qt:qt.git
9 years agoBring in WebKit revisions 139553 and 139921.
Raphael Kubo da Costa [Tue, 2 Jul 2013 10:48:28 +0000 (13:48 +0300)]
Bring in WebKit revisions 139553 and 139921.

They are primarily useful for getting rid of some libstdc++-specific
includes which break the build with libc++, and also for simplifying the
code and removing a Darwin-specific code path.

r139553 by ossy@webkit.org:
  Use sync_add_and_fetch instead of gnu_cxx::exchange_and_add
  https://bugs.webkit.org/show_bug.cgi?id=106729

  After r139514 we need atomicIncrement(int64_t volatile*) for all
  platform. Now the GCC implementation of atomicIncrement() is based on
  gnu_cxx::exchange_and_add, which doesn't support int64_t type, but
  sync_add_and_fetch does.

  Reviewed by Benjamin Poulain.

r139921 by benjamin@webkit.org:
  Use GCC's implementation of atomicIncrement/Decrement on Mac
  https://bugs.webkit.org/show_bug.cgi?id=106976

  Reviewed by Filip Pizlo.

  wtf/Atomics.h: GCC and LLVM have builtin for atomic ADD and SUB:
                 sync_add_and_fetch, sync_sub_and_fetch.

  Using them let the compiler just generate the atomic operations inline
  instead of generating a function call to LibC. It also simplify the
  code a bit.

Cherry-picked from qtwebkit/0baf197 and qtwebkit/801fc96. I'm cherry-picking
both revisions together to match what was done in 0de22e80.

Change-Id: Ie5c1067980662ff04e8e36d8cf6e9459b7c46aab
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@digia.com>
9 years agofix off-by-one index check
Mark Brand [Mon, 1 Jul 2013 12:45:37 +0000 (14:45 +0200)]
fix off-by-one index check

Change-Id: I6d6461eec24895e9aa0e77deccd39210958b7e2c
Reviewed-by: Andy Shaw <andy.shaw@digia.com>
(cherry picked from qtbase/96a22ed926cfd706c1b5aca1881f124bc9624f2f)

9 years agoQNX: Changing the way JSC retrieves the current stack base pointer
Fabian Bumberger [Thu, 13 Jun 2013 11:49:20 +0000 (13:49 +0200)]
QNX: Changing the way JSC retrieves the current stack base pointer

What Qt is calling the stackbase is the top of the stack.
The thread local storage [__tls()] area is at the top of the stack
and the stack pointer is initialized below this on thread
creation.
With this patch, the stack base pointer can be retrieved faster then in
the current implementation.

This patch is from Sean Boudreau.

Change-Id: I3d1ac58d5c43997cbf462424c66be0c7caafcf1b
Reviewed-by: Thomas McGuire <thomas.mcguire@kdab.com>
Reviewed-by: Alan Alpert <aalpert@blackberry.com>
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
9 years agoFix leak in QDragManager::drag on mac
Ivan Komissarov [Fri, 3 Aug 2012 12:30:05 +0000 (16:30 +0400)]
Fix leak in QDragManager::drag on mac

Change-Id: I467129f989b6e28078c9dd789cde7ff898faf1f5
Reviewed-by: Christoph Schleifenbaum <christoph.schleifenbaum@kdab.com>
Reviewed-by: Andy Shaw <andy.shaw@digia.com>
Reviewed-by: Morten Johan Sørvig <morten.sorvig@digia.com>
9 years agoUpdated changes-4.8.5 file
aksalova [Wed, 26 Jun 2013 14:50:22 +0000 (17:50 +0300)]
Updated changes-4.8.5 file

Change-Id: I7c5b349e2aecbe2939d2ec5e39c211ad56b7ef1a
Reviewed-by: Sergio Ahumada <sergio.ahumada@digia.com>
9 years agoQNetwork: Check AuthenticationReuseAttribute when using credentials from url
El Mehdi Fekari [Thu, 13 Jun 2013 14:27:57 +0000 (16:27 +0200)]
QNetwork: Check AuthenticationReuseAttribute when using credentials from url

Note: This is not needed in Qt5, as the authenticationRequired method in Qt5
has an allowAuthenticationReuse parameter (bool) that is checked before
the credentials are used.

Change-Id: I5a2734de615a1a96d1fe648bd251850f3b45e167
Reviewed-by: Peter Hartmann <phartmann@blackberry.com>
9 years agoDesigner: Fix morphing of widgets in QSplitters.
Friedemann Kleint [Fri, 21 Jun 2013 08:21:36 +0000 (10:21 +0200)]
Designer: Fix morphing of widgets in QSplitters.

They were not handled at all, children were just reparented
resulting in the morphed widgets being appended.

Task-number: QTBUG-31911
Change-Id: Idae8f66663416d514a2f3c299e58d726a3dd578a
Reviewed-by: Jarek Kobus <jaroslaw.kobus@digia.com>
(cherry picked from qttools/d251ee9dcbd015f2acbe76231902b72163cd05a2)

9 years agoQUrl stringprep: fix handling of U+0080: it's prohibited
Thiago Macieira [Sat, 8 Jun 2013 02:20:32 +0000 (19:20 -0700)]
QUrl stringprep: fix handling of U+0080: it's prohibited

Edge case: a > that should have been >=. Without it, we never ran the
rest of the IDN nameprepping.

(cherry-picked from qtbase commit 4d93393a6de2d6631979df2bc6d12aa43781dc6f)
Change-Id: I2276d660de3a70d0c561bb18816820d9a0f47e77
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
9 years agoQUrl stringprep: avoid recalculating the surrogates we already know
Thiago Macieira [Sat, 8 Jun 2013 01:58:11 +0000 (18:58 -0700)]
QUrl stringprep: avoid recalculating the surrogates we already know

(cherry-picked from qtbase commit 53388cd8e0451ea375ed250b59f9e89319fb3e1c)
Change-Id: Icac4e81fff6f7f7fa4f46ec2a08105f8d3d2b403
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
9 years agoQUrl stringprep: fix handling of prohibited characters
Thiago Macieira [Sat, 8 Jun 2013 01:56:58 +0000 (18:56 -0700)]
QUrl stringprep: fix handling of prohibited characters

RFC 3454 says about prohibited characters (section 2, "Preparation
Overview"):

   3) Prohibit -- Check for any characters that are not allowed in the
      output.  If any are found, return an error.  This is described in
      section 5.

In other words, we mustn't simply strip the output of prohibited
characters. We must generate an error if they are present. We do that by
clearing the data.

We already had tests for prohibited output, but they were
indistinguishable from being stripped. So instead add some extra
characters so that we can tell whether the label was cleared.

(cherry-picked from qtbase commit 736a052d93d9c75e51e8f3da733bc8e4a50c39ce)
Change-Id: I2d95217c27be5e2d54deed0036cb009e3b7f4886
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
9 years agoQUrl stringprep: fix case folding from non-BMP to BMP
Thiago Macieira [Sat, 8 Jun 2013 01:42:04 +0000 (18:42 -0700)]
QUrl stringprep: fix case folding from non-BMP to BMP

When uc > 0xffff (non-BMP character) and l == 1 (replacement is in the
BMP), we must use QString::replace so the correct number of characters
is replaced.

There's one case testing this in tst_qurl, but it is being
obscured by another bug (false positive).

(cherry-picked from qtbase commit 86312275197c3fde948035a59c0358162701f9f2)
Change-Id: I32388dd5bef32d4d6804aeeec4904bd5f563e9b9
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
9 years agoQUrl stringprep: recalculate the current position if the size changes
Thiago Macieira [Sat, 8 Jun 2013 01:39:12 +0000 (18:39 -0700)]
QUrl stringprep: recalculate the current position if the size changes

If the case folding operation results in either expansion or reduction
of the string, we need to adapt. Reduction happens most often when a
non-BMP character is case-folded to a character in the BMP (example:
mathematical signs at U+1D400-1D7FF). Expansion happens in the rare
case of symbols containing words, like U+2121 ℡ (this is part of the
unit test), and one common case: the German sharp S (ß) is expanded to
"ss".

