| Commit message (Collapse) | Author | Age | Files | Lines |
|\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
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
...
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The benchmark ran out of stack for large grids somewhere inside
QWidget::show(). Skipped show cases for large grids in Symbian,
since we cannot increase application stack any further.
Also removed some unnecessary whitespace.
Task-number: QTBUG-6693
Reviewed-by: Janne Anttila
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Deployment was handled similarly to WinCE.
Some whitespace was also removed.
Task-number: QTBUG-6690
Reviewed-by: Janne Anttila
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The postEvent function was causing problems in Symbian because it
couldn't handle multiple iterations of the benchmark.
- Moved event allocation inside QBENCHMARK so that each iteration will
have a fresh event object.
- Added resetting of the PingPong counters inside QBENCHMARK.
- Used QTestEventLoop::instance().exitLoop() instead of
QCoreApplication::quit() to make the cases beyond first
to actually test something meaningful.
- Added a "first time" case to postEvent_data, as the first time the
event loop is used it takes much longer that the subsequent times,
at least in Symbian, so that skewed the results.
Task-number: QTBUG-6688
Reviewed-by: axis
|
| |
| |
| |
| |
| |
| |
| |
| | |
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
|
| | | |\ |
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
SqlServer can throw custom error messages (via RaiseError) that can be
up to 8000 chars long. Fixed this with a pre-query as to the length of
the error string, then allocating enough space to retrieve the whole
error string at once.
Task-number: QTBUG-6618
Reviewed-by: Derick Hawcroft
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Using Symbian Math package.
Benhcmarks are showing that some Symbian native math functions
(trigonometric, exp, pow ...) are faster then their Open C counterparts.
Task-number: QTBUG-4894
Reviewed-by: Iain
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
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
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
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
|
| | |\ \ \ |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
NokiaX86 compiler doesn't like QCOMPAREs unless the compared pointers
are cast correctly.
Reviewed-by: Janne Anttila
|
| | | |\ \ \ |
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
These don't have to be platform specific.
RevBy: Trust me
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
When selection was deleted, the changes where not updated on virtual
keyboard. Added new connection to notify when text has changed.
Task-number: QTBUG-4847
Reviewed-by: Markku Luukkainen
Merge-request: 2149
Signed-off-by: axis <qt-info@nokia.com>
|
| | |/ / / /
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Currently QS60Style stores pixel metrics values (96 of them for each
layout) in a static lookup table. There is one "line" for each
screensize, orientation and layout direction (ten in total).
This fix removes the layout direction specific "lines" as there are
only two pixel metrics values that differ depending on direction.
We can handle those two inside QS60Style::pixelMetrics() and
thus remove half of the rows, and thus gain 16*96*5 ~ 8kB of RAM.
Task-number: QTBUG-6803
Reviewed-by: Alessandro Portale
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
- Added deployment of jpeg plugin
- Reduced deep stacking recursion from 1000 -> 200 in order to not
run out of stack.
- Bunch of whitespace fixes.
Task-number: QTBUG-6779
Reviewed-by: Janne Anttila
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
FEP indicator is shown in few cases in statusbar when it should not be.
Case a)
Application contains item view.
QAbstractItemView is initialized as having WA_InputMethodEnabled on,
irregardless of functionality of each separate cell. So we need
to check the model for the itemview cell about to receive focus,
is it editable. If it isn't then, we'll set the WA_InputMethodEnabled
as false.
This will fix the FEP indicators in a lot of cases: lists with cells,
empty lists, combobox lists, ...
Case b)
Combobox also initializes itself with WA_InputMethodEnabled active.
Even if the fix a) fixed the list inside combobox, the widget itself
when receiving focus needs also fixing. If we check the editable
flag of combobox when initializing the comment, we can set the
flag off.
Task-number: QTBUG-5705
Reviewed-by: axis
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Symbian devices typically have limited memory, so skipping the cases
that require tens of megabytes or more of memory.
