Summary:
ASAN found a heap-use-after-free when deleting the focused keyboard
surface in the client library. Keyboard did not track the lifetime of
the focused surface and thus one can access already freed memory.
Test Plan: Adjusted auto test to verify the variable gets cleared
Reviewers: #frameworks, #plasma
Subscribers: plasma-devel
Tags: #plasma_on_wayland, #frameworks
Differential Revision: https://phabricator.kde.org/D6741
Summary:
This is a change inspired by https://bugreports.qt.io/browse/QTBUG-61930.
When Qt closes a window due to a key press event it starts to repeat the
event as KWayland does not send a keyboard leave event. Weston on the
other hand does send out the keyboard leave. In my opinion it doesn't
make much sense to send out the keyboard leave in this situation and in
my opinion that is a client bug, but if it makes clients happy we can
send them the keyboard leave. Similar this should be done for pointer,
touch, etc.
BUG: 382280
Test Plan: Run the example added to the Qt bug and it worked fine
Reviewers: #frameworks, #plasma
Subscribers: plasma-devel
Tags: #plasma_on_wayland, #frameworks
Differential Revision: https://phabricator.kde.org/D6683
Using the new extra-cmake-modules module ECMAddQch (since 5.36.0)
this adds the option to automatically build and install a file
in QCH format with the docs about the public API, which then can be
used e.g. in Qt Assistant, Qt Creator or KDevelop.
Additionally the installed cmake config files will be extended
with a target KF5Wayland_QCH containing information about how to "link"
into the generated QCH file, which then can be used in the cmake build
system of other libraries building on this library, by
simply listing this target in "LINK_QCHS" of their ecm_add_qch() usage.
And a respective doxygen tag file with all the metadata about the
generated QCH file and used for the "linking" will be created and
installed.
Pass -DBUILD_QCH=ON to cmake to enable this.
Use the QProcess::start() variant with explicit (empty, in these cases)
arguments, so the program strings are not parsed as shell commands,
thus preserving paths with spaces as such.
Summary: client requests to toggle those states, to be used by libtaskmanager
Test Plan: setting keep above from the taskbar works
Reviewers: #plasma, hein, graesslin, #plasma_on_wayland
Reviewed By: #plasma, hein
Subscribers: graesslin, hein, plasma-devel, #frameworks
Tags: #frameworks, #plasma
Differential Revision: https://phabricator.kde.org/D5757
Summary: Verified it it send before the initial_state and adjust tests and docs accordingly
Test Plan: All unit tests pass
Reviewers: #plasma, graesslin
Reviewed By: #plasma, graesslin
Subscribers: graesslin, plasma-devel, #frameworks
Tags: #plasma_on_wayland, #frameworks
Differential Revision: https://phabricator.kde.org/D5887
Summary:
This patch adds a pid event to the plasma window management protocol. It
allows the compositor to tell allow a mapping between windows and processes.
Bumps the version number of the interface to 8 to indicate this.
Test Plan: autotest added, passed
Reviewers: #plasma, hein, graesslin
Reviewed By: #plasma, hein, graesslin
Subscribers: apol, davidedmundson, plasma-devel, #frameworks
Tags: #frameworks, #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D5747
Summary:
It's possible for the surface to be unbound when we send the leave
event; we've called Resource::unbind() of Surface, so the Surface has,
deleteLater called, but it's still a valid object, and the first check
passes.
We get in this situation because when a surface is destroyed, we're
handling text input from the same source event.
Sending a nullpointer is a protocol error, and wayland kindly closes the
connection.
This fixes my constant:
"Did the Wayland server die" error messages when running clients.
Test Plan:
Got errors after setting up qt virtual keyboard.
Had reproducible case.
Restarted kwin after this patch, now doesn't crash.
Reviewers: #plasma, graesslin
Subscribers: apol, graesslin, plasma-devel, #frameworks
Tags: #plasma_on_wayland, #frameworks
Differential Revision: https://phabricator.kde.org/D5712
Summary:
This extends the client side API to support creating popup ShellSurface
windows and the server side API to send out the popup_done request.
This is needed to properly support popup windows (e.g. context menus)
in KWin.
Reviewers: #plasma_on_wayland, #frameworks, #kwin
Subscribers: plasma-devel
Tags: #plasma_on_wayland, #frameworks
Differential Revision: https://phabricator.kde.org/D5174
The setRegion call allows a null region. This means nullptr is an
allowed value which can be passed to ConfinedPointer::setRegion and
LockedPointer::setRegion.
In that case we crash if we try to convert the Region into a wl_region.
Thus add proper nullptr check, just like in
PointerConstraints::lockPointer and ::confinePointer.
Auto test adjusted to cover the condition.
Summary:
The pointer constraints protocol is an unstable protocol and thus
the implementation follows the semantics of unstable protocols.
The protocol allows to create a constraint on the pointer - either a
lock or a confinement on a surface. Those are not activated at once, but
when the compositor actively grants it.
During lock no further pointer motion is emitted, during confinement the
pointer is kept in a certain area.
This implements T4451.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D3466
Summary:
In SeatInterface we need to get all PointerInterfaces related to a given
Surface (Client) and call a method on it. The implementation we had so
far went through all Pointers and put all PointerInterfaces into a new
temporary QVector. In most cases all we did then was iterating over the
returned vector.
Which means we created a temporary vector for nothing.
This change implements a kind of std::for_each with the constraints of
the previously used pointersForSurface which does the check that Surface
is not null and that the client matches. If a PointerInterface is found
for that, the passed in method is invoked on it.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D3295
With this change the generator is able to detect whether an interface
follows the unstable semantics. In that case the header file on server
side looks different. An enum needs to be generated containing the
interface version. Each of the generated classes has a new method
interfaceVersion returning that enum. The ctor of the class is protected
instead of private.
So far only the header side is adjusted. The implementation currently
generates not matching code.
* static void fooCallback definitions added to Private class
* static const foo_listener s_lister added to Private class
* Private::setup generates the foo_add_listner call
* implementation of s_listener added
* base implementation with a TODO marker added for the callbacks
Summary:
Pointer gestures are created for a pointer and there are two types of
gestures: swipe and pinch.
At a given time there can only be one active gesture. The implementation
in SeatInterface ensures that there can only be one active gesture.
Each gesture consists of a start event, 0 to multiple update events and
an end event. The end can also be a cancel. To better support this the
implementation doesn't follow the protocol and splits end and cancel
into dedicated methods in the server side and into dedicated signals in
the client side.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D3169
This is ancient code that is outright wrong most of the time and at best
just incredibly unnecessary.
It is also not present in the great majority of frameworks due to this.
Its wrongness comes from the fact that it hardcodes the installation path,
which breaks relocatability of the KF5 tree as it will always attempt to
find the include dir $PREFIX/KF5 (e.g. /usr/include/KF5), which may or may
not exist given that the tree was relocated.
Worse yet, in a cross-building scenario we maybe for example
build on ARM and install to /usr but for cross building take the entire ARM
tree and shift it into /arm/usr/. If we then crossbuild on that tree the
bogus include list in this framework will make sure that we always search
in /usr/include/KF5 and thus potentially load a !ARM header simply because
the relevant ARM header was not installed etc.. Similarly of course a
build in $HOME can pick up /usr/include/KF5 headers because the home ones
are missing, causing unexpected results.
This happens whenever the KDE_INSTALL_INCLUDEDIR_KF5 var is absolute, which
it usually is.
On top of all that the premise of the code in question is flawed. It seeks
to add $PREFIX/$KF5INCLUDES to the search paths (e.g. /usr/include/KF5).
This is unnecessary because the target itself is properly installed via
cmake's install(TARGETS ... EXPORT ...) function [1]. This function has
smart functionality built in which will add the passed INCLUDES destination
to the INTERFACE_INCLUDE_DIRECTORIES property of the targets (i.e. what
the useless code wants to do) [2].
So what happens is that we install the target to the
KF5 locations, which has "include/KF5" as INCLUDES location,
thus causing the correct path to be added to the includes list of the
Targets.cmake file.
