| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
| |
On Windows, filePath() was used for a function that was supposed to get
the bin directory for Qt. However path() should be used instead as this
will remove the filename from the originally obtained string.
Change-Id: I16f8bda91fe9666c935318f8d73c715d2ea908f5
Reviewed-by: Laszlo Papp <lpapp@kde.org>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
(cherry picked from commit 05dab891649b76fe7351d51950edf3caba2bfece)
|
|
|
|
|
|
|
|
| |
EGL property is type EGLint and is defined in eglplatform.h header as
typedef khronos_int32_t EGLint;
Change-Id: I3ace17447aa9a83655b6fa6b4246a6a16592936c
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
|
|
|
|
| |
VxWorks does not support to globally lock and unlock the scheduler from
process. In kernel mode (DKM) above functions disable preemption from
requested task and such functionality does not exist for process.
Change-Id: Id41eab4c1973e4181e82539d08707659e0780f99
(cherry picked from qtbase/49b8e21429c7ab785fb11e8ef84bd7e65c943861)
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
| |
Change-Id: Id71bf7fd8e7371284076247558cba7edb0307e13
(cherry picked from qtbase/00faa09aadfc6c6db9da639ba6b7866c72447e39)
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
VxWorks don't have gettimeofday function so we always use the one
implemented in Qt. VxWorks DIAB compiler is not supported building Qt
for VxWorks 6.9.2. Anyone using older VxWorks should also not be
affected so use new defines VXWORKS_DKM and VXWORKS_RTP for VxWorks
version 6.9.2 onward.
Change-Id: I2e9546a101256ea0557b65163b40cd0f28be8519
(cherry picked from qtbase/7e3d5a720752900b1ccbb3d8dcbf12cc02b01b63)
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
|
|
| |
Check if _WRS_KERNEL is defined so we are compiling for DKM mode.
Change-Id: I15801b0575d3fe6e543f81a177fd01d015d9085f
(cherry picked from qtbase/baad50e97946c4b99998f215aa5c023c17be29ea)
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
|
|
|
|
|
| |
The uniprocessor taskIdCurrent global variable (declared in taskLib.h)
does not exist in VxWorks SMP, because of concurrent execution on
multiple CPUs. Any uniprocessor code that reads taskIdCurrent should
make calls to taskIdSelf() instead.
Change-Id: I4e0efef32297f339d6121c7d4bca3820e0fc9294
(cherry picked from qtbase/608cd1ad84484448ac9b0b8b4d88e0e8ff4a9cf1)
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
|
|
|
|
| |
VxWorks has no concepts of users and groups, therefore no such functions
exists on VxWorks. Use stub functions from qfunctions_vxworks.h for
those, so we don't need to make major source code changes.
Change-Id: Iaad80ec18441e3d3e9c0f96e92ccc3766b27d976
(cherry picked from qtbase/ae229b91929955e7418da938bace6d0cc6a2cde5)
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
|
|
|
| |
Shared libraries cannot be used in kernel mode (DKM), only at process
mode (RTP).
Change-Id: I8cecc12461aa4417b16577db3bc9cd85a1aa7efa
(cherry picked from qtbase/2d44f879a2af258ce5bba75ab3a478e2066f3c10)
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
|
|
|
|
| |
In VxWorks set default values for scheduling priority to use
SCHED_FIFO_HIGH_PRI and SCHED_FIFO_LOW_PRI defines for other scheduling
policies than SCHED_RR or SCHED_FIFO.
Change-Id: If78b84cd9ef94d7712206e9442e96cdba727610f
(cherry picked from qtbase/eae8faabed151fd32219f2f333fb632c104c73a7)
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
|
|
| |
VxWorks compiler fails to compile *.c file with C99 style comments on it
Change-Id: Ib5c5ff14006c17f9392b77363232b8b7cc112d34
(cherry picked from qtbase/2aaffe1800db0d8a170b278ac9a43f2e00ef0e8f)
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
|
|
|
| |
VxWorks has defined variables with same name as in Qt's headers.
Undefine those variables to avoid naming conflict.
