Friedemann Kleint [Fri, 27 Apr 2012 09:06:41 +0000 (11:06 +0200)]
Windows: Fix warning about retrieving the handle from non-toplevels.
Change-Id: I967f218c98309e77371717aa16e46995a767caa3
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Friedemann Kleint [Fri, 27 Apr 2012 12:01:26 +0000 (14:01 +0200)]
QWindowsBackingStore: Flush to correct window.
Task-number: QTBUG-25435
Change-Id: I13f57a328a7ac3402fe0def9d724f7e700016c80
Reviewed-by: Jonathan Liu <net147@gmail.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Morten Johan Sorvig [Thu, 26 Apr 2012 08:29:16 +0000 (10:29 +0200)]
Remove deprecation tags from QWheelEvent.
0df1c9f1 enabled deprecation warnings from Qt,
causing warnings from QWheelEvent use.
Using delta() and orientation() in new code is still
not advised, but since QtWidgets uses them the
"deprecated" tag is not suitable.
Change-Id: I597e8ddf9fd2dc911d0e30044e96df744982ff81
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Friedemann Kleint [Thu, 26 Apr 2012 13:38:17 +0000 (15:38 +0200)]
Windows/Vista style: Fix painting of QGroupBox.
QGroupBox in Vista style does not show a frame, currently.
drawBackgroundDirectly should also be called if the paintDevice
is a backing store.
Task-number: QTBUG-25395
Change-Id: Ifa20802dea6c182904f68271ddd0d175007e4b0e
Reviewed-by: Miikka Heikkinen <miikka.heikkinen@digia.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Friedemann Kleint [Tue, 24 Apr 2012 14:48:40 +0000 (16:48 +0200)]
Fix warning about int-truncation (MSVC2010, 64bit).
Change-Id: Idfc483ccb80a8bbad128672052edfa04635bd532
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@nokia.com>
Rohan McGovern [Fri, 27 Apr 2012 02:45:22 +0000 (12:45 +1000)]
doc: removed wrong license headers from several .pro snippets
qmake can't parse /* C-style comments */. The license headers in these
files should use # as the comment delimiter. However, .pro files anyway
don't need license headers, so just remove them.
Continues from commit
8dea79701579f78023c6b40c4b5cbea278bc62c1, which
removed the header from one file.
Change-Id: Ib227aabb27224fc1fe89b7ced4fbbf7b1af8b193
Reviewed-by: Casper van Donderen <casper.vandonderen@nokia.com>
Kent Hansen [Sun, 22 Apr 2012 20:44:58 +0000 (22:44 +0200)]
Add QMetaMethod::fromSignal() function
Given a member function that's a signal, returns the corresponding
QMetaMethod. Inspired by the implementation of the template-based
QObject::connect().
The primary use case for this function is to have an effective and
exact (not subject to shadowing) way of checking whether a known
signal was connected to in reimplementations of
QObject::connectNotify(QMetaMethod), avoiding string comparisons.
Example:
void MyObject::connectNotify(const QMetaMethod &signal)
{
if (signal == QMetaMethod::fromSignal(&MyObject::mySignal)) {
// Someone connected to mySignal ...
}
}
Change-Id: I5e4de434275fe543c004d569dcaa9ceda3442f03
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
aavit [Wed, 18 Apr 2012 14:15:41 +0000 (16:15 +0200)]
Fix workaround for image readers that can scale but not clip
If both scaling and clipping was requested from an image reader that
only supports scaling, the code failed to apply the intended
workaround (i.e. read normal and do clipping and scaling afterwards).
Ref. the comment just above about "Only enable the ScaledSize option if
..."
Change-Id: I273edba60bf9f785f2082aa8f236505ec3ce0776
Reviewed-by: Kim M. Kalland <kim.kalland@nokia.com>
aavit [Wed, 18 Apr 2012 13:02:21 +0000 (15:02 +0200)]
Implement inline downscaling in png reader, to save memory
Instead of allocating and reading the entire image and then scaling it
down, this code reads only one line at a time and scales it on the
fly.
Change-Id: I61fde307146c11dcd90ca617cc2e7f85dd3b66c4
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Rohan McGovern [Fri, 27 Apr 2012 05:09:08 +0000 (15:09 +1000)]
Removed parallel_test from parallel-unsafe autotest.
This autotest passes when run alone, but sometimes fails when run
concurrently with other tests.
Change-Id: I327de7f4a9f8af385351e4c6b09a57311efd6eb6
Reviewed-by: Toby Tomkins <toby.tomkins@nokia.com>
Reviewed-by: Kalle Lehtonen <kalle.ju.lehtonen@nokia.com>
Christoph Schleifenbaum [Tue, 10 Apr 2012 15:17:41 +0000 (17:17 +0200)]
Respect the DontUseNativeDialog flag.
QFontDialog and QColorDialog were ignoring the DontUseNativeDialog. This
lead to a native (Cocoa) dialog created all the time.
Fix the testcase for QFontDialog. It needs the DontUseNativeDialog flag
set.
Task-number: QTBUG-24321
Change-Id: I159c1ad057bac38226f1e01a56b15f142650bfd8
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Rohan McGovern [Fri, 27 Apr 2012 01:06:37 +0000 (11:06 +1000)]
cmake: temporarily set tests/auto/cmake insignificant on mac
cmake was installed on CI macs _after_ this test was added, causing the
test to block unrelated changes. Temporarily disable it as previously
agreed with Stephen Kelly.
Change-Id: I079c0016a5e4a9a03ee1a0bae9a2e836c6b985d6
Reviewed-by: Toby Tomkins <toby.tomkins@nokia.com>
Thiago Macieira [Fri, 20 Apr 2012 11:24:05 +0000 (13:24 +0200)]
Mark that the Sun compiler includes the return type in the mangling
Source: http://developers.sun.com/solaris/articles/external_linkage.html
%dem __1cFgreet6F_pc_
__1cFgreet6F_pc_ == char*greet()
Source: http://www.oracle.com/technetwork/articles/servers-storage-dev/stablecplusplusabi-333927.html
Function Mangled Name
float f(float) __1cBf6Ff_f_
int f(int) __1cBf6Fi_i_
int T::f(int) __1cBTBf6Mi_i_
int T::f(char*) __1cBTBf6Mpc_i_
int T::U::f(int) __1cBTBUBf6Mi_i_
int N::T::f(int) __1cBNBTBf6Mi_i_
Change-Id: Ibe1934f1c137d1657fb122cbf1a1b1fbc3e4a202
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Casper van Donderen [Thu, 26 Apr 2012 08:37:53 +0000 (10:37 +0200)]
QDoc: Pretend that .qtx and .qtt files are .qdoc files.