In particular thanks to more internal magic in cmake it will do so with
automatically resolved root paths such that the installed tree is
relocatable and able to relatively find the other KF5/* headers. So it
does what the code in question wants to do, just correctly.
Since cmake automatically takes care of injecting $prefix/include/KF5 we
can simply get rid of the wrong custom inejection code. This makes the
generated cmake file find the correct include/KF5/ directory and stops it
from always expecting a /usr/include/KF5/ directory to be present.
[1] https://cmake.org/cmake/help/v3.0/command/install.html
[2] https://cmake.org/cmake/help/v3.0/command/install.html
> The INCLUDES DESTINATION specifies a list of directories which will be
> added to the INTERFACE_INCLUDE_DIRECTORIES target property of the
> <targets> when exported by the install(EXPORT) command.
REVIEW: 129273
CHANGELOG: Improved relocatability of CMake export
The test destroyed the connection prior to destroying registry and event
queue. Thus causing problems. Hopefully this change fixes the segfault
on build.kde.org.
The test has KWayland::Client objects like Registry as member variables
of the test object. This causes the objects to be destroyed with the
dtor after cleanupTestCase is run which destroys the connection and
Wayland server. At least on the CI system this seems to cause problem.
In general our tests do not keep any state around, especially not
KWayland::Client objects. The normal way is to have a new dedicated
client connection for every test method. This test doesn't follow this
approach at all.
In case that this change does not fix the test and still crashes on
build.kde.org the only option is to drop the test and replace it by a
new variant which follows the approach of other tests.
Reorder the cleanup code. It doesn't make sense to delete the client
side objects after deleting the server side objects. This might be a
reason for the failing tests on build.kde.org.
Summary:
In case the current selection does not have a DataSourceInterface
updating the focused keyboard surface resulted in a crash. The current
selection is sent to the DataDeviceInterface of the newly focused
client and thus a DataOfferInterface would be created for a null
DataSourceInterface.
This is a similar fix as D3148 and D3149.
Test Plan: Test case added which used to crash before
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D3150
Summary:
This is a similar condition as D3148. If a DataDeviceInterface is
created for the currently focused keyboard Surface the current selection
is sent to the new DataDeviceInterface. If the current selection does
not have a DataSourceInterface a DataOfferInterface for a null
DataSourceInterface would be created and result in a crash.
This change verifies that there is a DataSourcInterface on the current
selection prior to sending out the selection.
Test Plan:
A test case is added which simulates the condition by
using two clients.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D3149
Summary:
SeatInterface provides a way to set the current selection. This method
did not verify whether the new DataDeviceInterface actually has a
DataSourceInterface. If there is no DataSourceInterface on that
DataDeviceInterface the selection should not be sent to the current
selection owner. This results in a crash as DataOfferInterface
(correctly) doesn't expect the passed in DataSourceInterface to be null.
To ensure we don't hit this again the DataOfferInterface ctor gained an
Q_ASSERT to validate the DataSourceInterface.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D3148
The generator misdetected the destructor Requests by looking at the
arguments instead of the type of the request.
In addition the destructor handling changed in KWayland::Server since
the generator got created. There is now a shared implementation for the
Resource destruction. The generator is adjusted to generate the code for
that and implements the destruction for the Global resource destruction.
Summary:
By default a panel does not take focus. But there are panels which
should get keyboard focus. Examples in a Plasma session are the widget
explorer.
This change adds a new request to PlasmaShell interface to specify
whether a panel should get focus. The compositor can use this request to
decide whether to pass focus to a panel.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D3035
Summary:
This change adds support for auto-hiding panels. A PlasmaShellSurface
with Role Panel and PanelBehavior AutoHide can request to get
auto-hidden at a screen edge. The compositor will then not show the
surface although it is still mapped and will show it again once the
screen edge gets triggered.
The interface is extended by one new request to allow the client to
request the hiding of the surface. In addition two events are added to
inform the client when the surface got hidden and when it got shown
again.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D3007
Summary:
Especially for Xwayland windows the compositor might not have a themed
icon name. Resulting in a task manager not having dedicated icons for
Xwayland windows.
This change deprecates the way how a compositor is supposed to set the
window icon. Instead of passing the themed icon name, it is now supposed to
pass the QIcon. In case it's a themed icon the existing way to pass to
the client is used.
Otherwise a new event is used to inform the client that there is an icon
- no data is transmitted at this point. The client can then create a
file descriptor and pass it to the compositor. The compositor serializes
the icon into the file descriptor and the client can read from it. This
all happens transparently on client side there is no api change at all.
The writing and reading of the icon is done in a thread. Due to that
Qt5::Concurrent is now a required dependency instead of an optional
dependency.
Reviewers: #plasma_on_wayland, hein
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D3049
Summary:
Qt supports setting generic data on a QWindow. This change implements
the callback properly and forwards the property to the surface extension
instance. This allows the compositor to e.g. listen to
DynamicPropertyChangeEvents and get access to all additional set
properties.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D3045
Summary:
This is similar to what already exists for Surface. With this methods
it's possible to get a ShellSurface from an existing QWindow and to
perform low level native calls directly for the ShellSurface.
Similar calls will also be needed for XdgShellSurface, though the
xdg_shell_surface is not yet available through the native interface.
Test Plan: Used in breeze to trigger window moving
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D3018
Summary:
This is what Weston does. With this change clients can create
multiple wl_pointer instances and thus get events reported to all of them.
This will be needed to e.g. support drag on empty area in Breeze on Wayland.
A similar change was done for wl_keyboard already in 25dbc84d.
Test Plan: Seat auto test still passes
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D3016
Summary:
In the situation that a wl_client gets destroyed while still
wl_resources are around it can happen that one of them calls into the
ClientConnection during the cleanup handling which gets triggered at the
same time. This can then trigger a crash.
This change uses deleteLater for the ClientConnection and sets the hold
wl_client pointer to null instead of deleting directly. So the
ClientConnection is still around while the Resources gets cleaned up.
This is similar to the cleanup of Resource where on unbind the
wl_resource pointer is set to null and the Resource gets delete later.
BUG: 370232
FIXED-IN: 5.28
Reviewers: #plasma, bshah
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D3004
Summary:
This change implements the zwp_relative_pointer_v1 protocol which allows
to send relative motion events.
The (unstable) protocol consists of a RelativePointerManager which
creates RelativePointers for a given Pointer. This interface currently
only has one event to report the relative motion. It carries the delta,
the non-accelerated-delta and a timestamp in microsends granularity.
On the server side the implementation is mostly internal. Once a
RelativePointerManagerInterface is created one can send relative motion
events through the SeatInterface. The SeatInterface takes care of
sending it to the responding RelativePointerInterface. The protocol does
not restrict the sending of "normal" and relative motion events. Thus it
can be combined in any way one wants. This allows to have a rather
simple implementation. A user of the SeatInterface can just start to
feed the relative motion events (if the information is available) to the
SeatInterface together with the pointer events.
On client side a new RelativePointerManager and RelativePointer class
are added. The RelativePointerManager creates the RelativePointer for a
given Pointer. The event sent to RelativePointer is transformed in a
normal signal.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2978
Summary:
When changing the selection the previous selection needs to be
cancelled. This is already done in the "normal" updating of the
selection. If the previous selection doesn't get cancelled QtWayland is
not able to accept anything new from the clipboard. The setSelection
didn't cancel it yet, due to that copy from Xwayland to QtWayland
windows doesn't work in KWin as KWin uses the setSelection call for the
Xwayland clipboard.
With this change the cancelling of previous selection is moved into a
dedicated method and called from the normal way and the setSelection
way.
Reviewers: #plasma_on_wayland, bshah
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2997
Summary:
This is what Weston does. With this change clients can create multiple
wl_keyboard instances and thus get events reported to all of them. This
will be needed to e.g. support KModifierKeyInfo on Wayland.
Similar changes are probably also needed for pointer and touch.
Test Plan:
Auto test for seat still passes. A custom change to kscreenlocker
is able to report whether caps lock is on with this change.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2963
Summary:
This change makes use of the internal key state in better way. If a
key is not considered pressed, no key release is sent. This can happen
for example if the compositor grabs a key press (global shortcut) but not
the release. The Wayland client cannot do anything with the release as it
never got the press. Thus it doesn't make sense to send the release.
