When qdoc finds a .qdoc file in an exampledirs directory,
it only adds it to the source list if it isn't already in
the source list. This may be overkill, but it's safe.
Change-Id: I2c5714c968f06e90f9b29b3a5481f80469e19ced
Reviewed-by: Jerome Pasion <jerome.pasion@nokia.com>
If glXCreateContextAttribsARB does not succeed or is not available,
we should fallback to using glXCreateContext. But we should not just
create a context with glXCreateContext by default that is being thrown
away if glXCreateContextAttribsARB succeeds.
Otherwise glXMakeCurrent with context 0 might cause an unexpected
context change when dealing with multiple contexts.
Change-Id: I7627abbe2500b4006180653a1b3b074fe7aca1d3
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Moreover, exporting it causes the horrible side-effect on Windows
(with MSVC) that the compiler will not inline the functions, but
instead will place indirect calls to the functions in QtCore DLL,
even in release mode. For such a critical piece of code, inlining
is necessary.
Change-Id: Ib31c12f6bf8dc8ece1b51824716a480559753c24
Reviewed-by: Marius Storm-Olsen <marius.storm-olsen@nokia.com>
This is so we can insert valgrind (helgrind) annotation macros. They
require the actual address of the variable to work.
Change-Id: I988f6a46385ad58143c53ad34b6cf0f58be2cdb8
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
This can't realisitically be unit tested.
Task-number: QTBUG-26305
Change-Id: If7f56c44c472ff0ffbda4744b76ed2119bb64bf8
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Previously, a QVariant parameter would be wrapped inside a new
QVariant, and you would have to cast the QSignalSpy's QVariant to
a QVariant to get the actual value. This behavior was unintuitive
and undocumented.
Check if the parameter type is QVariant, and copy it directly if it
is. This makes the QSignalSpy's QVariant directly usable (no need to
"unwrap" the value in user code).
Existing tests that use QSignalSpy together with QVariant parameters
(such as tst_QPropertyAnimation::valueChanged()) and do cast the
QVariant parameter to a QVariant, continue to work after this change;
this is because qvariant_cast<QVariant>() returns its input value
(unchanged) when the type is not QMetaType::QVariant.
Task-number: QTBUG-21645
Change-Id: Ibfb171edd60c0d3f7ca1d5419e5c5f3d0380d5b3
Reviewed-by: Jason McDonald <jason.mcdonald@nokia.com>
There was a race where QGuiApplicationPrivate::processMouseEvent accessed
QWindowSystemInterfacePrivate::windowSystemEventQueue without holding
QWindowSystemInterfacePrivate::queueMutex.
There was a memory leak where QWindowSystemInterfacePrivate::windowSystemEventQueue
would not delete events contained in it when it was destroyed.
Fix both of these by properly encapsulating the QList/QMutex pair
in a small class, WindowSystemEventList, that allows only properly protected access
to the internal QList and calls qDeleteAll() in its dtor.
Change-Id: Ifaa9968c9272096df2f7109a7a6cf1c8e5fa736c
Reviewed-by: David Faure <faure@kde.org>
Reviewed-by: Jørgen Lind <jorgen.lind@nokia.com>
Those are defined below, after the list that describes the macros and
the papers.
Change-Id: I1f2df0e33c84eb17ebbb0147662f560defed182c
Reviewed-by: Marius Storm-Olsen <marius.storm-olsen@nokia.com>
This allows code using methods marked Q_DECL_NOEXCEPT to benefit from
optimisations before MSVC supports the C++11 keyword. Even MSVC 2012
doesn't have it yet.
Using throw() in other compilers is not a good idea because they might
actually be implementing the C++ standard -- which is broken.
Change-Id: Id07ab4fe40a641583d5285d5abb536998bc419ba
Reviewed-by: Marius Storm-Olsen <marius.storm-olsen@nokia.com>
QMutexLocker does not support being passed already-locked mutexes,
unless they are recursive mutexes. But in that case, it behaves as if
the mutex weren't locked in the first place.