Change-Id: I2b530e526d2c1f5102960dab4991ceade414f1eb
Reviewed-by: Ed Baak <ed.baak@nokia.com>
Reviewed-by: Peter Yard <peter.yard@nokia.com>
Konstantin Ritt [Thu, 26 Apr 2012 17:00:59 +0000 (20:00 +0300)]
fix build with mingw
since MinGW is GCC, invert the order and place the GCC constructs first
Change-Id: I98113aa77e51f9e01c3641987e915bf475053a60
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Ian Dean [Mon, 23 Apr 2012 15:10:57 +0000 (16:10 +0100)]
Fixed iOS compilation issues.
Put MacOS-specific code in #ifdef blocks so that it is not compiled for
iOS.
Change-Id: I303b7dbb63eb298e92351951731868df6a38bffa
Reviewed-by: Shane Kearns <shane.kearns@accenture.com>
Eskil Abrahamsen Blomfeldt [Thu, 26 Apr 2012 14:47:56 +0000 (16:47 +0200)]
Revert "Harfbuzz-shaper - fix incorrect logClusters being set in HB_OpenTypePosition"
This reverts commit
add629d4f16a536fc56d55727195e3247b621a54.
The commit caused crashes when laying out some text (e.g. ordinary
Arabic text) with certain fonts (it triggered the assert in
QTextLayout::addNextCluster()). The regressions will have to be
weeded out before the fix can be recommitted.
Change-Id: Iee457f138367d4cf8fcbd2e518271d8eaa95a62c
Reviewed-by: Jiang Jiang <jiang.jiang@nokia.com>
Shane Kearns [Wed, 25 Apr 2012 13:43:51 +0000 (14:43 +0100)]
Uncomment qnetworkproxyfactory autotest
It no longer requires hardcoded configuration, and it passes on
linux, mac 10.6 and windows desktops
Change-Id: Ibaa63520dade58ce13c23cf2aba3ddbc5be1c472
Reviewed-by: Jason McDonald <jason.mcdonald@nokia.com>
Reviewed-by: Sergio Ahumada <sergio.ahumada@nokia.com>
Honglei Zhang [Thu, 26 Apr 2012 10:17:40 +0000 (13:17 +0300)]
Fix memory leaks in 64-bits ODBC driver
On 64-bits Windows system, integers and longs are still 4 bytes values.
Several functions that were previously defined with SQLINTEGER and
SQLUINTEGER parameters have been changed where appropiate to use new
SQLLEN and SQLULEN typedefs. SQLGetStmtAttr() is one of these functions.
This fix replaces SQLINTEGER with SQLULEN in appropriate functions to avoid
memory leaks.
Task-number: QTBUG-25256
Change-Id: I744927f42b8578ece60815df360e3b337ebf452a
Reviewed-by: Andy Shaw <andy.shaw@digia.com>
Reviewed-by: Mark Brand <mabrand@mabrand.nl>
Konstantin Ritt [Thu, 26 Apr 2012 10:00:56 +0000 (13:00 +0300)]
get rid of Q_*_EXPORT_INLINE macros
> Girish:
> We should be able to remove the macro completely today,
> just mark all those functions as plain inline.
> With Qt5, we don't have to worry about bc yet.
this fixes "import attribute ignored" warnings on mingw with -fno-keep-inline-dllexport
Change-Id: I616e5de7c8d59953ce03a316b941a439fae56298
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Olivier Goffart [Thu, 26 Apr 2012 09:28:35 +0000 (11:28 +0200)]
Fix C++11 compilation
Fix compilation with compilers that support user defined literal
(such as GCC 4.7)
Change-Id: I31cd3d2177688f963ab33cf68cd6060a5fb8640f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Kevin Krammer [Wed, 11 Apr 2012 11:25:28 +0000 (13:25 +0200)]
Create interface for navigator calls and implement for BPS and PPS
Makes QQnxServices available on non-BPS systems by delegating to an interface
which is implemented on systems with BPS using the currenly used navigator API
and on systems without BPS by sending an appropriate message to the navigator's
PPS service interface.
Change-Id: I0e32fb11e6debb7b7b4693c0bc02af4f75ee2162
Reviewed-by: Rafael Roquetto <rafael.roquetto@kdab.com>
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
Stephen Kelly [Tue, 17 Apr 2012 17:45:20 +0000 (19:45 +0200)]
Take account of spanned items in QTreeView when dragging.
Also remove some code which has been unused since it was introduced in
32182d107fa75e5619ecc91a81f50626f429ebe1
Task-number: QTBUG-25140
Change-Id: Ic7053d68d8200f845c1ae330342d27af7275e057
Reviewed-by: Tarja Sundqvist <tarja.sundqvist@digia.com>
Reviewed-by: David Faure <faure@kde.org>
Casper van Donderen [Thu, 26 Apr 2012 12:56:13 +0000 (14:56 +0200)]
Doc: Use the proper way to find qdoc.
Previously $$QT.core.bins was used to find qdoc, the proper way is to
use qtPrepareTool().
Change-Id: I5d97f5517ae63253ccaf1fb1487034c3a312a074
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Casper van Donderen [Thu, 26 Apr 2012 12:08:36 +0000 (14:08 +0200)]
Doc: Add "make docs" command for QtXml.
Also fix a typo in the qhp/index URL.
Change-Id: I5774982b15f9de5b59966a2d932307b156eecc7f
Reviewed-by: Jerome Pasion <jerome.pasion@nokia.com>
Casper van Donderen [Thu, 26 Apr 2012 05:58:25 +0000 (07:58 +0200)]
QDoc: Disambiguate prev and next links in html source.