Similar if a key is already pressed, it doesn't make sense to send
another press event. This ensures that if the server sends in repeating
key presses they are filtered out. Key repeat is handled on client side.
Also if several physical keys send the same key code, pressing them at
the same time won't send double press/release event.
This change might cause regressions in KWin in case KWin does not handle
the situation correctly. But that would be a bug in KWin which needs to
be fixed there. If it causes regressions the bug might have shown in
other situations as well.
BUG: 366625
FIXED-IN: 5.27
Reviewers: #plasma_on_wayland, #kwin
Subscribers: plasma-devel, kwin
Tags: #plasma_on_wayland, #kwin
Differential Revision: https://phabricator.kde.org/D2786
Summary:
According to the Wayland documentation a data source needs to be
cancelled whenever it is no longer valid. A reason for no longer being
valid is that the data source has been replaced by another data source.
So far KWayland did not implement this aspect which resulted in clipboard
breaking in QtWayland applications. As soon as one copied once from an
application it was no longer possible to paste to it from another
application.
With this change the data source gets properly cancelled and also
ensured that the server code doesn't run into a possible crash condition
when trying to cancel an already unbound data source.
BUG: 368391
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2743
Summary:
This change implements support for the wl_surface events enter and
leave. Those events are emitted whenever a surface becomes visible on
an output by e.g. mapping the surface, moving or resizing it. Similar
the leave is sent whenever the surface is no longer on an output.
The server side is not yet fully complete yet. It also needs to emit
when the client binds the output another time and needs to send a
leave before destroying the output.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2528
Summary:
This change starts to track all Outputs and introduces a static
method to get an Output* for a wl_output* in case the wl_output* is
known to Output.
This is needed for the enter and leave events on wl_surface.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2526
The test started to segfault on build.kde.org. It looks like an event
survives the destruction of the server. Thus this change tries to
dispatchEvents once more before destroying everything.
It started to fail on build.kde.org with a segfault. While I cannot
reproduce this locally it looks like client objects survive to the next
test and thus cause issues.
This change attempts to reduce the risk by making sure that everything
is cleaned up correctly. If that fixes the issue we can look into a
proper fix.
Summary:
Instead of hard depending on the include of linux/input.h we check
whether that include file exists and properly ifdef all usages.
Unfortunately there is no replacement for those parts which do mapping
of input event codes.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2344
On build.kde.org the Surface is often already destroyed when returning
from waiting on client disconnected.
This change tries to handle the situation better: only if the signal
is not yet emitted, try to wait for it.
Newer Weston renamed the interface to the unstable naming scheme. As
KWayland does not support this interface yet, the tests are failing.
This change skips the test if the Weston is too new and doesn't provide
the expected interface any more.
If the client deletes the currently entered Surface the Pointer should
return null in Pointer::enteredSurface and not invalid memory.
At the same time a now incorrect assert is removed. After the client
deletes the Surface it might still get a left event for the Surface.
In that case a comparison between the enteredSurface and the Surface
from the leave event would fail.
Reviewed-By: bshah
Summary:
This allows the compositor to expose the absolute window geometry to
processes which need it.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2183
Summary:
If we try to send a pointer enter on a null resource of a Surface
(e.g. after unbound) we hit a marshalling error:
error marshalling arguments for enter (signature uoff): null value passed for arg 1
The added test case hits this error without the change and passes
cleanly with the change.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2159
Summary:
This change introduces support for the unstable xdg-shell interface in
the server. The implementation is based on version 5 of the unstable
interface. This is the version used by toolkits like e.g. GTK.
There is also a version 6 of the protocol under development which is
incompatible. This makes it difficult to implement it in a backward
compatible way.
Because of that the implementation is a little bit different to other
interfaces and inspired by the TextInput interfaces:
On client side an XdgShell class is exposed which does not represent
it directly. Instead it delegates everything to an XdgShellUnstableV5
implementation. For the Surface/Popup the same is done.
In the Registry it's possible to create an XdgShell and it accepts
the XdgShellUnstableV5 and in future will accept XdgUnstableV6, etc.
On server side it also follows the approach from TextInput. That is
there is a version enum which gets passed to the factory method in
Display. It currently supports only V5, but in future can be extended
for V6. As there is lots of similar code between wl_shell, xdg_shell
and in future xdg_shell_unstable_v6 a templated GenericShellInterface
class is added which combines the common parts.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2102
Summary:
When setting the keyboard focus the server needs also to send the current
selection to the client. So far KWayland only sent the selection if it was
set. That is if the last focused client cleared the selection it was not
updated and the client might have had an outdated selection.
To prevent this situation the server now explicitly sends the clear to the
client on enter if there is no selection. Also if the selection is cleared,
the SeatInterface now unsets it's current selection to make sure that the
next focused keyboard will get the clear selection sent.
Test Plan:
Existing test case adjusted and a new test case added which
simulates the interaction of two clients.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2091
Summary:
Even if the XDG_RUNTIME_DIR is not set, KWayland should not crash
if methods are invoked which only make sense after the server was
started successfully. This was not yet the case for
Display::dispatchEvents.
Also the wl_display was not destroyed in the situation that creating
the socket failed.
Test Plan:
Test case added which simulates the situation of no
XDG_RUNTIME_DIR being set.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2077
Summary:
This change adds a parent_window event to Plasma Window. From server
side it's possible to specify that a window is a transient for another
window - that is it has a parent window.
On client side this is exposed respectively with a new change signal.
Reviewers: #plasma_on_wayland, hein
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1779
Summary:
As Wayland doesn't have a proper ToolTip window type yet, we add it
to PlasmaShellSurface, so that we can use it for Plasma's tooltips.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2030
Summary:
On client side the newer wl_pointer_release is used which is a
destructor call. On server side the shared destroy callback is used
and it's ensured that KWayland doesn't crash if called into the
PointerInterface between unbound and destroyed.
Test Plan:
Test case extended to cover the condition of an unbound
PointerInterface.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2037
Summary:
On client side the newer wl_keyboard_release is used which is a
destructor call. On server side the shared destroy callback is used
and it's ensured that KWayland doesn't crash if called into the
KeyboardInterface between unbound and destroyed.
Test Plan:
Test case extended to cover the condition of an unbound
KeyboardInterface.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2036
Summary:
On client side use wl_touch_release to get into the proper destroy
handler on server side. There the shared destroy implementation is
used.
The test case is extended to verify the condition and ensure that
our code doesn't crash in case SeatInterface calls into the already
unbound TouchInterface.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D2035
Summary:
So far SeatInterface handled automatically which DataDeviceInterface
holds the current clipboard selection. While this works fine and is
correct it doesn't support use cases like a clipboard manager where
the clipboard is hold by a different ClientConnection than the one
from the focused keyboard.
This change allows to manually set the selected DataDeviceInterface
to override the automatic selection, though the automatic selection
is still in place. Thus the next update of a selection will override
the manually set selection again.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D1972
Summary:
A Resource might be unbound, but not yet destroyed. In that case this
return a Resource instead of nullptr.
This change adds an explicit nullptr check.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D1938
Summary:
Creates a PlasmaShellSurface and allows to set the different roles
through command line argument.
Needed to verify that KWin properly handles the notification type.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D1932
Summary:
Plasma needs to be able to mark windows as Notifications, so that KWin
can put them into the right layer.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D1928
Summary:
The QtExtendedSurfaceInterface might be unbound and thus the resource
might be null when calling into close. Thus we need to do a nullptr
check.
Hit a crash there.
Test Plan:
Unfortunately no test case as we don't have a client side
implementation for this.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D1961
Summary:
This is a gotcha moment:
1. Create Surface with id 1
2. destroy that Surface
3. Create another Surface
Now if in step 3 the id is by pure chance getting reused and also 1, the
wl_resource pointer of the SurfaceInterface of step 1 and step 3 are
the same. This is rather unexpected and causes problems.