Task-number: QTBUG-6780
Reviewed-by: Janne Anttila
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
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
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
In some places we do have direct math related calls. Almost all are
calling the double precision functions even if the args and results
are float.
This leads to unnecessary float->double and double->float transitions.
By using wrapper functions we can control which functrion variants
are effectively called.
Task-number: QTBUG-4894
Reviewed-by: Shane Kearns
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
These were missed from previous fix.
Reviewed-by: Alessandro Portale
|
| | |\ \ \ \ |
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
In the style demo, spinbox widget, we noticed that to go from 0 to 99
, with key Up, takes 35 seconds. This was too slow.
It turned out that no acceleration has been implementd when using
keyboard, although acceleration works when mouse used.
We fixed this by using timer events, as it is the case for mouse events.
We also needed to get what is the keyboard repat rate interval from system.
This value is used to set timer intervals in more natural way.
Task-number: QT-927
Reviewed-by: janarve
|
| | |/ / / /
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
3.1 does not have theme graphics IDs that we use in style, so it
cannot be compiled in 3.1. As a fallback, 3.1 will use the previous
method of drawing a slider.
Reviewed-by: Alessandro Portale
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Currently there is no visual clue that slider is "sunken" (pressed
state). There is S60 theme graphic for it, so lets use it.
Task-number: QTBUG-6722
Reviewed-by: Janne Koskinen
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Style uses just a line as slider groove, when S60 provides themed
graphics for it as well.
Task-number: QTBUG-6723
Reviewed-by: Janne Koskinen
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
- Temporary files need to be deleted after each test, as there is no
space to have several 40MB files on typical device.
- Read buffer must be allocated dynamically, as Symbian devices have
limited stack.
- Moved metatype declarations to proper place
- Changed _exit() -> exit()
- Removed assert from around mkdir - dir creation fails if dir exists,
and dir was not created in release builds.
- Added QDir::Files to readSmallFiles test directory filter to actually
find the files to read.
- Fixed filenames to absolute in readSmallFiles test so that it'll find
the files even if they are not in current dir.
- Write a linefeed to the end of each created file in createSmallFiles,
so that the files created have proper size
- Only create 1/10th of files in createSmallFiles for Symbian to
speed up the test to bearable level.
- Added missing ::flose() call to QFileFromPosixBenchmark of open().
- Skipped Windows specific tests on non-Windows platform as just
failing them left temporary files to disk
Task-number: QTBUG-6593
Reviewed-by: MariusSO
|
| | |\ \ \ \
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
Conflicts:
src/s60installs/bwins/QtGuiu.def
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
QS60Style does not draw highlight for a list that user has not navigated
in yet. Once user makes navigation attempt, highlight is drawn to
correct index.
This was due to that QS60Style only used State_Selected for deducing
where highlight should be drawn to.
Task-number: QTBUG-6357
Reviewed-by: Janne Anttila
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
- RVCT 2.2. couldn't compile static templated function f(), so changed
it to non-static, as it should make no difference for this test.
- Limited insert_Large tests to fit to typical Symbian device memory.
- Some whitespace fixed, too.
Task-number: QTBUG-6592
Reviewed-by: Janne Anttila
|
| | | |/ / /
| | |/| | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Some misc whitespace got fixed, too.
Reviewed-by: Janne Anttila
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
The old implementation had been hacked on for a while and needed
cleanup. The new code is heavily based on looking at the behavior of
other platforms. It also reuses more of the cross platform code,
which improves the handling of Enter and Leave events.
We also switched to letting Symbian grab the pointer automatically
when pressing down the mouse button, which improves things
considerably compared to doing it ourselves.
Popups should also work a lot better after this fix, since they were
not really handled at all in the old code.
The old code had calls to set the Symbian cursor sprite. This code
has been removed since that code is now being called from within
dispatchEnterLeaveEvents().