Change-Id: Iaa82b9b6de71b577a44eae86eb6847c07d068670
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@nokia.com>
Jan-Arve Saether [Wed, 25 Apr 2012 08:46:30 +0000 (10:46 +0200)]
Map QAccessible::Role to the proper IA2Role.
All enum values after the MSAA-specific roles needs a special
mapping.
Change-Id: I6c77abf09b68d2d3d772937be986d532db2e5292
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@nokia.com>
Sean Harmer [Thu, 26 Apr 2012 08:49:27 +0000 (09:49 +0100)]
Use non-blocking reads on virtual keyboard pps device.
Change-Id: Iad350c948049128f3014a6ad6536ac9dc2ee86d5
Reviewed-by: Kevin Krammer <kevin.krammer@kdab.com>
Reviewed-by: Giuseppe D'Angelo <dangelog@gmail.com>
Sean Harmer [Thu, 26 Apr 2012 08:51:02 +0000 (09:51 +0100)]
Fix compilation when debug output enabled on virtual keyboard
Change-Id: I467e9ef624bcd10cb5b61f6f665a7bbb0bb1f57a
Reviewed-by: Kevin Krammer <kevin.krammer@kdab.com>
Reviewed-by: Thomas McGuire <thomas.mcguire@kdab.com>
Reviewed-by: Giuseppe D'Angelo <dangelog@gmail.com>
Jason McDonald [Mon, 16 Apr 2012 14:12:48 +0000 (00:12 +1000)]
Make QTimeLine test significant.
Known failures in this test are now handled by QEXPECT_FAIL.
Task-number: QTBUG-24796
Change-Id: I12ba57370cf3df1a85a108fbbcdc9db2222491c1
Reviewed-by: Rohan McGovern <rohan.mcgovern@nokia.com>
Thiago Macieira [Mon, 23 Apr 2012 15:16:31 +0000 (17:16 +0200)]
Ensure that QUrl::{to,from}LocalPath encode/decode properly
Unlike path(), toLocalFile() isn't reporting a URL component, so it
should decode the percent-encoded characters fully. This extra
decoding pass is meant to catch %00 to %1F, %7F and %25 (the percent
sign itself).
It also catches %80 to %FF, which aren't decoded because they don't
form UTF-8 sequences. That means QUrl::toLocalFile() has undefined
behaviour if the path contained non-UTF8 sequences.
Task-number: QTBUG-25459
Change-Id: Iab5a0ba6afcfc4510e297984f2ffc208cedd752b
Reviewed-by: Shane Kearns <shane.kearns@accenture.com>
Qt4iOS [Fri, 16 Mar 2012 15:00:33 +0000 (15:00 +0000)]
Added mkspecs for iOS.
Modified common MacOS gcc config for iOS compatibility. Added iOS
mkspecs for iOS devices and simulator using G++, LLVM-G++ or CLANG
compilers.
Change-Id: If0a0b1a3ce4e0dc4db319a315e227a9d31549399
Reviewed-by: Morten Johan Sørvig <morten.sorvig@nokia.com>
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Olivier Goffart [Fri, 20 Apr 2012 13:08:11 +0000 (15:08 +0200)]
Remove deprecated conversion from QKeySequence to QString
Change-Id: I3844913c16b9d6222f48e66dddc1d680458ffa56
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@nokia.com>
Miikka Heikkinen [Tue, 24 Apr 2012 10:06:33 +0000 (13:06 +0300)]
Fix QRawFont autotest in Windows
Can't use style name with Windows native font engine, so do the
comparison using style in case multipleRawFontsFromData().
Also XFAILed the advances() case when using Freetype font engine in
Windows, as some of the advances returned are one pixel wider than
expected.
Task-number: QTBUG-24197
Change-Id: Ib5fecd83a93908e57a4c82ffb5495a92474ce45a
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Olivier Goffart [Fri, 13 Jan 2012 19:09:19 +0000 (20:09 +0100)]
Remove QMetaObjectExtraData and put everything into QMetaObject
QMetaObjectExtraData was added when support for QMetaObject::newInstance
was added. One needed a place to put the pointer to static_metacall in
the QMetaObject.
But as we break binary compatibility, one can change the size of
QMetaObject, and put everything back inside QMetaObject's own structure.
Meaning it is not required anymore to have one QMetaObjectExtraData
instance per QMetaObject anymore.
Change-Id: If0b8f586cbaf633eed10045adee3ba3366826c86
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Reviewed-by: Kent Hansen <kent.hansen@nokia.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@nokia.com>
Kent Hansen [Fri, 20 Apr 2012 12:45:21 +0000 (14:45 +0200)]
Split qobject_impl.h into qobjectdefs_impl.h
This is done to make QtPrivate::FunctionPointer available to the
QMetaMethod declaration in qmetaobject.h (which already included
qobjectdefs.h, since that's where QMetaObject is declared), so that
the new template-based QMetaMethod::fromSignal() function may be
implemented.
The logic for statically generating the array of qMetaTypeId (used by
the template-based QObject::connect()) remains in qobject_impl.h,
since it's not needed for QMetaMethod::fromSignal(). Moreover, moving
that code would introduce a circular dependency, since qmetatype.h
as of commit
194674044693d6b101c3dc2f4784718540d343a4 now includes
qobjectdefs.h.
Change-Id: I36c35041e0c6661c3cf523684177a0b6c19e2d35
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Kent Hansen [Sun, 22 Apr 2012 18:47:08 +0000 (20:47 +0200)]
Add comparison operators == and != for QMetaMethod
This is done in preparation of introducing the
QObject::connectNotify(QMetaMethod) function. Together with the
forthcoming QMetaMethod::fromSignal() function, which returns the
QMetaMethod corresponding to a Qt/C++ signal (member function), the
comparison operators provide an effective way of checking which
signal was connected to.
Change-Id: I2de48628c4884a7174fb8574895f272cb3fe5634
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Thiago Macieira [Fri, 20 Apr 2012 12:59:30 +0000 (14:59 +0200)]
Don't use the QRegExp methods that modify the object [QtWidgets]
QRegExp matching methods modify the object, which we don't want to. In
particular, when we receive a QRegExp from the user or we store in a
context that might require thread-safety, make sure we make a copy
before using it.