Since that's the case, there's no point in testing the low bit to see if
it's set or not. It's never going to be.
Change-Id: Ie4b81f7e2cca16e6db36f3cb51a5377dbdfc157d
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
The GL_IMG_texture_npot extension only provides partial support for
npot textures in that it allows use of npot textures but it does not
support the GL_REPEAT texture mode (needed for tiling fill modes in the
QQ2 image element).
Adding this new finer-grained feature check allows QQ2 to still use npot
textures where GL_REPEAT is not needed with only the IMG extension
present.
A follow-up commit will make a check for this in qtdeclarative.
Change-Id: Iff3dbdb955fb334d9e32f3abd49e90ff0ed9836c
Reviewed-by: Gunnar Sletta <gunnar.sletta@nokia.com>
This was a regression caused by acbfb4d777,
causing for instance qdoc to crash.
Change-Id: Id8daa19d467c4f1729e5a5a203a388f16ec4a6de
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
-Simple landing page which contains a link to the SQL guide and API.
-Gave a new title to C++ API page.
Change-Id: I54eca4f6933bafa0affd5825e7fe2e1a2522dad0
Reviewed-by: Geir Vattekar <geir.vattekar@nokia.com>
Make the test behave in Jenkins similarly as it behaves in Pulse:
- a test run in Jenkins is not an ad-hoc run
- the JENKINS_HOME environment variable implies we are running in
Jenkins
- the GIT_BRANCH environment variable, set by the Jenkins git plugin,
is equivalent to PULSE_GIT_BRANCH
- there is no equivalent to PULSE_TESTR_BRANCH, since testr is no
longer used
Change-Id: I89ffeec659b4adaab309d8b93ad793ce640029c7
Reviewed-by: aavit <qt_aavit@ovi.com>
Using a reference changes the value of oldState within this function,
which is undesired.
Change-Id: I9fb66e488015d6b3e586ffa2f0b05a40c095e16b
Reviewed-by: Christian Kamm <christian.d.kamm@nokia.com>
When using [ ] for tests in shell scripts, the ending ] must be a
separate parameter. Otherwise, it won't work. configure was reporting:
configure:5918: ']' expected
Change-Id: I38a843356ee0feb97edb8692a828306821045c77
Reviewed-by: Laszlo Papp <lpapp@kde.org>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
qdoc was mistakenly assuming that the output
was going into subdirectories. The base dir
variable was not being tested.
Task nr: QTBUG-26638
Change-Id: I9b331926f8954b58102f75fad3f233eaebb2bb4d
Reviewed-by: Jerome Pasion <jerome.pasion@nokia.com>
qmake -project was always outputting a project with subdirs template, because
Option::h_moc_mod was not being properly read, causing addFile() to misbehave.
Change-Id: I2c07aea132f9885eabf188de993b0fabfb352886
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
configure will now run qmake without -recursive, as on modern systems
one can get a lot more out of parallelization done by make, which qmake
cannot do.
use -fully-process to get back the old behavior. -dont-process is
unchanged.
Change-Id: I2874321a963175463ae8992f3ab2b01bc13c9922
Reviewed-by: Joerg Bornemann <joerg.bornemann@nokia.com>
i think this was a vestige from the old wince build proces
Change-Id: I13305de627c65fc1c93ba6d2383557ff5ea7a147
Reviewed-by: Andreas Holzammer <andreas.holzammer@kdab.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@nokia.com>
Fixed crash when resizing a scroll area with a native viewport
(typically a QGLWidget) causes the scroll bars to get hidden, while
there are still expose events in the queue for the scroll bar widget.
Task-number: QTBUG-26746
Change-Id: Ia77c8eb32a6730670333120af3f9a772c64807f1
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
In the case of a file dialog with a completer, the completion list view
is a popup that is not a transient child of the file dialog. Thus it
ends up getting its blockedByModal flag set due to the file dialog
window being a modal window.
