It can happen that the system is reporting a screen
width or height of zero, if the system does not know the
size instead of returning an error. You can either
set a define QQNX_PHYSICAL_SCREEN_WIDTH/QQNX_PHYSICAL_SCREEN_HEIGHT
or set it as environment variable QQNX_PHYSICAL_SCREEN_SIZE.
The preference order is:
1) Take detected size
2) Take size from env var
3) Take defined size if any define is set
Change-Id: I493a2c8988a225224587a6c323efeb9e9b7b5a54
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Do the same in QAccessibleTextWidget
I got a bit annoyed by this, should hopefully make life easier
when forward porting QPlainTextEdit
Change-Id: Ib4b90d1892974ea39ecfcbc5cad6ed0694207b58
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@nokia.com>
A new class called QAccessibleTextWidget was added.
This class should implement all methods of QAccessibleTextInterface and
QAccessibleEditableTextInterface which only need a QTextCursor, and it
defines two pure virtual methods, to obtain and set the text cursor, so
accessible implementations of widgets which use a text cursor can implement
these two methods.
QAccessibleTextEdit is now a subclass of QAccessibleTextWidget and most of
its methods were moved to QAccessibleTextWidget.
This is a forward port of ba5d7d608cc31fc63354fd74d85a1bad7780fc45 from
Qt 4.8, and is a prerequisite for forward-porting QPlainTextEdit
Change-Id: I6093c4fa7e0a77b84de779479c6074db006efec1
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@nokia.com>
This is a plugin that bridges the QAccessible world
to AT-SPI 2 on Linux.
Change-Id: I7af22621ee6a3cefc723b137b7f227a611cf6641
Reviewed-by: Jan-Arve Sæther <jan-arve.saether@nokia.com>
This does the job of keeping the terminal keyboard working when
the application segfaults but is somewhat unsafe because ioctl()
is not async-signal safe.
Change-Id: Ie6728a4252e18c29bba8f8308e6c00d4a1eb6a8e
Reviewed-by: Elvis Lee <kwangwoong.lee@lge.com>
Reviewed-by: Jørgen Lind <jorgen.lind@nokia.com>
QWindow::setParent() sets the parent to zero instead of the desired
parent, if platform window has not yet been created for the parent.
This caused QWindowsWindow::setParent() to skip setting the parent
later, when correct window was specified, as the QWindow parent-child
relationship hadn't changed. Fixed by changing the the check to use
native handles instead.
Task-number: QTBUG-26791
Change-Id: I292a1ddf746583a7268f2d07c20166995c0dd7d6
Reviewed-by: Joerg Bornemann <joerg.bornemann@nokia.com>
Followup to I5c925ae3e191244c7ab9415e4ba2fe49b93dd2af : touch was not
actually working by default. A major symptom was getting double-clicks
instead of clicks on buttons and such, making it quite hard to
interact with anything. There are some problems remaining.
Change-Id: I09fa5965036bac242ccc1f4db71f6d912445ca01
Reviewed-by: Laszlo Agocs <laszlo.p.agocs@nokia.com>
If QWindow::setGeometry() is called before QWindow::create(),
QKmsWindow::setGeometry() is not called at all. This is wrong
because there is some special size restricting happening there
which must never be skipped.
Change-Id: I321632d4fdb327b0a75e6791c019b3f4c2793888
Reviewed-by: Elvis Lee <kwangwoong.lee@lge.com>
Reviewed-by: Paul Olav Tvete <paul.tvete@nokia.com>
If glXCreateContextAttribsARB does not succeed or is not available,
we should fallback to using glXCreateContext. But we should not just
create a context with glXCreateContext by default that is being thrown
away if glXCreateContextAttribsARB succeeds.
Otherwise glXMakeCurrent with context 0 might cause an unexpected
context change when dealing with multiple contexts.
Change-Id: I7627abbe2500b4006180653a1b3b074fe7aca1d3
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Using a reference changes the value of oldState within this function,
which is undesired.