QRegularExpression has no such shortcoming.
Task-number: QTBUG-25064
Change-Id: I44cc44b75b3da10c5ece97cb6315c2c425327dc0
Reviewed-by: Alexis Menard <alexis.menard@openbossa.org>
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Holger Hans Peter Freyther [Wed, 25 Apr 2012 13:59:49 +0000 (15:59 +0200)]
lance: Document all available command line options.
Change-Id: I0e0d95cb5988574c1ecfbe7423ac7388022101fb
Reviewed-by: Girish Ramakrishnan <girish.1.ramakrishnan@nokia.com>
Holger Hans Peter Freyther [Wed, 25 Apr 2012 11:51:50 +0000 (13:51 +0200)]
baselineserver: Fix compilation by removing PI_BuildKey from the source
The buildkey support was removed in
7493ee1c44a16dc3. This commit removed
PI_BuildKey and most of the usages from the code.
Change-Id: I971d68cd4377c358381983c1edcb98360936b698
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Reviewed-by: Girish Ramakrishnan <girish.1.ramakrishnan@nokia.com>
Romain Pokrzywka [Wed, 25 Apr 2012 00:33:12 +0000 (17:33 -0700)]
Add makespecs for standard devices in the OMAP3 family
The beagleboard and Archos gen8 tablets are built on the same hardware
and OpenEmbedded platform build system.
The makespecs enable OpenGL/ES2 support and EGLFS as the default
platform.
Reviewed-by: Girish Ramakrishnan <girish.1.ramakrishnan@nokia.com>
Reviewed-by: Donald Carr <donald.carr@nokia.com>
Change-Id: I04938676f82f9136efd662ff3ede16221eadf275
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Reviewed-by: Girish Ramakrishnan <girish.1.ramakrishnan@nokia.com>
Friedemann Kleint [Wed, 25 Apr 2012 14:51:04 +0000 (16:51 +0200)]
Fix recursion in qwidget gdiPainting test.
Change-Id: If4881dfecc6fc7cebcd3ed896846d34e35b3e3ae
Reviewed-by: Jason McDonald <jason.mcdonald@nokia.com>
Thiago Macieira [Sun, 22 Apr 2012 14:05:31 +0000 (16:05 +0200)]
Add detection code for AVX2, HLE, RTM and AES to qsimd.cpp
AES is currently not enabled, since we don't use it for anything. The
code is here with the proper detection should we want to in the
future.
RTM and HLE (Transactional Memory Extensions) I'll use soon in the
locking code.
Also rename a few variables to make it easier to read later on.
Change-Id: I800c66d7e1ba86ec037692928d94f53ea5d5868f
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
Thiago Macieira [Tue, 24 Apr 2012 14:27:35 +0000 (16:27 +0200)]
Replace the x86 XGETBV instruction with its opcode bytes
This is a new instruction, present on the SandyBridge architecture and
later. Some older assemblers do not support it and produce:
{standard input}:225:no such instruction: `xgetbv'
The use of this instruction is protected by a CPUID check (function 1,
ECX bit 27), so we only run it in processors that do support it.
Change-Id: Ife7500c0deaab9539074835a4511e8c19602608e
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Thiago Macieira [Sun, 25 Dec 2011 22:36:14 +0000 (20:36 -0200)]
Update the x86 feature-detection code, unifying x86 and x86-64
Most of it is the same for both of them, so let's avoid doing
everything twice. Or more, since we may support x32 soon. For Windows,
use the intrinsics. For GCC, we'd like to use cpuid.h, but it only
exists since GCC 4.3, so we can't.
And properly detect AVX support: it's not enough to detect that the
processor supports them, we also need to check that the OS enabled
support for the 256-bit registers.
Change-Id: Ibb4872cdb774de5701b18c40f4e612330a266214
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
Thiago Macieira [Tue, 24 Apr 2012 16:06:44 +0000 (18:06 +0200)]
Fix atomics on SPARC
Inspired by
https://bugreports.qt-project.org/secure/attachment/26020/qt_atomic_sparc64.patch
Task-number: QTBUG-22479
Change-Id: Ie3275df96c639d6a75e05f70fe5745aeb34457f9
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Bradley T. Hughes [Wed, 21 Mar 2012 13:01:18 +0000 (14:01 +0100)]
Cocoa: support modal windows
Qt::WindowModal windows and dialogs are shown using [NSApp
beginSheet:modalForWindow:modalDelegate:didEndSelector:contextInfo:] as
long as they have a valid parent. Otherwise they are behave as
application modal.
Use the existing modal session support in the QCocoaEventDispatcher
(which was inherited from Qt 4) to support Qt::ApplicationModal
windows and dialogs. Some changes to this code are needed to ensure
proper behavior:
1. Window level modification is now done in
QCocoaWindow::recreateWindow() instead of in QCocoaEventDispatcher.
2. Make interrupt() use [NSApp abortModal] to stop a modal session
(previously we were freeing memory from under Cocoa's feet, causing
tools like valgrind and Instruments.app to complain)
3. Do not remove an item from a list and use a const reference to the
removed item immediately after (minor bug fix).
Also make sure that QCocoaEventDispatcher cleans up any modal sessions
and retained user input events on destruction (otherwise we leave
NSApplication in a weird state, which causes some autotest failures).
Change-Id: Iaeefa025400f324b5348b8c81a40384ef026efb4
Reviewed-by: Morten Johan Sørvig <morten.sorvig@nokia.com>
Bradley T. Hughes [Wed, 25 Apr 2012 09:55:43 +0000 (11:55 +0200)]
Skip tst_QInputDialog on Mac OS X
This test hangs in CI, but so far I've been unable to reproduce the
problem. Disable the test for now so that this test does not block
changes that bring in needed functionality.
Task-number: QTBUG-25496
Change-Id: I81faa574b6d7bcab2e32becc2af0f71006c7dd9c
Reviewed-by: Morten Johan Sørvig <morten.sorvig@nokia.com>
Reviewed-by: Jason McDonald <jason.mcdonald@nokia.com>
Thiago Macieira [Wed, 25 Apr 2012 10:48:02 +0000 (12:48 +0200)]
Revert QUrl::isRelative to its Qt 4 behaviour
Instead of trying to return whether the URL is relative to something
undefined, let's instead follow what the documentation was saying all
along and what the RFC says about "Relative References".
