| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
| |
Reviewed-by: Trust Me
|
| |
|
|
| |
Reviewed-By: Samuel Rødal
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Adds new NEON configure test and -no-neon configure option. NEON
implementations can also be turned off by setting the QT_NO_NEON
environment variable.
Performance improvements (in frames per second):
- Blending ARGB32 on RGB32/ARGB32, mostly opaque: 71 %
- Blending ARGB32 on RGB32/ARGB32, no opaque pixels: 108 %
- Blending ARGB32 on RGB32/ARGB32, with 0.5 opacity: 158 %
- Blending RGB32 on RGB32/ARGB32, with 0.5 opacity: 189 %
Task-number: QTBUG-6684
Reviewed-by: Gunnar Sletta
Reviewed-by: Paul Olav Tvete
|
| |\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
4.6-integration
* '4.6' of scm.dev.nokia.troll.no:qt/qt-s60-public: (31 commits)
Fixed qstylesheetstyle benchmark for Symbian
Fixed qdiriterator benchmark for Symbian
Fixed events benchmark postEvent function
Fix for QTBUG-4908 SVG transparency rendering problem.
Long informative texts causes messagebox to grow outside of screen area
qreal-ization
qreal-ization
qreal-ization
QS60Style: Remove layouts with mirrored information
Fixed qgraphicsview autotest build for winscw.
Enabled input method update code for all platforms.
(ODBC) Fixes segfault when error string is larger than 256 chars.
Fixed QGraphicsView benchmark for Symbian.
FEP indicator shown in status pane when it should not
Skipped the most memory intensive tests in QByteArray benchmark.
qreal-ization
qreal-ization
QS60Style: Theme graphics for QSlider in 3.1
QS60Style: Groove changes caused build break of S60 3.1
Slow spinbox on N95 when using keys Up/Down
...
|
| | |
| |
| |
| |
| |
| |
| |
| | |
QPixmap::copy discarded alpha channel.
This fix also removes usage of member variables CFbsBitGc and
CFbsBitmapDevice. Now those are used only in function scope.
Reviewed-by: Sami Merila
|
| | |\
| | |
| | |
| | |
| | | |
Conflicts:
src/s60installs/bwins/QtGuiu.def
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Using math wrapper functions instead direct call. This gives us top-level
control to what (single/double) precision we are effectively using.
Task-number: QTBUG-4894
Reviewed-by: janarve
Reviewed-by: Kim Motoyoshi Kalland
|
| | |/
|/|
| |
| |
| |
| |
| |
| |
| | |
Do the blur in half the resolution in software and then upload the
result as a texture and smooth-scale it on the GPU. This leads to stable
and decent performance regardless of the blur radius, and simplifies the
implementation quite a bit.
Reviewed-by: Bjørn Erik Nilsen
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The paper size determination requires a cupsGetPPD(), which can be a slow
HTTP download.
So doing it on demand makes QPrinterInfo::defaultPrinter() much faster,
because it doesn't actually need the paper sizes at all. Before this fix,
it was requesting the PPD file of every known CUPS printer, every time
it was called.
My battery of unittests for code that uses QPrinter went from 19 minutes
when the cups print server is switched off, to 13 seconds!
This fix also removes some code duplication, which is always nice.
Task-number: 232664, QTBUG-3033
Reviewed-by: Trond
|
| | | |
|
| | | |
|
| |\ \ |
|
| | | |
| | |
| | |
| | | |
Reviewed-by: Eskil
|
| | |\ \ |
|
| | | |/
| | |
| | |
| | |
| | |
| | |
| | | |
uses of QT_DEPRECATED must be protected by #ifdef
Task-number: QTBUG-6649
Reviewed-by: João Abecasis
|
| | | |
| | |
| | |
| | | |
Reviewed-by: Eskil
|
| | |/ |
|
| |/
|
|
| |
Reviewed-by: Tom
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
drawing vertically aligned text
When text is drawn in a font which has an odd-numbered height and it's
vertically aligned in a rectangle which has an even-numbered height
(or vice versa) we have to round the y-position, as the backend cannot
position the text in subpixels. The y-position in this case will be at
half a pixel and we can either round it up or down without increasing
the size of the error.
In Qt 4.5, the font height would be reported as one higher than what it
actually was. Therefore, in many widgets such as comboboxes, we would
appear to round the y-position down when faced with this issue. This is
apparently also what Windows does, probably because most latin fonts
have more glyphs extending into the maximum descent than the ascent.
