On OS X 10.11 (El Capitan) the system menu bar can be automatically
hidden, in which case the menu bar height is reported to be 0 when
using the menuBarHeight API.
This resulted in failing to prepare an image for the system tray
icon item, making the tray item "invisible".
Instead we now use the [[NSStatusBar systemStatusBar] thickness]
API, which returns the correct height regardless of the menu bar
being hidden or not.
Task-number: QTBUG-48960
Change-Id: I208fb8df13754964a6f254cadfbff06dd56c6bab
Reviewed-by: Markus Goetz (Woboq GmbH) <markus@woboq.com>
Reviewed-by: Jake Petroules <jake.petroules@theqtcompany.com>
These variables are accessed from both the executing
thread as well as the thread waiting for results.
Note for some variables which threads access them.
Change-Id: I1c84ddff92585abb32341c42072106066e485f7e
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Since 7c6625b105 native OpenGL widgets
are able to render, but performance is very low. It also slows down
rendering for top-level OpenGL widgets.
Render-to-texture widgets prevent triggering unnecessary backingstore
painting when only the OpenGL content changes.
This patch makes possible to use that feature for native and top-level
widgets by calling "markDirtyOnScreen()" function only when necessary.
This patch also prevents native and top-level OpenGL widgets from
flickering during resizing.
Task-number: QTBUG-50381
Task-number: QTBUG-49620
Change-Id: I0c16ea644fb869a00772fd74e1709611e627bff3
Reviewed-by: Laszlo Agocs <laszlo.agocs@theqtcompany.com>
Build the QPlatformAccessibility support on WinCE7, so
that QtQuickControls, which depend on the availability of
QPlatformAccessibility, can be used on WinCE7 as well.
Change-Id: Ice7370b891a2963483bbfe1af8b219717541e64b
Reviewed-by: Björn Breitmeyer <bjoern.breitmeyer@kdab.com>
Reviewed-by: Jan Arve Sæther <jan-arve.saether@theqtcompany.com>
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@theqtcompany.com>
Remove QPrintPropertiesDialog::showEvent(). It just accepts
QShowEvent, it makes no sense. At the same time this method
overrides QDialog::showEvent(), which centers dialogs.
Change-Id: I11510feb4a946f347c19d6ac59874b4cb931020e
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
This can happen with a suitably complex use case
(See QTBUG-50263)
Change-Id: Ie8fa7b5872a902e802fda5795ade3369399ddb54
Task-number: QTBUG-50263
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@theqtcompany.com>
operator< between pointers is undefined unless the two pointers
point in the same array, which is not what QOrderedMutexLocker does.
Change-Id: Ia6594900cfa807a73f20e157ce896b4321a3d746
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
On some drivers (e.g. NVIDIA, tested on Windows) the
status check is apparently expensive. A simple FBO
test app rendering into the FBO on every frame (and
thus calling bind()) shows a 19-21% CPU usage. With
the patch this gets reduced to 0.8-1%.
The check is fairly useless too - whenever creating
attachments, the status is checked, so d->valid is
up-to-date.
Task-number: QTBUG-50496
Change-Id: Ie9f0db34e5c2bc1a1141c734ca0b318419013c0d
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@theqtcompany.com>
This is aligned with our coding style and it should have been this way from
the start.
Change-Id: I23a00eb220dd9f17d9239c811b556885a2c0186a
Reviewed-by: Fredrik de Vibe <fredrik.devibe@theqtcompany.com>
Reviewed-by: Jan Arve Sæther <jan-arve.saether@theqtcompany.com>
When the remote peer closed the connection, a read notification needs
to always be emitted, otherwise the higher layer does not get the
disconnected signal. From the other side, underlying QAbstractSocket
object could temporarily disable notifications from the engine at
any time. To avoid possible blocking of the socket, take a pending EOF
into account when the read notifications are re-enabled.
Change-Id: Iac9d4e2f790530be3500baf5a2000f1f63df5cc2
Reviewed-by: Ulf Hermann <ulf.hermann@theqtcompany.com>
These QPairs are larger than a void*, so holding them in QLists
is needlessly inefficient. Worse, the code could come to depend
on the fragile property of (inefficient) QLists that references
to elements therein never are invalidated.
Fix by holding them in QVector instead.
Change-Id: I3c205f5326cfd96482563078bdca1747d718457f
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Merge in the changes up to revision 306 of the lp:qtubuntu repo.
Change-Id: I55dcb9e06e0a3503d6abe4b0894d5ef5a785b6bb
Reviewed-by: Christian Stromme <christian.stromme@theqtcompany.com>
We need to have reasonable paths for loading icons and such things, just
as in any other Unix GUI environment. However there is some concern that
it would be too much of a behavior change if there was a theme by default,
so for now it's required to set the env var: QT_QPA_PLATFORMTHEME=generic
That works because QGuiApplicationPrivate::createPlatformIntegration() reads
the env variable and passes platformThemeName to init_platform().
Step 3 in init_platform() does not find a theme plugin by that name
(because QGenericUnixTheme is statically linked via libQt5PlatformSupport.a).
Then in step 4 it iterates the given platformThemeName plus any which
were returned from QPlatformIntegration::themeNames() (which in our case
will be an empty list) and calls createPlatformTheme() with each of
those, until something is returned. So,
QEglFSIntegration::createPlatformTheme() will be called with the
value of the QT_QPA_PLATFORMTHEME env var, and
QGenericUnixTheme::createUnixTheme() will create the generic, KDE or Gnome
theme depending on that value.
Change-Id: Id16b881819ba872830b019ab147b32fbc2156520
Reviewed-by: Laszlo Agocs <laszlo.agocs@theqtcompany.com>
The type Fetching is larger than a void*, so holding it in QList
is needlessly inefficient. Worse, the code could come to depend
on the fragile property of (inefficient) QLists that references
to elements therein never are invalidated.
Fix by holding it in QVector instead.
Also optimize the append site by liberal use of std::move(). This
code would greatly benefit from emplace_back(), but we can neither
assume it's present in std::vector nor do we require the necessary
C++11 features that would allow us to implement it in QVector, yet
(uniform init and, less so, variadic templates).
Change-Id: I50da0ffd557adff57477245d0e8c1fc1fec1ebc1
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
... just so you can observe their absence with QHash::value()
returning nullptr.
Instead, use find() + comparison to end() to detect presence
or absence.
Simplifies quite a bit of code.
Change-Id: Ifd7921bfc8102677ea345ae37d38da31b8105426
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
It failed to copy 'g'.
Fix by letting the compiler generate one.
Change-Id: Ie19fdacb8f27aef821be58c0b727c802d71bfe64
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
It's undefined behavior. Use std::less, which has
defined behavior.
Change-Id: I990d197590cf535c1cb5c055d0b6798e602446dc
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Code like this makes it impossible to provide a templated
QSharedPointer<T>(X*) ctor, so proactively rewrite to use
the proper default ctor, which has the same effect.
Change-Id: I2572e92b12804f873fac4927e93db83f796729f5
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Like xcb does since 1bcfc3de75.
For TranslucentBackground (ARGB32_Premultiplied) there is no change
in practice. For RGB32 we will promote to ARGB32_Premultiplied
but everything should behave like before, except that we do not rely
on undefined behavior anymore when QOpenGL/QuickWidget is in use. For
RGB16 this will enable QOpenGL/QuickWidget since there the hole
punching mechanism needs an alpha channel always.
Change-Id: Id04ea548cee245ec91642f1358a5e501b62ff64c
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@theqtcompany.com>
This caused the filename in messages to be one of the other arguments
that were specified instead of the actual filename.
Caused by the fact that mutations are possibly made to `args` in
`parseCmdLine` and thus the amount of items in `args` does not always
match the amount of items in `argv`.
Change-Id: Ief3716dde39dfdc949a5192e7f83d93cf90130f0
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
On Andorid 64 bit O.S.s the libs are not located anymore on dataDir/lib/
but they are located on dataDir/lib/<arch>/. Using nativeLibraryDir instead
of dataDir we'll get the right folder.
