Forward declare some types in qaccessiblewidgetfactory_p.h instead of
including qaccessible.h. In turn, qapplication.cpp which relied on
transitive includes now needs to include qaccessible_base.h.
Change-Id: I8ac00d45a7ffccd84769fb210f29364a45bcd59e
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
qaccessible.h includes quite a few heavy headers. While QAccessibleEvent
and QAccessible can be forward declared, the enums in QAccessible
cannot. By moving QAccessible into its own lightweight header, we can
significantly reduce the cost of using the enums.
qaccessible.h still includes qaccessible_base.h, and the syncqt rules
are adjusted to ensure that this changes is source compatible.
Additionally, we no longer include cstdlib, as we only need cstring.
[ChangeLog][Potentially Source-Incompatible Changes] <QAccessible> no
longer includes <stdlib.h>. This might break code that relied on
transitive includes.
Task-number: QTBUG-97601
Change-Id: I15fbd9c85f5746885f9e89eabfc6d07b9bb1f968
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
The C++ code, which is called from Java, was executed on Java thread.
However Qt has its own main GUI thread, where all GUI elements and their
accessibility instances are created.
As a result we have threading issues when accessing A11Y objects from
Java thread.
This patch uses QMetaObject::invokeMethod calls to dispatch all the
critical parts of the C++ code to the main thread.
It uses BlockingQueuedConnection, so that Java thread can still use
these methods synchronously.
The proper context is based on the m_accessibilityContext object, which
is created as a child of the base accessibility QObject of the
application (which is the QGuiApplication instance in most cases).
Task-number: QTBUG-95764
Pick-to: 6.3 6.2 5.15
Change-Id: Iff4f3f2645657f6aca426fa19ccc86a2cbe4d4d0
Reviewed-by: Assam Boudjelthia <assam.boudjelthia@qt.io>
Reviewed-by: Jarkko Koivikko <jarkko.koivikko@code-q.fi>
This commit amends 850a7f1238.
We can't extract the parentId for the hidden object on Java side,
because the Java call is executed in a separate thread, so the
original hidden object can be destroyed somewhere in the middle of
parentId() call.
As a workaround, we get the parentId in advance, on C++ side, and pass
it as a parameter to JNI function.
Task-number: QTBUG-95764
Pick-to: 6.3 6.2 5.15
Change-Id: Ied2ab4ab39b947f3f582575cf77cc76fbac9e274
Reviewed-by: Jarkko Koivikko <jarkko.koivikko@code-q.fi>
Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
Reviewed-by: Rami Potinkara <rami.potinkara@qt.io>
Reviewed-by: Assam Boudjelthia <assam.boudjelthia@qt.io>
Some properties and commands were mistakenly assigned to Qt 6.3.
Pick-to: 6.3
Task-number: QTBUG-100212
Change-Id: I504021863ce628628af469f0ac12881c2eb07dbb
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
qtdeclarative has a call like
qt_internal_add_example(imagine/automotive)
This causes issues when trying to pass that as the name of the
external project. Use the last part after the last slash as the name
of the project.
Pick-to: 6.2 6.3
Change-Id: Ifc074e50e537f07f3636699ed255d2561930d873
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
Reviewed-by: Alexey Edelev <alexey.edelev@qt.io>
It was using a faster divide by 256 instead of an accurate divide by
255. This meant black screen wasn't a null operation as it should be.
This also fixes alpha mixing for a few other compositions.
Pick-to: 6.3 6.2 5.15
Fixes: QTBUG-100327
Change-Id: I149ad39147176e00ce753979d55dc8633704dc1a
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
Yield a warning instead of an error that -skip and -submodules have
no effect in a per-repo build.
Amends acaba63260
Adjusts to 11ae0e772c
Pick-to: 6.2 6.3
Change-Id: Ic26b0ba29e48be5162f13ad75308f8358d8878e3
Reviewed-by: Alexey Edelev <alexey.edelev@qt.io>
Reviewed-by: Jörg Bornemann <joerg.bornemann@qt.io>
Having just spent another afternoon wrapping my head around why this
code is correct, I chose to save the next person (quite possibly me)
faced with it the slowest half of the effort that went into doing so.
