The webkit AtomicallyInitializedStatic and Qt's Q_GLOBAL_STATIC can
deadlock on the Mac, as the mac compiler inserts calls to
__cxa_guard_acquire and __cxa_guard_release around initialisation of
local statics.
In Q_GLOBAL_STATIC case, this is the QGlobalStaticDeleter local static
Whereas webkit AtomicallyInitializedStatic is a local static variable
in any case.
Problem is triggered because webkit constructs QNetworkConfigurationManager
inside the constructor of a local static - networkStateNotifier
And the generic bearer plugin calls QNetworkInterface::allInterfaces
in the bearer thread, which needs an initialised Q_GLOBAL_STATIC.
Reviewed-by: Laszlo Agocs
(cherry picked from commit 38db40d9a2db44e47b0aabd9487284cd1106b353)
Task-number: QTBUG-24554
Change-Id: Id2c7818faefb46e66b0bbc30ce30595d46d53016
Reviewed-by: Zeno Albisser <zeno.albisser@nokia.com>
Config checks are done in the module.
We pick up the pkg-config stuff also in the module. There shouldn't be a
need to do this in configure anyway
Change-Id: I9ef73760511c6b684c6cd5dd13e7e581c588e7aa
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Remove its dependency on QtSvg by converting the icons to
(optimized) png files.
Change-Id: If63257938dc816d7ab76a3042b9ac88ef3d51b07
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
No longer base the implementation on a QWidget which is not
necessary when all that is required is a message window listening
to task-tray messages. Export a service function creating a message
window from the Windows native interface and use that.
Task-number: QTBUG-20978
Change-Id: I01d0faeac777df4eee802c51d2bc722fce814080
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
This variable is there for documentation, but its not being used anymore.
There are two libraries from wayland, libwayland-client and
libwayland-server. Add two variables that will be picked up in the
QtWayland module.
Change-Id: I16219092a4ed2e48080921ce39f61192b6abb711
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Change-Id: Id565fa1eb8fe13c62a93a5afa39a5701ce7b20ea
QPixmap::grabWidget is deprecated, which calls QWidget::grab() at present.
Reviewed-by: Gunnar Sletta <gunnar.sletta@nokia.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Added the error strings from PCRE to be picked up by lupdate,
to enable translations.
Change-Id: Iaeabde5d7a17f9a0273511e0741e67a097d23a98
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Doing so works fine on Linux as g++ version there silently ignores
this. However, the qcc toolchain for QNX barfs.
Change-Id: Ia236910adc09dc1653e4169e20476b69c2de62ab
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
GCC complained about undefined behaviour when
deleting subclasses of AbstractTester through
pointers to AbstractTester, and it's doing so
correctly...
Change-Id: Ie641281d8aafe32c5c9784e8aa39672ff0b699c7
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Commit b979956ec4 introduced a
distinction between rows that have a pending INSERT operation and
rows that have already been inserted in the database but still are
in the change cache. Both cases are rows that are not in the underlying
query. Unfortunately, we overlooked a case where the point of the test
is whether the row is in the query.
Change-Id: I0f58bed232d9336fed6e67c3d140fd580ec35868
Reviewed-by: Honglei Zhang <honglei.zhang@nokia.com>
Considering a subtable to be invalid when it was empty meant that it
returned an error causing all further hinting to be aborted. This means
that we get no hinting in certain cases with certain fonts.
Change-Id: I840c016dc85935b4f2c4373c66a79b4bd8b3a30d
Reviewed-by: Adrian Yanes <ext-adrian.yanes@nokia.com>
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@nokia.com>
Allow it to take advantage of QList optimisations for pointer types.
Change-Id: Iddbb9741efef43604e38fc3eeb08b522c0414e21
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@nokia.com>
Changing the naming scheme from Blackberry to QNX in line with pattern
of using platform names.
Change-Id: I048a6a18010bc932311d63c8dde19ccab97894c8
Reviewed-by: Robin Burchell <robin+qt@viroteck.net>
Reviewed-by: Nicolas Arnaud-Cormos <nicolas@kdab.com>
This was never a good idea since it has issues:
It doesn't really work with more than one update without
the client fetching the data. The same is unlikely to work reliable
since it involves two ipc roundtrips.
Instead we should have an extended QAccessibleEvent that contains
the data so that the bridges can decide how to pass on the data if
needed.
Change-Id: Iaf6b74f49586f7155909a6fe1a17137b71b31175
Reviewed-by: Jan-Arve Sæther <jan-arve.saether@nokia.com>
These unit tests do not depent on QtWidget.
Change-Id: I95526125c563885c0531da7ebfee06bca9a87b1c
Reviewed-by: Gunnar Sletta <gunnar.sletta@nokia.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Subclass QAccessibleEvent to give details what changed in the
state change.