Change-Id: Ia8ca04a66acece683cd6c7f71df7e5a2800ec98d
(cherry picked from qtbase/f7bd8652caab2f53ced739ce90c640924d4962dc)
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
|
|
|
| |
Vxworks does not have gettimeofday function, use function
implementation from qfunctions_vxworks.h/cpp instead.
Change-Id: Iad0b811049b974c555febe1e47398b7df9ddda34
(cherry picked from qtbase/1309d9a9d68af902821d0ba5f76aa6e994aa6963)
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
|
|
|
|
| |
VxWorks does not have concept of users and therefore has no function
getpwnam. Use QDir::homePath() which returns actually QDir::rootPath()
if there is no HOME env variable set.
Change-Id: I0786ace2c05a14380dd725384972a19ccc07d916
(cherry picked from qtbase/2d8a4c2d3f86e3ae40b4a388de57021b98b9778d)
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
|
|
|
|
| |
In VxWorks qfunctions_vxworks.h file includes correct time
related headers and it is included already at file
qplatformdefs.h for all mkspecs targets.
Change-Id: Icf970e28bf6deba9201fa33d7f728cc2251f7e5c
(cherry picked from qtbase/2d2da44f6381fdeaa195319c0dc66e3909c2ca7b)
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
|
|
|
| |
When show/hide is called on a widget on Mac then it needs to send an
enter/leave event. But it should only do that if the widget under the
mouse is actually the one that is being shown/hidden.
Change-Id: I53c782f5deee740cbc34e1fcc7b551a108230081
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@digia.com>
|
|
|
|
|
|
| |
Change-Id: Id088effb7b38353c05b4f0baba89ddaa637aa016
Reviewed-by: Casper van Donderen <casper.vandonderen@gmail.com>
Reviewed-by: Jerome Pasion <jerome.pasion@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch fixes a critical bug in the qsqlpsql driver where
notifications aren't delivered when received. Any blocking libpq
function(specifically PQexec) will read all the incoming data
from the socket, including any pending notifications. This would
cause the socket notifier to never be fired for incoming
notifications that are already queued inside libpq. The qsqldriver
test case was skipping the postgres notification test because of
this bug, now its enabled and passing. In order to fix this
bug I made a wrapper function for PQexec in QPSQLDriverPrivate
that calls _q_handleNotification via QMetaObject::callMethod
QueuedConnection in order to deliver pending notifications
when control returns to the event loop. I also added a flag
to ensure only one call is made each time the event loop is
entered.
This backport omits the changes to the NotificationSource version
of the "notification" signal that were included in the original
commit, since NotificationSource itself has not yet been introduced
into Qt4.
(Backport of qtbase aea684506945a12312fc05fb3bb4f549da93f7f5)
Change-Id: I19f5297094ae7ae46bfb0717e4fca744d69f7b92
Reviewed-by: Mark Brand <mabrand@mabrand.nl>
|
|
|
|
|
|
|
|
|
|
|
| |
Fix MinGW compile breakage introduced by
68331c5436506b6f9b4c2a49692a09020e4eef27 .
Task-number: QTBUG-8764
Task-number: QTBUG-10032
Change-Id: I26bcefc382010eee7d79b806b3d9f08b6cb2ff8d
Reviewed-by: Rafael Roquetto <rafael.roquetto@kdab.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
qmake tries to replace all occurrences of the install prefix with a
literal ${prefix}. of course this would mess up build paths which happen
to live under the prefix (this is untypical, but still), which would
break the sed magic in the subsequent install step. as qmake itself has
no use for the .pc files, it's ok to target them directly for the
install dir. we do the same with the include and lib dirs already.
qt5 is already doing the right thing.