When creating a ShellSurface in step 1 and step 3 it would fail. KWayland
finds a ShellSurface which was already created for the Surface. The same
can happen with QtSurfaceExtensionInterface and PlasmaShellInterface which
also go into error state.
On client side this would trigger a protocol error and terminate the
application. An easy way to reproduce is opening the file open dialog
from within Kate multiple times.
This change addresses this problem by setting the surface pointer in
those classes to null when the parent Surface gets destroyed. Thus
creating a new ShellSurface won't find the old referenced Surface any
more.
For ShellSurface and PlasmaShellSurface a test case is added which hit
the condition without this change. For QtSurfaceExtension we don't have
the client side, so we cannot really simulate the condition.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D1937
Summary:
So far for internal cleanup we mostly listen to QObject::destroyed.
When a Resource gets unbind the wl_resource is set to null and the
Resource gets deleteLater. This creates a short time frame when the
Resource is still there, but the wl_resource is null. For most internal
usages the Resource is completely useless at that point and should no
longer be considered. So far it was still considered and could hit
crashers, if a code path did not nullptr check. Unfortunately
libwayland-server is not nullptr safe: if called with a null value it
tends to crash.
So this check introduces a new signal unbound which can be listend to
in addition to the destroyed signal. It's used in SeatInterface for
DataDeviceInterface, where we experienced a crash related to that.
A test case is added which exposes the crash, but it already needs
the unbound signal to get into the crashy condition. The actual crash
is fixed twice - with the help of the unbound signal, but also by
introducing the nullptr check where it's needed.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Tags: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D1868
Summary:
The fact that the buffer is still referenced - that is used - when the
BufferInterface gets destroyed is an error, but it is not fatal.
Unfortunately KWin/Wayland is still hitting this assert from time to
time and the assert is not helping to find the cause as a backtrace
does not show where a reference is still hold.
This change removes the hard assert by a soft warning. The advantage
of the warning is that the compositor is not killed and that we can
observe the reason and find a usage pattern which triggers the condition.
With that we will hopefully be able to find the case where the buffer
is still referenced when being destroyed and fix that.
Reviewers: #plasma_on_wayland
Differential Revision: https://phabricator.kde.org/D1783
Summary:
This improves the cleanup of a shadow from client side. The server now
notices when the client destroyed the shadow.
Reviewers: #plasma_on_wayland
Subscribers: plasma-devel
Differential Revision: https://phabricator.kde.org/D1789
Sometimes the test is failing due to a wait on the signalspy not
working and then the cleanup triggers a heap-use-after-free.
This change tries to address the problem by using a QTRY_COMPARE
instead of a wait on signalspy.
If a PlasmaShellSurface got already created for a Surface, the
PlasmaShell::createSurface should return that one. This change justs
adds a check to an autotest to verify this condition.
Summary:
The event is sent to the client once all initial state is transmitted.
This means the client is able to see the PlasmaWindow completely created
and not in the intermediate state with further updates being pushed after
being created.
The client side API is adjusted to emit the windowCreated signal after
the initial state event is received. In addition if the window is already
unmapped, the signal will never be emitted which means the not valid
windows are not exposed to the client at all.
The tests are adjusted to reflect the new reality, which in most cases
just means removing the comment that this needs to be improved.
There is one kind of unrelated change included: when an empty icon is
set, the client side now creates a QIcon() instead of going through
QIcon::fromTheme. This wrong behavior was exposed now by the auto tests.
Reviewers: #plasma, hein
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1773
Summary:
If the surface passed in the transient request is to the surface of the
shell surface, KWin aborts. It's an obvious invalid request, so handle
it properly by sending an error.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1755
Summary:
So far ConnectionThread dispatched events without checking for (protocol)
errors. But Wayland errors are considered fatal. We do need to consider
them.
This change introduces a way to detect errors and expose them in
ConnectionThread. Errors are handled gracefully, they are exposed, but
not considered application fatal. E.g. a new ConnectionThread could be
created afterwards.
This allows to add tests to verify error conditions.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1753
Summary:
The idea of shipping a test server is to have something like Xvfb for
Wayland. To be able to run a test application with a fake Wayland server.
To make this super easy the test server binary is installed into libexec
directory and provides a cmake function to run a test application with
the test server.
The test server takes full control over the process. It's a guiless
application and starts the passed test application once it is fully
set up. The environment is setup to have the test application connect
to the fake server (WAYLAND_SOCKET env variable and QT_QPA_PLATFORM).
When the started application finishes the test server goes down and
exits with the exit value of the test application. This allows a good
integration with ctest.
The test server is a virtual server which supports the following
interfaces:
* Shm
* Compositor
* Shell
* Seat
* DataDeviceManager
* Idle
* SubCompositor
* Output (1280x1024 at 60 Hz with 96 dpi)
* FakeInput
This is sufficient to bring up a QtWayland based application and
allows some basic interactions from a test application (e.g. fake
input).
So far the server fakes a repaint every 16 msec, but does not yet
pass events to the test applications.
To integrate this into an application for testing use:
find_package(KF5Wayland CONFIG)
add_executable(myTest myTest.cpp)
target_link_libraries(myTest Qt5::Gui Qt5::Test)
kwaylandtest(myTest)
When now running ctest in the build directory the test server gets
started and will start the myTest binary and report the passed/failed
in the expected and normal way.
This way a test case can easily be run against both X11 and Wayland.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1726
Changelog: Virtual framebuffer server for auto tests
Summary:
So far we directly destroyed the resource on the server side. But this
causes a wayland error when the client tries to cleanup. This results in
the client being destroyed. A problem which brings down plasmashell
regularly when short living windows are shown. This happens e.g. in
Dolphin with the adress auto complete.
This change addresses the problem by creating a temporary
PlasmaWindowInterface for the already unmapped window. It doesn't get
added to the list of known windows and it's only purpose is to properly
handle the unmap and the destroy of the just created resource.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1734
Summary:
Destroying the SlideInterface on the server side before the client has a
chance to cleanup results in a protocol error:
wl_display@1: error 0: invalid object 7
Which would terminate the client. If we would not destroy the resource,
but only delete the SlideInterface it could result in heap-use-after-free.
So just don't do anything, the client needs to cleanup which will result
in the SlideInterface being deleted.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1714
Summary:
The destructor was not properly implement. Let's use the generic one
from Resource.
Test case is adjusted to verify that the SlideInterface gets cleaned up.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1713
Summary:
Unlike the other cases this one is not as dangerous as the shadow
protocol doesn't have a destructor request (yet).
Once that is added the problem would be the same: destroying the
ShadowInterface when the parent SurfaceInterface gets destroyed would
result in a protocol error on client side.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1711
Summary:
Destroying the ServerSideDecorationInterface on the server side before
the client has a chance to cleanup results in a protocol error:
wl_display@1: error 0: invalid object 7
Which would terminate the client. If we would not destroy the resource,
but only delete the ServerSideDecorationInterface it could result in
heap-use-after-free.
So just don't do anything, the client needs to cleanup which will result
in the ServerSideDecorationInterface being deleted.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1710
Summary:
Destroying the ContrastInterface on the server side before the client has
a chance to cleanup results in a protocol error:
wl_display@1: error 0: invalid object 7
Which would terminate the client. If we would not destroy the resource,
but only delete the ContrastInterface it could result in
heap-use-after-free.
So just don't do anything, the client needs to cleanup which will result
in the ContrastInterface being deleted.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1709
Summary:
Destroying the BlurInterface on the server side before the client has a
chance to cleanup results in a protocol error:
wl_display@1: error 0: invalid object 7
Which would terminate the client. If we would not destroy the resource,
but only delete the BlurInterface it could result in heap-use-after-free.
So just don't do anything, the client needs to cleanup which will result
in the BlurInterface being deleted.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1708
We saw it fail on build.kde.org a few times and it looks like a
entered spy gets triggered timing dependent (no surprise it's on
client side). So better check whether the entered spy got delivered
already and if not wait for the event.
Summary:
This change standardizes the behavior regarding the destructor request.
The destructor should destroy the resource and nothing else. The
Wayland library invokes the static unbind method once the resource is
destroyed. The implementation provided by Resource::Private::unbind
triggers a delete later on the Resource. So there is no need to trigger
a deleteLater from the destructor request callback.