Change-Id: I40ef9513eb5b51f81032bd41f9943ab7b959ad51
Task-number: QTBUG-50394
Reviewed-by: Christian Stromme <christian.stromme@theqtcompany.com>
Fixes a regression appearing in 5.5 where QtWidgets-based file dialogs
showed up too small on Windows.
Change-Id: I5f74ab01d4368d67ab79d3a1b145fb63bdd69ca0
Reviewed-by: Shawn Rutledge <shawn.rutledge@theqtcompany.com>
These QPairs are larger than a void*, so holding them in QLists
is needlessly inefficient. Worse, the code could come to depend
on the fragile property of (inefficient) QLists that references
to elements therein never are invalidated.
Fix by holding them in QVector instead.
Change-Id: Ie8d8eff3448ada7aef5dfba7fc701a59821f5b54
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
When called from outside of qfontdatabase, the qfontdatabase mutex will
not be locked upon entry to qt_fallbacksForFamily, so we need to add
it to the exported version.
Change-Id: I16a21708d7cca15730c12de3dbeeaab13ffbd315
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@theqtcompany.com>
Recursion is alredy prevented by the emittedReadyRead member.
Disabling the read notifications also occurs when:
- a new chunk of data has arrived on an unbuffered socket;
- the buffer size limit has been reached on a buffered socket;
- pauseSocketNotifiers() has been called
Subsequent calls to read() or resumeSocketNotifiers() should re-enable
the notifications independently from possible nesting.
Change-Id: I4587265b8d3ed137516e08fbe92dce2f5eab508c
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Markus Goetz (Woboq GmbH) <markus@woboq.com>
Add a function to allow the users to easily run asynchronously Runnables for any thread
directly on Andoroid UI thread.
Change-Id: I631bf8a2c602e038039fec621ec01272af20a400
Reviewed-by: Christian Stromme <christian.stromme@theqtcompany.com>
QFileSelector::select(QUrl) will use ":" for scheme "qrc" to
call QFileSelector::select(QString). Scheme "assets" needs
to remain "assets:" for select(QString), otherwise it won't
recognize the file in "assets".
Following failed because it was passed as ":/qml/example.qml"
to select(QString):
select(QUrl("assets:/qml/example.qml"));
This will call select(QString) to:
select("assets:/qml/example.qml");
Change-Id: I6bdeed6bb67992498ae3b8e1273c20e70049381a
Task-number: QTBUG-50435
Reviewed-by: BogDan Vatra <bogdan@kdab.com>
From Qt 5.7 -> LGPL v2.1 isn't an option anymore, see
http://blog.qt.io/blog/2016/01/13/new-agreement-with-the-kde-free-qt-foundation/
Updated license headers to use new LGPL header instead of LGPL21 one
(in those files which will be under LGPL v3)
Change-Id: I046ec3e47b1876cd7b4b0353a576b352e3a946d9
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
Optimizes composition on platforms that may use 10-bit per color channel
formats.
Change-Id: Ib303c391d47795c79a4ba55d78dbb1c3c702d90a
Reviewed-by: Laszlo Agocs <laszlo.agocs@theqtcompany.com>
QJsonDocument stores the entire JSON tree in its binary format. Since
the binary format is the same on-disk as in-memory, it has a fixed
endianness. But when converting from QString to the little-endian
format, the code accidentally converted twice (from little endian, to
little endian), which resulted in a no-op and the string got stored as
big-endian.
It's like encrypting with double-ROT13.
No new testcase because tst_QtJson::removeNonLatinKey was already
failing and gets fixed by this commit.
Task-number: QTBUG-50419
Change-Id: I408dcb81ba654c929f25ffff1428cc79472bbe13
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
Detect premultiplied backing store and upload it without conversion
and adjust blend to not multiply alpha again.
Task-number: QTBUG-50381
Change-Id: I51939c4f82fbb0c48eaca6e2475bf4cf3722bc2d
Reviewed-by: Laszlo Agocs <laszlo.agocs@theqtcompany.com>
Reviewed-by: Gunnar Sletta <gunnar@sletta.org>
... use QByteArray::append(char*,int) instead.
Also cache the return value of the out-of-line function
xcb_get_property_value_length().
Saves ~120b in text size, and a heap allocation.
Change-Id: I4d1deafdcd3345f2b7dfbf8c45702cfee733a269
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Required to change an erase() loop into std::remove_if to
avoid running into quadratic behavior.
While at it, made QXcbXSettingsCallback a proper struct,
used aggregate initialization, and ported another loop to
C++11 range-for.
Saves ~0.5KiB in text size on optimized GCC 5.3 Linux AMD64
builds, and a lot of heap allocations.
Change-Id: I228bb853519ed2590375dc511e527f47bb8daa34
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Since 1bcfc3de75 every image has alpha
in XCB even if it is not used by window. This patch prevents clearing
the alpha channel on every repaint when window is not transparent.
Change-Id: Icfdd39f8636401906ace77ea3c2fa25386979d5f
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@theqtcompany.com>
Previously the hardware and camera button handler were guarded inside a
Q_OS_WINPHONE which does not apply to Windows 10.
Instead use WINAPI_PARTITION_FAMILY like on other places, this covers
Windows Phone 8.1 as well as Windows 10.
To find windows.phone.ui.input.h at build time the Mobile Extension
directory needs to be added to the include paths inside qmake. On
runtime we need to check whether we have hardware buttons or not. In
case they exist, register the handlers, otherwise skip registration.
Skipping also helps to keep WACK succeeding.
Task-number: QTBUG-50427
Change-Id: Ibeae15dbde12553cebd2b73b1a40b754c014f426
Reviewed-by: Oliver Wolff <oliver.wolff@theqtcompany.com>
This patch prevents visibility of native widgets which for some reasons
has width or height of 0.
Due to async expose event in Qt5 we must force "hide_sys()" during
resizing for widgets with WA_OutsideWSRange flag. This avoid problems
for non-native and visible widgets which are converted to native and
then resized:
child->winId();
child->resize(0, 0);
Task-number: QTBUG-49445
Change-Id: Ied62a4f253f89447941b2dc03316db8c168f4bb5
Reviewed-by: Ulf Hermann <ulf.hermann@theqtcompany.com>
Use ExposeCompressor for mapping the window to prevent duplicating
expose event when window is mapped.
Remove sending expose event from configure notivy event.
Change-Id: I6836c06047ffebd65775aa8aca90750afba6efe8
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@theqtcompany.com>
This adds a way to resume reading from a stream after a ReadPastEnd error.
This is done by introducing a stream read transaction mechanism that keeps
read data in an internal buffer and rolls it back on failure.
[ChangeLog][QtCore] Added QDataStream startTransaction(),
commitTransaction(), rollbackTransaction(), abortTransaction()
functions to support read transactions.
Task-number: QTBUG-44418
Change-Id: Ibf946e1939a5573c4182fea7e26608947218c2d9
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Irrelevant once we get to C++11 (so we can revert this in 5.7), but
division's rounding direction is implementation defined when either
operand is negative [0]. The prior code assumed C++11's truncation
(a.k.a. round towards zero), but rounding may be downwards instead.
[0] http://en.cppreference.com/w/cpp/language/operator_arithmetic#Multiplicative_operators
Change-Id: I2b6b27e1cf629def48b25433e81b9ed8230d8795
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Avoid returning fractional scale factors, which we
do not support. This matches the pixelDensity()
implementation for the Xcb and Windows platform plugins.
Change-Id: I79156e802a0a436b9344f12d35d1f4861e20e7fa
Reviewed-by: Mitch Curtis <mitch.curtis@theqtcompany.com>
There are three deprecated language codes that Java still uses for the
locale so we need to account for these inside QLocale by mapping them to
the right language.