Change-Id: I32722321a6b36c6e3480669ad769390e4c6f7d1c
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Thiago Macieira [Wed, 19 Oct 2011 21:51:17 +0000 (23:51 +0200)]
Warn if we use deprecated methods from Qt modules
This change will cause any use of methods and classes declared with
Q_DECL_DEPRECATED to produce a warning in the compilation
output. Those warnings need to be silenced before the Qt 5.0 release,
preferably before the beta.
Change-Id: I28ce133867cc5918e177c57f6e5f7ab9421989ad
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Thiago Macieira [Tue, 24 Apr 2012 17:15:49 +0000 (19:15 +0200)]
Make configure.exe not overwrite its own Makefile in -fast mode
configure.exe's Makefile is created by configure.bat. While severely
lacking, it is what gets executed when the user configures, so
developers modifying configure.exe should use that too.
Change-Id: Ib216889594d08c980bfae6d2dbbb27ee9901375a
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Thiago Macieira [Tue, 24 Apr 2012 17:14:10 +0000 (19:14 +0200)]
Make configure.exe create files with Windows line-endings
Except for the perl sources, just in case.
This helps if you need to open the output files in, say, notepad.exe.
Change-Id: Ibf46b74af2efd6936586650e1405750bfef17952
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Jan-Arve Saether [Wed, 25 Apr 2012 08:17:08 +0000 (10:17 +0200)]
MSAA bridge should not return any IA2-specific roles from get_accRole()
Small cleanup/improvement. This patch just does what the removed ###
says.
Change-Id: I0d16541d2e4b8f948f32734ef0138fde5517932a
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@nokia.com>
Jan-Arve Saether [Wed, 25 Apr 2012 07:13:07 +0000 (09:13 +0200)]
Do not map LayeredPane and Terminal to the wrong MSAA role
Microsoft did at some point extend the roles in MSAA with two extra
roles (ROLE_SYSTEM_IPADDRESS and ROLE_SYSTEM_OUTLINEBUTTON).
These are defined in oleacc.h as:
#define ROLE_SYSTEM_IPADDRESS ( 0x3f )
#define ROLE_SYSTEM_OUTLINEBUTTON ( 0x40 )
This means that LayeredPane will map to ROLE_SYSTEM_IPADDRESS
and Terminal will map to ROLE_SYSTEM_OUTLINEBUTTON, which is
obviously wrong.
We now reserve some valuespace for more roles to reduce the likelyness of
any collisions in the future. Having the reserved area also serves
as a way of indicating the boundary between the "MSAA" enums and other
enums.
Change-Id: Ic67a1a7200382fed3040e69b3e8856376ba642ac
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@nokia.com>
Konstantin Ritt [Tue, 24 Apr 2012 15:38:57 +0000 (18:38 +0300)]
get rid of QT_NO_IMAGE_TEXT switcher
Change-Id: Ie72f907ffb959f629af6a414959348a992c4c941
Reviewed-by: aavit <qt_aavit@ovi.com>
Casper van Donderen [Mon, 23 Apr 2012 14:47:43 +0000 (16:47 +0200)]
QDoc: Implement -installdir CLI option for module cross-linking.
QDoc needs to know the final location of the installed documentation to
generate correct relative links between the modules. Normally you can
use QLibraryInfo::DocumentationPath for this, but since QDoc gets compiled
during Qt bootstrapping QLibraryInfo is not available yet.
The -installdir option still allows us to specify
QLibraryInfo::DocumentationPath on the command line.
Change-Id: Ic4729f4daad112f0d175931467cf09cfcf5145a3
Reviewed-by: Martin Smith <martin.smith@nokia.com>
Jason McDonald [Mon, 23 Apr 2012 09:32:43 +0000 (19:32 +1000)]
Mark known test failures in Windows developer build.
Previously, the CI system has not been testing on Windows with the
-developer-build configure flag. Mark known failures for this
configuration so that tests can be run in enforcing mode.
Change-Id: I5fbbbe09a7b400d626107c66dcbd5c5469a45b20
Reviewed-by: Sergio Ahumada <sergio.ahumada@nokia.com>
Jędrzej Nowacki [Tue, 24 Apr 2012 08:07:55 +0000 (10:07 +0200)]
Use macro for QByteArrayData initialization.
Change-Id: Ib3f906dc5f313c7f9669efda26a93a76af7d7f80
Reviewed-by: João Abecasis <joao.abecasis@nokia.com>
Konstantin Ritt [Wed, 11 Apr 2012 08:06:21 +0000 (11:06 +0300)]
support UTF-8 encoded file names in QZip*
Change-Id: Iba0457b55ef9fa3673f78b4d30d018a146e8fb1c
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Reviewed-by: Harald Fernengel <harald.fernengel@nokia.com>
Laszlo Agocs [Tue, 24 Apr 2012 13:31:18 +0000 (16:31 +0300)]
Update QTest::QTouchEventSequence docs
Change-Id: I2a8f88929f985b1543ec7c223266e8387f0a8a48
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@nokia.com>
Peter Hartmann [Sun, 22 Apr 2012 19:20:27 +0000 (21:20 +0200)]
QHttpMultiPart: supply new line at the end of the body
... to conform to RFC 2046 (section 5.1.1). Apparently IIS had problems
without the new line.
Task-number: QTBUG-25429
Change-Id: Ia619bbdcebd407b2716bc467323634e4c8d77bcd
Reviewed-by: Shane Kearns <shane.kearns@accenture.com>
Konstantin Ritt [Mon, 23 Apr 2012 04:00:28 +0000 (07:00 +0300)]
optimize QChar::decomposition()
* by not using QString::fromUtf16() as we know for sure that
the data is 'raw' UCS-2;
* it's safe to avoid a check for > UNICODE_LAST_CODEPOINT
as GET_DECOMPOSITION_INDEX macro already does a similar check
Change-Id: Ifb660efc51c664d06733ac8ed46d54278520da06
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Debao Zhang [Mon, 16 Apr 2012 20:16:11 +0000 (13:16 -0700)]
Using QLatin1String instead of QLatin1Literal
QLatin1Literal is just a typedef of QLatin1String.