This change adds a generic implementation to Resource::Private which is
now used by all inheriting classes replacing the custom implementations.
Test Plan:
For a few Resources the test is extended to ensure that the Resource
gets deleted on server side.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1679
Summary:
Not needed as the dtor of the Resource ensures that the wl_resource
gets destroyed.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1680
The text input change creates an additional serial, so the check
for last generated serial on the Display fails. The test is now
adjusted to the new semantics. A more reliable way would be to
verify the serial on the SeatInterface, though.
Summary:
This change introduces support for text input. Text input allows to
compose text on the server (e.g. through a virtual keyboard) and sent
the composed text to the client.
There are multiple interfaces for text input. QtWayland 5.6 uses
wl_text_input, QtWayland 5.7 uses zwp_text_input_v2.
wl_text_input is from pre Wayland-Protocols times and considered as
UnstableV0 in this implementation. The other interface is UnstableV2.
Unfortunately the V2 variant is not yet part of Wayland-Protocols, but
used in Qt.
The implementation hides the different interfaces as good as possible.
The general idea is the same, the differences are rather minor.
This means changes to how interfaces are wrapped normally. On client
side in the Registry a manager is factored which represent either of
the two interfaces. Similar on the server side Display's factory method
takes an argument to decide which interface should be factored. This
way a user of the library can expose both interfaces and thus be
compatible with Qt 5.6 and Qt 5.7 onwards.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1631
Summary:
When destroying a SurfaceInterface all callbacks are getting destroyed.
This used to iterate over the callbacks and performing
wl_resource_destroy on them. This triggered the destroy handler which
removes the resource from the callback list. Which means removing from
the list we are iterating on. This could result in a double delete or
accessing invalid memory.
This change copies all callbacks to a temporary list and clears the
normal lists. So the destroy handler does no longer modify the lists
currently being iterated on.
Test Plan: Added a test case which crashed with previous code
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1677
The ShellSurface might have the resource destroyed, before being deleted,
so there is a short time frame where resource might be null.
Crash was caught by KWin.
Reviewed-By: sebas and notmart
A downstream KWin test shows a possible heap-use-after-free if we
access the wl_client pointer. Basically we get the client disconnected
just before the focused surface gets unbind. Thus for a short moment
the ClientConnection pointer is gone. This needs to be extended with
a test case, but for the moment it should be good enough to get KWin
green again.
Summary:
So far the server component performed manual cleanup in some cases
when a client disconnects. But this is not needed: the Wayland library
calls the static unbind methods which do cleanup. If we cleanup ourselves
this can result in double deletes in the worst case, so let's only use
the Wayland functionality.
Adjusted:
* RegionInterface
* SurfaceInterface
* ShellSurfaceInterface (doesn't take a parent anymore)
* DpmsInterface
* QtSurfaceExtensionInterface
* KeyboardInterface
* PointerInterface
* TouchInterface
* DataOfferInterface
* PlasmaShellSurfaceInterface
For each adjusted case a test case is added to verify that the cleanup
works. Exceptions are DpmsInterface as the actual Resource is not exposed
at all in the Server component and DataOfferInterface as that is server
side created.
Reviewers: #plasma
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1640
From time to time the test is failing on build.kde.org with an ASAN
heap-use-after-free error. From my investigation this seems to be caused
by the OutputDevice being constructed on the stack and being destroyed
while handling Wayland events, but before all are handled.
The test mostly operates on the changed signal. There is also a done
signal emitted later on. Wayland sends the done after a set of changes
is transmitted. Thus the test is adjusted to wait for done instead of
changed. So we can ensure that all events are handled before the object
gets destroyed.
I have never been able to reproduce the problem locally, so I cannot
guarantee that the issue is solved for good. If it still happens more
investigation will be needed.
Summary:
When the model gets destroyed the lambda connections were still invoked
and could cause crashers.
Test Plan: Test case added and each one verified that it crashed
Reviewers: #plasma, hein
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1624
Summary:
There is a possibility that a PlasmaWindow is unmapped when the
PlasmaWindowModel gets created. In this situation the unmapped
PlasmaWindow will be deleted in the next event cycle. So far
PlasmaWindowModel didn't handle this situation and the model might
hold deleted objects due to this.
This change addresses this potential problem and ensures the model
gets updated when a PlasmaWindow is deleted.
Test Plan: Test case which exposes the problem is added
Reviewers: #plasma, hein
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1622
Summary:
So far when the active PlasmaWindow got unmapped or destroyed, the
PlasmaWindowManagement didn't update the activeWindow. This means it
could expose a deleted object through it's API which could result in
a crash.
This change addresses the problem by updating the active window when
a window gets unmapped or destroyed.
Test Plan: Tests added which exposed the problem
Reviewers: #plasma, hein
Subscribers: plasma-devel
Tags: #plasma
Differential Revision: https://phabricator.kde.org/D1621
Summary:
The protocol is extended by a dedicated destructor request. When a
PlasmaWindow is umapped we no longer destroy the resource directly,
but only send the unmap. The client is then supposed to clean up
(which it already did in that case) and will invoke the destructor.
The PlasmaWindowInterface object will be automatically deleted after
the unmap once all resources bound for it are destroyed.
The tests are extended by two new test cases which triggered protocol
errors on the client side prior to this change.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1594
Summary: To use pid_t one should include sys/types.h -- else the build fails on FreeBSD.
Reviewers: graesslin
Reviewed By: graesslin
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1579
Coverity complains (in CID 1335150) about `Generator::parseInterface()`
since the default copy-ctor for `Interface` will end up copying
uninitialized memory (the unset `m_factory` member). Fixed by
initializing the m_factory.
REVIEW:127836
Summary:
Exposes closable state in the window model and adds tests. This was
included in the protocol and interface, but missing from the model.
Test Plan: Autotest extended.
Reviewers: graesslin
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1435
Summary:
Analogous to NET::ActionChangeDesktop.
KWindowInfo::actionSupported(NET::ActionChangeDesktop) is hardcoded
to return true in kwin, but that's not how it should be; as this will
be fixed later the Wayland protocol needs this state bit as well for
parity.
Test Plan: PlasmaWindowModel test is extended.
Reviewers: graesslin
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1448
Summary:
Adds states and request methods to determine whether a window can be put into
move or resize modes, and request move and resize modes, respectively.
request* naming pre-discussed with Martin. I chose to add *Mode to be more
explicit as well as avoid namespace conflicts with possible other Move/Resize
methods in the future.
Since these are not toggleable states, there is no requestToggle* methods.
Protocol version is not bumped (also pre-discussed with Martin) since we have
pending changes already bumping to 3.
Depends on D1417.
Test Plan: PlasmaWindowModel test extended.
Reviewers: graesslin
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1432
Summary:
This adds support for 'shadable' and 'shaded' states to the protocol and
to the client and server classes, as well as the window model.
Test Plan: The PlasmaWindowModel test has been extended to test the new states.
Reviewers: graesslin
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1417
Summary:
Surface/SubSurface and Shadow handle the case that an attached buffer
gets destroyed by the client. So far we didn't have this code covered,
but it's rather important as incorrect reference counting can hit
asserts.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1403
Summary:
KWayland does not have a client implementation of the QtSurfaceExtension
protocol. Thus the test is different: it starts a helper binary which
creates a QWindow. The test closes that one which should terminate the
started applications.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1388
Summary:
Basic functionality is covered. Changing of shadow elements not covered,
there seems to be lacking server API for that - no change signal.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1398
Summary:
Now also tests:
* requesting mode from client side
* update supported change on server side
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1357
If the focused pointer and keyboard surface is the same we use pointer
clicks as a hint to which child surface should have keyboard focus.
Keyboard focus handling for sub surfaces is rather limited overall.
We just don't have a good model on how to determine which child surface
should get the keyboard focus. When passing focus to a surface there
is no way to know which of the sub-surfaces should get the focus.
Ideally the client should handle this, but that's just not the case.