(cherry-picked from qtbase commit 2ccf4c32cc593f568581dd237e1e27d39fd965a2)
Change-Id: I1bdbdc908b958a89bf30e4bb648d65dfdd9097f8
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
9 years agoQSystemLocale: Fix the time format on BlackBerry 10
Mehdi Fekari [Fri, 14 Jun 2013 10:56:06 +0000 (12:56 +0200)]
QSystemLocale: Fix the time format on BlackBerry 10

The time format should depend on the device settings (24 hour format)

cherry picked from qtbase/686c94c892006a3e72291ca8971aabf18ea1ec85

Change-Id: I6226486ee1feedc27fab49802a0043619e83659c
Reviewed-by: Rafael Roquetto <rafael.roquetto@kdab.com>
9 years agoDoc: remove the duplicate for in QStyle documentation
Liang Qi [Thu, 20 Jun 2013 05:49:57 +0000 (07:49 +0200)]
Doc: remove the duplicate for in QStyle documentation

Task-number: QTBUG-31739

(cherry picked from commit 0a55499c9af9a4b0a98a12a8aba82c3764fa5994)

Change-Id: I1d2d1bb5ca7a19a8c42f11e944e8339e56df6e4b
Reviewed-by: Liang Qi <liang.qi@digia.com>
9 years agoFix various NTLM/Digest multi-threading and usage issues
Aron Rosenberg [Wed, 25 Apr 2012 16:57:43 +0000 (09:57 -0700)]
Fix various NTLM/Digest multi-threading and usage issues

- Fix multi-threading bug where NTLM/Digest auth would fail when
concurrent requests were on the wire. The fix for this is too not
internally share QAuthenticationPrivate pointers, since the challenge
values would get overridden in different threads. This was failing
because the internal QAuthenticationPrivate members would have been
set with the status/values of the current request which would mess
up the state of the new request. As currently implemented, the helper
functions inside QAuthenticationPrivate can't call detach to implement
proper copy on write symantics.
- Fix issue where if user was set via constructor, the NTLM domain
parsing would not occur. Parsing of DOMAIN\user is now redone if
proxy type is determined to be NTLM.

Task-number: QTBUG-15472
Task-number: QTBUG-17322
Task-number: QTBUG-18794
Task-number: QTBUG-13063
Task-number: QTBUG-16585
(cherry picked from commit 6ccbfd6ca498da04e4ef02102c4ded9768225b5a)
Cherry-picked-by: Peter Hartmann <phartmann@blackberry.com>
Change-Id: I58ae6ae362baadfb25eaa7b29556e6e8250d5d95
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
9 years agonetwork requests: do not access 1st byte of empty byte array
Peter Hartmann [Thu, 13 Jun 2013 15:43:56 +0000 (17:43 +0200)]
network requests: do not access 1st byte of empty byte array

... because otherwise this would crash.
Apparently there are cases where the header name is empty.

Task-number: QTBUG-31667

(cherry picked from commit dd050d35d77e2f6ac37bd1266a843df55848022f)

Change-Id: I4c99094d2bad3ebb7cf44f83f9ed6a5f73b73016
Reviewed-by: Richard J. Moore <rich@kde.org>
9 years agoRecognize separator item in QMenu.
Leonard Lee [Mon, 17 Jun 2013 08:47:06 +0000 (10:47 +0200)]
Recognize separator item in QMenu.

The mouse over event on upper level menu separator should
automatically close the sub menu listing. Manual test is not
needed for this commit since it is easy to test it against common
examples.

Task-number: QTBUG-31664
Change-Id: I6632d0bce27ca064ddfdd05743df969b3e6e02ee
(cherry-picked from qtbase/f730cd7542548d7d5fb7cd25f631e471697960c8)
Reviewed-by: J-P Nurmi <jpnurmi@digia.com>
9 years agoImplement apostrophe in windows cmd args.
Leonard Lee [Fri, 14 Jun 2013 10:56:19 +0000 (12:56 +0200)]
Implement apostrophe in windows cmd args.

Task-number: QTBUG-15379
Change-Id: I113692983a8c16c56f0f684146d0d32e25d4869d
(cherry-picked from qtbase/c1e24f883824ad894aa121568e4fde8553028d95)
Reviewed-by: Andy Shaw <andy.shaw@digia.com>
9 years agoFix crash when re-creating QThreadData after initially destroying it
Tor Arne Vestbø [Tue, 11 Jun 2013 13:05:13 +0000 (15:05 +0200)]
Fix crash when re-creating QThreadData after initially destroying it

We destroy the thread data for the main thread when the QCoreApplication
is destructed, and then delete the pthread key for the thread data in
the global static destructor function 'destroy_current_thread_data_key'.

The user may have its own Q_DESTRUCTOR_FUNCTION though, which may or may
not run after we've destroyed the key. If it runs after we've destroyed
the key, we'll end up trying to re-create the tread-data, as expected,
but set_thread_data() will fail to persist it, as pthread_setspecific
is called with an invalid key. The result is an infinite recursion:

...
6  in QThreadData::current () at qthread_unix.cpp:216
7  in QObject::QObject (this=0x48e1b30, dd=@0x48e1b40, parent=0x0) at qobject.cpp:703
8  in QThread::QThread (this=0x48e1b30, dd=@0x48e1b40, parent=0x0) at qthread.cpp:396
9  in QAdoptedThread::QAdoptedThread (this=0x48e1b30, data=0x48e1af0) at qthread.cpp:120
10 in QAdoptedThread::QAdoptedThread (this=0x48e1b30, data=0x48e1af0) at qthread.cpp:130
11 in QThreadData::current () at qthread_unix.cpp:219
12 in QObject::QObject (this=0x48e1a20, dd=@0x48e1a30, parent=0x0) at qobject.cpp:703
...

To solve this, we reset current_thread_data_once when destroying the key,
so that subsequent calls to pthread_once to potentially create the key
will call create_current_thread_data_key once more. This means we'll leak
the key for this particular use-case, since we don't end up calling
pthread_key_delete a second time, but this leak is small and happens
typically only for a short duration during application shutdown.

Cherry-picked from qtbase/46667d604fb2ae11a87c0c075a3d2468d02f7bdb.

Change-Id: Iffc372ca530a486cd3efc2237ab02468bdcb5c81
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
9 years agoMake sure QThreadData and QAdoptedThread object is destroyed at app exit
jian liang [Thu, 21 Mar 2013 15:24:36 +0000 (23:24 +0800)]
Make sure QThreadData and QAdoptedThread object is destroyed at app exit

Task-number: QTBUG-30134

Restore the lines in qcoreapplication.cpp removed by commit
950b35cf97ad398f97883efd2a18ee97994a8a9c to ensure QThreadData and
QAdoptedThread object of main thread is destroyed at application exit.
We don't set QCoreApplicationPriavte::theMainThread to 0 as before since
it will be set to zero in QThreadData::~QThreadData()

Cherry-picked from qtbase/7dc622290bb8e81af634034f443e25be0d6d48a3.

Change-Id: I6bc62235d0aff7818ca9a9cad45625b2777afc55
Reviewed-by: jian liang <jianliang79@gmail.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
9 years agoClear the current thread data for the main thread
Thiago Macieira [Fri, 18 Jan 2013 06:39:00 +0000 (14:39 +0800)]
Clear the current thread data for the main thread

This avoids crashes accessing deleted memory when creating a QObject
after the last QObject had been deleted, like a qDebug() in global
destructors.