Task-number: QTBUG-49632
Change-Id: Ib66b3f2763e085f7384228f2490b048bb56be259
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
Same as the SSE4.2 implementation: use the (optional) ARMv8 crc32[bhwd]
instruction to calculate hashes for strings. For Aarch64, support for
the instruction is dynamically detected. For a 32bit ARM binary, dynamic
detection is only done when the compiler is explicitly told to target
ARMv8. When telling the compiler to target an other/older version, the
crc32 code is not compiled.
Change-Id: I51ebc1a4545dede4988247e75043f29a64c2a6c5
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Added a new overload function that allows the developer to
specify the desired precision. Until 6.0, it will require
the symbol and precision to be passed to it. Once Qt is at
version 6.0, it will replace the overload function that
requires a value and optionally a symbol.
[ChangeLog][QtCore][QLocale] Added an overload for toCurrencyString()
that allows the decimal precision to be specified.
Change-Id: I1fb7dde3583f46de2ed20ec2a7abaeca23a903ef
Task-number: QTBUG-46595
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Passing nullptr as the second argument of memcpy/memmove
constitutes undefined behavior, even if the length argument
is zero at the same time.
Fix by protecting mem{cpy,move,chr} from nullptrs.
Found by UBSan:
qtbase/src/corelib/io/qiodevice_p.h:105:33: runtime error: null pointer passed as argument 2, which is declared to never be null
qtbase/src/corelib/io/qiodevice_p.h:175:53: runtime error: null pointer passed as argument 2, which is declared to never be null
Change-Id: I979158b0a74169ca4eb459928398ebc40f77dfb5
Reviewed-by: Alex Trotsenko <alex1973tr@gmail.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The code-path used by non-OSX platforms were not passing on the native
scan-code, virtual key, and modifiers, breaking QKeyMapper::possibleKeys
in the process.
Task-number: QTBUG-50360
Change-Id: Idc36380a234f0a37d016dbeaca594aeb82c496d2
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
Reviewed-by: Błażej Szczygieł <spaz16@wp.pl>
When closing a popup dialog using a shortcut, when the popup
was originally opened using a shortcut, the closing-shortcut
would interfere with the state of the first shortcut, and we
ended up sending a key event for the original shortcut.
Task-number: QTBUG-50360
Change-Id: I62e5ddb9ca43b28519ede629775bc0d7598dccc4
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
... so we can drop the user-defined copy special member
functions and the destructor.
As a side-effect, enables the move special member functions,
which were previously inhibited by the presence of the
now-removed functions, and makes the copy constructor safe
for self-assignment.
Change-Id: I430f83a6a08b1f5ee94b52f52e4d80fa1139d1c1
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
GCC 6 is able to detect when you use a placement new of an object in a
space that is too small to contain it.
qvariant_p.h: In instantiation of ‘void v_construct(QVariant::Private*, const T&) [with T = QRectF]’:
qvariant_p.h:142:9: error: placement new constructing an object of type ‘QRectF’ and size ‘32’ in a region of type ‘void*’ and size ‘8’ [-Werror=placement-new]
new (&x->data.ptr) T(t);
^~~~~~~~~~~~~~~~~~~~~~~
This happens even for the false branch of a constant expression (the
enum). So split the v_construct function in two pairs, one pair for
being able to use the internal space and one pair not so.
Change-Id: Ibc83b9f7e3bc4962ae35ffff1425ed898f279dea
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@theqtcompany.com>
v_construct does what we want, so use it. This is required for the next
commit, which solves a GCC 6 warning issue.
Change-Id: Ibc83b9f7e3bc4962ae35ffff1425ed5f035f631a
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@theqtcompany.com>
Otherwise it's possible to get garbage for primitive types (trivially
constructible) under some conditions.
Change-Id: I408dcb81ba654c929f25ffff142885fc62395948
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@theqtcompany.com>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Make a private member mutable to enable default assignment operator.
Change-Id: I1216875c186ed800e07c6b41a5bae18c3b71b2fa
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Introduce a transaction mechanism that gives the ability to read the
data atomically. Current implementation supports transactions for both
types of devices. For sequential devices, it records the whole input
stream during transaction. For random-access devices, device position
is saved when transaction starts. If an error occurs, the application
may be able to recover the input stream by rolling back to the start
point.
Also, QIODevice::peek() was rewritten to make use of transactions
internally. The replacement of QIODevicePrivateLinearBuffer by
QRingBuffer is closely entangled with that, which makes it unfeasible
to do separately.
Bump the TypeInformationVersion field in qtHookData, to notify the
Qt Creator developers that the offset of QFilePrivate::fileName was
changed and dumpers should be adapted.
[ChangeLog][QtCore] Added QIODevice's startTransaction(),
commitTransaction(), rollbackTransaction(), isTransactionStarted()
functions to support the read transactions.
Task-number: QTBUG-44418
Change-Id: I3564b343ebeeaaf7c48a1dcdb7ef0a7ffec550f2
Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
the main objective was to fix the bootstrap modules in framework builds.
bootstrapped modules which "borrow" headers from "proper" modules can
specify this in a clean way now.
a side effect of this is that the bootstrap-dbus module now has its own
syncqt call.
most includepath-related setup from qt_module_pris.prf was moved to
qt_module_headers.prf.
Change-Id: Ie0d8192cfac1a8cdae0ddd0bc0cd8c3092b1e85b
Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
This fixes the problem as long as we have C++ 11's semantics for
in-function statics. The Q_GLOBAL_STATIC_WITH_ARGS() is a non-trivial
wrapper whose destructor gets unloaded with the plugin, leading to a
crash-on-exit if the global it wraps outlives the plugin. The plain
(in-function) static manages to avoid this by avoiding the wrapper.
As the static was only needed in one place, this proves a sufficient
solution to the plugin-unload problem *in this case*.
Task-number: QTBUG-45891
Change-Id: I599fbee0b55ece4dceb4bf7202db374b507f5388
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
When QDateTime::addDate() and friends sanitize their end-state, they
were using the DST status of their start-state (if known) to control
it. This lead to misguided results and, in particular, inconsistent
results given that a raw-constructed QDateTime comes into being
ignorant of its DST, while a .toLocalTime() one knows its DST.
Furthermore, the code to do this was triplicated, tricky and poorly
explained. So pull it out into a local static function and explain
what it's doing, and why, more clearly and only once.
Task-number: QTBUG-49008
Change-Id: Ia4bb3c5e9267fff8bb963ea705267998218ed623
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@theqtcompany.com>
The same message is already printed on UDP and buffered TCP.
Change-Id: I533baf97fe7e1359a38db3c3eb31d0463fe158e5
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The code was checking the write buffer in a branch for unbuffered sockets.
Change-Id: I4dc722e7f182562332c3d167e6ebbab7f53f51e7
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
... such as GCC.
truncate(0) was chosen because it is already used throughout
the function. Using clear(), say, which is inline, produces
slightly more code.
As is, saves 160b in text size on optimized GCC 4.9 Linux
AMD64 builds.
Change-Id: I415d09ad2b4547f1d69f78d85e2aa1c1f9a17ed3
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Saves just 168b in text size on optimized GCC 4.9 Linux
AMD64 builds, but most for loops are in non-Linux code.
Change-Id: I4f20a65c2e4953011308ff831c9e8fa37a25274b
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
It was already movable, so it's BC. Only needed to add
nothrow member-swap and nothrow move assignment.
Change-Id: Iefedb877078da8ee075eb67185eef221143ddec1
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
For HSL colors, the (heavily over-parenthesized) code compared hslHue
values mod 36000, _and also_ for equality. Of course, the second
comparison is dead code.
This was a cut'n'paste error from the non-HSL comparison.
So, remove it. And also some of the over-parenthesization.
There are no tests for operator==, and the HSL color comparison is
completely opaque for me, so I'm not going to write any for this
trivial change.
Change-Id: I74572273730cb5cc9f427c524c268ba3f90304c1
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Saves a bit more than 0.5KiB in text size on optimized
GCC 4.9 Linux AMD64 builds.