Task-number: QTBUG-28807
Change-Id: I82c17284d1dbac4a96845db2ddec9680e251e7a5
Reviewed-by: Davide Pesavento <davidepesa@gmail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
|
|
|
|
|
|
| |
Change-Id: I7119ddfd8fb135adc7561451d5b5a25f3f4f7d4c
Reviewed-by: Franklin Weng <franklin@goodhorse.idv.tw>
Reviewed-by: Liang Qi <liang.qi@digia.com>
|
|
|
|
|
|
|
|
|
| |
QWSSharedMemSurface is not available if we don't have multiprocess QWS,
which requires QSharedMemory support. On VxWorks DKM mode, we don't
support QSharedMemory and there no multiprocess QWS.
Change-Id: I647a576ee4d718b3dcddb0dc4d629ffcd2e163f2
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
emitCursorPositionChanged won't emit cursorPositionChanged if the
cursor position hasn't changed but that doesn't mean the micro focus
hasn't changed, so emit updateMicroFocus changed when
cursorPositionChanged isn't.
Task-number: QTBUG-21017
Task-number: QTBUG-21011
Task-number: QTBUG-20719
Reviewed-by: Joona Petrell <joona.t.petrell@nokia.com>
(cherry picked from commit 6e0e834e0398192a6da11d1e1bca6b74769fb75d)
Change-Id: Ia0d244eea051f2a870a084742c4c22f5a45a87bc
Reviewed-by: Robin Burchell <robin+qt@viroteck.net>
|
|
|
|
|
|
| |
Change-Id: I62fa2443cc7cfd187458eb99d370472d553600f8
Reviewed-by: Sergio Ahumada <sergio.ahumada@digia.com>
Reviewed-by: Johanna Äijälä <johanna.aijala@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Depending on the implementation updating an input method can be
expensive and various widgets will at times call setInputMethodHints
with unchanged hints. QGraphicsView being a notable offender due to
the complexity of the circumstances in which the hints can change.
Skipping the update here ensures the input method isn't updated
unnecessarily for all widgets.
Task-number: QTBUG-19854
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
(cherry picked from commit 40e6b1d0b8c1804dfb33032fa70b6604b91f6f90)
Change-Id: I36ae35585ee20a4e01ca0d62c71e896dbdb51a3f
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
libmng delivers the last animation frame with a 1ms delay, and delivers
an extra blank frame after that with the proper delay time, but this
confuses QMovie when QMovie::cacheMode() == QMovie::CacheAll.
This was commited as 313d60c107d9659efd5e93ba5a036054f6b1ebba in qt5
Task-Number: QTBUG-28894
Change-Id: I1dd585cd6f0ced8c777bd1f15ac2c9f36ca63f41
Reviewed-by: aavit <eirik.aavitsland@digia.com>
|
|
|
|
|
|
|
|
| |
Nothing is used from the above header files so these dead includes can
be removed.
Change-Id: I381ea62e5a0bc623bb984c764be55c7a2f2ee317
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
|
|
|
|
|
|
| |
Change-Id: I0d679638d5a9564f612ae5e70717f164b74aefb0
Reviewed-by: David Faure (KDE) <faure@kde.org>
(cherry picked from qtbase/52e14e05fa83f19f64cedda2291b3a055bdead06)
|
|
|
|
|
|
|
|
| |
Task-number: QTBUG-26394
Change-Id: Idbaa4bad71125b74d8407bb56ee9fdbf83831fe0
Reviewed-by: Jarek Kobus <jaroslaw.kobus@digia.com>
(cherry picked from qttools/a7d61534acdb90b925d7f250665dcd30c3f37668)
|
|
|
|
|
|
|
|
| |
Task-number: QTBUG-29234
Change-Id: I5762c75efc24bf5322ff767a9dbecfdfc02c3377
Reviewed-by: Jarek Kobus <jaroslaw.kobus@digia.com>
(cherry picked from qttools/8695c1753964eeb06b5d9306010b56be5fc151c0)
|
|
|
|
|
|
|
|
|
|
| |
Fix compilation issue when QT_NO_IM is defined.
Also fixed references in tests. But fixing
plugins/examples seem not necessary at this point.
Task-number: QTBUG-26109
Change-Id: I2067051951616012117efa6716640cbf198fdb2a
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The char array 'dash_o' is an automatic variable. The string "-o" is copied into this array.