==41000== Invalid read of size 4
==41000==    at 0x5F01ED5: bool QBasicAtomicOps<4>::ref<int>(int&) (qatomic_x86.h:208)
==41000==    by 0x5F01309: QBasicAtomicInteger<int>::ref() (qbasicatomic.h:147)
==41000==    by 0x5F24051: QThreadData::ref() (qthread.cpp:100)
==41000==    by 0x614A984: QObject::QObject(QObject*) (qobject.cpp:681)
==41000==  Address 0x6ee73f0 is 0 bytes inside a block of size 152 free'd
==41000==    at 0x4A0736C: operator delete(void*) (vg_replace_malloc.c:480)
==41000==    by 0x5F240BF: QThreadData::deref() (qthread.cpp:109)
==41000==    by 0x6113F6B: QCoreApplicationData::~QCoreApplicationData() (qcoreapplication.cpp:268)

The comment right above the change in qthread.cpp looks eerily similar
to the problem I'm trying to fix. However, the actual change that
introduced the change is not in the Qt public history, so we can't
know for sure what the problem was then.

Cherry-picked from qtbase/950b35cf97ad398f97883efd2a18ee97994a8a9c.

Change-Id: Ic4072c15529e2ae94ea36fbd0340cf5ee61413d2
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
9 years agoclucene: Compile with clang's libc++
Bradley T. Hughes [Wed, 30 May 2012 09:41:06 +0000 (11:41 +0200)]
clucene: Compile with clang's libc++

When using clang's libc++, the 'using namespace std;' in CLucene's
StdHeader.h causes ambiguity between wcschr() from wchar.h and
std::wcschr() from cwchar (which is automatically included by libc++).

The only symbols used from the std namespace are std::min() and
std::max() so only pull these in, not the whole std namespace.

Cherry-picked from qttools/bc35938d0d49b04bf98feda3e85da186ecc8184d.

Change-Id: Ife6647df795f59a5feebd5c1aa3eebe7446ec720
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
9 years agoBring in WebKit revisions 139553 and 139921.
Raphael Kubo da Costa [Tue, 11 Jun 2013 16:25:08 +0000 (19:25 +0300)]
Bring in WebKit revisions 139553 and 139921.

They are primarily useful for getting rid of some libstdc++-specific
includes which break the build with libc++, and also for simplifying the
code and removing a Darwin-specific code path.

r139553 by ossy@webkit.org:
  Use sync_add_and_fetch instead of gnu_cxx::exchange_and_add
  https://bugs.webkit.org/show_bug.cgi?id=106729

  After r139514 we need atomicIncrement(int64_t volatile*) for all
  platform. Now the GCC implementation of atomicIncrement() is based on
  gnu_cxx::exchange_and_add, which doesn't support int64_t type, but
  sync_add_and_fetch does.

  Reviewed by Benjamin Poulain.

r139921 by benjamin@webkit.org:
  Use GCC's implementation of atomicIncrement/Decrement on Mac
  https://bugs.webkit.org/show_bug.cgi?id=106976

  Reviewed by Filip Pizlo.

  wtf/Atomics.h: GCC and LLVM have builtin for atomic ADD and SUB:
                 sync_add_and_fetch, sync_sub_and_fetch.

  Using them let the compiler just generate the atomic operations inline
  instead of generating a function call to LibC. It also simplify the
  code a bit.

Cherry-picked from qtscript/184fc178f4c753727445694c4830c250f53da588.

Change-Id: Id785e35944682691725947e0f329668d17ff6903
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
9 years agoQGtkStyle: add getIconThemeName() helper function
Dmitry Shachnev [Sun, 19 May 2013 11:49:47 +0000 (15:49 +0400)]
QGtkStyle: add getIconThemeName() helper function

Add QGtkStyle::getIconThemeName() static helper function
(which uses GtkSettings to retrieve the theme name) and use it in
QGuiPlatformPlugin::systemIconThemeName() when desktop is GNOME.

In modern distributions like Debian and Ubuntu, the
/desktop/gnome/interface GConf schema is no longer installed,
so the old method for retrieving icon theme name no longer works.

Qt 5 already uses GtkSettings for this thing, although the code
is in a different place.

Change-Id: Ib13d4809b55dd55dc1fa0bc2c4b1552f13feb030
Reviewed-by: J-P Nurmi <jpnurmi@digia.com>
Reviewed-by: Jens Bache-Wiig <jens.bache-wiig@digia.com>
9 years agoQDockWidget: Keep position when undocking.
Friedemann Kleint [Tue, 11 Jun 2013 13:35:22 +0000 (15:35 +0200)]
QDockWidget: Keep position when undocking.

Initialize undockedGeometry to roughly the current position.

Task-number: QTBUG-31044
Change-Id: I03cbe280d1215bb58ab721b60e29b45359cde76d
Reviewed-by: Andy Shaw <andy.shaw@digia.com>
(cherry picked from qtbase/0d459619a902e547d8c74745453b3bdb8f67a660)

9 years agoDocument the fact that QPA is experimental in Qt 4.8
Andy Shaw [Tue, 11 Jun 2013 12:59:53 +0000 (14:59 +0200)]
Document the fact that QPA is experimental in Qt 4.8

Task-number: QTBUG-26912

Change-Id: I8544a970f58faf9fd9811603d6640ad2b99ebe1d
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Jerome Pasion <jerome.pasion@digia.com>
9 years agoCleanup the QDrag when doing a drag with the right mouse button on Mac
Andy Shaw [Fri, 7 Jun 2013 07:48:56 +0000 (09:48 +0200)]
Cleanup the QDrag when doing a drag with the right mouse button on Mac

When doing a drag with the right mouse button on Mac then it will be
ignored straightaway because it will not be accepted. Therefore we
need to cleanup the QDrag object as normal so that it is deleted as
appropriate.

Change-Id: I781fcea7b7c75775adb814090e917363fa7c9189
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@digia.com>
9 years agoOnly use Vista style for Wizards when on Vista or later
Andy Shaw [Mon, 10 Jun 2013 08:49:48 +0000 (10:49 +0200)]
Only use Vista style for Wizards when on Vista or later

If the application is running in XP compatibility mode then we don't
want to show the Vista style for the wizard as this will include a back
button that should not exist. Therefore we check the platform it is on
before trying to resolve the symbols so that it is correct.

Change-Id: Ic7249e06461068fe08184a22bb750ab8736ebc3f
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
9 years agoCompile with clang's libc++
Bradley T. Hughes [Wed, 30 May 2012 09:49:40 +0000 (11:49 +0200)]
Compile with clang's libc++

An overload for std::swap() what works with std::pair<A,B> is included
in libc++ (but not libstdc++). Avoid ambiguity in HashTable.h by
omitting the version there and using the std::swap() version.

Collector.cpp needs to include pthread.h directly, as it is not
automatically included by libc++'s headers (libstdc++ does pull in
pthread.h though).

Cherry-picked from qtscript/48baeef0766ba145d26d374561e152b40245f3b8.

Change-Id: I4dc88638ce8553bb97037b841fc6cc5e8c9911e2
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
9 years agoFix for OS X QFileDialog showing bundle as directory
Samuel Gaist [Wed, 5 Jun 2013 16:47:39 +0000 (18:47 +0200)]
Fix for OS X QFileDialog showing bundle as directory

Also check if the selectedFile is a bundle to correctly
set currentDir.

This fix does not apply to Qt 5 because QFileDialog has
been reimplemented differently and this file does not
exist anymore.

Task-number: QTBUG-31562

Change-Id: I145108647dd613cb529bb65bc6c7191154dab5b8
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@digia.com>
9 years agoCorrect error in Makefile generation for Windows MinGW. v4.8.5
Michael Brüning [Wed, 5 Jun 2013 12:11:06 +0000 (14:11 +0200)]
Correct error in Makefile generation for Windows MinGW.

Remove the build_pass check that was added for vcproj
generation as it is no longer deemed necessary.

See http://trac.webkit.org/changeset/62264 and
http://trac.webkit.org/changeset/59266 for context.