Change-Id: I9fb66e488015d6b3e586ffa2f0b05a40c095e16b
Reviewed-by: Christian Kamm <christian.d.kamm@nokia.com>
This solves a new corner case resulting from the recent refactoring of
QQnxWindow::setGeometry() that occurs when using a plain OpenGL backed
window in a single threaded environment and the user code calls
QOpenGLContext::makeCurrent(this) in the QWindow subclass ctor.
Change-Id: Iaf07fbff1cb2743f5706020a8d657a82cb9f2255
Reviewed-by: Kevin Ottens <kevin.ottens.qnx@kdab.com>
Reviewed-by: Kevin Krammer <kevin.krammer@kdab.com>
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
For Windows CE the taskbar needs to be hidden
manually.
Change-Id: Ife69a2a91457ba0c162e4e1be88f87ad22c61190
Reviewed-by: Björn Breitmeyer <bjoern.breitmeyer@kdab.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@nokia.com>
Clipboard should reacquire the clipboard whenever the content or metadata
(e.g the list of supported targets) changes. Patch enables us to monitor
changes to the clipboard through help of XFixes extension.
Cleanup xlib xa_* naming conventions
Task-number: QTBUG-26709
Change-Id: I9d47766ad9859b5628b0358b1c47e8af8fecef73
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
In Xlib, the 'length' variable gets updated with the the actual number
of 8-bit, 16-bit, or 32-bit items stored in the returned data, but xcb
returns the actual number of bytes read through xcb_get_property_value_length,
therefore the logic of calculating offset was broken.
Task-number: QTBUG-26709
Change-Id: I04de3b5c5631cfaf9b3c2c3d4513be73c569f61f
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
We explicitly do not send geometry change events to Qt from QQnxWindow
from the constructor. This prevents us from ending up in resizeEvent()
reimplementations from the QWindow ctor.
Change-Id: I045b35aa7eb23890772fe131c3d19314252f6a5a
Reviewed-by: Kevin Krammer <kevin.krammer@kdab.com>
Reviewed-by: Kevin Ottens <kevin.ottens.qnx@kdab.com>
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
If QWindow::create() gets called before resize() or setGeometry() then
the rect argument passed into QQnxWindow::setGeometry() by the ctor
is null.
This could potentially result in 0 sized buffers and EGL surface
creation being attempted. We bail out in this case and tell
user how to fix their application.
Change-Id: Ia709dba617fa266bd66b61409bdb1c980acfb1a2
Reviewed-by: Kevin Krammer <kevin.krammer@kdab.com>
Reviewed-by: Kevin Ottens <kevin.ottens.qnx@kdab.com>
Reviewed-by: Thomas McGuire <thomas.mcguire@kdab.com>
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Change-Id: I52027752630beee81a97644983dbccf973c0c2c2
Reviewed-by: Kevin Krammer <kevin.krammer@kdab.com>
Reviewed-by: Kevin Ottens <kevin.ottens.qnx@kdab.com>
Reviewed-by: Thomas McGuire <thomas.mcguire@kdab.com>
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
QWindow::setWindowState is not supposed to set the window active.
The method requestActivateWindow() should be used for that.
When switching from and to fullscreen mode we're always passing
SWP_NOACTIVATE to SetWindowPos to not change the activation state
of the window. This is inverse to the old behaviour, which did not
have an effect.
Change-Id: I339337935cdad76b3ef252202e92177f37543038
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
If the setWindowState is called in a early stage,
where the window is not yet shown, the fullscreen
flag is cached in the QWindow. Then in the call to
setVisible(true) the window gets generated, but the
fullscreen is not handled there. So set the state
while window is generated.
Change-Id: I0ea53ed8a2465da6a9973d84fb4579381543e89b
Reviewed-by: Joerg Bornemann <joerg.bornemann@nokia.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Previously synchronous window system events were
implemented by bypassing the queue and processing
the event immediately. This is not ideal since the
event order is not preserved - there might be "happened
before" events waiting in the queue.