Change-Id: I9005d311e85a3c8bfa6e062833fa6a8a7dc6a4a4
Reviewed-by: Jan-Arve Sæther <jan-arve.saether@nokia.com>
The reconstituteFieldMemberExpression static function was not usedf
anywhere.
Change-Id: Ide23045d7f7c3194a9725ae658df85562ba29ad5
Reviewed-by: Martin Smith <martin.smith@nokia.com>
qmake uses the variable name verbatim, which makes the generated
makefile look "interesting" without this cleanup.
Change-Id: If6c4f12563a0ee6429513fbfac534f40bb7c0b97
Reviewed-by: Rohan McGovern <rohan.mcgovern@nokia.com>
Commit 8e5eb1bddc introduced the assumtion that mkspecs are
immediately below the mkspecs directory itself. This is not true for
e.g. unsupported/blackberry-armv7le-qcc.
This commit restores qmake's ability to find the "root" of the mkspecs
collection no matter how deeply the actual mkspecs are nested.
Task-number: QTBUG-24665
Change-Id: I98faaf8e6ae7b8524277aea6c17e685e507e37b3
Reviewed-by: Sean Harmer <sh@theharmers.co.uk>
Reviewed-by: Joerg Bornemann <joerg.bornemann@nokia.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
this makes the "sysrootable" properties more magic, with the raw
versions being omitted from the qmake -query output and automatically
falling back to the "cooked" variant if there is no sysroot set.
this makes the "normal" qmake -query less noisy. this will become even
more obvious when i add more "overloads" of the properties.
Change-Id: I08000986427264ec6238c8fe0a77f5cecdbf1201
Reviewed-by: Joerg Bornemann <joerg.bornemann@nokia.com>
the feature is rather obscure and unlikely to be used by anyone.
Change-Id: I2dfb4ca4d5d1f210d385c013f46bc6389fd6ea2d
Reviewed-by: Joerg Bornemann <joerg.bornemann@nokia.com>
This changeset allows Qt to be built for QNX platforms in light of
commit d59e85d909.
Change-Id: Idf8e89cf1b0a5625ef7ee6397c223137fa151cdc
Reviewed-by: Rafael Roquetto <rafael.roquetto@kdab.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Add an implicit #include back in to avoid
needless SC and compilation breaks.
Add some deprecated method for codecForTr, until
all other modules are ported to mot use it anymore.
Change-Id: I5334b47a0c32819b9eb6b7203cc98ce4e6073a64
Reviewed-by: Kent Hansen <kent.hansen@nokia.com>
The app name has changed in Growl 1.3, but the id
stays the same. Also don't send notifications if
Growl is not running to follow recommendations.
Change-Id: I31ff7df272b4af1b4f1e4db80c47e7ba75038dec
Reviewed-by: Morten Johan Sørvig <morten.sorvig@nokia.com>
The test was assuming that "data()" is a special function in autotests,
but that hasn't been the case since early prototypes of testlib.
Change-Id: Ic24cf5dc539b55d12eba0a6ab17173e2ed698f21
Reviewed-by: Rohan McGovern <rohan.mcgovern@nokia.com>
- Use const_cast to avoid "deprecated conversion from string constant to
'char*'" warning when building argv arrays from string literals.
- Use Q_UNUSED to avoid warnings on unused local variables.
Change-Id: Idd2c8279adc102b6ebc6af7486ba26fe9ed4e7c1
Reviewed-by: Rohan McGovern <rohan.mcgovern@nokia.com>
Previously, property setting was randomized in order, which means that things
like: qproperty-foo: 4; qproperty-bar: 5; where foo may affect bar worked by
chance - or not at all - depending on the hash function.
Change-Id: Ifb9813ee72842cefb278cbedb644f24b91113f3f
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Make the various versions of the failure message align consistently so
that it's a little easier to compare the actual and expected values. Of
course, the value won't align nicely unless the "actual" and "expected"
strings are the same length, but at least this commit makes that
consistent across all versions of the message.
Change-Id: If9ce231df3b5d279a06f6458fdb5da0aa4586068
Reviewed-by: Rohan McGovern <rohan.mcgovern@nokia.com>
Add Q_IS_ENUM() macro to determine if a given type is an
enumeration. Use information from that in QMetaType::registerType()
to store whether custom registered metatypes are enums or not.
This information can then be accessed by calling
QMetaType::typeFlags(int type). This is used by the declarative
code to determine whether a custom type in a variant can be safely
cast to an integer, which is required to allow passing non-local
enums as signal/slot params.
Change-Id: I9733837f56af201fa3017b4a22b761437a3c0de4
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
The function is public, so it should validate input instead of crashing
Change-Id: Id67463b0b61ab74a76c1ede7f052bdbed37822b6
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
