| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
Setting a pixmap brush when painting to a 32-bit target might cause the
source pixmap to be converted to 32-bit. We should detach the pixmap if
we need to convert it.
Reviewed-by: Trond
|
|
|
|
|
|
|
|
|
| |
This test already uses qWaitForWindowManager(), but on X11, reshowing
a dialog can take long as the window manager has to do more work
before remapping the window. we should give it more time (hence the
change to QTRY_VERIFY)
Reviewed-by: Rohan McGovern
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The default cache_limit in different platforms have different value.
For example in Symbian the default is currently 1024 KB where as for
desktop platforms the default is 10 MB.
The purpose of modified qpixmapcache test cases was to do operations
until cache_limit was reached. This was achieved by hard coded 40000
iterations. However this hard-coded value is fargile for cache limit
changes, and in addition it unnecessarily made the test exectuion to
take very long time on platforms which had smaller cache limit.
This patch changes the test so that number of expected items to fit
in cache is calculated and then 1000 extra items is tried to put
in cache to make sure limit is exceeded.
Reviewed-by: Alexis Menard
|
|
|
|
|
|
|
| |
The test tst_QListView::task254449_draggingItemToNegativeCoordinates was failing
in cocoa because of this. (on, cocoa, the call to show was doing the first paintEvent)
Reviewed-by: Thierry
|
|
|
|
|
|
|
|
|
|
|
|
| |
When setting a large font using style sheets, the whats this
popup size calculation would be incorrect resulting in
half visible lables. By calling ensurePolished before showing the
label, we ensure that this will be properly handled.
We also added a new test case for whatsThis in tst_qtooltip
Task-number: QTBUG-2416
Reviewed-by: ogoffart
|
|
|
|
|
|
|
|
|
| |
It also invokes the destructor directly, even on forward-declared types.
That must be an ABI convention.
This test isn't reliable.
Reviewed-by: TrustMe
|
|
|
|
|
|
|
| |
The device can't cope, so let's keep only the small thread count tests
there.
Reviewed-by: Trust Me
|
|
|
|
| |
Reviewed-by: Trond
|
|
|
|
|
|
|
|
|
|
|
|
| |
Nokia X86 compiler is not able to resolve correct templated qCompare
functionfor derived types. So implicit casting does nto work corectly.
I have tried to add template specilations for qCompare also, but then
compiler complainf about unambiguous functions.
The promlem for now is fixed by adding explicit cast to same type
against which the comparuion is done.
Reviewed-by: TrustMe
|
|
|
|
|
|
|
| |
Without this fix we got the following error:
"line 776: Error: #254: type name is not allowed"
Reviewed-by: TrustMe
|
|
|
|
| |
Reviewed-by: trustme
|
| |
|
|
|
|
|
|
|
| |
Reset the mouse position to a known state at the start, instead of
depending on what the previous test did.
Reviewed-by: Jeremy
|
|
|
|
|
|
|
|
| |
It turns out that we need to let singleStep keep its value
in qtextedit so that pushing the up/down buttons on the slider
works as they should. Moreover, overriding the behaviour in
abstract slider also makes QGraphicsView work out of the box.
Also added in the test fix suggested by Olivier.
|
|
|
|
|
|
|
|
|
| |
This reverts commit ee51f0ee63fe5c6388cea1a1043963303ee3810b.
The selftest should "see" the git settings as they are to detect
if they are wrong on CRLF platforms.
Reviewed-by: Aleksandar Sasha Babic <aleksandar.babic@nokia.com>
Reviewed-by: Peter Hartmann
|
|\ |
|
| |\
| | |
| | |
| | |
| | |
| | | |
Conflicts:
src/gui/graphicsview/qgraphicsscene.cpp
src/gui/graphicsview/qgraphicsscene_p.h
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
If you set the flag itemClipsChildrenToShape to true on a parent, an
optimization was made in 4.5.0 to not add children of this parent in
the index. But when you set the flag back to false all the sub-tree
of the parent should be re-added to the index otherwise the index will
never find all children. This code is not relevant for 4.6 since
the index part of QGraphicsView has been refactored and handle this case
with itemChange in QGraphicsSceneIndex.
Reviewed-by:andreas
|
| | | |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
It is much better if test case is self contained, not referring
to the files in some other projects.
Alos, this makes life easire when it comes to deployement on
the embedded devices.