Add QWindowSystemInterface::flushWindowSystemEvents
and change all handleSynchronous* to 1) queue the
event 2) call flushWindowSystemEvents.
flushWindowSystemEvents is almost identical to the
already existing sendWindowSystemEvents with the
exception that it does not call QApp::sendPostedEvents.
Move the common implementation to a new private function.
Task-number: QTBUG-20778
Change-Id: Ie98a83875bc0a14e335e36bed0dd9e0ed4a1dea0
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
The accessible table/list implementation would ignore
when a root index is set, resulting in trying to use row/column
for the top level index, leading to accessing invalid indexes.
(cherry picked from commit def2ee829435c705055733a0f99c0ca44a012c50)
Conflicts:
src/plugins/accessible/widgets/itemviews.cpp
Change-Id: Ic2745ab3e262ccee2d43e0d532e165d2958f519e
Reviewed-by: Jan-Arve Sæther <jan-arve.saether@nokia.com>
Use QObject::inherits() to check if the current input focus object is
a numerical input instead of trying to cast.
Change-Id: Ie476fba3d2f9f20a1ce0328043430cb6a96a8a3d
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Using the nullary version has the advantage that multiple calls
during a program run are much more efficient, since an inlined
atomic is used to store the result. It also ensures that
Q_DECLARE_METATYPE(T) has been used, whereas qRegisterMetaType<T>("T")
will happily register anything. So I've added the macro where it
was missing, or moved it to a central place when it existed
hidden.
In tst_qnetworkreply, this became a bit tricky, because a private
header is conditionally included, so moved the Q_DECLARE_METATYPE()
into a conditional section, too.
Change-Id: I71484523e4277f4697b7d4b2ddc3505375162727
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
Set QT_KMS_TTYKBD=1 to prevent this and keep vt switching, ctrl-c,
input appearing on the tty, etc. working.
Change-Id: I0071685205ae9e02536dcb71f6fef6495c48f8fb
Reviewed-by: Jørgen Lind <jorgen.lind@nokia.com>
Do not upload the entire (typically fullscreen) image
all the time.
Taken from eglfs' backingstore implementation.
Change-Id: I53db5cf54577ebea715d6cbd7215b9d2154e8960
Reviewed-by: Jørgen Lind <jorgen.lind@nokia.com>
If the SHAPE extension is not available and we try to use it, the xcb connection
will shut down and go into an error state. This is bad because there would be no
error message that would give a hint on what went wrong.
Change-Id: I1a6734bd146ca179a990b6ce896813bf76007b66
Signed-off-by: Uli Schlachter <psychon@znc.in>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
This functionality does not need libXext. The bindings for the shape extension
from libxcb-shape is enough and is available without Xlib.
Change-Id: I9f1927024c64df00ace1666998c6c6deda2ef782
Signed-off-by: Uli Schlachter <psychon@znc.in>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
This behavior is now implemented in a finer way in QtQuick and
QtWidgets, it's not needed anymore in the platform plugin.
Change-Id: Iacdcd313253627fbba094abe4e9a9c1d23a431e1
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Reviewed-by: Kevin Krammer <kevin.krammer@kdab.com>
No reason to keep a virtual method for Windows when all other similar methods
(macEvent and x11Event) have been removed, and when installNativeEventFilter
provides a much nicer solution (no need to derive from QApplication).
Change-Id: Ia2a7960e320fcbd04cef91f467900861dbb377c1
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
This allows using the kms plugin in QWidget apps having a single
(preferably fullscreen) top-level widget.
Based on eglfs' implementation. Dirty rectangle tracking is missing,
should be added later.
There is no composition so multiple TLWs will not work nicely.
Change-Id: Ia78589d1a375925ebdcc46aa20fc1619ec14d6cc
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
Reviewed-by: Elvis Lee <kwangwoong.lee@lge.com>
Previously, if the drivers' swap behaviour was single-buffered
it would fallthrough (just like DefaultSwapBehavior) and remain
single-buffered.