Change-Id: I3b7e4751c4799c3e2c9f8f23b769e1659d863579
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
It just prevents the compiler from synthesizing move
special member functions, something that is very much
desired, seeing as there's a QVector member.
Change-Id: I4daabb380cd73dcacf3f514827b84562767a7a20
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This partially reverts commit e486d69133.
It broke too many users, even though all of them deserved to be broken.
The new functionality will be provided by differently-named functions,
where possible (problem: equality operators).
I did not revert the fix for the off-by-one error in
tst_qtextdocumentfragment.cpp.
I also didn't revert the change in the inequality relational operators,
since for all strings s1, s2 and s2' where s2' is s2 truncated at the
first NUL, s1 < s2 ⟺ s1 < s2' (since NUL < c for any c != 0), and,
trivially, for ≤, >, ≥, too. This does not hold for = and ≠, of course,
since "foo\0bar" ≠ "foo".
[ChangeLog][Important Behavior Changes][EDITORIAL] Reverted: All
conversions from QByteArray to QString now preserve embedded NULs...
Change-Id: If4b47048b39ae5be6ed08e6d91809626a67ea7f5
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Passing nullptr as the 2nd argument of memcpy
constitutes undefined behavior.
Fix by protecting the block with 'if (len)',
which, presumably, is the only valid case
where 'data' may be nullptr.
Change-Id: I7647d7e0808b1f26444ea3cf8bbf5cda9ddc9e6c
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Apparently, it is considered valid to call the function with
'data' set to nullptr, and 'len' to zero. But doing so
invokes undefined behavior because nullptr is passed to
fwrite().
Fix by protecting the loops with 'if (len)'.
Found by UBSan:
qtbase/src/corelib/io/qfsfileengine.cpp:732:84: runtime error: null pointer passed as argument 1, which is declared to never be null
Change-Id: Idfe23875c868ebb21d2164550de3304d2f01e9df
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
... each time around the loops. Cache it.
Also use QStringLiteral, since the string will never be
modified.
Also saves 96b in text size on optimized GCC 4.9 Linux
AMD64 builds.
Change-Id: I0269586235da18f3073a553739561ea7db6356e8
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
For almost the whole of its life time, 'realname' has the same
prefix, but the old code used the same string-builder expression
to construct it over and over again. If QStringBuilder would
re-use the capacity of the LHS object, that would boil down to
just a bit more copying of characters than necessary. But
assigning a QStringBuilder expression to a QString works by
implicit conversion of the QStringBuilder expression to a
QString, followed by move-assigning the new QString into the
old.
The new code keeps the common prefix around, resetting
'realname' to that prefix with truncate, only appending the
varying suffixes. In this way, one memory allocation per
assignment is saved (ignoring a potentially required capacity
increase in one of the appends here), and also some out-of-line
QString dtor calls, since op+=(QString&, QStringBuilder...)
doesn't create a temporary QString.
Also saves ~1KiB in text size on optimized GCC 4.9 Linux AMD64
builds.
Change-Id: I0872a69c9111d7218567f06f8fefb010f2430532
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
It turns out that f558bde was not enough to stop a crash when
trying to access forceTouchCapability of traitCollection. The
reason is that traitCollection is available on UIScreen starting
from iOS 8, while forceTouchCapability is available on
UITraitCollection starting from iOS 9. So only checking the former
will cause a crash when running on iOS 8.
Change-Id: I44f9fb785349694004fbf2f48fe3b85bb01d9a5a
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@theqtcompany.com>
'currentType' was not sanitized before being used as a shift.
Fix by checking for a valid shift amount before shifting.
Also change the shifted value from 1 (int) to 1U (uint).
It's just the right thing to do.
Found by UBSan:
qtbase/src/corelib/kernel/qvariant.cpp:3131:59: runtime error: shift exponent 1114 is too large for 32-bit type 'unsigned int'
Change-Id: Id3910d6d7f166fd7c80adf5ce1699f0eeb453562
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@theqtcompany.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Whether or not to call XInitThreads() doesn't apply since Qt-5.0's switch to XCB.
Change-Id: I5f1e5e664a251c98af6357c87fc9a6bb03a46ce3
Reviewed-by: Martin Smith <martin.smith@theqtcompany.com>
Reviewed-by: Topi Reiniö <topi.reinio@theqtcompany.com>
The code says (for Q_OS_UNIX)
if ((mode & QAbstractSocket::ShareAddress) || (mode & QAbstractSocket::ReuseAddressHint))
socketEngine->setOption(QAbstractSocketEngine::AddressReusable, 1);
so clearly ReuseAddressHint does the same as ShareAddress, which is: setting SO_REUSEADDR.
Change-Id: Ic2ab4d139c3f58c3c63723fc609a9d4f71bac97a
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Topi Reiniö <topi.reinio@theqtcompany.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
... instead of using erase in a loop, with quadratic complexity.
Change-Id: I9686d117e092f5d74c6e74a462adf503a7b7ae79
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
This has only been identified by WACK for Windows 10.
QWinRTWindow::setVisible adds a Window to the screen and immediately
tries to set the native visibility. This only works when the system
events are handled immediately. While this is the case most of the time,
certification tests revealed that this is not always the case. We have
to flush before setting the element visibility.
Change-Id: Ifce4c045c185c57bc386a4e832074fb84f5d0053
Reviewed-by: Andrew Knight <andrew.knight@intopalo.com>
Reviewed-by: Oliver Wolff <oliver.wolff@theqtcompany.com>
According to MSDN Tls* is inline replaced by Fls* on Windows (Phone) 8.1
and beyond.
However, this does not seem to be the case for Windows 10. An
application links against Tls* and the certification step fails due to
using non-allowed APIs.
Hence we do the inline replacement manually. QThreadStorage and QThread
tests continue to work, so it seems to be an oversight by Microsoft.
Task-number: QTBUG-50292
Change-Id: Ice1b6e54fcee238c94af5c6fb1753d903db7476d
Reviewed-by: Oliver Wolff <oliver.wolff@theqtcompany.com>
When shutting down an application on Windows Phone the SceneGraph tries
to create an offscreen surface to render into. If there is no offscreen
surface available, it creates a new native window and tries to hide it.
As the native event loop is about to shut down, creation fails and
exceptions are raised. Instead we create a vanilla
QPlatformOffscreenSurface. The SceneGraph recognizes it as such and can
handle a proper cleanup on its own.
Furthermore removing the suspend/resume handler in the destructor of
QWinRTIntegration fails for Windows Phone as the application object
itself does not accept this anymore. Hence skip this part for this
platform.
Task-number: QTBUG-49310
Change-Id: I02acdd5a635ef0b9d6ef8199376537b8f0f1a8fb
Reviewed-by: Oliver Wolff <oliver.wolff@theqtcompany.com>
Do not create a dummy eglDisplay when the global static is constructed.
This causes ANGLE to wrongly set some internals, which breaks usage of
EGL_PLATFORM_ANGLE_ENABLE_AUTOMATIC_TRIM_ANGLE as the attribute map
might be empty after calling eglGetPlatformDisplayEXT. Furthermore
initialize() assigns a new display to it without terminating the old
one. This way, the internal suspend handler in ANGLE (Trim11.cpp) will
be added to the application.
The suspend handler is not invoked when an application suspends though.
Reason being that the handler needs to be added from inside the Xaml
thread. As we cannot control this inside ANGLE, we will call
eglInitialize inside the Xaml thread and hence get the suspend event
properly.
Task-number: QTBUG-50337
Task-number: QTBUG-50292
Change-Id: I3867377254c37084bf24f18e78af467f6237db57
Reviewed-by: Oliver Wolff <oliver.wolff@theqtcompany.com>
Note the excessive escaping around "GNU".
Otherwise this leaks the -fPIE flag to the target.
Change-Id: I340df5d5bce00ebec4e1ff3a557ade67022ba23b
Reviewed-by: Stephen Kelly <ske@ableton.com>
glXGetCurrentDisplay can return NULL so need to check
for that.