Task-number: QTBUG-27413
Change-Id: Ie82032cf4b183b31c66defc2d49cc8095989ec8f
Reviewed-by: Jocelyn Turcotte <jocelyn.turcotte@digia.com>
9 years agoDon't overwrite QMAKE_QT_CONFIG with empty value
Jonathan Liu [Thu, 6 Jun 2013 02:42:13 +0000 (12:42 +1000)]
Don't overwrite QMAKE_QT_CONFIG with empty value

If the mkspec sets QMAKE_QT_CONFIG, QMAKE_QT_CONFIG may be overwritten
with an empty value from .qmake.cache. Avoid this by first checking
if the value from .qmake.cache is not empty before assigning it to
QMAKE_QT_CONFIG.

This was already addressed differently in Qt 5.

Change-Id: I95fa630139b8798156a2fb15d0dde630a0a53a0a
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
9 years agoRemove Nokia references.
Leonard Lee [Tue, 4 Jun 2013 11:53:31 +0000 (13:53 +0200)]
Remove Nokia references.

These Nokia references removed as they are not appropriate.

Task-number: QTBUG-31400
Change-Id: Ib80d2c2d28353e74c46e9ead9cde9807dad7022d
(cherry-picked from qtdoc/019f96bc47cf56170c125ae90f5fc00059f6665d)
(cherry-picked from qtwebkit/d7b8c69ccdd5547b5660be5876dbdedc3ac29548)
Reviewed-by: Jerome Pasion <jerome.pasion@digia.com>
9 years agoRe-enable compilation of idc for MinGW
Kai Koehne [Tue, 4 Jun 2013 07:26:50 +0000 (09:26 +0200)]
Re-enable compilation of idc for MinGW

Re-enable compilation of idc for MinGW to fulfill the activeqt dependency
on idc.

idc and activeqt compile fine with MinGW, but most examples won't work
because of the missing midl compiler. This was the reason activeqt and idc
was disabled originally after Qt 4.8.4. Anyhow, 330ca76b then re-enabled
compilation for activeqt, but not idc, which broke the build. This patch
re-enables compilation of idc, but still by default does not compile
activeqt examples.

Qt 5 ships with activeqt / idc for MinGW too, btw, so excluding it in Qt
4.8.5 is inconsistent.

Task-number: QTBUG-31517
Change-Id: I7d4e7ed56c1954996b7291f56143ff83fcdcb0a5
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
9 years agoMake sure idc links only against QtCore
Kai Koehne [Tue, 4 Jun 2013 10:43:16 +0000 (12:43 +0200)]
Make sure idc links only against QtCore

Do not link against QtGui (which might not have been compiled yet).

Task-number: QTBUG-22260
Change-Id: I42d926cc30a0bc278fc796f32652281c03f140d6
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
9 years agoDisable precision timers when running MSVC2012 code on pre-Windows 8.
Friedemann Kleint [Thu, 2 May 2013 15:04:08 +0000 (17:04 +0200)]
Disable precision timers when running MSVC2012 code on pre-Windows 8.

Precision timers can cause the event loop to lock up
when running MSVC2012 code on pre-Windows 8.

Task-number: QTBUG-27266

Change-Id: Idd73731e82843d0d140859bab825bc1a54eccf1a
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
(cherry picked from qtbase/aa1b4c0943187d82e0c313b93559e99226a9c75a)

9 years agoUpdate LICENSE.PREVIEW.COMMERCIAL license
Sergio Ahumada [Mon, 3 Jun 2013 11:42:23 +0000 (13:42 +0200)]
Update LICENSE.PREVIEW.COMMERCIAL license

Change-Id: I8af7630ec54e1a9d4b063e8278df5e75d2abf97e
Reviewed-by: Akseli Salovaara <akseli.salovaara@digia.com>
Reviewed-by: Iikka Eklund <iikka.eklund@digia.com>
9 years agoDoc: Fixed typo in QDialog.
Jerome Pasion [Mon, 3 Jun 2013 10:37:35 +0000 (12:37 +0200)]
Doc: Fixed typo in QDialog.

Task-number: QTBUG-31493

Change-Id: Ie112f601d629c794842d746e7b9c96849c74bdf6
Reviewed-by: Sergio Ahumada <sergio.ahumada@digia.com>
9 years agoFix Qt WebKit build for non-default Xcode setup / Xcode 4.3+
Danny Boelens [Wed, 17 Apr 2013 13:18:37 +0000 (15:18 +0200)]
Fix Qt WebKit build for non-default Xcode setup / Xcode 4.3+

There's no need to test the full path, the actual name of the SDK will do.
This fixes the build in cases where Xcode is not installed in /Developer
or when building with Xcode 4.3 or higher (where Xcode is a bundle), because
in those cases the SDK's full path does not start with /Developer.

Change-Id: I598ba2a5ebaae813f576dc72102d05fe8804a86f
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@digia.com>
9 years agoTools: unnecessary to make qmlplugindump a bundle
Liang Qi [Mon, 27 May 2013 14:04:58 +0000 (16:04 +0200)]
Tools: unnecessary to make qmlplugindump a bundle

Embed Info.plist in qmlplugindump executable in a special way.

Change-Id: Ia83cf15a1dc8c7e37560a280bd711ff2e4468ad8
(adpated from qtdeclarative/4d173d5f16c103f640d4089feed44a74b14e9004)
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
9 years agopostgresql driver: escape backslashes only when server requires it
Mark Brand [Thu, 7 Mar 2013 10:52:52 +0000 (11:52 +0100)]
postgresql driver: escape backslashes only when server requires it

Task-number: QTBUG-30076

Change-Id: I408cda941884f01484d0edfa82c91fc19cb8718c
Reviewed-by: Israel Lins Albuquerque <israelins85@yahoo.com.br>
Reviewed-by: Sergey Blagodarin
Reviewed-by: Andy Shaw <andy.shaw@digia.com>
Reviewed-by: Andras Mantia <andras@kdab.com>
(cherry picked from qtbase/e3c5351d06ce8a12f035cd0627356bc64d8c334a)
Reviewed-by: Mark Brand <mabrand@mabrand.nl>
9 years agoDoc: corrected LGPL Exception header
Nico Vertriest [Thu, 23 May 2013 12:05:58 +0000 (14:05 +0200)]
Doc: corrected LGPL Exception header

Task-number: QTBUG-31271

Change-Id: Ifdb36c4bc01f88f743f128e46272b5c9c20fde4e
Reviewed-by: Jerome Pasion <jerome.pasion@digia.com>
Reviewed-by: Topi Reiniö <topi.reinio@digia.com>
Reviewed-by: Caroline Chao <caroline.chao@digia.com>
Reviewed-by: Sergio Ahumada <sergio.ahumada@digia.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
9 years agorebuild configure.exe
Oswald Buddenhagen [Fri, 24 May 2013 09:41:13 +0000 (11:41 +0200)]
rebuild configure.exe

583f6e1: Disabling largefile support for WinCE
2bba0ea: Avoid macro re-definition after 65bd0bd1f04
45a277c: Fix url about evaluation license
e80d7bb: unmess CONFIG/QT_CONFIG static/shared
e5735c9b: Update copyright year in Digia's license headers
f7e9417: fix -inotify option
e5e5238: add inotify command line switch
69b918f: Fix syntax error on configure.exe source code.
deca010: Use slogger2 for logging on Blackberry instead of writing to stderr

Change-Id: I5ed9540aa3841be05926669cdc40afd608f69a37
Reviewed-by: Akseli Salovaara <akseli.salovaara@digia.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
9 years agoFix creating new QWS surface only when it does not exist already
Pasi Petäjäjärvi [Thu, 23 May 2013 07:46:48 +0000 (10:46 +0300)]
Fix creating new QWS surface only when it does not exist already

Calling createSurface(const QString &key, const QByteArray &data)
function is needed when starting/creating new server/client
application, and changes to region sizes (and memory allocation)
after that is handled by QWS surfaces themself inside
setGeometry(const QRect &rect) function.