In Qt 4.6, the font height bug was fixed, and since we use qRound() we
now round half-pixels up, thus moving all the text faced with this issue
one pixel down. This looks odd and inconsistent, both with the platform
look and feel and with older versions of Qt.
The patch reverts to the old look and feel by rounding the half pixels
down rather than up for the particular case in question.
Task-number: QTBUG-6578
Reviewed-by: Samuel
|
| |\ |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
In the paintEvent of the widget used in a QGraphicsProxyWidget, the worldMatrix was wrongly used by the painter instead of the deviceMatrix.
Similar problem in the WindowsXP Style, the worldMatrix was used instead of the deviceMatrix for determining if the widget is transformed (reviewed by eblomfel).
Task-number: QTBUG-5939
Reviewed-by: bnilsen
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@nokia.com>
|
| |\ \
| |/ |
|
| | |\ |
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The first time a path is drawn we call makeCachable on the path, which
means that if it is drawn again, we start caching it. This is a bit of
a trick to avoid caching paths that are drawn once and discared while
at the same time cache paths that are reused automatically.
The GL engine owns the vertex information and is responsible for cleaning
it up. If the vectorpath is destroyed first, it will call the cleanup function.
if the engine dies first, we still require some hooks to clean up the cache
in the path. More to come. When VBO's are used, these will be a leaked if the
path is destroyed after the engine.
Reviewed-by: Samuel
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Fixed potential bug where you could end up taking the square root of a
negative number in drawTextItem() in the raster and OpenGL paint
engines.
Task-number: QTBUG-6327
Reviewed-by: Trond
|
| |/ /
| |
| |
| | |
Reviewed-by: Samuel Rødal
|
| | |
| |
| |
| |
| | |
Since qreal is a float on these targets, it was having trouble picking
the right version of qMax.
|
| |\ \
| |/
|/| |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The WaveUnderline used to be a series of 180 degree arcs, that didn't
always align very well between different text items and that could be
clipped at the bottom since the font height got reduced. Now it uses
quad beziers so that it is more like a sinus curve, and the alignment
and clipping issues have been fixed.
The painting of the WaveUnderline has been optimized using QPixmapCache.
A 100 pixels wide sample of the wave is now cached as a pixmap and
repeated to draw the WaveUnderline. This is an order of magnitude faster
than rasterizing a QPainterPath.
The QPlainTextEdit needs to set the brush origin since it draws the text
items in local widget coordinates, rather than relying on a painter
translation like the QTextEdit.
Done with mae.
|
| |\ \
| | |
| | |
| | | |
4.6-staging2
|
| | |/
| |
| |
| |
| |
| |
| |
| |
| |
| | |
In Symbian we save memory by deleting widget backingstore from
non-visible windows. 12-key FEP is fullscreen and causes the
backing store to be deleted but still continues to scroll the view
to show current caret position in focused lineedit control.
Task-number: QTBUG-5922
Reviewed-by: Jason Barron
|
| |/
|
|
|
|
|
| |
The problem was that the images added to the PDF were not clipped to the expected area.
Task-number: QTBUG-3412
Reviewed-by: Trond Kjernaasen
|
| |
|
|
| |
Over src/ tools/ examples/ and demos/
|
| |\ |
|
| | |\ |
|
| | |\ \ |
|
| | | |\ \ |
|
| | | | | | |
|
| | | | | | |
|
| | | | | | |
|
| | | | | | |
|
| | |_|_|/
|/| | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
The left bearing of the glyph has to be rounded before truncated,
otherwise the subpixel positioning of the glyph will be lost. The
visible consequence of this was on Mac Cocoa where certain pairs of
glyphs would get an extra pixel of space between them when running with
graphics system raster. With small fonts it would look pretty bad.
Task-number: QTBUG-5430
Reviewed-by: Trond
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Removing some unreferenced variables that were left overs after
revert&merge operation.
Reviewed-by: Kim Motoyoshi Kalland
|
| |\ \ \ \
| |/ / / |
|
| | |\ \ \
| | | | |
| | | | |
| | | | | |
4.6-staging2
|
| | | | | |
| | | | |
| | | | |
| | | | | |
Reviewed-by: Trust Me
|
| |/ / / /
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Increased the precision of the output to 9 decimals. Not ideal but
better than just 6.
Reviewed-by: Trond
|
| |\ \ \ \ |
|
| | |\ \ \ \
| | |/ / /
| |/| | /
| | | |/
| | |/| |
|
| | | |\ \
| | | |/
| | |/| |
|