In addition, there was code to check whether the up key event had
been left out by the platform. This was solved a bit differently now,
instead putting the code in the section that handles virtual mouse,
since that is where the problem occurs.
Task: QTBUG-4990
RevBy: Shane Kearns
AutoTest: N/A, Platform specific code that an autotest cannot catch.
Lots of manual testing was done on normal examples as well
drag'n'drop examples and it seemed to work fine.
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
If QS60Style lets QCommonStyle to draw PE_PanelScrollAreaCorner,
common style draws transparent small rect to the end of scrollbar.
This is due to our style's palette's Base is Qt::Transparent.
Error is only visible when two dialogs/windows are on top of each other
and their background is such that a hole in the upper dialog's
background will make the hole "visible"
(i.e. background is not similar).
As a fix, QS60Style should not let QCommonStyle draw the primitive,
but instead handle itself (and do nothing with it).
Task-number: QTBUG-6657
Reviewed-by: Alessandro Portale
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
QStringList benchmark uses some std classes which indirectly
need pthread library.
Task-number: QTBUG-6594
Reviewed-by: Janne Anttila
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
The image can get corrupted if access to it is not properly
surrounded with beginDataAccess and endDataAccess calls.
Task-number: QTBUG-6050
Reviewed-by: Jani Hautakangas
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Currently if we launch QMessageBox::aboutQt(), the dialog grows
too large to fit into any kind of device display.
To fix this, we have added a QTextEdit instead of QLabel into
QMessageBox. QTextEdit automatically adds scrollbars to its content,
if the text won't fit into the designated area.
This change causes error QTBUG-6853. This is due to that QTextEdit
does not support activating external links.
Task-number: QTBUG-3232
Reviewed-by: Alessandro Portale
|
|\ \ \ \ \ \
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
4.6-integration
* '4.6' of scm.dev.nokia.troll.no:qt/oslo-staging-2:
Fix moc generated code with dummy Q_PROPERTY
Fix autotest which checks for sample buffers
Fix EGL surface leaks when re-parenting QGLWidget on X11/EGL
Add autotest for creating a QGLWidget with alpha channel
Fix QGLWidgets created with an alpha channel on X11/EGL
Add QT_DEBUG_X11_VISUAL_SELECTION to aid debugging X11/EGL
Attempt at fixing compilation on Symbian.
Got rid of unused variable compiler warning.
Fix vertical text centering on Mac OS X
Clicking on a selected item wouldn't reset the selection anymore.
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
If there is properties, we cannot skip the code that substract
the property number from the id
Task-number: QTBUG-5590
Reviewed-by: Brad
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
Change c4d66e27 made sample buffers on EGL platforms to default
to off, matching desktop GL. This patch fixes the autotest which
was still testing for the old behaviour.
Reviewed-By: TrustMe
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
When a QGLWidget is re-parented, it's native X11 window usually
gets destroyed and re-created. This also happens when you set a
window attribute or flag. On EGL, we must destroy the surface
for the window before destroying the window itself, otherwise
we can leak the surface. This also fixes lots of BadDrawable
errors when running the autotests (which were due to surface
leaks!).
Reviewed-By: TrustMe
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
Task-number: QT-2602
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
If the QGLWidget's QGLFormat says it should have an alpha
channel, try to find an ARGB Visual.
Reviewed-By: Trond
Task-number: QT-2602
|
| | | | | | | |
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
The Symbian compiler seems to have a problem with forward declarations
of template functions.
Reviewed-by: Bjørn Erik Nilsen
|
| | | | | | | |
|
| |\ \ \ \ \ \ |
|
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
Before commit 88ecc8c8250505129ccff2660c60412996e2fd85, this case was handled
during the mouse release event, but was responsible for selections being made
twice sometimes. We now check whether a selection actually happened during
the mouse press event. If not, we will try to select again during the mouse
release event. Auto-test included.
Reviewed-by: Thierry
Task-number: QTBUG-6753
|