Task-number: QTBUG-31254
Change-Id: I97c78ebe83b2aa9ab9e4ffc8a9987ab2528f8cf1
Reviewed-by: Paul Olav Tvete <paul.tvete@digia.com>
9 years agoFix QLocale::standaloneMonthName when d->m_data == systemData()
El Mehdi Fekari [Thu, 2 May 2013 10:30:49 +0000 (12:30 +0200)]
Fix QLocale::standaloneMonthName when d->m_data == systemData()

At the moment if d->m_data == systemData() it calls
systemLocale()->query but forgets about the standalone part
so you get the wrong data

This patch introduces the new enums so that backends can implement
properly the standaloneMonthName feature properly. At the moment the Windows
and Mac ones still return the monthName, the Unix and Blackberry ones return
the data we store in months_data.

cherry picked from qtbase/b1de018e9a1efcd0f9e298ae4c865197feb47895

Change-Id: I1054420dc696ed2be0d52bdc6567ebaf202faf12
Reviewed-by: Albert Astals Cid <albert.astals@canonical.com>
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
9 years agoQNX: Enabled ALSA support
Bernd Weimer [Fri, 17 May 2013 09:05:09 +0000 (11:05 +0200)]
QNX: Enabled ALSA support

If ALSA version >=1.0.10 is available on QNX it will be used by QAudio
classes. In Qt5 this is enabled already.

Change-Id: Ia4e3952ee147eaf78be7cec3c8a9fcfc97da332c
Reviewed-by: Fabian Bumberger <fbumberger@rim.com>
Reviewed-by: Thomas McGuire <thomas.mcguire@kdab.com>
9 years agoConvert the new filename to native separators before checking it
Andy Shaw [Mon, 15 Apr 2013 11:18:11 +0000 (13:18 +0200)]
Convert the new filename to native separators before checking it

If a native separator was put in the new name when renaming a file name
via the file dialog then it would correctly fail. But if a non-native
one was used on Windows then it would cause the file to be moved
instead if the directory existed.

Change-Id: If01760b8c54a69b600c9a44c7509017be70d33e3
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
(cherry picked from qtbase/9ec493fa41d06a618d3983c02d9a1141d9d306df)

9 years agoRespect specified minimum height for menuBar also if it has HFW.
Jan Arve Saether [Tue, 14 May 2013 07:45:53 +0000 (09:45 +0200)]
Respect specified minimum height for menuBar also if it has HFW.

If the menu bar is subject to height for width (HFW) we should of
course respect that, but in addition we should ensure that the HFW is
within the minimum and maximum height. This also is consistent with how
QGridLayout calculates the effective minimum row height.

This fixes a regression because change 4780f94e391b5e881497c5228661dead
turned QTabWidget into a proper height-for-width citizen, and when
setting a QTabWidget as a menuwidget, the buggy codepath for HFW was
suddenly hit in menuBarHeightForWidth().

Task-number: QTBUG-31057

Reviewed-by: Paul Olav Tvete <paul.tvete@digia.com>
(cherry picked from qtbase/27e690e163408dec1e9c56ffc07131354d4b2c8a)

Change-Id: I915d37c69152c1804925000303e82a3fae5a9a47
Reviewed-by: Paul Olav Tvete <paul.tvete@digia.com>
9 years agoQGV: fix items not to be selected on right mouse button release
J-P Nurmi [Wed, 15 May 2013 12:31:47 +0000 (14:31 +0200)]
QGV: fix items not to be selected on right mouse button release

Task-number: QTBUG-30990
Change-Id: Iaf2dd7ed496625097daa05d5dc92ef5957574ee9
(Cherry-picked from qtbase/0b862e067756132225e33be09670631edd50d944)
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Andreas Aardal Hanssen <andreas@hanssen.name>
9 years agoPrevent crash due to giving QWidget::update() a large region.
Samuel Rødal [Fri, 26 Apr 2013 08:36:04 +0000 (10:36 +0200)]
Prevent crash due to giving QWidget::update() a large region.

Similar to what change a298216bb4383dbe96 does for update(QRect) we clip
the update region against the widget's rect and return if it's empty.
Otherwise we risk ending up with update rects that are larger than
INT_MAX due to multiple update rects being merged.

Task-number: QTBUG-30876
Change-Id: Idf695b1fdca50449a1e5ddf37500653de290590c
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
(cherry picked from commit 69ee30260a667b2b977a0d4b52abf6537521cce8)

9 years agoPrevent crash due to giving QWidget::update() a large rect.
Samuel Rødal [Fri, 26 Apr 2013 08:36:04 +0000 (10:36 +0200)]
Prevent crash due to giving QWidget::update() a large rect.

We can simply clip the update rect against the widget's rect and return
if it's empty. Otherwise we risk ending up with update rects that are
larger than INT_MAX due to multiple update rects being merged.

Task-number: QTBUG-30876
Change-Id: I23bd0149fbe8d1a007a60b228e6bddb45dc4fc32
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
(cherry picked from commit a298216bb4383dbe96688dfb80da0cd875766de0)

9 years agotest: Mark tst_qinputdialog as insignificant only on Mac OS X
Sergio Ahumada [Thu, 9 May 2013 12:53:39 +0000 (14:53 +0200)]
test: Mark tst_qinputdialog as insignificant only on Mac OS X

This test has been passing lately for Linux and Windows.

Task-number: QTBUG-31101
Change-Id: Iccb8169dc513df7c0c34cd15d21d11e59cb44f11
Reviewed-by: Simo Fält <simo.falt@digia.com>
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@digia.com>
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agoWidgets: avoid integer divide by zero in QProgressDialog
Liang Qi [Wed, 15 May 2013 10:34:47 +0000 (12:34 +0200)]
Widgets: avoid integer divide by zero in QProgressDialog

Autotest is included.

Task-number: QTBUG-31046

Change-Id: Ief7d71b58e7a5416f3659f19445e5d729849b3b6
(cherry picked from commit 69c05bbef47fb9d30d70e594bd5942add8b135fe)
Reviewed-by: J-P Nurmi <jpnurmi@digia.com>
Reviewed-by: Liang Qi <liang.qi@digia.com>
9 years agoBlackBerry: Revised error handling in event dispatcher
Bernd Weimer [Tue, 14 May 2013 12:28:41 +0000 (14:28 +0200)]
BlackBerry: Revised error handling in event dispatcher

Added an upper bound check for socket notifier file descriptors. Too
high FDs have been a real source of failure.
Added compiler hints to allow error-free code path run faster (errors
are highly unlikely) and adjusted warning messages (some were misleading
and too long).
Backported from qtbase: ebd5de1

Change-Id: Ib982f853ee803447bdcaf22aed310697a36b4ccf
Reviewed-by: Fabian Bumberger <fbumberger@rim.com>
Reviewed-by: Thomas McGuire <thomas.mcguire@kdab.com>
9 years agoBackport qmlRegisterUncreatableType with version template
Alan Alpert [Fri, 3 May 2013 18:11:27 +0000 (11:11 -0700)]
Backport qmlRegisterUncreatableType with version template

Otherwise it is not possible to extend or add uncreatable types safely.

Change-Id: I22e81bb46a69efcec3522454d5ad465099b2555b
Reviewed-by: Christopher Adams <chris.adams@jollamobile.com>
Reviewed-by: Alan Alpert (Personal) <416365416c@gmail.com>
9 years agoDisabling largefile support for WinCE.
Bjoern Breitmeyer [Wed, 27 Jun 2012 08:40:39 +0000 (10:40 +0200)]
Disabling largefile support for WinCE.

This was disabled before, but
2bba0eadc1570f29331285738b9d319687f84d4c
enabled it per default.