Change-Id: I1b85a5c575d1deb24dcce4719b6f737283adeabe
Reviewed-by: Błażej Szczygieł <spaz16@wp.pl>
Reviewed-by: Laszlo Agocs <laszlo.agocs@theqtcompany.com>
... instead of using erase in a loop, with quadratic complexity.
Change-Id: If30c6c99a775aec07eef9ddf953e944dc916b5a2
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Not just fuzzily equal. The fuzzy comparison fails for infinities, since
the expression p1 - p2 where p1 = p2 = infinity is NaN. And NaN
comparisons are always false.
As a nice side-effect, we don't do the more expensive computation of a
multiplication if the two numbers really are equal.
Task-number: QTBUG-50036
Change-Id: I11f559ef75544c50b3f8ffff1420cec7c7273295
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
This reverts commit 618e2cc081. The
original commit has a section of code that I failed to review properly
and is of questionable functionality.
Change-Id: I61c53d7b8b2aa7c3312292b017a18aba7da11bc5
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The extra length has been used for a sentinel null value.
By checking the index before indexing into the array, that
extra element can be dropped.
Change-Id: I0b309420d8d89f75e2381a8065c8e68a1a33427c
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
Off-by-one error: we should have calculated whether the current year is
leap, not the next year. This affected any 53-week leap years.
Task-number: QTBUG-50273
Change-Id: I134ce5db2f82468585ffffff14264cb9f12998fd
Reviewed-by: Martin Klapetek <mklapetek@kde.org>
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
... instead of using erase in a loop, with quadratic complexity.
Change-Id: Ia5201e0bedca7abec6d485505f61f3f76a028bd1
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
We forgot to setQuery() after the query was modified.
Task-number: QTBUG-49963
Change-Id: I11f559ef75544c50b3f8ffff1420148ba3092200
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: David Faure <david.faure@kdab.com>
Fix the signature of QWindowsShcoreDLL::GetProcessDpiAwareness and output the
value in the warning about failures of SetProcessDpiAwareness and in debug
output. Failures occur when Qt is embedded into another application, for
example Active Qt servers.
Task-number: QTBUG-41186
Task-number: QTBUG-50206
Change-Id: I3fd6cba26826ee8bbfa0a34f129deb64797c947f
Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
It is equal to m_geometry since c173a50719
(Remove QT_DEVICE_PIXEL_RATIO support from xcb).
Also remove the declaration for QXcbWindow::screenForNativeGeometry(),
which was forgotten to be removed in that commit.
Change-Id: I8ec425a3bc111ec579e2e25c4fda8a02f1c8d09c
Reviewed-by: Błażej Szczygieł <spaz16@wp.pl>
Reviewed-by: Paul Olav Tvete <paul.tvete@theqtcompany.com>
We depend on being able to punch holes in the backing store when
integrating with FBO elements. To do that we need a format with an alpha
channel. This was only working previously because RGB32 didn't mask when
filling or when converting to ARGB32_PM, but other formats didn't.
Also unifies the logic for getting alpha versions of QImage formats.
Task-number: QTBUG-50281
Change-Id: Ied1325f60e2b67d9ea2dfa9701b06fc2231ebfca
Reviewed-by: Laszlo Agocs <laszlo.agocs@theqtcompany.com>
The wrapper batch files were called wrapper.bat for both, flex and
bison. Resolve this conflict by providing proper tool names.
Change-Id: I7987f94976abc29302c9f6a696faa5c7d1fb030d
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
std::stack, based on std::vector, tends to be more efficient that
QStack, based on QVector, but, crucially, it is an actual stack,
so doesn't allow, say, iteration or insertion in the middle, which
QStack does allow.
Since a lot of places where QStack is used use at least some of
those non-stack features, protect this instance which doesn't, yet,
from becoming another such site.
Since the use of std::stack with the default std::deque increases
text size by almost 1K, and a vector makes a perfectly good stack,
use std::vector as the backing container (which std::stack, being
a container adapter, allows easily).
Saves ~0.5KiB in text size on optimized GCC 5.3 Linux AMD64 builds.
Change-Id: I51c8fb1dc4e9907ae00ed1cee8f320304321b322
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
This is a false positive because the only offset that can be outside the
bounds was the last one (-1), which could not be reached in this line
because of the qBound on the line before limiting the maximum value.
The -1 wasn't generated by the Perl script embedded in the file anyway.
qdbuserror.cpp:142:64: error: offset outside bounds of constant string [-Werror]
Change-Id: I24a735698d3c4a719fc9ffff1425f8aad5e5978e
Reviewed-by: Alex Blasche <alexander.blasche@theqtcompany.com>
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@theqtcompany.com>
this is just an optimization/clarification: variables which are known to
be never empty (like PATH) can be extended with less convoluted code.
Change-Id: Ib365bbec8301673ed1c874979b4de19bc983dab1
Reviewed-by: Romain Pokrzywka <romain.pokrzywka@gmail.com>
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
Render-to-texture widgets that are native children may have a non-zero
offset telling the position within the top-level widget. The size is
usually different (smaller) than the tlw's size as well.
An unfortunate optimization attempt led to ignoring the size for native
children with offset (0, 0). This resulted in interesting rendering
artifacts whenever such widgets were placed in the top-left corner of
the toplevel. This is now corrected.
Task-number: QTBUG-48130
Change-Id: If449610c3f3333cb73ae149efaefd7ef42953951
Reviewed-by: Paul Olav Tvete <paul.tvete@theqtcompany.com>
Setting a screen to a child window causes a warning message. This patch
prevents the debug message by catching incorrect behavior made by QPA
earlier.
Change-Id: Idf3ac02290a18afc134f19c2b91348dc009ca4ca
Reviewed-by: Shawn Rutledge <shawn.rutledge@theqtcompany.com>
Find the top level window on the primary virtual desktop first to avoid
obtaining a window which doesn't belong to the primary virtual desktop
when screen geometry is similar.
Change-Id: I78fdfa0b5146d0ba9b912338adeb612c102c4ac3
Reviewed-by: Adam Majer <adamm@zombino.com>
Reviewed-by: Shawn Rutledge <shawn.rutledge@theqtcompany.com>
Qt3D is making heavy use of this, causing the QList node allocations
to be among the top 10 per frame allocation sources. Switching to
QVector fixes that.
Change-Id: I3b4df329710f82bf8d6797ea1f0c79b288a08063
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
The reply serial is displayed for method call returns and errors,
while the serial is displayed for all message types.
To see a message serial it is required to dump messages after
sending, not before.
Task-number: QTBUG-44490
Change-Id: I859f50d739ed059d5b2dfe1a2efdf04b906891a7
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This patch includes setup of class member 'msg' in
QDBusMessagePrivate::toDBusMessage() to be able to get the
serial after message sending.
Testcases for comparing the 'reply serial to' with the 'serial'
are included.
Task-number: QTBUG-44490
Change-Id: Iae7c48f5b0c70a6c5ae500904072b38b46dfd876
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Change the QWindowsInputContext::composition if it is called
with lParam has flags GCS_RESULTSTR and GCS_DELTASTART,
it doesn't call endContextComposition.
Task-number: QTBUG-49955
Change-Id: I49de3f239bf8a77414e433b255db08f227141158
Reviewed-by: Tasuku Suzuki <stasuku@gmail.com>
Reviewed-by: Takumi ASAKI <takumi.asaki@gmail.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Andy Shaw <andy.shaw@theqtcompany.com>
Reviewed-by: Liang Qi <liang.qi@theqtcompany.com>
(cherry picked from qtbase/786984e7e47a63094ad64ec86a4892cc5c0ad6d4)
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
On Android, back button will send a CloseEvent to the focused window.
This change will activate (focus) the Popup windows when they'll become
visible.
Task-number: QTBUG-46419
Change-Id: I6787435b24502bbf9a78662a0c6fd159d4ee2e8d
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@theqtcompany.com>
Implicitly add the minimum required `-std=...` to Qt5 module targets.
Use the newly introduced INTERFACE_COMPILE_FEATURES property for this,
which is available starting from CMake version 3.1.
