There was a race condition in the test, since the window
might be exposed before the event to activate the
application has been received. And the short cut is not
triggered before the application is active, which
means it could just be discarded and the window would
never be closed.
This happened consistently when testing on Wayland.
Task-number: QTBUG-91418
Change-Id: I40cb143985175f5f2b5405e9502a48475c93074a
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
The test assumed that dialog.move() would actually move
the dialog box to the requested location, which is an invalid
assumption on Wayland.
Since the objective of the test is not to check if move() works
this way, but to check whether the dialog box shows up in the
same location when it is re-shown, we actually fetch the pos
after showing it the first time, rather than assume this is the
same as the requested one.
Task-number: QTBUG-91418
Change-Id: Ifa21fa08429f198988f90d7ee328e4f35a4764c2
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
This test depends on XCB, but is only protected by a build-time
flag, so it will be executed as long as XCB is available at
build-time. Do as with the offscreen backend and just skip it
when a different platform plugin is in use.
Task-number: QTBUG-91418
Change-Id: Ida076dc81d0740af2cec164bd40ca10d16345f05
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
If we only wait until the window is shown before testing the
event delivery, we may be getting a WindowActivation event
and subsequent paint event after we start testing. This can
cause a shift in the paint events, so that we end up with an
unexpected paint event at the end.
We should wait until the window is active before we start
checking the response to updates, to be sure nothing is pending
on the queue still.
Note that you'd expect QTest::qWaitForWindowActive() to do this,
but this actually falls back to qWaitForWindowExposed() when
the platform does not have WindowActivation capability. While
there is no real link between WindowActivation capability and
waiting for a window to be active, changing the behavior of that
function would be too scary, so we just implement an explicit
wait in the functions that depend on this.
Task-number: QTBUG-91418
Change-Id: Iee40dcfa1377f543ea05042cc5a972270b346708
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
Since Qt 5.5, brush textures are stored as QImages. If the texture has
been set as a QPixmap, the conversion to QImage may change its pixel
format, depending on the native system format. That is acceptable.
Fixes: QTBUG-69193
Task-number: QTBUG-91418
Change-Id: Ic4dbeaa552b1f2183774a011e38414855a9dc4b1
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
Includes both minimum deployment targets and minimum sdk
versions.
As per supported Apple platforms versions which was done
in qt/qtdoc at
8807fdedce29cbbd7662fcd745234da30eace3fb
For Qt for iOS 6.0.x we only bump the minimum
deloyment target because applications seem to crash with iOS 12.4+,
and it's better to have a build error than a runtime error.
The minimum required sdk will not be bumped for 6.0.x, so we don't
accidentally break someone's existing build, given that 6.0 is already
released.
Pick-to: 6.1
Task-number: QTCREATORBUG-23574
Change-Id: I3046384164f2d7fdbd0cfd16dcb85e0d60bc56ce
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Otherewise, we hit an assertion because the name is empty.
Fixes: QTBUG-91709
Pick-to: 6.0 6.1
Change-Id: I03a530d64ea8dead3efc5fcb8c00909388a387d0
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Otherwise the benchmark crashes as the data tag for the row would be
empty.
Fixes: QTBUG-91708
Pick-to: 6.0 6.1
Change-Id: I484ded5b8670571b80012e64d67846d3b8db5320
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
...when it comes to the output size.
This mirrors what all other backends do. For example, with Vulkan
the only source of size is the surface (VkSurfaceKHR), never the
QWindow, even though we'd expect that the surface size equals
to window_size * dpr, and that's almost always true, but there
are exceptions. (e.g. we have seen bugs on Windows with some drivers
in high DPI situations where the Vulkan surface did not fully match
the window size, yet it is the surface, and only the surface, that
matters for rendering, i.e. viewports and such must match the surface,
not the native window)
With Metal we hit a similar problem on iOS: the QWindow's size*dpr
and what we calculate from the CAMetalLayer have a height difference
of 1.