This amends commit d3ff95dcb8
Change-Id: If0fcd2fb1fff5e99b1046d07039169e928fda9e4
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
moc, qmake, rcc, and uic tests execute corresponding tools inside
the test executables. This adds dependencies explicitly, to execute
tests using _check targets, without the need to manually build
dependencies.
Pick-to: 6.3 6.2
Fixes: QTBUG-98576
Change-Id: I8b989276e21199b867591f7b0f6af9204695ac13
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
With Ninja Multi-Config generator it's necessary to provide the
'-C' option matching the build configuration when running
ctest. This patchset adds the missing option to the command line
of the '_check' targets.
Pick-to: 6.2 6.3
Change-Id: I42acfba26a09877a6d3a5be9230860f15f95a0cd
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
In a single-config build on Linux, we usually don't want a debug
postfix in library names (as opposed to Windows/macOS).
But when doing a multi-config build on Linux, assigning no postfixes
causes CMake to generate rules for a single config only, the first one
specified in CMAKE_CONFIGURATION_TYPES.
This leads to being unable to build all configurations from the same
build.ninja file as well as other obscure issues like race-conditions
when generating prl files.
To address this, when doing a multi-config build, always assign a
postfix for each config except the first release-like one, while
preserving the existing rules we had for debug postfix names.
Pick-to: 6.2 6.3
Fixes: QTBUG-100493
Change-Id: Ie9c88e074abdcf2961d7b3dee19a5694292717b8
Reviewed-by: Alexey Edelev <alexey.edelev@qt.io>
To clarify its purpose and separate it from the others
Pick-to: 6.3
Change-Id: Id9efcf474ac0cd8777393380bb85c83780454005
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
CI tests for QNX are run in QEMU which does not have enough RAM
to run this test successfully.
Pick-to: 6.2 6.3
Change-Id: Idab3e8b6a1e0ae3eddf5aedb82e6784a74ae9a3a
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
The headers are already included in QtGui/QtGui. This also fixes
an issue when compiling with the respective features disabled.
Pick-to: 6.2 6.3
Fixes: QTBUG-100290
Change-Id: I5e7c6578952535315bacde218a26d2133c486b2c
Reviewed-by: Jörg Bornemann <joerg.bornemann@qt.io>
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
My understanding is that the two bullet list items should be
treated as one item.
Pick-to: 6.3
Change-Id: Ibdd2bbc8677e925e71e6de7b3f076a1fab3914bf
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
This link command simply links back to the same page. Also, none of the
other modules put their module name inside a link command at the
beginning of the document, for example: The Qt QML module provides...
Pick-to: 6.3
Change-Id: I2299b92f44ee736d181f34d9e71a31ba210a34e8
Reviewed-by: Topi Reiniö <topi.reinio@qt.io>
There is a TODO for this in Qt Quick from the 6.0 times. To decide
upfront if Metal can be expected to function, or if a fallback to
OpenGL needs to be triggered (especially important with macOS virtual
machines, where, unlike any real macOS system, Metal may not be
present at all), the scenegraph calls create() and then drops the
result. The idea to make this less wasteful was back then to add a
dedicated probing function which can, possibly, perform the checks in
a more lightweight manner than full initialization. Implement this
now, focusing on Metal.
Brought to attention by QTBUG-100441: printing warnings about not
having an MTLDevice is confusing in a Metal-less macOS VM, because it
is not an actual error, only part of the probing at scenegraph
initialization. We can now avoid printing confusing warnings there.
Change-Id: Ie52c36af9224bedc3f5e4c23edb486d961c9f216
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Andy Nichols <andy.nichols@qt.io>
Those platform specific classes are not covered by our
binary/source compatibility guarantee between minor releases.
Pick-to: 6.3
Change-Id: I2a69c7ea07feea62fdf9899839e47fc540f89c12
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
The auto-repeat timer of QScrollBar kicks in after 50ms, so if the test
takes too long and the timer fires during the qWait call or during event
delivery, then we might get multiple valueChanged emissions.
Detect that scenario and allow the test to XFAIL if things take
a significant time (more than 40ms vs the expected 1ms).
Pick-to: 6.3 6.2
Change-Id: Ie90aadc62372397db695fd2b34fe1f5252ce8d37
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Qt 6 automatically registers all types used in properties. Thus, the
functionality is mostly obsolete since Qt 6. We can simply return the
stored metatype's id instead.