The best we have is a reference through the pointer. But that's of
course also limited. Keyboard focus passed to the surface for another
reason (Alt+Tab) cannot select the proper sub-surface without interaction
from another input device.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1330
Summary:
The idea behind this change is to make the existance of sub-surfaces
an implementation detail for pointer events. The user of the library
does not need to care about on which sub-surface the pointer is on.
It only needs to care about the main surface and passes the focus to
the main surface.
Internally the PointerInterface takes care of sending the enter to
the sub-surface at the current pointer position. Also whenever the
pointer position changes, the PointerInterface evaluates whether it
triggered a change for the focused sub-surface and sends enter/leave
events accordingly. If the focused sub-surface does not change, it
sends motion events as normally, but of course under consideration
of the sub-surface position.
Overall this means that from pointer usage perspective a user of the
library doesn't need to care about the fact that there are sub-surfaces
at all. The library does the correct thing for it.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1329
Summary:
The new method returns the (child) surface at a given surface position
taking care of stacking order, whether surfaces are mapped, etc.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1319
Summary:
This change introduces a damage tracking feature in SurfaceInterface.
So far the SurfaceInterface only exposes the damage compared to the
last attached buffer. But this is not always usefull for the user of
the library. E.g. if:
* server renders
* client damages buffer and commits
* client damages buffer and commits
* server wants render
In this situation the last damage information is not helpful to the
user of the library. It would need the combined damage information over
all attached buffers.
The new API combines the damage of the two commits in the example above.
The user of the library can then call resetTrackedDamage once it
processed the current damage (e.g. by updating the OpenGL texture).
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1281
Summary:
QtWayland doesn't map the parent sub-surfaces in a sub-surface tree.
According to the spec this would mean also the child sub-surface is not
mapped. But being strict according to the spec will make applications
like SystemSettings fail badly. Embedded child windows will not be
rendered and QtWayland is going to hard freeze. This is not acceptable,
thus we need to workaround this QtWayland bug till it's fixed.
It's worth mentioning that Weston as the reference compositor also
doesn't handle this situation according to spec and renders the
sub-surface. See https://bugs.freedesktop.org/show_bug.cgi?id=94735
The difficult part for the workaround is to determine whether a surface
should be considered unmapped. E.g. when the parent gets unmapped we need
to really unmap it. But what's the difference between an unmapped parent
surface which should be considered mapped and an unmapped parent surface
which should be considered unmapped?
The implementation goes with considering a new sub-surface always as
mapped - independently of whether it ever got a buffer attached. As soon
as it had a buffer attached and it gets unmapped again, it will go back
to a standard conform way.
The behavior now is not standard conform, thus the autotest is adjusted
to have QEXPECT_FAIL for the now no longer standard conform areas.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1250
Summary:
If a Surface doesn't have a buffer attached and a null buffer gets
attached the buffer state doesn't really change. Thus neither the
unmapped signal nor the damaged signal should not be emitted.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1261
Summary:
New test case which verifies the behavior when a surface is considered
mapped or unmapped in a sub-surface tree.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1248
Summary:
In a SubSurface tree a Surface is only considered mapped if the Surface
has a buffer applied and the parent Surface is mapped.
The added method implements this check. It's useful for the compositor
to easily check this condition as it allows to easily figure out whether
a SubSurface needs to be rendered and it's also useful for implementing
the input handling as a not mapped sub-surface should not get any input
events.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1247
Summary:
This is a workaround for https://bugreports.qt.io/browse/QTBUG-52118
It should take effect whenever the parent surface's state is applied,
but QtWayland never commits the parent surface. Thus the position is
always wrong.
Having a workaround for that in our server code is not good, but better
than completely broken applications such as Systemsettings.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1212
Summary:
From spec:
If the wl_surface associated with the wl_subsurface is destroyed,
the wl_subsurface object becomes inert. Note, that destroying either
object takes effect immediately.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1211
Summary: Fixes errors in API of methods which should be const not being const.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1214
Summary:
See: https://bugreports.qt.io/browse/QTBUG-52092
Freeze happens if a sub-surface is rendered to before the main surface
is rendered. The compositor has no chance to know that this is a window
which needs to be rendered, thus the application might freeze without
ever becoming visible.
Famous example applications being affected: all kcms with a nested
QQuickView. E.g.: kcmshell5 kwineffects
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1208
Summary:
Recursively go up to the main surface, that is the top level surface
which doesn't have a parent.
This is useful to know to which surface tree a sub-surface belongs to.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1207
QtWayland doesn't commit the parent surface when creating a sub-surface.
This results in a QtWayland application to freeze as it renders to the
surface and waits for the frame rendered, which it will never get as the
Compositor waits for the commit on the parent prior to mapping the
sub-surface.
To work around this behavior, we apply the adding/removing directly.
The behavior around this is actually not fully documented, so QtWayland
is not wrong per se. See:
https://lists.freedesktop.org/archives/wayland-devel/2016-March/027540.html
Once this is properly clarified and implemented in the Client, we should
revert this change.
Differential Revision: https://phabricator.kde.org/D1191
The idea behind this signal is to notify whenever the tree of sub
surfaces changes in a way that a repaint of the Surface is required.
Possible situations are:
* surface damaged
* surface unmapped
* subsurface added/removed
* subsurface moved (position changed)
* subsurface stacking changed
Ideally it would be possible to provide the actual area which needs
repainting, but due to the possible complexity of the tree, synced
and desynced changes this doesn't look worth the effort. A user of
the signal might trigger too many repaints with it, but if it really
wants to be only notified about the actual changes, it can just track
the individual sub-surfaces.
When committing the state of a sub-surface, the state should not
be immediately applied if the sub-surface is in synchronized mode.
Instead it should be cached and only applied after the parent surface's
state is applied.
To implement this the Surface::Private has now a third cached state
buffer. When committing the state is either swapped between pending and
current or pending and subSurfacePending. Once the parent state is
applied the state is swapped between subSurfacePending and current.
The logic for applying state changes is changed. Instead of copying the
complete state object, the individual state changes are now copied and the
source gets completely reset to default values. Only the children tree is
copied back, as that list needs to be modified.
The mode is not sufficient to determine whether a SubSurface is in
synchronized mode.
Quoting spec:
"Even if a sub-surface is in desynchronized mode, it will behave as in
synchronized mode, if its parent surface behaves as in synchronized mode.
This rule is applied recursively throughout the tree of surfaces.
This means, that one can set a sub-surface into synchronized mode, and
then assume that all its child and grand-child sub-surfaces are
synchronized, too, without explicitly setting them."
The test application creates a sub-surface tree consisting of overall
three surfaces:
* blue main surface
* red sub surface
* green sub surface to the red sub surface
All surfaces are in synchronized mode. There is a timer to turn the
green surface into yellow after five seconds.
Summary:
The wrapper for wl_surface::set_buffer_scale was still missing.
Main reason for implementation is the need for the added auto-test.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1188
Summary:
The ServerSideDecorationManager gains a new event which gets sent
to the client when it binds the manager. The event indicates the
default server decoration mode used by the server. This allows the
client to know the decoration mode when it creates a decoration and
thus does not need to roundtrip to the server to get the mode.
Reviewers: #plasma, sebas
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1129
DataDevice exposes new methods to get the DataOffer provided through
drag'n'drop and also signals to indicate the various changes on the
DataDevice during drag'n'drop.
This allows to add a first basic auto test for the drag'n'drop
functionality covering both server and client side.
Summary:
How drag'n'drop works on Wayland:
When a surface has a pointer grab and a button pressed on the surface
(implicit grab) the client can initiate a drag'n'drop operation on the
data device. For this the client needs to provide a data source
describing the data to be transmitted with the drag'n'drop operation.
When a drag'n'drop operation is active all pointer events are interpreted
as part of the drag'n'drop operation, the pointer device is grabbed.
Pointer events are no longer sent to the focused pointer but to the
focused data device. When the pointer moves to another surface an
enter event is sent to a data device for that surface and a leave
event is sent to the data device previously focused. An enter event
carries a data offer which is created from the data source for the
operation.