This also fixes the Qt5 CMake tests, which previously try-compiled projects
using Qt 5.7 headers in non-C++11 mode, thus failed.
Change-Id: I54fc0e07403b646b5d6ac3ab6b8b47119b4feef6
Reviewed-by: Stephen Kelly <steveire@gmail.com>
On X11, QXcbVirtualDesktop represents an X11 screen while QScreen
represents an X11 output. In the case that there are multiple screens
(possibly with multiple outputs), calculate the screen number correctly:
Find the screen index on the primary virtual desktop first to avoid
obtaining a screen index which doesn't belong to the primary virtual
desktop when screen geometry is similar.
Change-Id: I4cbb29b7aa7cd2125759ffbbbe3db4e934feaeae
Reviewed-by: Shawn Rutledge <shawn.rutledge@theqtcompany.com>
Find the root widget only when more than one virtual desktop exists
and find the screen index using virtual siblings from this root
widget.
Use intersecting rects instead of middle point to obtain the screen.
This can help to get the screen index when the middle point is outside
the screen geometry, but part of the window is still on the screen.
If the widget is completely outside the screen geometry, -1 is
returned.
This commit amends: a6b2a4642f
Change-Id: I80247fc1956a82c487ee6f728d1576bf48b28748
Reviewed-by: Shawn Rutledge <shawn.rutledge@theqtcompany.com>
This is a partial revert of eaa3a9d010
Task-number: QTBUG-50081
Change-Id: Ic3dc4daa90d7a968a4ebf45d3029c99a12985686
Reviewed-by: Shawn Rutledge <shawn.rutledge@theqtcompany.com>
Use QSharedPointer::create(), to fuse control block and
payload into a single memory allocation.
Change-Id: I8d87c73566c3be960e957ec0b5419a77da2fc8c9
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
GCC 5.3's undefined-behavior sanitizer checks that the
declared type of the object is a base class of the
dynamic type of the object on each access to a member
of a class type.
It therefore requires the typeinfo for these types,
which for polymorphic types is emitted in the TU where
the vtable is emitted, too.
QFileDialogPrivate is a polymorphic non-exported class,
so this failed at link-time. Ditto for the other cases.
Fix by autotest-exporting the classs.
Also, where applicable, de-inline the dtors, so the
vtable (and typeinfo) are pinned to one TU, and the
ctor, just because it's the correct thing to do.
Change-Id: I2b7dba776282a2809c80eb2bc36440d7d698f926
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Remove a superfluous loop and thus the usage of QIncrementalSleepTimer
from QProcess::waitForFinished. We just wait for the process handle.
There's no need for a loop that checks multiple wait conditions.
This enables us to remove QWindowsPipeWriter from the Windows CE port in
a subsequent commit.
Change-Id: If6a82405227cf145263dba3726bae959e6871d0e
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
...even if reserve() hasn't been called before.
[ChangeLog][QtCore][QString] resize() will no longer shrink the
capacity. That means resize(0) now reliably preserves capacity().
Change-Id: If499a20990bbf3a20553da14e50a42918d310c9f
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
Added check for traiCollection selector.
Task-number: QTBUG-50159
Change-Id: Ie3efafe9e22e59aef862a681ad733f2bb0a5ba49
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@theqtcompany.com>
Instead of converting a QVector to a QList to a QSet just to
be able to look up some indexes, take the original QVector,
sort it, and use std::binary_search().
Change-Id: If56e9371972b9aaebb033a8a499be306163266e2
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
There's no variant with bool first argument.
Change-Id: Ib18b50ed6271d21253d075dc72b6e2b8744cf131
Reviewed-by: Topi Reiniö <topi.reinio@theqtcompany.com>
The function is not available on Windows XP.
Task-number: QTBUG-50188
Change-Id: I4ded1a4eb37af84979c3ce1a9af0d95da60b56ac
Reviewed-by: Kai Koehne <kai.koehne@theqtcompany.com>
This code seems to be hit three times per frame in a hybrid QQ2/Qt3D app,
quickly generating thousands of allocations here via the detach in begin().
Heaptrack confirms they are all gone now.
Change-Id: I7d09f3f5c2b24a2590b701216715055751306e82
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
This patch fixes drag and drop operation on XCB platform when window
will be hidden. The window can be hidden during dnd operation by
switching virtual desktops or by minimizing all windows (show desktop)
using key shortcut.
The ShapedPixmapWindow must grab mouse before dnd operation if mouse is
not grabbed by other window (like in Qt4).
Task-number: QTBUG-46243
Change-Id: I807bc842719a2d0ea0f4dcb733c06c1fd08813e1
Reviewed-by: Shawn Rutledge <shawn.rutledge@theqtcompany.com>
... instead of rolling your own.
qLoadPlugin() expects the create() method of the Plugin
class to take at least the QString key. Since this plugin
create() method doesn't take one, supply a wrapper.
Change-Id: I690fecd3390cc706a45f40ba06dda2b952ff0d85
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Laszlo Agocs <laszlo.agocs@theqtcompany.com>
... instead of rolling your own.
qLoadPlugin() expects the create() method of the Plugin
class to take at least the QString key. Since this plugin
create() method doesn't take one, supply a wrapper.
Change-Id: I5b90b4b87e83f2e8a2e8942b792bb39b87d5f2de
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Laszlo Agocs <laszlo.agocs@theqtcompany.com>
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
If we select XInput events, then when the mouse is clicked, there will
not be a fallback to a core pointer event. But a typical Qt application
doesn't own the root window. If the window manager (such as OpenBox,
Awesome or fvwm) relies on receiving core pointer click events, e.g.
to show a desktop menu, then each time a device is hotplugged while a
Qt application is running, we would select XI2 events and thereby
prevent the window manager from receiving them.
QDesktopWidget's native window is added to m_mapper, even when
it isn't mapped. Then after hotplugging there's a hierarchy event,
and that calls xi2Select for every window in m_mapper. The assumption
with this patch is that the root window does need to be in m_mapper
in case the QDesktopWidget is shown (that was done already in Qt 5.1:
fca94fa5ed), but xi2Select must avoid
selecting XI2 events on it to fix this bug.
Task-number: QTBUG-49952
Change-Id: I5c160e879d93fadfce14120ef2e89a4f71d4f599
Reviewed-by: Uli Schlachter <psychon@znc.in>
Reviewed-by: Paul Olav Tvete <paul.tvete@theqtcompany.com>
Each QGtk2*DialogHelper class owns a QGtk2Dialog. It is possible that
QGtk2Dialog's parent window is destroyed before QGtk2*DialogHelper.
Then the QGtk2Dialog is destroyed, because of the QWindow parent/child
relationship. Before destroying QGtk2*DialogHelper,
QQuickPlatformFileDialog calls QGtk2*DialogHelper::hide, which will
then operate on the already destroyed QGtk2Dialog.
Break the parent/child relationship when QGtk2Dialog's parent is
destroyed.
Change-Id: Ie2bda234f759e7bf4d0642accd82fdc38e0379c2
Task-number: QTBUG-49203
Reviewed-by: J-P Nurmi <jpnurmi@theqtcompany.com>
While a native dialog is open, the application message queue is
handled by the native event loop which is external to Qt. In this
case, QEventDispatcherWin32::processEvents() does not run and socket
notifiers will not be activated. So, this patch moves the notifier
activation code into the window procedure, which enables socket
event processing with native dialogs.
Task-number: QTBUG-49782
Task-number: QTBUG-48901
Change-Id: Icbdd96b2e80c50b73505f4fe74957575b83d6cf1
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Kai Koehne <kai.koehne@theqtcompany.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
Windows 10 requires apps to handle the PrelaunchActivated property. This
enables a faster startup by launching an app and immediately suspending
it. This happens by the system and at the time the user launches the app
no library loading or initialization is required.