Later the address of the array dash_o is assigned twice to lpargs[++i].
After leaving the block, the array dash_o is gone and lpargs[...] contains an illegal reference.
This was discovered in a release mode when compiled with gcc version 4.7.2
Patch doesn't apply for Qt5 where 'lpr' support has been removed.
Change-Id: I4f99badfa380ad3b29893a350f0d699bfb934c68
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
ABI compatibility dictates that these members cannot go. On the other hand, new
Clang warns about these being unused.
Konstantin Tokarev suggested adding Q_UNUSED to any method; this one looks as
good as the other one.
Change-Id: Icfb47bff7662d1f31f191a702416983776ce42e3
Reviewed-by: Konstantin Tokarev <annulen@yandex.ru>
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
Reviewed-by: Jason McDonald <macadder1@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Qt 4.8 shows frequent crashes in runMultiple apparently caused
by the QMutex construction in the free functions by different
threads. Use a common QMutex class member instead.
Change-Id: I851d4e2d3637a7b4f404ed843f5360c10caa21f5
Reviewed-by: Mitch Curtis <mitch.curtis@digia.com>
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
(cherry picked from qtbase/ced523af7a3559b0f30dd2a4406e05418ff7ec89)
|
|
|
|
|
|
|
|
| |
- Indonisian locale (in.xml): to return correct AM/PM text and correct time Fomat
- Catalan locale (ca.xml): to return correct time Format
Change-Id: I65bbabdc6311dc3c958605cbc04e7a034243e9db
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
|
|
|
|
|
|
|
| |
Change-Id: Iadb89137ec017b9dcd4d1588fd582ea46a9d7cc1
Reviewed-by: Aurelien <aurelien.lourot@gmail.com>
Reviewed-by: David Faure (KDE) <faure@kde.org>
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
|
|
|
|
|
|
|
|
|
| |
Fixes performance issues in apps which register and deregister objects
very frequently (like nepomukstorage).
Change-Id: Id7d6ef508336758c55fa894868241f143b3e30e0
Backport-Of: ac9ab9703ff299c94dca7585d5a12ecde28931bb
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
|
|
|
|
|
|
|
|
|
| |
Drop the read and write permissions for group and other users in the
system.
Change-Id: I8fc753f09126651af3fb82df3049050f0b14e876
(cherry-picked from Qt 5 commit 856f209fb63ae336bfb389a12d2a75fa886dc1c5)
Reviewed-by: Richard J. Moore <rich@kde.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This definition was missing. It's clear MSVC supports this feature
because it has been in use in qtconcurrentrun.h under an #ifdef
Q_COMPILER_DECLTYPE, which is defined for MSVC.
This is needed in order to re-enable lambdas and other functors in
QtConcurrent after the bug fix for QTBUG-28984.
Task-number: QTBUG-28984
Change-Id: I8c3994096dad120a47ad5f0a7dac6bdbe304fcae
(partly cherry-picked from Qt 5 commit 932034b3df322e7a9dc3d7979584019b46277e50)
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
|
|
|
|
|
|
|
|
|
| |
... by trying more often to get dates close enough to each other.
(backport of commit 170bf2a59aa863e724097825e5a0c8e8148ce109)
Change-Id: Ib713ed13cf5b8bcf469ade148026d5bfa239b304
Reviewed-by: Sergio Ahumada <sergio.ahumada@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This task is valid for Qt 4 only.
Under Qt 5, the "Configuration Option for Qt" documentation page removed
by https://codereview.qt-project.org/40852 commit. Use 'configure -help'
to see all available options.
Task-number: QTBUG-2069
Change-Id: Idd98208338549a0c62d80c35b5eac45b6bfd6db4
Reviewed-by: Jerome Pasion <jerome.pasion@digia.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
QPerformanceTimer was a copy of QElapsedTimer with some additional functionality
which has since been subsumed into QElapsedTimer, so remove the forked code.