To prevent this we make sure popups don't get blocked by modal windows,
as they were special cased and processed before modal windows in Qt 4.x.
Change-Id: Id5688cfb1534541cd1678b1e53d75e34a68f9b8e
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
enum values in the Qt namespace were being printed
as. e.g.: Qt::Qt::AlignLeft. This was due to the
presence of a collision node that was not handled
properly.
Task nr: QTBUG-26628
Change-Id: I54adaba72410b2838f0922f181846bb7e76b61d4
Reviewed-by: Jerome Pasion <jerome.pasion@nokia.com>
This function is called in OOM situations and when other exceptions
are still in flight, so it really shouldn't throw, indeed.
Change-Id: I50cda699ffd74f3710c3bafd15af356ff410bc47
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
The naming suggests that this variable should be set to true here (instead of leaving the expression result unused). Also, the variable isn't written to anywhere else.
Change-Id: I8aae904f6e4456ce0bdd053ce89c2721168dccd3
Reviewed-by: Laszlo Papp <lpapp@kde.org>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
The function showed up in the profiler when doing
heavy logging involving timestamps in Qt Creator.
Change-Id: I81c4e22981a5de91df3da6a0f1df2a08552fa71a
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
they are an unreadable and unmaintainable mess. the options are properly
documented below.
Change-Id: If2ec683fb7c3740b19798979f8a1f9cd8d84f457
Reviewed-by: Joerg Bornemann <joerg.bornemann@nokia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QT_PRIVATE is meant to be for something different.
only jsondb uses that so far. as it has little activity, i feel free to
break it temporarily.
Change-Id: Ifaa054af7234594ee534120c99506b76aebbf59f
Reviewed-by: Joerg Bornemann <joerg.bornemann@nokia.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
the paths may be explicitly added before some other paths, so it would
be wrong to remove them.
Change-Id: I68ae93fd307afe14a07a0f24de952783950b5bea
Reviewed-by: Holger Freyther <holger+qt@freyther.de>
Reviewed-by: Joerg Bornemann <joerg.bornemann@nokia.com>
QDBUSXML2CPP_{INTERFACES,ADAPTORS}_{HEADER,SOURCE}_FLAGS are now
understood.
Change-Id: I26492cec5ecb7ff4a6b83755ee8a0c17859b2a68
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@nokia.com>
it's two times the same code, once with "adaptor" (and an -a option) and
once with "interface" (and a -p option).
the de-duplicated code looks a bit nasty, as qmake cannot deal with
braces on the LHS yet.
Change-Id: I199a07947661ab9abe7a736b56c437f3f6dc4fa4
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@nokia.com>
The QIcuCodec::defaultCodecUnlocked function is not thread-safe
because it calls codecForNameUnlocked. In turn, that one accesses and
even modifies a QHash stored in the QCoreGlobalData singleton, which
is what makes it non-threadsafe.
In order to call the Unlocked function, we need to lock the mutex
first.
Change-Id: I915570110229f4c0929986aa26731244317ef6ab
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Mark the thread-safe functions with the \threadsafe doc marker. This
includes public API, which should be thread-safe anyway.
The thread-unsafe functions are marked "\nonreentrant" already. In
addition, I renamed the functions that must be called with locked
mutexes to Unlocked, following the convention in other libraries like
libdbus-1.
Change-Id: Ibd93d1266149767f546c8e82959b73c138008469
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
This solves a new corner case resulting from the recent refactoring of
QQnxWindow::setGeometry() that occurs when using a plain OpenGL backed
window in a single threaded environment and the user code calls
QOpenGLContext::makeCurrent(this) in the QWindow subclass ctor.
Change-Id: Iaf07fbff1cb2743f5706020a8d657a82cb9f2255
Reviewed-by: Kevin Ottens <kevin.ottens.qnx@kdab.com>
Reviewed-by: Kevin Krammer <kevin.krammer@kdab.com>
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>