By default we opt-out of this and return early. The reason is that we
cannot know the type of application written in Qt and whether it breaks
the guidelines by using prelaunch. For further details check here:
https://msdn.microsoft.com/en-us/library/windows/apps/mt593297.aspx
Task-number: QTBUG-50292
Change-Id: I4d0b0b95a03f93b99211d042895404a88ea7cb9d
Reviewed-by: Oliver Wolff <oliver.wolff@theqtcompany.com>
qstrtod() used to accept trailing junk until libdouble-conversion was
introduced and we need this behavior in order to implement EcmaScript's
parseFloat() correctly. The QString and QByteArray methods should not
accept trailing junk, though.
Task-number: QTBUG-50131
Change-Id: Ide922da0d65b2576be2c9f47f6053057eff77994
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
- Use perfect forwarding for the additional argument.
- Provide a variadic template version
- Deprecate qLoadPlugin1() — there's no reason for a different name
— and fix all callers in qtbase.
- Provide non-variadic overloads for up to three additional args
(QPlatformIntegration rolled its own function because it needs
three args).
Change-Id: I72fb2dd9a021de704cbf5e4b6ea31c80447fb3b1
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
The use of QPair made the return type of initializeAnimation()
so complicated, that the original author opted to declare the
pair on one line, then assign to it in the next, to keep below
the line length limit.
She also copied a member of the pair just so as to give it a
descriptive name.
Fix both by introducing a small result struct. It has a nicer
name, compared to the pair, but still port callers to use 'auto'.
The member names are descriptive enough now.
Saves more than 0.5KiB in text size on optimized GCC 4.9 Linux
AMD64 builds, too.
Change-Id: I7ed007ffa0fb16e182e38cd405cfd54da4e363fb
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Erik Verbruggen <erik.verbruggen@theqtcompany.com>
When a high-priority event is posted in overrided
'QStateMachine::beginSelectTransitions', the event may be remained in
event queue, and be not dispatched until another event posted.
Change-Id: Ifda288d9c00ac7985e426b9cc02bda382ebaac35
Reviewed-by: Erik Verbruggen <erik.verbruggen@theqtcompany.com>
This reverts commit 56aad2ad60.
QWidget::mapFromGlobal() does not work correctly when the widget is
a child widget of another widget embedded into a QGraphicsView with a
transformation (scaling/rotation). It starts applying offsets going
up the widget tree (just as mapToGlobal) until it hits the embedded widget
not taking into account the transformation.
It would need to go in from to top to bottom or better be reimplemented
such that a QTransform for mapping coordinates from/to global is determined
which is then applied in reverse.
Task-number: QTBUG-50030
Task-number: QTBUG-50136
Change-Id: Iadeb891d793be1938c64942bfbf38d541a281c33
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Instead of using a "custom wait function" in cases,
where a timeout is needed (like in qhostinfo_winrt.cpp)
we should have the timeout as part of our await
function.
By having one common place to handle this, we can avoid
unnecessary warnings, that might be caused by custom
functions. The current implementation in qhostinfo
for example causes at least 1 "originate error"
exception per call.
Change-Id: I7b6cfdfd861af2b0d271465eecaefe4a93e3109b
Reviewed-by: Maurice Kalinowski <maurice.kalinowski@theqtcompany.com>
By using await instead of using a custom wait function we
get rid of a lot of "function was called at an unexpected
time" warnings and use the default way of waiting for an
asynchronous function to complete.
Change-Id: I6c5bf73ccf68a219b8f3facd3531873d7e8d4c67
Reviewed-by: Maurice Kalinowski <maurice.kalinowski@theqtcompany.com>
That API has been available for WinRT and Windows
Phone for some time now. By using it to get the
machine name and for hostname resolution we can get
rid of some winrt-only code and use qhostinfo_win.cpp
on WinRT and Windows phone as well.
Additionally the required capability was added to
tst_qhostinfo so that this auto test can be run without
any manual editing.
Change-Id: I63fa5521bf8cdb0c919bd5a0100ea977c865622a
Reviewed-by: Maurice Kalinowski <maurice.kalinowski@theqtcompany.com>
Windows Phone 8.1 provides access to the camera button and press/release
events get passed as Key_CameraFocus and Key_Camera. Unfortunately a
release does not provide what has been pressed before, hence this
information needs to be cached when the press happens.
Done-with: Maurice Kalinowski<maurice.kalinowski@theqtcompany.com>
Task-number: QTBUG-39115
Change-Id: I6ce58a1f07a6bf7183b8d99a26e5cd7b0d32d6db
Reviewed-by: Topi Reiniö <topi.reinio@theqtcompany.com>
Reviewed-by: Samuel Nevala <samuel.nevala@intopalo.com>
Reviewed-by: Oliver Wolff <oliver.wolff@theqtcompany.com>
_panTimer has a target - for example, widget. 'reset' should stop this timer.
The problem (quite subtle and difficult to reproduce) found in tst_QTouchEvent:
widgets created/destoryed by different tests but we still can have a timer
waiting for event dispatcher to 'processEvents', firing with now-invalid
dangling pointer - 'target'.
Change-Id: Iccaf3368a8ee6a0a2f60e9dcdf5d40fb7392ca21
Task-number: QTBUG-49844
Reviewed-by: Morten Johan Sørvig <morten.sorvig@theqtcompany.com>
Mimicking what QTableView is already doing correctly, the header's
height needs to be constrained by its own minimum/maximum height.
Task-number: QTBUG-49277
Change-Id: I695a4398991b738c4b4c924716176b9ad2152e87
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
Reviewed-by: David Faure <david.faure@kdab.com>
Saves just under 5.5KiB in text size on optimized GCC 4.9 Linux
AMD64 builds.
Change-Id: I6d868a7d2e469cf0564127b0e66cd7b272a0c9cd
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
A QStringList is almost as efficient as a QVector<QString>.
More importantly, the QStringList case can share code with
the myriad of other QStringList users, in particular
because std::stack is but the thinnest of wrappers around
its underlying container.
Saves 1400b in text size on optimized GCC 4.9 Linux AMD64
builds.
Change-Id: If37ed55802aa8a529ca772df465990ded0aaba7f
Reviewed-by: David Faure <david.faure@kdab.com>
... by not re-creating the same string in every loop iteration.
Change-Id: If17c5980bec8276b531267b309578dac8901c93e
Reviewed-by: David Faure <david.faure@kdab.com>
Trying to set data for a compressed cubemap texture face fails, because
the code incorrectly tries to bind to a cubemap face target.
This was already fixed for uncompressed textures. Instead of duplicating
the code, moved the texture binding/unbinding code to a RAII class.
Change-Id: I8f1cb7fc661b919200f85c3465d6e6e3c22d0871
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
When using cipher algorithms with forward secrecy an ephemeral key is
used to generate the symmetric session key. Beside the SSL certificate's
key, this ephemeral key is of cryptographic interest.
The ephemeral key is chosen by the server side - currently statically in
the Qt implementation - so it is only of interest on the client side to
check it. Therefore the ephemeral key is the null key if the connection
is set up in server mode or a cipher without forward secrecy is used.
Change-Id: If241247dbb8490a91233ae47f2b38952c6591bf4
Reviewed-by: Markus Goetz (Woboq GmbH) <markus@woboq.com>
QApplication::wheelScrollLines() was always 1 due it being a bool.
Task-number: QTBUG-49561
Change-Id: I83e97da3507c31dfd024db7bf543c4e6a3b4a98b
Reviewed-by: Oliver Wolff <oliver.wolff@theqtcompany.com>
There is no need in special OS dependent code now, because the
socket notifiers got the identical behavior on all platforms.
Change-Id: I53ee51cd5eeff328e0b73f7325d2c9e883c5b213
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Markus Goetz (Woboq GmbH) <markus@woboq.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
Qt uses QHash as the container for faking environment variables on
Windows Runtime and CE. Environment variable manipulation functions are
protected by mutex. Accessing the QT_HASH_SEED environment variable
inside QHash can lead to situation where qputenv() call leads to
qgetenv() call and that leads to a deadlock. Change the container
from QHash to QVector to avoid deadlock.