Reviewed-by: Peter Hartmann
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
In the qhttpnetworkconnection.cpp the defaultChannelCount is
hardcoded to 3, for Symbian platform, and that fact has to be
reflected in test case as well.
Reviewed-by: Markus Goetz
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Since the GC looks for pointers in the C stack, try to kill those
pointers before calling collectGarbage().
Reviewed-by: Simon Hausmann
|
| | | |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
- clock() should return valid value except in stop state, return 0.
- Added clock() function testing to unit tests.
Reviewed-by:Justin McPherson
|
|/ /
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This code is just causing more problems, than helping us to solve
autotest problems. So initial purpose of this code is against us, why:
qWarning used in this code sto print debug info is not visible in HW
since global static objects are initialized before the QTestLib
initializes the output channel. Running hostname resolving in globab
static object constructir is not very trivial, requiring IAP setup and
all network related code to be executed. Since application environment
is no yet completely running at the time when gloabl static are
costructed, the operation is deemed to fail.
Failing means calling qFatal whcih in essence is causing panic in very
early phase of test execution. This happens also for some QtCore
autotests which have minor dependency to QtNetwork. making the autotest
result to look worse than they actually are.
Reviewed-by: Aleksandar Sasha Babic
|
| |
| |
| |
| |
| |
| | |
for the Q3Table one, it sometimes fail with XPASS.
There is no point trying to stabilize a failure on a test for a class
we are not likely to touch anyway.
|
|\ \
| |/ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
When the editor had been created inside the QtPropertyEditorView (inheriting
QTreeWidget), the subsequent show sent a synthetic mouse move event down to
the QLineEdit, and a new selection was made on the text because the mouse
button was marked as pressed in the event.
QApplicationPrivate::sendSyntheticEnterLeave() now sends a mouse move event
without any button pressed. Auto-test included in tst_QWidget.
Task-number: QTBUG-4055
Task-number: 253159
Task-number: QT-659
Task-number: 245398
Reviewed-by: bnilsen
|
| |
| |
| |
| |
| |
| | |
added file to preserver the correct line endings for the test file
Reviewed-by: Aleksandar Babic
|
| |
| |
| |
| |
| | |
Task-number: QT-2243
Reviewed-by: thartman
|
| | |
|
| |
| |
| |
| |
| | |
Also improves a bit the quality of the code by sending the event
instead of calling directly the event handler
|
| |
| |
| |
| |
| |
| |
| | |
Add a testcase that checks that high priority requests are actually
high priority.
Reviewed-by: Peter Hartmann
|
| |
| |
| |
| | |
Reviewed-by: Jesper
|
| |
| |
| |
| | |
Reviewed-by: TrustMe
|
| |
| |
| |
| |
| |
| |
| | |
SRCDIR problem fixed and deployment augmented to accomodate image and
icon tests.
Reviewed-by: Joerg
|
| |
| |
| |
| |
| |
| |
| | |
These test will require more work on WinCE which we will invest
shortly.
Reviewed-by: Joerg
|
| |
| |
| |
| |
| |
| |
| | |
Test does not really make sense on WinCE as it is.
Lists with items size over 5000 are not useful there.
Reviewed-by: Joerg
|
| |
| |
| |
| |
| |
| |
| | |
If you let top level widget hanging around with a broken state it
may crash when a process event is triggered.
Reviewed-by:ogoffart
|
| |
| |
| |
| | |
The test relies on wall-clock time.
|
| |
| |
| |
| |
| |
| |
| |
| | |
The test tries to load a file from an unsuported format.
The problem is that tga is a supported format if KDE plugins are installed
It is unlikely that cpp will ever be a supported image format
Reviewed-by: paul
|
| |
| |
| |
| |
| |
| |
| |
| | |
Introduced a helper function in our custom source provider,
columnNumberFromOffset(), that maps an absolute offset in the source
input to a relative column number.
Reviewed-by: Jedrzej Nowacki
|
| |
| |
| |
| |
| |
| | |
Pulse machine runs a 16bit display, some comparisons fails.
Reviewed-by: Samuel
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Don't issue this warning:
comparing floating point with == or != is unsafe
There is many of them and the Qt code is correct (it has to be fast comparison
to know if fast code path must be taken)
Reviewed-by: Jeremy
|
| |
| |
| |
| |
| |
| | |
Calling rm on . remove the dir on Windows.
Reviewed-by:TrustMe
|