During pointer motion there is a feedback mechanism. The data offer
can signal to the data source that it can or cannot accept the data
at the current pointer position. This can be used by the client being
dragged from to update the cursor.
The drag'n'drop operation ends with the implicit grab being removed,
that is the pressed pointer button which triggered the operation gets
released. The server sends a drop event to the focused data device.
The data transfer can now be started. For that the receiving client
creates a pipe and passes the file descriptor through the data offer
to the sending data source. The sending client will write into the
file descriptor and close it to finish the transfer.
Drag'n'drop could also be initiated through a touch device grab, but
this is not yet implemented.
The implementation in this change focuses on the adjustments for pointer.
For the user of the library drag'n'drop is implemented in the
SeatInterface. Signals are emitted whenever drag is started or ended.
The interaction for pointer events hardly changes. Motion, button press
and button release can still be indicated in the same way. If a button
release removes the implicit grab the drop is automatically performed,
without the user of the library having to do anything.
The only change during drag and drop for the library user is that
setFocusedPointerSurface is blocked. To update the current drag target
the library user should use setDragTarget. Sending the enter/leave to the
data device gets performed automatically.
The data device which triggered the drag and drop operation is exposed
in the SeatInterface. The user of the library should make sure to render
the additional drag icon provided on the data device. At least QtWayland
based applications will freeze during drag and drop if the icon doesn't
get rendered.
The implementation is currently still lacking the client side and due to
that also auto test. It's currently only tested with QtWayland clients.
Reviewers: #plasma, sebas
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1046
Summary:
The Cursor wasn't properly initialized. E.g. the damage signal didn't
get connected resulting in the server not noticing when the cursor
changes. The damage only got connected if a new cursor got instelled by
the client on the same pointer.
This change ensures that the Cursor is properly initialized by calling
into the same method as when the cursor changed.
The tests are extended by a new test case for damaging the surface.
Reviewers: #plasma
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1022
Summary:
The signal gets emitted whenever the focused PointerInterfaces gets
newly set or reset to nullptr. This is needed to better track the
current cursor image in the compositor.
Reviewers: #plasma, sebas
Subscribers: plasma-devel
Projects: #plasma
Differential Revision: https://phabricator.kde.org/D1007
Summary: Convenient API to get the absolute executable path for the pid.
Reviewers: sebas, mart
Subscribers: plasma-devel
Differential Revision: https://phabricator.kde.org/D858
This patch transports the EDID data base64-encoded over the wire.
Apparently, we can't just send random QByteArrays as "strings" over, it
has to be encoded and decoded. So...
* base64-encode the data before sending to the client
* base64-decode it on the client side
* document the above, fix documentation woes in the xml definition
* change test accordingly
The test data used was actually invalid, it's a base64 string of the
actual data, so fix the tests (which actually breaks it), and encode on
the server-side and decode on the client side.
REVIEW:126380
So far we only supported mapping global to surface-local coordinates
using a 2D-offset. With this change it's possible to register a
QMatrix4x4 to describe the transformation for going from global to
surface-local coordinates in a full 3D space.
The existing 2D-offset is transformed to use the new matrix based
variant describing a translation.
REVIEW: 126271
A system isn't idle by definition below 5 sec. Before it's not possible
to properly determine whether the user is still interacting with the
system or not. Thus such a short timeout is not sufficient to determine
whether the system is idle.
Furthermore allowing short timeouts like 1 msec can be used to gather
information from the system. It would allow to reconstruct the timestamp
changes on the seat very reliable, which we do not want a Client to know.
And also this can be used to get events on each key press and key release
and the time between each of the events. This can be used to gather
statistics which would in worst case allow to reconstruct what the user
is typing. Determining where a word starts and ends should be relatively
straight forward if you know the timing. With the length of the word and
the statistics of alphabetic distribution it becomes possible to
reconstruct some words. And also individual letters. With enough
statistic and some known words one can determine how long it takes to
press a certain letter. At that point the idle interface would be a
keylogger.
To prevent such attacks the timestamp is now modified to be at least
5 sec. Why 5 sec? Because it's the smallest timestamp used in the
KIdleTime example application which I do not want to break.
5 sec are long enough to destroy this possible attack.
REVIEW: 126220
After deleting an OutputInterface the resources are not necessarily
destroyed, so unbind might still be called. The existing code just
casted the resource's user data which could then point to invalidated
memory.
This change verifies that we still have a Private* for the resource.
If not, it doesn't have to do any cleanup anyway.
REVIEW: 126097
Wrapper around wl_client_destroy. In case the ClientConnection got
created through Display::createClient we need to destroy the
ClientConnection again. The exiting client will not cause it to be
destroyed.
Reviewed-By: Bhushan Shah
This implements the server part of the screen management protocol. The
protocol is implemented as a wayland protocol.
It provides the following mechanisms:
- a list of outputs, close to wl_output, with additional properties for
enabled, uuid, edid, etc.. These OutputDevices correspond to a
connected output that can be enabled by the compositor, but is not
necessarily currently used for rendering.
- a global OutputManagement, which allows creating config objects, one
per client. The client can make changes to the outputs through
setScale(outputdevice*, scale) for example.
- an OutputConfiguration resource, that can be handed to a client and
used for configuration. Changes are double buffered here. Only after
OutputConfiguration.apply() has been called, the changes are relayed
over the global OutputManagement.
The compositor is responsible to handle changes.
For a more detailed description, see the API docs in especially
outputconfiguration.h.
REVIEW:125942
this exposes the geometry of taskbar entries in
plasma-windowmanagement, in order to make the
minimize effects possible.
unlike on X11, it takes relative positions and
it has one geometry per panel, making possible
to have multiple taskbars working.
REVIEW:125871
A transient surface can indicate through the flags that it does not
want to accept keyboard focus. This is now exposed through a dedicated
method.
REVIEW: 125552
So far transient was a mutual exclusive mode causing a transient window
to not be able to be fullscreen. This seems wrong. Let's have transient
still as a dedicated mode allowing the window to be maximized and/or
fullscreen. Only popup stays a dedicated mode.
REVIEW: 125468
On client side a setTransient method is added which wraps the semantic
of wl_shell_surface_set_transient.
On server side both set_transient and set_popup are implemented, though
for popup only the transient part is implemented. In particular the grab
is not yet handled and also no popup done is provided.
For the transient on server side the flags are ignored. Main reason is
that Qt does not use the flag, so testing whether it works is tricky
(needs a test application).
REVIEW: 125223
* wl_data_device_manager -> version 2
* wl_data_device -> version 2
* Wayland -> version 1.7
Unfortunately the client side is not yet completely correct. We
need to call wl_data_device_release only if we have a data device
of version 2. But there is no easy way to test this. To change we
will need to introduce a client side version tracking.
BUG: 352663
Follows a useful change added in the kwaylandScanner tool the
s_version becomes part of the Private class.
Also fixes the related generation in the tool.
* fix initializer type -- was hardcoded, seems like an oversight
* add semicolons after class declarations -- one less thing to fix in the
resulting code
REVIEW:125110
Uses QWidgets to render a very small UI.
* lists each Output
* for each output shows whether dpms is supported
* and which dpms mode it's in
* and provides a button to change the mode
Adding support for a org_kde_kwin_dpms interface. On server side
Dpms is mostly bound to the OutputInterface exposing just a very
small manager. Whether Dpms is supported and which mode is used is
tracked directly on the OutputInterface.
In the header for each request a method is created, with following
features:
* ignores requests working as destructor
* adjusts return type for requests containing newId
* adds QObject *parent = nullptr for requests containing newId
* adds the arguements
* turns arguments from wayland into Qt types
In the cpp the same logic is applied. Methods are not properly
implemented and won't compile in the case of returning.
* parses the protocol
* can derive file name to generate from protocol name
* detects which interface is a global and which one is referenced
* generates client code for global and for resource
* comes with a mapping.txt file to translate wl -> KWayland
The small tool added here allows to automate the process of generating
the boiler plate for an interface. Currently it's able to generate the
Client side code representing a wl_global.
The tool does not get installed, so run it from build directory:
Example usage:
./kwaylandScanner -c Foo -f foo -g wl_foo
This will generate a wrapper for the wl_foo interface in files:
foo.h and foo.cpp
The class is called KWayland::Client::Foo.