The code in moc is kept in place for compatibility with older code,
though there seem to be no remaining users of the metacall in Qt itself
anymore.
Change-Id: I3c654150988ae7780b4e9ea1f16191fdadf8d791
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Qt 6's metatype does no longer care whether the type is from Core, GUI
or a user metatype.
Change-Id: I4f41d44a61a9839f58b957219c2404512630587a
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
[ChangeLog][Potentially Source-Incompatible Changes] The qmetatype.h
header no longer includes qvarlengtharray. It might be necessary to
include the header explicitly if code previously relied on it being
implicitly included.
Change-Id: If58245836715ec8a323ec8cbadc67d73a742b15f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
Create QAbstractSpinBox test
Update test for QSpinBox
Add tests for QDoubleSpinBox, QDateEdit, QTimeEdit
Add test for QDateTimeEdit incl. calendar popup
Task-number: QTBUG-99677
Change-Id: Ia74686f47a1520ab4c7e878b19fc774e00db88dd
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
During a touch update, this looked questionable since I don't have any
barcode readers or buttonboxes attached:
XI2 touch event type 19 seq 1481 detail 193 pos 116.3, 77.8 root pos 6476.3, 2941.8 on window 3e00006
valuator BARCODE value 47796.365791 from range 0.000000 -> 65535.000000
valuator BUTTONBOX value 59504.858830 from range 0.000000 -> 65535.000000
touchpoint 193 state QEventPoint::Updated pos norm QPointF(0.729326,0.907986) area QRectF(...
After the fix it's like this:
XI2 touch event type 19 seq 1235 detail 201 pos 154.0, 64.3 root pos 6168.0, 2794.3 on window 5000006
valuator Abs MT Position X value 45520.589958 from range 0.000000 -> 65535.000000
valuator Abs MT Position Y value 56519.790001 from range 0.000000 -> 65535.000000
touchpoint 201 state QEventPoint::Updated pos norm QPointF(0.6946,0.862437) area QRectF(...
ValuatorClassInfo's QXcbAtom::Atom is a Qt enum, which has to be
converted to xcb_atom_t before we can call xcb_get_atom_name_name.
Originally in ce2fe90faa vci was of type
XIValuatorClassInfo; then 7499d4404f
changed it to our ValuatorClassInfo struct, but didn't add the lookup
indirection to this qCDebug.
Pick-to: 5.15 6.2 6.3
Change-Id: Ib6107f17d6c6c209573e3df54149d6cfffc0b8b6
Reviewed-by: Liang Qi <liang.qi@qt.io>
...for now anyway. The plan is to have C++ API for custom handlers
later on; and we'll see if we find any other "must-have" use cases.
Task-number: QTBUG-68110
Pick-to: 6.2 6.3
Change-Id: I5e624a500af37797345d25bb5f54b146b13a45ca
Reviewed-by: Venugopal Shivashankar <Venugopal.Shivashankar@qt.io>
Since this is only used by qtpaths and qmake, let's not make every
application pay the price of dynamically initializing a QString whenever
QtCore is loaded (which 100% of Qt applications do). Instead,
initializing a null pointer costs zero and is one third the size of
QString. Even the assignment in qmake and qtpaths is faster this way.
Pick-to: 6.3
Change-Id: I6fcda969a9e9427198bffffd16ce8d1eb8dc19da
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
As with QHoverEvent, it's better to require globalPos rather than
"initialized to QCursor::pos(), which may not be appropriate" as the
docs have pointed out for many years now. This removes the remaining
calls to QCursor::pos() in event constructors.
Task-number: QTBUG-52430
Task-number: QTBUG-69433
Task-number: QTBUG-100324
Change-Id: I076dae56f37abaad7085cc95dddee453a80a45f3
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Clang-cl supports the clang detections, and only approximates specific
MSVC version.
This works as the MSVC standard library is ahead of the clang standard
library. Where this logic was written 7 years ago when it was not the
case.
Pick-to: 6.3
Change-Id: Ic13e2f2fbcde011a0d8831cc48c0c9041bfe3b99
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Change ref() and deref() in qatomic_cxx11.h to use acq_rel ordered
semantics as documented instead of the stricter sequential consistency
that was previously used implicitly.
