance on Mac (second try).
I10n/German: Update translations for 4.6.2
I10n: Use 'real' Linguist comments in Phonon::EnvironmentalReverb
|
| | | | | | |
|
| | | |\ \ \
| | | |/ /
| | |/| |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.6-integration
* '4.6' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
Another fix for the non unified title+toolbar regarding text under icons
My 4.6.2 changes
Added my changes to changes file.
Doc: Correcting image
My changes for 4.6.2
My changelog for 4.6.2
My Changelog entries for 4.6.2
Add the QMAKE_FILE_EXT variable to extra compilers generation.
Doc: Removed promisse to fix a problem
Add some entries to the change log
Change log updated
Improve raster graphics system performance on Mac (second try).
I10n/German: Update translations for 4.6.2
I10n: Use 'real' Linguist comments in Phonon::EnvironmentalReverb
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This makes the case where there is text under the icons in toolbuttons
and the title and toolbar is not unified on Mac. This was a regression
against Qt 4.5.x
Reviewed-by: jbache
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The problem is out of scope and will not be fixed.
Task-number:QTBUG-4377
Reviewed-by:Trust Me
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This fixes the qmlviewer "sluggish animations and lost
mouse events" issue by making sure we don't block and
wait for for the screen refresh when flushing the
backing store to the screen.
NB: This commit fixes build issues found in f5f62c0bed.
Review: msorvig
Details:
- Don't force repaints, flush the backingstore in response
to a Cocoa paint/display events only.
- Flush once per window.
- Get the CGContext from the window (don't create a new one)
- Don't call CGContextiFlush on the context.
|
| | | | | |
| | | | |
| | | | |
| | | | | |
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
|
| | |\ \ \ \
| | |/ / /
| |/| | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.6-integration
* '4.6' of scm.dev.nokia.troll.no:qt/berlin-staging-1:
uic: Fixed code generating setCentralWidget()-calls of QMainWindow.
|
| | | |\ \ \
| | | |/ /
| | |/| |
| | | | |
| | | | |
| | | | |
| | | | | |
4.6-integration
* '4.6' of scm.dev.nokia.troll.no:qt/berlin-staging-1:
uic: Fixed code generating setCentralWidget()-calls of QMainWindow.
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Regression introduced by QTBUG-5824 change
6809bd020e3307324e136707f32f6f17f77b9591. Do not generate
setCentralWidget() for Qt3Support toolbar/dock widget classes.
Reviewed-by: ogoffart <olivier.goffart@nokia.com>
Task-number: QTBUG-7612
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
For graphics systems that use EGL surfaces in the backing store
destroying the surface does not guarantee that the memory is
immediately freed because this command does not cause a flush. This
implies that a manual flush is instead needed. We do this in 2 places;
the first is when the surface is destroyed due to a visibility changed.
The second case is just after the window has been destroyed. At this
point the backing store has already been deleted so the deletion of
both the surface and window can happen atomically in WSERV.
Task-number: QT-2506
Reviewed-by: Iain
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The support for multiple screens is not implemented yet in Qt on the
Symbian platform, so there is really no need to query the HAL for the
value since the additional screens can't be used yet anyway. The crash
here occured when the HAL returned more than 1 screen, but the arrays
were resized to contain one element. The loop in resizeEvent() was
iterating past the end of the array because it thought there were more
screens than the arrays did.
Reviewed-by: axis
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The logic for resolving the icon size in landscape mode for S60 5.0
(5800XM) is somewhat fragile, but better way is not yet know.
Also the 50% transparent mask what CBA implementation tries to create
for pressed down CB Abuttons fails for some reason. When button is
pressed down there are drawing artifacts in softkey images.
These issues will be tried to resolve with later commits
Task-number: QTBUG-7314
Review-By: Sami Merila
|
| | |\ \ \ \
| | |/ / / |
|
| | | | | |
| | | | |
| | | | |
| | | | | |
This reverts commit 7bc18035816a2eac15dfac4d987eb9bf43f90ef6.
|
| | | | | |
| | | | |
| | | | |
| | | | | |
This is already included in the Q_DECLARE_PRIVATE macro.
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
We can't add new symbols to QPainter for Qt 4.6.x, as we would not be
able to remove them again if we regretted the API. Hence, I've made
removable symbols instead, a private global function and a
drawStaticText() in QPainterPrivate. In order to tie things together,
I needed a static private-getter in QPainterPrivate, and hence it
had to be a friend of QPainter.
Reviewed-by: Trond
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Follow pattern from qpaintengine_opengl.cpp and listen for the
destroyed signal of the GL context. If the context that owns the VBOs
is deleted, then we delete the VBOs as well.
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
No reason to look through all the commands, since we have the transform
saved in the state.
Reviewed-by: Trond
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
A small hack to turn off the current matrix when drawing QStaticText,
since its coordinates are already transformed.
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
If you draw the QStaticText object to a new QGLWidget, then we need to
recreate the VBOs (if the vertex arrays are cached.) It would be
possible to share this across contexts, which can be implemented later
as an extra optimization for switching between different target
devices.
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Some changes were made to QTextEngine for the sake of QStaticText,
but as the implementation of QStaticText changed, these are no longer
needed. Although they are harmless, I'll remove them to minimize the
diff.
|
| | | | | |
| | | | |
| | | | |
| | | | | |
Revert some unnecessary changes from the QStaticText diff.
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Removed an commented-out, empty functions which was previously there
for testing purposes.
|
| |\ \ \ \ \
| | |_|_|/
| |/| | |
| | | | | |
qstatictext-4.6
|
| | |\ \ \ \
| | |_|/ /
| |/| | |
| | | | |
| | | | |
| | | | |
| | | | | |
Manual resolve of conflict in changes-4.6.2 in staging-2.
Conflicts:
dist/changes-4.6.2
|
| | | | | |
| | | | |
| | | | |
| | | | | |
This reverts part of commit d53315d30b38352db11063096ee3867a40bdc234.
|
| | | |\ \ \
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Conflicts:
dist/changes-4.6.2
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Task-number: QTBUG-7864
Reviewed-by: Sarah Smith
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Task-number: QTBUG-7791
Reviewed-by: Sarah Smith
|