Change-Id: Ic5a8e50d116cf6b15719668faf055967a6f34e06
(cherry picked from commit 45750b18187ef257a986cf6ab8b2c97ef3ba6e52)
Reviewed-by: Andreas Holzammer <andreas.holzammer@kdab.com>
Reviewed-by: Miikka Heikkinen <miikka.heikkinen@digia.com>
Reviewed-by: Björn Breitmeyer <bjoern.breitmeyer@kdab.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
9 years agoqmake/vcxproj: resurrect support for /we compiler switch
Joerg Bornemann [Thu, 2 May 2013 15:43:17 +0000 (17:43 +0200)]
qmake/vcxproj: resurrect support for /we compiler switch

The vcxproj format supports the /we compiler switch as
TreatSpecificWarningsAsErrors tag.
Support for this switch got lost during refactoring of the
vc(x)proj generators for Qt 4.8.0.

Task-number: QTBUG-29098

Change-Id: I607e907dcb9c1bc7d261053b6d757941ec7db822
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
(cherry picked from commit a3d1d41d1d88b989fe7c460ef32b4f8606aafe0c)

9 years agoqmake/vcxproj: fix parsing of /RTC* options
Joerg Bornemann [Thu, 2 May 2013 16:42:09 +0000 (18:42 +0200)]
qmake/vcxproj: fix parsing of /RTC* options

/RTCsu and /RTCus must be handled as full runtime check options.

Task-number: QTBUG-30711

Change-Id: I783bf49f2ab1d4fd9636dca8e434bccb54844c8c
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
(cherry picked from commit 3c62f4d0c65f1a60bd8d4e6b985d79cb0aa8342b)

9 years agotest: Mark tst_qsharedpointer as insignificant only on Mac OS X
Sergio Ahumada [Thu, 9 May 2013 09:44:20 +0000 (11:44 +0200)]
test: Mark tst_qsharedpointer as insignificant only on Mac OS X

This test has been passing lately for Linux and Windows.

Task-number: QTBUG-31100
Change-Id: I925accd0637250ba8e7a0173c51a2171b4e960b3
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
9 years agoSpecial-case the forward declaration of STL types for libc++.
Raphael Kubo da Costa [Fri, 10 May 2013 01:08:38 +0000 (04:08 +0300)]
Special-case the forward declaration of STL types for libc++.

This is somewhat of a follow-up to 5210d47aa66214e3cb16f394d0510a91f770c1b1.
libc++ declares the STL types in an inline namespace within the "std" one if
clang is used.

If one includes a header such as <iterator> before a Qt one and builds with
QT_NO_STL, the following ends up happening:

  // <iterator>
  namespace std {
      inline namespace __1 {
          struct random_access_iterator ... ;
      }
  }

  // qiterator.h
  #ifdef QT_NO_STL
  namespace std {
     struct random_access_iterator;
  }
  #endif

qiterator.h's declaration shadows the original one, and the compiler
complains random_access_iterator lacks an actual declaration.

Solve this by checking for libc++ and forward-declaring the iterator types
we need within the same inline namespace.

Not backported from qtbase because QT_NO_STL does not exist in Qt 5.

Change-Id: I6742d540f6538a30aa060a4447c288cfb9cd781d
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
9 years agotest: Remove insignificant_test from tst_qaudioinput
Sergio Ahumada [Thu, 9 May 2013 16:20:25 +0000 (18:20 +0200)]
test: Remove insignificant_test from tst_qaudioinput

This test is now passing on all CI platforms.

Change-Id: I51df22c00d4e13a0b83230bc25e1b810ed183539
Reviewed-by: Simo Fält <simo.falt@digia.com>
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Mark tst_qgraphicsscene as insignificant only on Embedded (QWS)
Sergio Ahumada [Thu, 9 May 2013 16:15:39 +0000 (18:15 +0200)]
test: Mark tst_qgraphicsscene as insignificant only on Embedded (QWS)

This test has been passing lately for Linux, Windows and Mac OS X.

Change-Id: I994fe4c69c3520fe1eabf2d92224c8dd040ba3d9
Reviewed-by: Simo Fält <simo.falt@digia.com>
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Remove insignificant_test from tst_qdbusmarshall
Sergio Ahumada [Thu, 9 May 2013 16:05:44 +0000 (18:05 +0200)]
test: Remove insignificant_test from tst_qdbusmarshall

This test is now passing on all CI platforms.

Change-Id: I19f1b4bd4969bf9e396d696a273d214af16ebd9a
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Simo Fält <simo.falt@digia.com>
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Remove insignificant_test from tst_qgraphicsanchorlayout
Sergio Ahumada [Thu, 9 May 2013 15:19:22 +0000 (17:19 +0200)]
test: Remove insignificant_test from tst_qgraphicsanchorlayout

This test is now passing on all CI platforms.

Change-Id: Idd094cdbc2a21f3e5840366f8f47b9a187331023
Reviewed-by: Simo Fält <simo.falt@digia.com>
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Mark tst_qgraphicsanchorlayout1 as insignificant only on Mac OS X
Sergio Ahumada [Thu, 9 May 2013 14:59:35 +0000 (16:59 +0200)]
test: Mark tst_qgraphicsanchorlayout1 as insignificant only on Mac OS X

This test has been passing lately for Linux and Windows.

Task-number: QTQAINFRA-574
Change-Id: I6f21bf40d93adc8e359cf9fa0feea7fd4140736e
Reviewed-by: Simo Fält <simo.falt@digia.com>
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Mark tst_qgraphicswidget as insignificant only on Mac OS X
Sergio Ahumada [Thu, 9 May 2013 13:02:46 +0000 (15:02 +0200)]
test: Mark tst_qgraphicswidget as insignificant only on Mac OS X

This test has been passing lately for Linux and Windows.

Task-number: QTQAINFRA-574
Change-Id: Ie658d90cbc9eb2b68a963b1f1de06ca37f44047e
Reviewed-by: Simo Fält <simo.falt@digia.com>
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Mark tst_qinputcontext as insignificant only on Mac OS X
Sergio Ahumada [Thu, 9 May 2013 12:58:20 +0000 (14:58 +0200)]
test: Mark tst_qinputcontext as insignificant only on Mac OS X

This test has been passing lately for Linux and Windows.

Task-number: QTQAINFRA-574
Change-Id: Ib1d44ce284e3d0dcfa539376928cb41516a2059a
Reviewed-by: Simo Fält <simo.falt@digia.com>
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Remove insignificant_test from tst_qitemdelegate
Sergio Ahumada [Thu, 9 May 2013 12:45:36 +0000 (14:45 +0200)]
test: Remove insignificant_test from tst_qitemdelegate

This test is now passing on all CI platforms.

Change-Id: I06b659a750dcc76858287eb8336aa5228298d6e5
Reviewed-by: Simo Fält <simo.falt@digia.com>
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Remove insignificant_test from tst_qmacstyle
Sergio Ahumada [Thu, 9 May 2013 12:39:56 +0000 (14:39 +0200)]
test: Remove insignificant_test from tst_qmacstyle

This test is now passing on all CI platforms.

Change-Id: I421a7f541e24f1f1ebad8a68cc599bbdd38a767a
Reviewed-by: Simo Fält <simo.falt@digia.com>
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Mark tst_qpainter as insignificant only on Mac OS X
Sergio Ahumada [Thu, 9 May 2013 12:31:09 +0000 (14:31 +0200)]
test: Mark tst_qpainter as insignificant only on Mac OS X

This test has been passing lately for Linux and Windows.

Task-number: QTQAINFRA-574
Change-Id: I7b92c105bfa4fc00c92a00d4d731cabc878bb073
Reviewed-by: Simo Fält <simo.falt@digia.com>
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Remove insignificant_test from tst_qscriptextensionplugin
Sergio Ahumada [Thu, 9 May 2013 11:41:21 +0000 (13:41 +0200)]
test: Remove insignificant_test from tst_qscriptextensionplugin

This test is now passing on all CI platforms.