(backport of cd6ebb4912a5d96c788ea0c838d8a1a06ff4029e from qtdeclarative)
(backport of d733f6a942e40e58ebc09f1d4414e2f8fa3f6a28 from qtquick1)
Change-Id: I261f0e957ffe3fc5fe42cd50672d2dd99b080f1e
Reviewed-by: Christopher Adams <chris.adams@jollamobile.com>
Reviewed-by: Alan Alpert <aalpert@rim.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Since Qt4, there is a bug which causes Qt to drop dead key modifiers
(like graves and acutes) if the user types enough fast on MS Windows.
This happens because of an extrange behavior of Windows, which drops
dead keys on ToUnicode() calls.
This patch tries to workaround that.
Task-number: QTBUG-8764
Task-number: QTBUG-10032
Conflicts:
src/plugins/platforms/windows/qwindowskeymapper.cpp
Change-Id: Ifdde25817743194fd5c0b7533c27f46a7a108ca4
(cherry picked from commit 5d2bb24cc90194a3458f8741e30ae7afe0b45f5c)
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Avoid using realpath(X,0) on Mac OSX at all, since
even on versions of OSX where realpath(X,0) is supported,
we still get the legacy version due to our compiler flags.
If we were to change the -mmacosx-version-min to 10.5 or
higher then this patch would be safe but unnecessary.
Task-number: QTBUG-28282
(cherry picked from commit 67195dc394001ee1c9adc7db97fd6ff9d88e9b0c)
Change-Id: I0e763321aed71b03e6bf256d112cf4b743411626
Signed-off-by: Peter Hartmann <phartmann@rim.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
|
|
|
|
|
|
|
|
| |
Backport from qtbase d5257644302e349fec8df5750d100fd4918ddd6b
Change-Id: Ib46a8aecf5c4abe83d0161f72c480094002719e3
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Reviewed-by: Rafael Roquetto <rafael.roquetto@kdab.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When two NSToolbars in an application share the same identifier, Cocoa
tries to synchronize their state (item position, view mode, etc.), and
this behvavior interferes with and crashes Qt applications in certain
cases when toggling the unified toolbar on and off. Therefore it is
necessary to use a unique identifier for each native toolbar created.
Does not affect Qt 5. This functionality has been removed and an
equivalent will be provided by QtMacExtras which uses a completely
different approach, though I7d5c8d58e45a504480a1ce67065add15cb35ad8f
could be considered Qt 5's rough equivalent to this.
Task-number: QTBUG-19207
Change-Id: I9030d56941d12a2c1f7a328337bf986f1b51eeb8
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@digia.com>
|
|
|
|
|
|
|
|
|
|
| |
Restrict disconnect() to the clicked() signal, leaving connections
to destroyed() (as used by QStyleSheetStyle) intact.
Task-number: QTBUG-20292
Change-Id: I7471b4d1262ec0684e4446b5c17513717c502749
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
(cherry picked from qtbase/273713b81f5e580748c281c17e08e8b3e2e8ee70)
|
|
|
|
|
|
|
|
|
|
| |
If a tool window is ordered out when there is a full screen widget
around then it would cause a crash because of a new animation from
Lion. This seems to be provoked by the fact that we are faking the
fullscreen functionality in Qt.
Change-Id: I07cdfddc6c2bb766df7809912fb97c4f1ba8873b
Reviewed-by: Morten Johan Sørvig <morten.sorvig@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The internal parser can read values outside this range (and cannot
report an error), but QUrl::port() must not return something outside
that range. The correct solution would be to report an error, like in
Qt 5, but that cannot easily be done.
The rewritten parser in Qt 5 is not affected by this issue.
Task-number: QTBUG-28985
Change-Id: I3cf595384f14272197dcfb85943213c8f8ddeba0
Reviewed-by: David Faure (KDE) <faure@kde.org>
|
|
|
|
|
|
|
|
|
|
| |
Task-number: QTBUG-20403
Change-Id: I2a5fe00dd16e9dc1ec0d742a8f48083fc2954996
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Jarek Kobus <jaroslaw.kobus@digia.com>
(cherry picked from qtbase/b6963f109ad54c6f03b3391d57db9a536d49ba7d)
|