Change-Id: If20ca225e57a7fb45a7775f0fc81aedb6da88c96
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Stephen Röttger [Tue, 24 Apr 2012 03:29:07 +0000 (05:29 +0200)]
Fix locale issue on unix systems
QSystemLocale::query() was missing the LanguageId QueryType. Therefore
QSystemLocale::fallbackLocale() was always used as default language,
which reads environment variables in the order: LC_ALL -> LC_NUMERIC ->
LANG. The correct behaviour is to read LC_ALL -> LC_MESSAGES -> LANG.
This leads to problems for users that want to use english language, but
non-english localization features (date, number formats etc.)
Change-Id: I4310537dac8622a3dd79231fbad58e22f20ca262
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Reviewed-by: John Layt <jlayt@kde.org>
Konstantin Ritt [Mon, 23 Apr 2012 03:00:16 +0000 (06:00 +0300)]
qunicodetables generator: improve the output and the generated code
better memory usage report;
an additional asserts with conditions the implementation is depends on;
a namespace for the internal static data;
styling fixes
Change-Id: Id4048ff6104c56b5f590f9ac6fbf7c0bce79ec47
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Konstantin Ritt [Mon, 23 Apr 2012 15:11:57 +0000 (18:11 +0300)]
workaround issue where casing diff overflows signed short
there are two such codepoints were added in the Unicode 5.1:
U+1D79 LATIN SMALL LETTER INSULAR G
U+A77D LATIN CAPITAL LETTER INSULAR G
two more of them were added in the Unicode 6.0:
U+0265 LATIN SMALL LETTER TURNED H
U+A78D LATIN CAPITAL LETTER TURNED H
and two more were added in the Unicode 6.1:
U+0266 LATIN SMALL LETTER H WITH HOOK
U+A7AA LATIN CAPITAL LETTER H WITH HOOK
we map them like special cases with length == 1
(note: all are in BMP which is checked explicitly in the generator)
Change-Id: I8a34164eb3ee2e575b7799cc12d4b96ad5bcd9c6
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Konstantin Ritt [Mon, 9 Apr 2012 14:56:44 +0000 (17:56 +0300)]
QChar: reduce code duplication by inlining some methods
using templates for toLower()/toUpper()/toTitleCase()/toCaseFolded()
makes the upcoming patches smaller and thus easier to review
Change-Id: Ideb23e8669dbc2fe9ea3f129bf0137e1805ece11
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Jan-Arve Saether [Tue, 24 Apr 2012 11:20:07 +0000 (13:20 +0200)]
Make NVDA read aloud stuff from a Qt app again.
Previously, NVDA did not read aloud anything from a Qt app.
Change-Id: Ie738e7e7d7acc54f45fab9195adc09bfde930174
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@nokia.com>
Jan-Arve Saether [Tue, 24 Apr 2012 07:52:52 +0000 (09:52 +0200)]
Fix a regression where events were not delivered on windows
Regression caused by
a17907829e6b180f2bb4af9a8594996b2a0e531a
Change-Id: I8a2ca35cf176b4db47f29d848cbc2cd8180596d0
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@nokia.com>
Jan-Arve Saether [Tue, 24 Apr 2012 07:45:43 +0000 (09:45 +0200)]
Add QDebug operator<< for QAccessibleEvent
(Keep it internal)
Change-Id: I9e99ccbd16cc595d2aff97a26181e8d8c3d9d7ae
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@nokia.com>
Yan Shapochnik [Fri, 20 Apr 2012 12:41:18 +0000 (08:41 -0400)]
Provide access to RTLD_NODELETE flag on Unix.
Introduce a new QLibrary::PreventUnloadHint to support the
RTLD_NODELETE flag support by dlcompat on Unix platforms.
Change-Id: Ib1327e968a2a888850ad1086a102a143f86c5090
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Giuseppe D'Angelo <dangelog@gmail.com>
Jeff Kehres [Fri, 23 Mar 2012 14:43:44 +0000 (10:43 -0400)]
Integrate Blackberry Platform Services (BPS) with Qt event loop.
This ensures interoperability between the Blackbery C and C++ APIs
and makes it easier to expose platform services in C++ that are
exposed in BPS - since events from both APIs can be processed on
the same thread.
Change-Id: I7270adc64c26396f66d9126141500d5e58be51e7
Reviewed-by: Kevin Krammer <kevin.krammer@kdab.com>
Reviewed-by: Giuseppe D'Angelo <dangelog@gmail.com>
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Thiago Macieira [Mon, 23 Apr 2012 15:06:08 +0000 (17:06 +0200)]
Ensure that QUrl::toLocalFile returns decoded paths
QUrl::path() already decodes almost everything, but let's pass the
formatting flag to be sure.
Note: decoding of control characters from U+0001 to U+001F is not
implemented. Non-UTF8 sequences are also not representable.
Change-Id: I9a0ae2282ec3d48cc0e70e5b2d3824fb120709ed
Reviewed-by: Shane Kearns <shane.kearns@accenture.com>
Thiago Macieira [Fri, 20 Apr 2012 12:59:30 +0000 (14:59 +0200)]
Don't use the QRegExp methods that modify the object [QtGui]
QRegExp matching methods modify the object, which we don't want to. In
particular, when we receive a QRegExp from the user or we store in a
context that might require thread-safety, make sure we make a copy
before using it.
QRegularExpression has no such shortcoming.
Task-number: QTBUG-25064
Change-Id: If119e06221ca99e57c5ad1a1d4cc6468e9f68c7b
Reviewed-by: Gunnar Sletta <gunnar.sletta@nokia.com>
Thiago Macieira [Mon, 23 Apr 2012 15:02:17 +0000 (17:02 +0200)]
Don't strip the leading slash of letter+colon paths on Unix
It's perfectly valid to have a path of /c:/a.txt on Unix, so don't
strip the leading slash unless we're on Windows.