Change-Id: Icd24090b782738c3ab2635d41560eaf0d6b7ba1a
Reviewed-by: Simo Fält <simo.falt@digia.com>
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Mark tst_qsettings as insignificant only on Windows
Sergio Ahumada [Thu, 9 May 2013 10:08:05 +0000 (12:08 +0200)]
test: Mark tst_qsettings as insignificant only on Windows

This test has been passing lately for Linux and Mac.

Task-number: QTQAINFRA-574
Change-Id: I9dfc0c5b79a019e70fba04502050b35997384928
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Mark tst_qstylesheetstyle as insignificant only on Mac OS X
Sergio Ahumada [Thu, 9 May 2013 09:28:52 +0000 (11:28 +0200)]
test: Mark tst_qstylesheetstyle as insignificant only on Mac OS X

This test has been passing lately for Linux and Windows.

Task-number: QTQAINFRA-574
Change-Id: I7a3001381df805c9424b694197a92b467d69c1bc
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agoFix jump and property changes on first move when dragging a Flickable.
Andrew den Exter [Tue, 12 Mar 2013 06:04:54 +0000 (16:04 +1000)]
Fix jump and property changes on first move when dragging a Flickable.

Fixes a regression whereby on the first mouse move the contentItem was
moved the total distance from the touch point to where the drag
distance was exceeded. For large drag thresholds this causes a
noticeable jump.

Task-number: QTBUG-30032

(cherry picked from qtquick1 3a80424aeae19e838be03aa12a5243911ec3f020)

Change-Id: I92c119d27dc2e22203484f9ada5978697d171957
Reviewed-by: Martin Jones <martin.jones@jollamobile.com>
Reviewed-by: Alan Alpert (Personal) <416365416c@gmail.com>
9 years agoCheck that 3DNow! instructions are supported before including mm3dnow.h.
Raphael Kubo da Costa [Thu, 9 May 2013 22:40:23 +0000 (01:40 +0300)]
Check that 3DNow! instructions are supported before including mm3dnow.h.

Include mm3dnow.h the way we already include the (S)SSE{3,4} and AVX
instruction headers: by checking that the compiler is set to support those
instructions besides verifying that QT_HAVE_3DNOW evalues to true (which
only means the compiler supported them when the build was being configured).

This has not caused a problem in most cases so far because GCC protects its
mm3dnow.h header with an #ifdef __3dNOW__ (contrary to what it does on, say,
pmmintrin.h). clang's mm3dnow.h, on the other hand, does not have that check
and can fail if -march is set to, for example, i386 or i486.

Not backported from qt5 because qtbase because qsimd_p.h does not include
the 3DNow! headers after commit a1b30b49ef09bef2e97b9a0622bf7ad622678fee.

Change-Id: I15ab5e936c71c55f89c3f25777ab27fbd262e9cd
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
9 years agoUkrainian translation update
Victor Ostashevsky [Thu, 4 Apr 2013 15:34:49 +0000 (17:34 +0200)]
Ukrainian translation update

Tiny fixes, mostly about copyright years.

Change-Id: I9223eaabccb1ce93f1688f3512344dbcfb05aa6a
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
9 years agoFix clang warning about bool to pointer conversion.
Jędrzej Nowacki [Thu, 17 Nov 2011 13:42:48 +0000 (14:42 +0100)]
Fix clang warning about bool to pointer conversion.

The operation is legal from C++ perspective but only for "false" as it
is guaranteed to be 0. Anyway returning 0 instead of "false" is
logical and it follows coding style used in the modified functions.

(cherry-picked from qt5 commit 50995e6936c5ca75a0a8a48c29bbbedd2cc60567)
Change-Id: Ia09758e8d28599097f5c40eb24722890508afdbc
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
9 years agotest: Remove insignificant_test from tst_qsslsocket_onDemandCertificates_member
Sergio Ahumada [Tue, 7 May 2013 21:51:20 +0000 (23:51 +0200)]
test: Remove insignificant_test from tst_qsslsocket_onDemandCertificates_member

This test is now passing on all CI platforms.

Change-Id: I526a92a2bd94fff03b58f63df0119db0d124ba18
Reviewed-by: Simo Fält <simo.falt@digia.com>
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@digia.com>
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Mark tst_qtableview as insignificant only on Windows
Sergio Ahumada [Wed, 8 May 2013 09:02:05 +0000 (11:02 +0200)]
test: Mark tst_qtableview as insignificant only on Windows

This test has been passing lately for Linux and Mac.

Task-number: QTQAINFRA-574
Change-Id: I4e1a95d66596b29a926a92dae325fd2fbb114d81
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Remove insignificant_test from tst_qtimeline
Sergio Ahumada [Wed, 8 May 2013 08:44:36 +0000 (10:44 +0200)]
test: Remove insignificant_test from tst_qtimeline

This test is now passing on all CI platforms.

Task-number: QTQAINFRA-574
Change-Id: I6a536c7c4c33817544bff1d933d6f4a05f1631c2
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Mark tst_qwizard as insignificant on Windows and Mac OS X
Sergio Ahumada [Wed, 8 May 2013 08:32:50 +0000 (10:32 +0200)]
test: Mark tst_qwizard as insignificant on Windows and Mac OS X

This test has been passing lately on Linux.

Change-Id: I3c7e5cf4214899a65f4eccd15a0fd49e822675d8
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Mark tst_xmlpatterns as insignificant only on Windows
Sergio Ahumada [Wed, 8 May 2013 08:24:06 +0000 (10:24 +0200)]
test: Mark tst_xmlpatterns as insignificant only on Windows

This test has been passing lately for Linux and Mac.

Task-number: QTQAINFRA-574
Change-Id: I507a6b4220a3e3ded7bcfed2aded1e15bf53958a
Reviewed-by: Janne Anttila <janne.anttila@digia.com>
9 years agotest: Remove insignificant_test from tst_qlocalsocket
Sergio Ahumada [Tue, 7 May 2013 22:35:24 +0000 (00:35 +0200)]
test: Remove insignificant_test from tst_qlocalsocket

This test is now passing on all CI platform

Change-Id: Ib16a0e1e914ae85271f751f66bdece7e4ec851b6
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Simo Fält <simo.falt@digia.com>
9 years agotest: Mark tst_qsocketnotifier as insignificant only on Mac OS X
Sergio Ahumada [Tue, 7 May 2013 22:20:55 +0000 (00:20 +0200)]
test: Mark tst_qsocketnotifier as insignificant only on Mac OS X

This test has been passing lately for Linux and Windows.

Task-number: QTQAINFRA-574
Change-Id: Ic4d96c3645d6d10227ad1ce957f27a2079e9b834
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Simo Fält <simo.falt@digia.com>
9 years agotest: Remove insignificant_test from tst_qsocks5socketengine
Sergio Ahumada [Tue, 7 May 2013 22:16:10 +0000 (00:16 +0200)]
test: Remove insignificant_test from tst_qsocks5socketengine

This test is now passing on all CI platforms.

Change-Id: I0d669849e0b17f42e78c1864ea3ba9f070560093
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Simo Fält <simo.falt@digia.com>
9 years agotest: Mark tst_qtcpserver as insignificant only on Windows
Sergio Ahumada [Tue, 7 May 2013 22:10:19 +0000 (00:10 +0200)]
test: Mark tst_qtcpserver as insignificant only on Windows

This test has been passing lately for Linux and Mac.

