| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
| |
The filesystem autotests are not paralized (because i was scary)
For reference, script I used:
cat corelib.pro | grep -v "=" | grep \\\\ | cut -f1 -d\\ | sed "s/\(.*\) /\1/" | grep -v file | xargs -I{} echo "echo CONFIG += parallel_test >> {}/{}.pro" > t
. t
Reviexed-by: Joao
|
|
|
|
|
|
|
| |
On 64-bit systems, len + sizeof(QByteArray::Data) could overflow and become 0
In this case, qRealloc could succeed and return 0, leading to a double free.
Reviewed-by: Joao
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously subpixel positioning was only used when subpixel rendering
is enabled (glyphType == Raster_RGBMask), however, it does not
necessarily require that and for Mac OS X it will keep using
subpixel positioning even when LCD font smoothing (subpixel rendering)
is turned off. To maintain consistency when switching to raster,
we should support subpixel positioning in this case.
Task-number: QTBUG-5053
Reviewed-by: Jørgen Lind
|
|\
| |
| |
| |
| |
| |
| |
| | |
master-integration
* 'master' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
Add more test entries for built-in JS types
Update generated QScriptValue test files
|
| |
| |
| |
| |
| |
| |
| | |
In particular the "wrapper" types (Boolean, Number, String) were
not tested at all.
Reviewed-by: Jedrzej Nowacki
|
|/
|
|
|
|
|
|
|
| |
A couple of bugs were introduced in the generator when it was
refactored; fixed now.
The generated tests are now in a separate directory/class, so
the test function qualifier had to be updated.
Reviewed-by: Jedrzej Nowacki
|
|\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
master-integration
* 'master' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
Add autotests for setting non-Object as prototype value
Revert new flaky test
compile fix for namespaced Qt
Ensured that the document nodes are sorted in the generated index.
Make qsTr() work with Unicode (non-Latin-1) strings
Add autotest that checks return value of QScriptContext::throwXXX()
Improve autotest coverage of QScriptEngineDebugger
Add benchmark for emitting signals from QtScript
Split monolithic QtScript autotest into smaller tests
|
| |
| |
| |
| |
| |
| | |
Currently it's possible to set the prototype to be any value,
but this doesn't make sense; see QTBUG-15154.
Mark test as expected failure until it's fixed.
|
| |
| |
| |
| |
| | |
Apparently this crashes on Mac. Remove for now, pending further
investigation.
|
| |\
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
master-integration
* 'master' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
compile fix for namespaced Qt
Ensured that the document nodes are sorted in the generated index.
Make qsTr() work with Unicode (non-Latin-1) strings
Add autotest that checks return value of QScriptContext::throwXXX()
Improve autotest coverage of QScriptEngineDebugger
Add benchmark for emitting signals from QtScript
Split monolithic QtScript autotest into smaller tests
|
| | | |
|
| | |
| | |
| | |
| | | |
Task-number: QTBUG-13828
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Converting the source string/context/comment to Latin-1 is very
broken, since JS strings are UTF-16.
The strings should be converted to UTF-8, which should also be
the default encoding for qsTranslate().
Effectively, this bug meant that only Latin-1 characters could
be used in source strings; the translations themselves could
have non-Latin-1 characters. But there was data loss in the
case where you passed a source string for which no translation
was found (since the Latin-1-ized string would be returned).
Task-number: QTBUG-14989
Reviewed-by: Jedrzej Nowacki
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The documentation says that they should return the error/value
that was thrown. When throwError() is called from a native function
call, the return value of the function will be ignored, so we need
a test that calls throwError() outside of a function context.
Reviewed-by: Jedrzej Nowacki
|
| | |
| | |
| | |
| | |
| | | |
The internal coverage is still poor, but at least the public API
has 100% coverage now.
|
| | |
| | |
| | |
| | | |
Reviewed-by: Jedrzej Nowacki
|
| | |
| | |
| | |
| | | |
Reviewed-by: Jedrzej Nowacki
|
|\ \ \
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
master-integration
* 'master' of scm.dev.nokia.troll.no:qt/qt-s60-public:
Fixed deployment for symbian/linux-armcc when using shadow builds.