It was pointed out in the discussion on atomic reference counts that
acq_rel semantics should be sufficient for reference counts. It also
turned out that that is all that the definition of "Ordered" in the
documentation really guarantees.
Original-patch-by: Kevin Kofler <kevin.kofler@chello.at>
Change-Id: I6cd50b34bdc699ff8b7c8fe6d303b28d71494731
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
The string was used as the return value of QLibraryInfo::build() and
also string-pasted into the format string in qt_core_boilerplate(). As
a proper substring (not suffix) of the format string, QT_BUILD_STR
could not be shared among the two users.
Fix by interpolating QT_BUILD_STR into the format string, so
QT_BUILD_STR appears by itself in both functions. The string should
now be shared, but compilers have quite low limits on lengths of
strings they will share, so turn the macro into an static function so
the actual string literal appears only once, guaranteeing
de-duplication.
It also feels a bit safer to interpolate the build string instead of
pasting it into the format string; it's not immediately obvious that
it cannot contain %s itself, even if we could expect a compiler
warning in that case.
Change-Id: I69d90c37c3a3dcf8ac46d4f36e7c0d2cf3b947af
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
It's literally
return true; // or
return false;
so it can be marked as noexcept and (the stronger form of) pure.
Change-Id: Id55fee1db5d18201be2ea6c6778ece3de68b05f2
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Even if it's unlikely any of those would exceed 2 billion entries.
Pick-to: 6.3 6.2
Change-Id: Ib0d4d4e7a64fcde03b7f24bc0667d63c4a737deb
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
Avoid allocating all these strings on load. Part 1: tables.
Pick-to: 6.3 6.2
Change-Id: Ibf7af9aea0088efe68b247c68022b98143634616
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
The cachedRelaySlotMethodIndex started out as -1, assuming no slot
could have that index.
But 0 is just as good a marker, because we know that method index 0
will be one of QObject's (destroyed() or deleteLater()).
So start out with a zero initial value, which means the variable can
be placed in the BSS segment now and doesn't need to be stored in
DATA.
Pick-to: 6.3 6.2
Change-Id: I40b96d54f11e60348f465cafa15af98d41038c95
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Apple Clang 12:
qcalendar.cpp:456:22: error: loop variable '[key, value]' is always a copy because the range of type 'QFlatMap<QString,
QCalendarBackend *, (anonymous namespace)::CaseInsensitiveAnyStringViewLessThan, QStringList, std::vector<QCalendarBackend *>
>' (aka 'QFlatMap<QString, QCalendarBackend *, (anonymous namespace)::CaseInsensitiveAnyStringViewLessThan, QList<QString>,
vector<QCalendarBackend *> >') does not return a reference [-Werror,-Wrange-loop-analysis]
for (const auto &[key, value] : byName) {
^
Change-Id: I54f205f6b7314351b078fffd16d05e5cd3ef4c22
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Common module uses the qt-cmake-private-install to install.
This ensures all configurations are installed.
Pick-to: 6.2
Change-Id: I4be97685c25392838dbc7ab7c3526f55987c08e6
Reviewed-by: Iikka Eklund <iikka.eklund@qt.io>
The option name should be passed instead of it's value.
Pick-to: 6.2 6.3
Change-Id: I74529c36c438f5d40ecd4dcf689b3ea2a100e5fb
Reviewed-by: Iikka Eklund <iikka.eklund@qt.io>
Because it's BiC.
From the API review.
Pick-to: 6.3
Change-Id: I69f1fe58428241a876ae9058e8a0108a8e496c22
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
During delivery of a pointer event, another type might be synthesized
from it; and we are not yet "done" with all access to the original event
until the delivery of the synthesized event is completed. Further,
there could be a nested event loop if exec() is called on a dialog.
If QPointingDevicePrivate::pointById() causes activePoints to grow
(for example if we handle a fresh incoming touch event in the nested
event loop), reallocation could cause the persistentPoint reference near
the top of processMouseEvent() to be dangling by the time we get to the
bottom; so we need to call queryPointById() again to be sure to get the
persistent QEventPoint that is actually used in the current mouse event
to ungrab.
Fixes: QTBUG-97157
Pick-to: 6.2 6.3
Change-Id: Ib3b90eef5db691675b03474fd75981e972d11d2d
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