Mitigate this by making QRhiSwapChain::surfacePixelSize() and the
calculation for currentPixelSize() done via the same route (the
CAMetalLayer). Otherwise, if there is a mismatch between what the
QWindow and the layer says, Qt Quick will think that there is a
resize happening (has happened) whenever starting a new frame, and that
has far reaching consequences (suboptimal performance, increased
memory usage by buffers, etc.)
Change-Id: I114df92bf35622c99f2747420fdce401db7705a6
Pick-to: 6.1 6.0
Fixes: QTBUG-91438
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
If there is a large (3/5 of the screen in portrait or 2/3
in landscape) TextEdit on the screen, and it gets focus
on click, it will be shrank to fit the screen. Next click on this
TextEdit will restore its normal height, the next will shrink.
Pick-to: 5.15
Fixes: QTBUG-91056
Change-Id: I3dbf085cbfdc2739d537a304c16e28c58a6e01ce
Reviewed-by: Rami Potinkara <rami.potinkara@qt.io>
Reviewed-by: Ville Voutilainen <ville.voutilainen@qt.io>
Associated changes in the qtdeclarative repo now ensure that arguments
for qml plugins are handled on the calling side. This reduces the
qml-specific logic needed in qtbase and gives qtdeclarative clearer
control over qml build and install locations.
As part of that work, the INSTALL_LOCATION keyword used in
qt_internal_record_rcc_object_files() has been renamed to
INSTALL_DIRECTORY to make it consistent with the keyword used for the
same concept in other commands.
Pick-to: 6.1
Change-Id: Iebd319899f63d79fbe15ce965b84ce324c28a508
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Reformat a couple of examples correctly.
Change-Id: I2f0897267b4e3c4d7d2925f2d20cc45687278b0b
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Features that are not emitted in the current
configuration (e.g. plugin-manifests on Linux) should not yield an
error. Instead, print a warning message like Qt5's configure did.
Set insignificant feature to OFF.
Remove the now unneeded emit_if parameter from
qt_feature_set_cache_value.
Pick-to: 6.1
Fixes: QTBUG-88305
Change-Id: I0f2ce152fca5f08417038c9bd2c07639ff6a3df4
Reviewed-by: Alexey Edelev <alexey.edelev@qt.io>
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
When the source models don't have the same number of columns, the proxy
keeps only the smallest number of columns across all source models.
Afterwards, if a source model emits dataChanged in a column past
that number (a "hidden" column), the proxy needs to ignore it rather than
assert.
But also, if the source model emits a dataChanged signal across both
visible and hidden columns, then the last column number needs to be
adjusted so that the signal is correctly processed and forwarded.
Task-number: QTBUG-91253
Pick-to: 6.1 6.0 5.15
Change-Id: I939e8ec0faf41370472f86785851292e4372f72c
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Instead of having so much manual pointer manipulation and duplicated
code.
Change-Id: Ic90d8429a0eb4837971dfffd1664f8f63753440a
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
Following patterns from the other backends is insufficient with OpenGL
because we do not use real uniform buffers. There is currently a
possibility that a shader program will be bound without following it
with setting uniforms. Correct this by having a second level of tracking
of the associated srb object in the pipelines.
Pick-to: 6.0 6.1
Fixes: QTBUG-91630
Change-Id: I74a012daade826dd22c436bde06381c1233bad11
Reviewed-by: Andy Nichols <andy.nichols@qt.io>
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
A protected context is required to allow the GPU
to operate on protected resources, including protected
surfaces and protected EGLImages.
For example, GPU can post-process on protected content
like DRM protected content so that complex, nonlinear
video effects or mapping onto textures can be used.
The surface format option may be relevant for DirectX
and Vulkan in the future:
https://microsoft.github.io/DirectX-Specs/d3d/ProtectedResources.htmlhttps://www.khronos.org/registry/vulkan/specs/1.1-khr-extensions/html/chap12.html#memory-protected-memory
Change-Id: I2d155f0e68b830276690b4833b22a2bc452cdcad
Reviewed-by: Elvis Lee <kwangwoong.lee@lge.com>
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
On Gnome, we will not default to Wayland at the moment, even
if the desktop is run in Wayland mode. In this case, we warn
users that they can still select Wayland manually using
QT_QPA_PLATFORM.