Task-number: QTBUG-49529
Change-Id: I550ead4ab12e7abebc044f52339063a44fcf0170
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Maurice Kalinowski <maurice.kalinowski@theqtcompany.com>
The latter is much faster as it doesn't have to deal with
time zones.
This change is safe because QNetworkHeadersPrivate::fromHttpDate(),
used as the QDateTime source in other, nearby, code paths, also
returns only UTC date-times. So not only is this change faster at
the call site, it also avoids inconsistent-timespec comparisons
down the line.
Credits to Milian Wolff, from whose QtWS15 talk this advice is
taken.
Change-Id: I26d308a2763cc45d28bc96871e651f30b17a6b85
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Saves more than 1KiB in text size on optimized GCC 4.9 Linux
AMD64 builds.
Change-Id: I1a974e903bc8352af800fa26edae0194c1e51705
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The old code asked each thread to stop, and then waited for
that one thread to finish before deleting it and turning
to the next thread in line.
Split the three actions, ie. first ask all threads to stop,
then wait for all threads to finish, and only then delete
all of them.
Apart from being more icache-friendly, this enables more
concurrency, as stopped threads' cleanup work can overlap
with other's, thus reducing the wait() time per thread to
potentially zero.
Did not replicate the inefficient foreach loop, but went
with C++11 range-fors directly.
Change-Id: I53e598e31999a772c0f81ff5885490216e13d492
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Saves more than 4KiB in text size on optimized GCC 4.9 Linux
AMD64 builds.
Change-Id: Ice3fe915622a399d1ed10e0ce364cf54b4b835fe
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
Saves a bit over 2.5KiB in text size on optimized GCC 4.9 Linux
AMD64 builds.
Change-Id: I18a635d65a3c20c4bf5ec2cbfa4ed81a8153cab3
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
Saves just shy of 4KiB in text size on optimized GCC 4.9 Linux
AMD64 builds, iow: ~0.07% of the total QtCore library size.
Change-Id: I87fdcc8ee25c6bb5dabddb9a694ab4496b1538fa
Reviewed-by: David Faure <david.faure@kdab.com>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Saves just shy of 6KiB in text size on optimized GCC 4.9 Linux
AMD64 builds, iow: ~0.1% of the total QtCore library size.
Change-Id: I087e1e34a64944affb1211e16371e314012c4b2c
Reviewed-by: David Faure <david.faure@kdab.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Not many left. Still 344b saved in text size on optimized
GCC 4.9 Linux AMD64 builds.
Change-Id: I51ba49fe065854f84e26d25d7da5e26427ea2c0b
Reviewed-by: Laszlo Agocs <laszlo.agocs@theqtcompany.com>
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Saves 327b in text size, but done mainly because such
functions shouldn't be inline in the first place.
Change-Id: Ieea97f13620e4411dc01cb0d390bd668c13cfa7a
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
...to turn mutable lvalues into const lvalues.
Like the proposed std::as_const, it should not apply to
rvalues to avoid lifetime issues in code like
for (auto x : qAsConst(someFunc()))
// dangling
At a more basic level, qAsConst isn't useful for rvalues,
because one can always store them in an lvalue first, with
no loss in performance (the object is created by the
compiler silently anyway). So the correct way to write the
above is:
const auto funcResult = someFunc();
for (auto e : funcResult)
To fail compilation when passing rvalues, I used the
const-&& pattern also employed by std::cref(), and the
proposed std::as_const.
Intended as internal API, but not put into the QtPrivate
namespace to make it simpler to use.
We could wait for std::as_const, but that is far, far
away (just entered the current C++17 draft as of this
writing), and the Qt containers with their tendency to
detach are a problem _now_.
Change-Id: I8824a59d2274de5c5cd642f117212322e4648025
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
It makes tst_QMutex::contendedQMutex with no msleep 8 times faster
Change-Id: Ic300e7618b4467e4e08b30f0213bd23c06d4d90a
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
When a datagram is sent to a closed host/port combination, the host will
likely send back an ICMP error message. In the regular socket API, there
isn't a good way of actually receiving this error, so some Windows kernels
wired that message back up to the call to WSARecvFrom() as a synthetic
datagram. Reading this datagram results in a WSAECONNRESET error code,
which should reported to the user as a refused connection attempt.
To make the errors a bit more informative, the native error strings for
WSAECONNRESET and WSAENETRESET were also added.
Task-number: QTBUG-49301
Change-Id: If659be54ba1b39965b5f481f0c0cb9eeea0a06d2
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Markus Goetz (Woboq GmbH) <markus@woboq.com>
The 'itemList' is a QList and is not maintained in sorted order.
Change-Id: I8f221281803ff0e9caa054007bee3058214f2f30
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The old code re-used a QFileInfo object, repeatedly
setting new file names with QFileInfo::setFile(QString).
But QFileInfo::setFile() simply assigns itself a
new QFileInfo object:
*this = QFileInfo(...)
so it's just as efficient to re-create a new QFIleInfo
object for every file name under test.
To avoid extra {} for object lifetime scoping, factor
the repeated evaluation of isReadable() && isFile()
into a small helper function, which also creates and
destroys the QFileInfo object.
The deeper significance of this change is that it
avoids implicit sharing of 'realname', which is
permanently modified. A later patch will make
changes that make 'realname' re-use its capacity
through the lifetime of the find_translation()
function, and sharing the variable implicitly will
nip any auch attempts in the bud.
Force the compiler to not inline the new function.
There's really no point in spending ~0.5KiB in text
size on inlining the code; the miniscule speed
improvement is dwarfed by the memory allocation of
the QFileInfo ctor, anyway.
As a consequence, this change even saves 96b in text
size on optimized GCC 4.9 Linux AMD64 builds, even
though that wasn't even the goal.
Change-Id: I08c5cbb7b6f1ba59440a1597e28d962ce63a7c65
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
... and remove misleading comments (these are overloads, not specializations).
The QList overloads do nothing different from the generic container
overloads. Remove them.
Only leave the QVariantList overload, because that converts to
QDBusVariant before serializing. Which means that this should
probably be templated on the container type, otherwise you get
different behavior for QList<QVariant> and, say, QVector<QVariant>,
which is surely wrong.
Change-Id: I215ba9891235b51304c2ed4041d3dbd003d69581
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The system call is not present on earlier releases, and since the 9.x
series will be supported until the end of 2016, add a check for the
__FreeBSD_version macro and only enable pipe2 support if the value is
high enough.
Change-Id: I5633531cec7e95d42ff5f4b14afe772ae8d7d66d
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This is a follow-up to c8c4ad0 ("forkfd: Define __BSD_VISIBLE and
_NETBSD_SOURCE").
Defining those macros in forkfd.c is not enough: forkfd_qt.cpp also sets
_POSIX_C_SOURCE, and sys/cdefs.h can be included implicitly via Qt's
headers (<algorithm> ends up pulling unistd.h that leads to sys/cdefs.h
and sys/types.h with both libstdc++ and older libc++ versions). In this
case, __BSD_VISIBLE/_NETBSD_SOURCE are not defined, _POSIX_C_SOURCE is,
several type definitions are omitted and by the time we include
sys/time.h in forkfd.c the build fails. On FreeBSD < 11, the error looks
like this:
In file included from io/../../3rdparty/forkfd/forkfd.c:36,
from io/forkfd_qt.cpp:80:
/usr/include/sys/time.h:94: error: 'u_int' has not been declared
Change-Id: I01fa2f5861027d99936d3026faeee9f0db3ecabd
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
- populate the widget with addItems(QStringList) instead of
looping over addItem(QString)
- Use findText() instead of looping over itemText(int).
QComboBox::findText() delegates searching to the model
(via QAbstractItemModel::match()), so is potentially
much faster. I say potentially, because match() isn't
properly reimplemented in most models, yet. But that is
something to fix in the models.
Change-Id: I6e52cf5af810ab7869c0270504a241868a5ca281
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