Task-number: QTBUG-20322
Change-Id: I721bd0a65b41048bc735d4eaa0d536174164fe64
Reviewed-by: Shane Kearns <shane.kearns@accenture.com>
Jason McDonald [Tue, 24 Apr 2012 11:10:54 +0000 (21:10 +1000)]
Remove insignification from qabstractitemmodel test on Mac OS X.
Since this test began to run again on Mac, it has been passing
consistently.
Task-number: QTBUG-22748
Change-Id: Ia81cf60b11d45fb331b5eca5d13df00556c18e07
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Jason McDonald [Tue, 24 Apr 2012 11:20:04 +0000 (21:20 +1000)]
lancelot test: Skip crashing test function on Mac OS X.
Task-number: QTBUG-22792
Change-Id: Iee70dafe9b4aced1a7c6cea63da88a92dff7d43d
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Laszlo Agocs [Tue, 24 Apr 2012 07:57:57 +0000 (10:57 +0300)]
Add flag in testlib to disable committing the simulated touch
Some declarative tests (e.g. 24319) need more fine-grained control
over processing the simulated touch events.
Change-Id: Ib163cfd29254016ea16e8d739ad97369b6ccdc90
Reviewed-by: Jason McDonald <jason.mcdonald@nokia.com>
Laszlo Agocs [Mon, 23 Apr 2012 13:18:01 +0000 (16:18 +0300)]
Add a QWindow autotest case for reentrant mouse/touch events
Declarative used to have some issues when the MouseArea event handlers
spinned the event loop: subsequent mouse events were all lost. Let's
add an autotest also for QGuiApplication to make sure the same problem
won't occur there.
Change-Id: If5c3b4f58dad609efb302d2ca932493680ffdf13
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
Stephen Kelly [Mon, 23 Apr 2012 17:24:13 +0000 (19:24 +0200)]
Run DBus tests only if QtDBus is found.
It's not enough to check for UNIX because Qt may have been configured
without dbus, or we could be running on APPLE (on which DBus is not
commonly installed).
Change-Id: Ic57200dad42c0c12b292d1066f799c827294fe9a
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Thiago Macieira [Tue, 24 Apr 2012 11:33:16 +0000 (13:33 +0200)]
Update the QUrl documentation concerning the encoding
Looks like I failed to update this earlier, when the behaviour changed.
Change-Id: Ic020c2a14d4e9153f2bc9d22d943a3a380c0851c
Reviewed-by: Shane Kearns <shane.kearns@accenture.com>
Thiago Macieira [Fri, 20 Apr 2012 12:59:30 +0000 (14:59 +0200)]
Don't use the QRegExp methods that modify the object [QtTest]
QRegExp matching methods modify the object, which we don't want to. In
particular, when we receive a QRegExp from the user or we store in a
context that might require thread-safety, make sure we make a copy
before using it.
QRegularExpression has no such shortcoming.
Task-number: QTBUG-25064
Change-Id: I7c5f5ebf4521c32337c9ea9aeeef50e1e8690bf8
Reviewed-by: Rohan McGovern <rohan.mcgovern@nokia.com>
Reviewed-by: Jason McDonald <jason.mcdonald@nokia.com>
Thiago Macieira [Thu, 29 Dec 2011 18:36:46 +0000 (16:36 -0200)]
Use the C99 keyword "restrict" in compilers that support it
This allows the compiler to produce better / smaller code due to not
being required to deal with two pointers possibly aliasing each
other. This also enables vectorisation where otherwise the compiler
wouldn't be able to do.
Change-Id: Ie7db3a54a1ba22f87ac5283d5c43905707957549
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
Thiago Macieira [Thu, 29 Dec 2011 17:12:59 +0000 (15:12 -0200)]
Mark some functions as always_inline with GCC and ICC.
These functions are too critical to be called. They should always be
inlined.
Change-Id: I698dc6d83e02c323759c6d066f17993340bd53a5
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
John Layt [Mon, 26 Mar 2012 16:23:21 +0000 (17:23 +0100)]
QLocale: Clean up QLocalePrivate implementation
In Qt4 QLocalePrivate is a struct returned by a d() method. This will be
unsuitable for the planned change to use ICU and may cause BIC issues.
This change makes QLocalePrivate a class and creates a new struct
QLocaleData to hold the data index. Further clean-ups are possible
but are left for later.
Change-Id: Ie316a07790f74674a3b520b735dff72695cc4060
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Reviewed-by: Jason McDonald <jason.mcdonald@nokia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Jędrzej Nowacki [Tue, 24 Apr 2012 10:44:34 +0000 (12:44 +0200)]
Cleanup QMetaObject string access.
We do not need operate on raw QByteArrayData, QByteArray gives us
better, stable API, without any performance penalty.
Change-Id: Idf47af2333c3694e81dcd31cd07d1a4720046df7
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Stephen Kelly [Mon, 23 Apr 2012 19:55:14 +0000 (21:55 +0200)]
Fix creation of config files in debug mode.
f88212c22f7e4bec261130a6f82294adfc75abca was not complete.
Change-Id: Ie5b1bb4832176ec8a72ace473f7dbc6c4de10ff2
Reviewed-by: Marius Storm-Olsen <marius.storm-olsen@nokia.com>
Jason McDonald [Fri, 20 Apr 2012 15:06:48 +0000 (01:06 +1000)]
Re-enable network tests on Mac OS.
Switch on most of the tests (all but the bearer tests), marking the two
known failures as insignificant tests.
Change-Id: I17f228a938de1a23eddf897f494bfa4e54338dae
Reviewed-by: Sergio Ahumada <sergio.ahumada@nokia.com>
Reviewed-by: Shane Kearns <shane.kearns@accenture.com>
Shane Kearns [Mon, 23 Apr 2012 14:29:41 +0000 (15:29 +0100)]
Fix tst_qtcpserver::serverAddress on Mac OSX
While linux returns the local address of dual mode sockets as "::"
i.e. the IPv6 address, OSX returns "::FFFF:0.0.0.0" i.e. the IPv4 address.
In the unix socket engine, treat both of these as candidates for
being a dual mode socket. This should hopefully cover other unixes
which we haven't tested.