Started using qtmain.lib for all Qt applications, also corelib ones.
|
| | | |
| | | |
| | | |
| | | | |
RevBy: Miikka Heikkinen
|
| |/ /
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This commit also fixes the issue that 564058a1bb didn't, namely that
config tests broke on Symbian. It does this by continuing to use the
Open C libcrt0.lib library for non-Qt applications. This is also more
correct, since you should not have to compile the Qt libs to compile
a non-Qt app.
Task: QTBUG-14735
RevBy: Miikka Heikkinen
|
|\ \ \
| |/ /
|/| |
| | |
| | |
| | |
| | | |
master-integration
* 'master' of scm.dev.nokia.troll.no:qt/oslo-staging-2:
Really fix getting qmakespec on windows
|
|/ / |
|
|\ \
| | |
| | |
| | |
| | |
| | |
| | | |
scm.dev.nokia.troll.no:qt/qt-integration into master-integration
* 'qt-master-from-4.7' of scm.dev.nokia.troll.no:qt/qt-integration:
Task-number: QTBUG-15059
|
| |\ \
| | |/
| |/| |
|
| | |\
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
Task-number: QTBUG-15059
|
| | |/
| | |
| | |
| | |
| | | |
Removed implicit QtGui linking from qsymbianbearer plugin as it
is not needed.
|
|\ \ \
| |/ /
|/| |
| | |
| | |
| | |
| | | |
master-integration
* 'master' of scm.dev.nokia.troll.no:qt/oslo-staging-2:
Get the qmakespec correctly on all platforms
|
| | | |
|
|\ \ \
| |/ /
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
master-integration
* 'master' of scm.dev.nokia.troll.no:qt/oslo-staging-2:
Revert "QPointer: assert when using QPointer on destroyed QObject"
Silence warning in qvector with MSVC
QStyleSheetStyle: Fix crash that occurs with several instance of QStyleSheetStyle
QStyleSheetStyle: add a way to style the up arrow
QPointer: assert when using QPointer on destroyed QObject
Add missing Q_ENUMS to QAbstractSocket
Fix QMenu rendering in the unified toolbar
Add even more platform info, and make it expandable.
Hide nonfunctional links, and improve html
|
| | |
| | |
| | |
| | |
| | |
| | | |
This reverts commit 96e873d4ed59bd9feef1c5dacc28f77377e53c47.
Many tests where failing. (including the QWidget autotest)
|
| |\ \
|/ / /
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
master-integration
* 'master' of scm.dev.nokia.troll.no:qt/oslo-staging-2:
Silence warning in qvector with MSVC
QStyleSheetStyle: Fix crash that occurs with several instance of QStyleSheetStyle
QStyleSheetStyle: add a way to style the up arrow
QPointer: assert when using QPointer on destroyed QObject
Add missing Q_ENUMS to QAbstractSocket
Fix QMenu rendering in the unified toolbar
Add even more platform info, and make it expandable.
Hide nonfunctional links, and improve html
|
| |\ \
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
master-integration
* 'master' of scm.dev.nokia.troll.no:qt/oslo-staging-2:
Silence warning in qvector with MSVC
QStyleSheetStyle: Fix crash that occurs with several instance of QStyleSheetStyle
QStyleSheetStyle: add a way to style the up arrow
QPointer: assert when using QPointer on destroyed QObject
Add missing Q_ENUMS to QAbstractSocket
Fix QMenu rendering in the unified toolbar
Add even more platform info, and make it expandable.