Problem was: We never checked if they had actually explicitly
selected Wayland, so even in the cases where we would end up
using that QPA plugin, people would still see the warning.
This moves the check to the end of the platform selection
algorithm, after we have collected all the information.
Pick-to: 6.0 6.1
Change-Id: I0d734bd0782c5e58d6dc63f69b7d531a479ad942
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
With the existing types we can only display a list of features or
whether a feature is on or off, but we cannot display arbitrary values.
Using the new message type, it is possible to show paths or versions as
summary entries.
Change-Id: I5d16cb4b30923f3566755bd4d7440bdd1ece82f5
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Wayland does not support WindowActivation capability, which
will trigger a warning in qWaitForWindowActive(). Since the
selftests depend on matching the exact output of the tests,
it gets confused by this, even though the tests are passing.
So we suppress the warning when the capability is not
supported.
Change-Id: I1e1f530a7b237183ee4ef123618e868b746ba863
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
There is a duplicate condition judgment in QComboBox::setPlaceholderText, just delete it.
Change-Id: I570415c5930372866f290216a89260353d3992b7
Reviewed-by: Christian Ehrlicher <ch.ehrlicher@gmx.de>
Quoting from https://en.wikipedia.org/wiki/ISO_3166:
"ISO 3166 is a standard published by the International
Organization for Standardization (ISO) that defines
codes for the names of countries, dependent territories,
special areas of geographical interest, and their
principal subdivisions (e.g., provinces or states).
The official name of the standard is Codes for the
representation of names of countries and their subdivisions."
In order to prevent the word "Country" from causing ambiguity, the
document should clearly state that this enumeration refers to a country
or region.
Change-Id: I05543faa4193b3ddaf31d02c670b4b5f46e94389
Reviewed-by: Paul Wicking <paul.wicking@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
When using XQuartz on macOS, the virtual pointer device is detected as
a touchpad, not a mouse; but QXcbConnection::xi2HandleScrollEvent()
expects the device to be an instance of QXcbScrollingDevice for storage
of some state. A touchpad that has the scrolling capability must be
that type, not a plain QPointingDevice.
Fixes: QTBUG-91402
Pick-to: 6.1
Change-Id: I1b82766d4a3f87f656e56c0d8904def26fb0979a
Reviewed-by: Liang Qi <liang.qi@qt.io>
We send window id in decimal, however, it is expected to be send in hex.
This causes a mismatch and makes portal dialog to show in background.
Pick-to: 5.15 6.0 6.1
Change-Id: Ibd77199bbb4a2ad4782a0457ddc5506c6b5608fe
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QTreeView's drawTree implementation performs lazy layouting
when calling itemDecorationAt. If animations are enabled,
this can change the list of items, and invalidate the copy
made earlier.
Don't copy the list of items, use a reference instead so that
code iterating over the items later operates on valid data.
Add an assert in the private itemHeight method, it must not
be called with an index that is out of bounds.
Fixes: QTBUG-42469
Pick-to: 6.0 6.1 5.15
Change-Id: Ifdb782881447912e00baffd1c407de10a1d8d0d4
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Cleanup the mariadb driver a little bit:
- use nullptr instead NULL
- use unsigned long instead int for blob handling now that QByteArray
can handle more than 2^31 bytes
Change-Id: I79639a42a739aabdf776cc688cb5d664248aa610
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
This function returns the size in device independent
pixels, and should be used when calculating user
interface sizes.
Change-Id: I528123f962595a3da42438ca560289a29aca4917
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Draws to a pixmap contents of the requested window. On Android it works
only for QtWidgets application.
Qml application contents could not be grabbed due to lack of access to
the contents of OpenGL framebuffer from QAndroidPlatformScreen
Fixes: QTBUG-63275
Pick-to: 5.15
Change-Id: Ief7de1b709387d798138387bb735f4b48cf2928d
Reviewed-by: Ville Voutilainen <ville.voutilainen@qt.io>