Task-number: QTQAINFRA-574
Change-Id: I5c4cbea784af6e879ccd6a02aaf77f9debad594f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Simo Fält <simo.falt@digia.com>
9 years agotest: Remove useless insignificant_test mark tst_qudpsocket
Sergio Ahumada [Tue, 7 May 2013 22:00:31 +0000 (00:00 +0200)]
test: Remove useless insignificant_test mark tst_qudpsocket

The mark is actually in qudpsocket/test/test.pro

Change-Id: I4b4d9843b5643e15d988ceabfb9abca4f8cf77a2
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Simo Fält <simo.falt@digia.com>
9 years agotest: Remove insignificant_test from tst_qsslsocket_onDemandCertificates_static
Sergio Ahumada [Tue, 7 May 2013 21:54:53 +0000 (23:54 +0200)]
test: Remove insignificant_test from tst_qsslsocket_onDemandCertificates_static

This test is now passing on all CI platforms.

Change-Id: Ia3f4e9dd863e04758e1c794f29142e4417d044d2
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Simo Fält <simo.falt@digia.com>
9 years agotest: Remove useless insignificant_test mark
Sergio Ahumada [Tue, 7 May 2013 21:24:13 +0000 (23:24 +0200)]
test: Remove useless insignificant_test mark

The mark is actually in qnetworkreply/test/test.pro

Change-Id: I391c3863b6a98f485b8e6b585f346bf06820999e
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Simo Fält <simo.falt@digia.com>
9 years agotest: Remove insignificant_test from tst_qhttpnetworkconnection
Sergio Ahumada [Tue, 7 May 2013 21:20:18 +0000 (23:20 +0200)]
test: Remove insignificant_test from tst_qhttpnetworkconnection

This test is now passing on all CI platforms.

Task-number: QTQAINFRA-574
Change-Id: I8a08beef2f9876d88b7b903cda445ec5decce706
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Simo Fält <simo.falt@digia.com>
9 years agotest: Mark tst_qnetworksession as insignificant only on Windows
Sergio Ahumada [Tue, 7 May 2013 21:04:29 +0000 (23:04 +0200)]
test: Mark tst_qnetworksession as insignificant only on Windows

This test has been passing lately for Linux and Mac.

Task-number: QTQAINFRA-574
Change-Id: I6ed79e9bf166a17226b3196a90b47a3fd48912b2
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Simo Fält <simo.falt@digia.com>
9 years agoCompile in C++11 mode: narrowing conversion problems
Thiago Macieira [Fri, 20 Jan 2012 18:46:06 +0000 (16:46 -0200)]
Compile in C++11 mode: narrowing conversion problems

Matches qtbase commit a13b17e42a7ca146ec9503285df546d4ade22c91
Change-Id: Ie554499f40c03c715d527f2c1624b97fafe5ce8c
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
9 years agoMake QContiguousCache with zero capacity not crash
Thiago Macieira [Tue, 25 Sep 2012 14:30:48 +0000 (16:30 +0200)]
Make QContiguousCache with zero capacity not crash

These containers don't make sense and will just result in no action
being taken (all items added will simply be discarded), but it
shouldn't crash due to a division by zero.

Task-number: QTBUG-27339
Change-Id: Ib9acf5c0a9a826e6853e7beaf5e56511fde98dc6
(cherry-picked from qt5 commit 320c4e31e124f99601399d00935362b587c77510)
Reviewed-by: Robin Burchell <robin+qt@viroteck.net>
9 years agoDon't bypass overwritten [set]data() methods in the proxy.
Volker Krause [Sat, 4 May 2013 16:26:25 +0000 (18:26 +0200)]
Don't bypass overwritten [set]data() methods in the proxy.

By calling itemData() of the source model directly, the result cannot
contain data provided by the proxy model itself. The base class
implementation however will call data() on the proxy instead.

Cherry-picked from qtbase/96e3c2bcbfedc8b5cb8fc099229a02a1fa335c21.

Change-Id: I7e8b65ab045382089c577d9832edc1555b71419e
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
9 years agoMake Qt 4.8 also compile with the LSB SDK on Linux
Thiago Macieira [Sun, 5 May 2013 21:24:13 +0000 (14:24 -0700)]
Make Qt 4.8 also compile with the LSB SDK on Linux

Generate worse code.

This is similar to Qt5 commits 46bfd84fdc24fa3e3e721a5dda6cfbebe75be073
and c69106d1bf66d64e046c4f639542457151187dfc.

Task-number: QTBUG-25336
Change-Id: Idd61c23d2b528db9d8c7895df008b9c1b0e2a2a4
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
9 years agoFix QtScript crash on 64bit with JIT.
Friedemann Kleint [Mon, 28 Jan 2013 14:15:34 +0000 (15:15 +0100)]
Fix QtScript crash on 64bit with JIT.

During back-trace generation we calculate a code offset towards JIT generated
code. Using JITCode::offsetOf() will crash/assert if the offset doesn't fit
into 32 bits, because the generated code can only encode relative offsets
in 32-bits and not 64-bits. However in this context - backtrace generation -
we just want to calculate the offset and are not interested in this
architectural limitation, therefore we can just calculate the offset ourselves
using the fully sized uintptr_t.

Initial-patch-by: Simon.Hausmann@digia.com
Task-number: QTCREATORBUG-8629
Task-number: QTBUG-23463

Change-Id: I0efadd5ed20855409122e1fcc9236fdfbc4f62a4
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
(cherry picked from qtscript/3b26f6f6643978c9d041a6267dc88e2a59ced763)
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
9 years agoQLocale: Add fr_MA locale
El Mehdi Fekari [Thu, 2 May 2013 14:04:09 +0000 (16:04 +0200)]
QLocale: Add fr_MA locale

French_Morocco locale was introduced in CLDRv22.1 and already exists in Qt5

Change-Id: I4f0452f421cf074eb2c4fae22351d430d5f02e42
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
9 years agomake sure qmake finds itself in CWD on windows
Oswald Buddenhagen [Thu, 25 Apr 2013 14:44:30 +0000 (16:44 +0200)]
make sure qmake finds itself in CWD on windows

and consequently that it finds qt.conf.

Task-number: QTBUG-30583
Change-Id: I48441477e941d9609270d6e5e1b405127c0c0aca
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
(cherry picked from qtbase/e7c3b21b7c1ed9247cb2f9eeddc154534a038549)

9 years agoBring in WebKit r74029 to fix the build with clang.
Raphael Kubo da Costa [Sun, 28 Apr 2013 18:58:04 +0000 (21:58 +0300)]
Bring in WebKit r74029 to fix the build with clang.

------------------------------------------------------------------------
r74029 | cwzwarich@webkit.org | 2010-12-14 18:57:14 +0200 (Tue, 14 Dec 2010) | 12 lines

Reviewed by Eric Seidel.

Clang fails to build the JSC interpreter
https://bugs.webkit.org/show_bug.cgi?id=51016

Clang does not allow indirect gotos out of scopes with cleanup. GCC 4.2 allows
them, but it does not correctly generate the cleanup, causing a leak if the
cleanup decrements a reference count.

* interpreter/Interpreter.cpp:
(JSC::Interpreter::privateExecute): Put an Identifier into its own scope.
------------------------------------------------------------------------

Cherry-picked from qtscript/c85662167123b433a69a662212c67ce03e44a25a.

Change-Id: I21ee882961264d131cebdee04cb641c0d0de3d18
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
9 years agoFixed -dbus-linked when cross-compiling from Windows
Matt Fischer [Thu, 28 Feb 2013 16:52:12 +0000 (10:52 -0600)]
Fixed -dbus-linked when cross-compiling from Windows

Since Windows doesn't have pkg-config, it doesn't get -ldbus-1
added to the command line automatically like Linux does.  Code
was present to deal with this case, however it was only configured
to work when native-compiling Qt.  The flag was not added when
cross-compiling, meaning that -dbus-linked did not work correctly
in that case.  This patch changes the code to add the flag properly
in both cases.

Change-Id: I67881643bd658161f4929f3932859ccf636ca7a9
Task-number: QTBUG-29984
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
(cherry-picked from qtbase/c82b0815ec80b3750a110d47c68142bb448ddf56)
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>