Task-Number: QTBUG-25445
Change-Id: If1251018d54da98618223ba538c00ba54ac4cb4e
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Martin Smith [Tue, 24 Apr 2012 11:10:40 +0000 (13:10 +0200)]
qdoc: Further changes to the QML specialization
1. Body elements to include type, property, method, signal, enum info, so:
qmlDetail to
- qmlTypeDetail
- qmlPropertyDetail
- qmlEnumerationDetail
- qmlMethodDetail
- qmlSignalDetail
2. Same thing for a section that contains elements specialized for QML, so
qmlDef to
- qmlTypeDef
- qmlPropertyDef
- qmlEnumerationDef
- qmlMethodDef
- qmlSignalDef
Rationale for both items above is that otherwise the body and section
content models would have to be exactly the same for a type, property,
method, etc.
3. For the import statement use:
<qmlImportModule>
<apiItemName>QtGraphicalEffects</apiItemName>
<apiData>1.0</apiData>
</qmlImportModule>
So we have an element indicating that this is for the import
statement, not an outputclass. And no need to specialize the elements
inside it as we know from the ancestor (qmlImportModule) what they're
for.
<qmlImportModule> to be specialized from apiDef.
4. The 'description' section name doesn't need to be specialized further from the API ref. DTDs so
<qmlDesc> should be <apiDesc>
The content model for the 'description' section would be similar to that of a normal DITA section, specialized into apiDesc in the API ref. DTDs, so we don't need to rename it for QML.
5. We need a separate element for the since information.
<qmlSince>
<apiItemName>QtGraphicalEffects</apiItemName>
<apiData>1.0</apiData>
</qmlSince>
6. Public identifier should be like this for a type:
-//NOKIA//DTD DITA QML Type//EN
Not like this:
-//NOKIA//DTD DITA QML API Type Reference v0.1.0//EN
Change-Id: Ie9047965bf32a2e3e9339eb743f2019cb38c88e4
Reviewed-by: Casper van Donderen <casper.vandonderen@nokia.com>
Jason McDonald [Thu, 19 Apr 2012 16:03:55 +0000 (02:03 +1000)]
Remove insignification of QMdiArea test on Windows.
This test had two stable failures, caused by including code not meant
for Windows.
Task-number: QTBUG-25298
Change-Id: I43d9d62ecf5a3c6eec240fafbc43a625c00f45fa
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Jason McDonald [Wed, 18 Apr 2012 17:37:37 +0000 (03:37 +1000)]
Re-enable the QMacStyle test.
This test has only stable failures. Mark those failures with
QEXPECT_FAIL and re-enable the test.
Task-number: QTBUG-25296
Change-Id: I5615700ec8119a827c30e43ae651a684e78170fe
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Markus Rathgeb [Mon, 16 Apr 2012 20:31:51 +0000 (22:31 +0200)]
Fix QMutex::tryLock timeout regression (integer overflow)
The timeout given in milliseconds should be converted to a timespec
struct.
To separate the seconds and nanoseconds the milliseconds are first
multiplied to represent the whole value in an int64 variable.
The calculation is done on integers, so we get an overflow if the
milliseconds are bigger then 2148.
If we cast the given value to an int64 we can avoid this problem.
Fix the used cast.
Task-number: QTBUG-24795
Change-Id: I864ae227cf7dda16a6f30aa4db74acc49e20f6eb
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Niels Weber [Fri, 20 Apr 2012 12:31:34 +0000 (14:31 +0200)]
Improved detection of libcrypto and libssl.
The previous solution didn't work on systems where the libraries
aren't in the same location.
Now we search for both libcrypto and libssl and load them if their
versions match, even if they are in different directories.
Task-number: QTBUG-25398
Change-Id: I37164638890586947d07670d8a59fc53a84f9c42
Reviewed-by: Joerg Bornemann <joerg.bornemann@nokia.com>
Reviewed-by: Shane Kearns <shane.kearns@accenture.com>
Reviewed-by: Richard J. Moore <rich@kde.org>
Donald Carr [Mon, 23 Apr 2012 21:15:14 +0000 (21:15 +0000)]
Add N9 device mkspec to device mkspecs folder
Change-Id: Ie8b0e8a014316b218c86cfdc0c58cad1ae4d3015
Reviewed-by: Girish Ramakrishnan <girish.1.ramakrishnan@nokia.com>
Friedemann Kleint [Tue, 24 Apr 2012 08:34:46 +0000 (10:34 +0200)]
Windows: Fix for frameless widgets.
Let Qt::FramelessWindowHint take precedence over
Qt::WindowTitleHint (which enforces WS_CAPTION and thus
WS_DLGFRAME).
Change-Id: I2c0248d8a3ee3ed0f04b926acdef3cbeb98ca571
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Friedemann Kleint [Tue, 24 Apr 2012 08:21:10 +0000 (10:21 +0200)]
Fix windowgeometry test for Windows.
- Specify main.cpp with full path to avoid obscure link
errors (grabbing main.cpp from windowflags)
- Set proper flags on the test window.
Change-Id: I9bbaf9cb5c7e210f53a51b58bd48a827d6104fe2
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Jason McDonald [Thu, 19 Apr 2012 16:41:39 +0000 (02:41 +1000)]
Remove insignification from QDateTimeEdit test on Mac OS X.
This test only has stable failures. Mark those with QEXPECT_FAIL and
re-enable the test.
Task-number: QTBUG-23674
Change-Id: I0e6cfe4f1992410bbc27b266f7bdafa14451aff7
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Jason McDonald [Thu, 19 Apr 2012 16:25:18 +0000 (02:25 +1000)]
Remove insignification of QSizeGrip test on Mac OS X.
The test has one stable failure. Mark it with QEXPECT_FAIL and
re-enable the test.
Task-number: QTBUG-23681
Change-Id: Iade17cbbd1570863bb628572d8da52ec9c0f62ba
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Jason McDonald [Tue, 24 Apr 2012 07:54:20 +0000 (17:54 +1000)]
Skip unstable tst_QMdiArea::subWindowActivated2 test on Mac OS X.
This test function recently became unstable in parallel with the change
to re-enable the test. Skip the unstable function until it can be
fixed.
Change-Id: I0fb962aa3294a0c9b799acee52ab1fb1cd1c6924
Reviewed-by: Sergio Ahumada <sergio.ahumada@nokia.com>