Change-Id: I4b93ad7a49094aa992d0b8fb3429c163bbbf655d
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Reviewed-by: Kim M. Kalland <kim.kalland@nokia.com>
No need to pass the dispatcher. Get rid of Windows logic to maintain
a stack of dispatcher associated with flags.
Change-Id: Ic2daad4b6762a46fac3274937effc188af436c9a
Reviewed-by: David Faure <faure@kde.org>
Cocoa lacked implementation of FrameStrut events, and also frameMargins on
QPlatformWindow. Fix both of these issues. Unfortunately QDockWidget also
contains a tangle of #ifdef MAC behaviour which I am unclear about. What's
included here disables some logic on Mac that seems definitely wrong -
while moving a window on Mac we now generate NonClientArea events
(as intended, I believe), but this should not cause dock-widget dragging to end.
Note the window titlebar is the only frame-strut/non-client area on Mac (as far as I can see)
Task-number: QTBUG-26296
Change-Id: Id0c6e954db64b9f9f71d16355cb92922877e5ebe
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
The previous API was hard to use (global function, no type safety,
manual chaining), and confusing (app vs dispatcher split only made
sense on Windows). Installing and removing out of order would have
the risk of setting back a dangling pointer (crash). Meanwhile QPA
added type safety, and this new API models the QObject::installEventFilter
API for ease of use. The virtual method is in a new interface,
QAbstractNativeEventFilter.
QPA was even calling the dispatcher event filter with QPA-private event
classes, which made no sense (refactoring leftover from when the code
was in the dispatcher). Now the QPA plugins trigger the qcoreapp event
filters with the actual native events directly.
Change-Id: Ie35e47c59c862383bcaf857b28d54f7c72547882
Reviewed-by: Marius Storm-Olsen <marius.storm-olsen@nokia.com>
With the change I69f4d5d504e2341555d9991c68e82beed2e8129c
IsZoomed function is no longer needed.
Change-Id: I07239f3e81509f8a966e64ee9705a04ea41cb38d
Reviewed-by: Björn Breitmeyer <bjoern.breitmeyer@kdab.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Do not let the cursor appear at some random position and then jump
back to the top-left corner when first moved.
Change-Id: I0b77763bdfa86c23435d6cc07bdc13dfcc682567
Reviewed-by: Jørgen Lind <jorgen.lind@nokia.com>
The page flip handling in QKmsScreen cannot currently cope with
contexts on multiple threads.
Change-Id: I9492095fec52fd1ad0197db00b3d6a2f58016e6d
Reviewed-by: Jørgen Lind <jorgen.lind@nokia.com>
Under Windows CE a resize event gets fired
when the style is set, this will produce
geometry change events inside Qt, which will
end up in changing the geometry back and forth.
Change-Id: I7f834997abc53760ccfe2ad97012bb74dc5c705c
Reviewed-by: Miikka Heikkinen <miikka.heikkinen@digia.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@nokia.com>
Public QtTest headers require it, so all unit tests would have to use private Qt
headers otherwise, which is not practical.
Change-Id: I5d4466ec30b6a57ebdfc34413e716e657eb51368
Reviewed-by: Girish Ramakrishnan <girish.1.ramakrishnan@nokia.com>
When turning off fullscreen mode and restoring the widget's geometry
we must inform the QWindow about the geometry change synchronously.
Otherwise QWidget::geometry() will return the old value.
Using the same technique for the state transition to fullscreen mode
without sending a separate resize event.
Autotest: tst_QWidget::saveRestoreGeometry
Task-number: QTBUG-26421
Change-Id: I869e36cd302d9a94e398f48949ab3cb7ee9cdf51
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
There is no need to query each time the window state
from windows, as we already know in which state it is.
We are also getting state change notifications. This fixes
issues under Windows CE.
Change-Id: I69f4d5d504e2341555d9991c68e82beed2e8129c
Reviewed-by: Joerg Bornemann <joerg.bornemann@nokia.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>