Hide nonfunctional links, and improve html
|
| | |\ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
master-integration
* 'master' of scm.dev.nokia.troll.no:qt/oslo-staging-2:
Silence warning in qvector with MSVC
QStyleSheetStyle: Fix crash that occurs with several instance of QStyleSheetStyle
QStyleSheetStyle: add a way to style the up arrow
QPointer: assert when using QPointer on destroyed QObject
Add missing Q_ENUMS to QAbstractSocket
Fix QMenu rendering in the unified toolbar
Add even more platform info, and make it expandable.
Hide nonfunctional links, and improve html
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Conversion from size_t to int loose precision
Reviewed-by: Robin Burchell
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
QStyleSheetStyle
The problem is that when a widget is destroyed, it was not properly removed
from the cache. We connected the destroyed signal to a slot in the
QStyleSheetStyle for the first QStyleSheetStyle that handle a widget.
but if this QStyleSheetStyle is destroyed, that connection is lost.
Solution: Create a new QStyleSheetStyleCaches that will be responsible
to clean the caches. This objects is not destroyed as long as there is
a QStyleSheetStyle instance, so the connection is not lost.
I took the oportunity to move all the caches to this object.
Reveiwed-by: Gabriel
Task-number: QTBUG-11658
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Patch originaly from Martin Petersson
Task-number: QTBUG-13491
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Uses the same assert as in QWeakPointer
Reviewed-by: Thiago
Task-number: QTBUG-9398
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Task-number: QT-1635
Reviewed-by: Markus Goetz
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
with the raster engine.
Task-number: QTBUG-12818
Reviewed-by: Samuel Rødal
|
| | | |\ \ |
|
| | | | | | |
|
| | | | | | |
|
|\ \ \ \ \ \
| |/ / / / /
|/| | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
scm.dev.nokia.troll.no:qt/qt-integration into master-integration
* 'qt-master-from-4.7' of scm.dev.nokia.troll.no:qt/qt-integration:
tst_qnetworkreply: New HTTP test with AuthenticationReuseAttribute
tst_qnetworkreply: One more HTTP test using setReadBufferSize()
Hide the cursor in the tested text edit
|
| |\ \ \ \ \
|/ / / / / /
| | | | | _ |
|
| |\ \ \ \
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-multimedia-staging:
Hide the cursor in the tested text edit
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Hiding the cursor will make the image comparison more robust. This fixes
the sporadic test fail we saw on some machines - every time the blinking
cursor was visible, the test failed.
Reviewed-by: Robert Griebl
|
| |\ \ \ \ \
| | |/ / / /
| |/| | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
tst_qnetworkreply: New HTTP test with AuthenticationReuseAttribute
tst_qnetworkreply: One more HTTP test using setReadBufferSize()
|
| | | | | | |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Task-number: QTBUG-13431
|
|\ \ \ \ \ \
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
master-integration
* 'master' of scm.dev.nokia.troll.no:qt/oslo-staging-1: (45 commits)
mkspecs: Move QMAKE_CC and QMAKE_CXX from gcc-base.conf to g++-base.conf
mkspecs: Make linux-x86-openkode-g++ not include g++.conf directly
mkspecs: Include gcc-base-unix.conf from g++.conf to match old behavior
configure: Deal with multiple redefinitions of qmake variables
Split QScriptValue::call autotest into smaller tests.
Optimize and clean up QClipData::fixup()
There is no predefined qmake x11 scope for x11 builds.
Refactor tst_QScriptContext::throwError test.
Fixed whitespace.
Don't use scopes in mkspecs as configure does not support them
Fixed destination path for the S60 feedback plugin.
Don't put Objective-C sources in SOURCES
Fix regression in tst_maketestselftest::naming_convention.
Split monolithic QScriptEngine autotests into smaller tests
Add SunSpider and V8 benchmarks to benchmarks/script.pro
Push and pop the thread-default context for the current thread
Silence preprocessor warnings about __TARGET_ARCH_ARM not being defined.
Import SunSpider benchmarks for QtScript
Import V8 benchmarks for QtScript
Add more microbenchmarks for QScriptValue
...
|