The tool will be extended. Following things are on the TODO list:
* parse xml to get the Wayland names
* add listener (if needed)
* generate code for Wayland interfaces representing Resources
* generate the Server boiler plate
CCMAIL: mart@kde.org
CCMAIL: sebas@kde.org
* Raises wl_seat supported version to 4 in both server and client
* Raises wl_keyboard supported version to 4 in wl_keyboard
* wl_pointer and wl_touch are still on version 3
* Raises minimum Wayland version to 1.6
a protocol to activate the blur behind windows and to
optionally set a sub region of the window where to apply
the blur to, in case the window is shaped
REVIEW:125015
This change addresses the sometimes cumbersome to use API of Registry.
So far to bind an interface one needed to connect to the signals when
setting up the registry. There was no other way to get the name and
version at a later point. This made it difficult to use one Registry
in different parts of an Wayland application.
This change adds a convenient:
struct AnnouncedInterface { quint32 name; quint32 version;};
AnnouncedInterface interface(Interface) const
QVector<AnnouncedInterface> interfaces(Interface) const
The first one is meant for interfaces like wl_shm which is normally only
announced once. The latter one is meant for interfaces like wl_output
for which normally multiple interfaces can get announced.
Those methods can be used without having to connect to the announced
spy.
REVIEW: 124734
The shadow protocol is inspired by the KWin's X11 protocol and the
DecorationShadow in KDecoration2.
A shadow is attached to a surface (after a commit on surface) and
consists of several image parts (represented through a buffer):
* top-left
* top
* top-right
* right
* bottom-right
* bottom
* bottom-left
* left
In addition there is an offset on each side.
For more information see also the X11 protocol described at [1].
Note: the protocol is currently missing documentation and changing
the shadow is not yet properly delegated to the server API.
[1] https://community.kde.org/KWin/Shadow
This interface allows a client to fake input events and the server
might use them. There is an authentication mechanismn in place which
requires the server to mark the client as authenticated in order for
any events to be emitted at all.
This interface is intended for use cases like kdeconnect which allows
to remote control a device.
The idle time interface is modelled for the use cases of the KIdleTime
framework to allow providing a Wayland specific implementation.
It supports registering idle timeouts which are triggered on server
side if there has not been any user activity on the seat for the
requested amount of time. Once user activity resumes a resume from idle
signal is emitted.
In additon there is the possibility to simulate user activity which
simulates the resume from idle.
Sending a new PlasmaWindow as an event is quite problematic. The main
problem here is that it sends an id of 0 and this doesn't make sense
when the interface gets bound and all existing windows are send to the
client.
This changes the interaction to have the client create the resource for
the PlasmaWindow instead of the server creating it. The server generates
"window ids" and sends those to the client. The client now binds a
PlasmaWindow with the "window id" in the normal way.
In case the server doesn't know the id any more, it directly sends an
unmapped and destroyes the newly created resource again. It is not a
protocol error as that can happen (common example: file open dialog of
Qt 5 applications on Xwayland).
on xwayland clients, send mouse press, move and release
corresponding to touchDown, move and release
this gives a very basic touchscreen support on
xwayland clients
reviewed-by: Martin Graesslin <mgraesslin@kde.org>
Server can set a themed icon name, client gets it directly turned into a
QIcon. That only works with QGuiApplications as can be seen in the
adjusted paneltest.
Sends an umap to the client and destroys the resource afterwards.
The PlasmaWindowInterface gets automatically destroyed once there
are no more resources bound to it.
Check whether the resource for the Keyboard/Pointer/TouchInterface
is still valid before trying to use it.
This was hitting crashes in wayland during the wl_foo_send_enter
calls.
The Qt surface extension is a small protocol to allow exchanging
additional data between QWindows and the compositor. What we are
currently only interested in is the possibility to close a surface
from the Compositor.
Protocol description is copied from QtWayland 5.4.2 branch.
Before emitting the new damage we need to restrict it to the geometry
of the surface. That's required by the documentation and at least
QtWayland/OpenGL emits non-sense damage events.
If a resource is null, because e.g. the surface got already destroyed,
wayland will create an error while marshalling arguments causing
the connected client to in worst case abort.
Apparently we need to add a small delay after starting weston otherwise
it fails to connect on the CI system.
Additionally we switch to using headless-backend.so which should also
make the setup more robust.
Running real world applications (Xwayland) showed that it doesn't like
at all that we do a dispatch when we are going to flush. This caused
in a very reliable manner a "Connection closed" error in XWayland, taking
down the client and in return the (xwayland-enabled) server.
BufferInterface used to have a SurfaceInterface as parent. This could
result in easily hitting the abort condition that the BufferInterface
was still referenced when it gets deleted by just having one other
user referencing the BufferInterface.
There is no need to have the BufferInterface deleted when the
SurfaceInterface to which it belongs gets deleted. The BufferInterface
will get deleted once it's completely unreferenced and also has a
destroy listener.
Basically just wraps the size of the attached buffer. Convenience to
not have to go down to the Buffer to read the size.
TODO: apply buffer scale and transform.
Uses eglQueryWaylandBufferWL (if available) to determine the size of the
buffer. In order to do so, the server library links against egl (1) and
one needs to register the EGLDisplay in Server::Display by the user of
the library. For this a new method Display::setEglDisplay is added.
1: not using epoxy as it doesn't wrap the Wayland interfaces yet.
This change demonstrates how one can install a custom event dispatcher
before creating the QGuiApplication and how to process Wayland events
using the event dispatcher to e.g. start an Xwayland server prior to
creating the application.
It also allows to use
--platform wayland
on the test application. The wayland QPA plugin will connect to the
KWayland server started prior to creating the application.
Please note that such a setup will freeze unless QtWayland has commit
1e32e71 [1].
[1] 1e32e71403
This allows to start the Wayland server before the QCoreApplication is
created with an eventDispatcher being installed through
QCoreApplication::setEventDispatcher.
Instead of performing dispatchEvents for a hard coded number of
expected events from Xwayland, we use select on the pipe to figure
out whether Xwayland has started writing the display number. As long
as select does not indicate that the pipe is ready we need to dispatch
further events.
Anologous to the implementation in Pointer:
* signal entered and left
* Surface *enteredSurface()
The passed in keys are not yet passed along and
currently ignored.
Now the code handles correctly the attaching of a null buffer
(emits a signal unmapped) and if a commit doesn't have a pending
buffer it won't be reset. Damage requests are ignored if no buffer
has been attached yet.
Properly reset focus when surfaces get added/removed. E.g. when
writing in Kate additional surfaces are created for which we do
not want to pass focus.
If one passes StartMode::ConnectClientsOnly the socket in
XDG_RUNTIME_DIR is not created and instead one can only connect to
this server using the Display::createClient method. A very nice
solution to just have a Server and one dedicated Client talk to each
other.
The selection is supposed to be sent to the DataDeviceInterface just
before getting keyboard focus. In order to do that the SeatInterface
keeps track of the DataDeviceInterface which is the current selection
and the DataDeviceInterface of the focused keyboard client.
SeatInterface friends DataDeviceManagerInterface so that the latter
can register each created DataDevice for the SeatInterface.
The button state is a seat-global state and not a per pointer state.
All pressed/released and axis events are moved to the SeatInterface
and just invoke the related method on the focused surface pointer.
Makes PointerInterface more like other Interface classes wrapping
wl_resource. The most important change is the handling of the
focused surface. This is now kept in the SeatInterface and can also
be set if there is no PointerInterface for the client yet.
The unit tests had to be adjusted and some are also disabled as the
button events are not yet moved into SeatInterface.
This method is supposed to return the PointerInterface for the current
focused surface. At the moment it just creates the one global
PointerInterface. The existing SeatInterface::pointer method got
removed as that is actually wrong usage.
There can only be one focused surface per Seat, thus the information
should be hold in the seat.
This only adjusts the API, the actual data is still hold in the
PointerInterface. This still needs adjustment.
The ClientConnection is managed by Display. Whenever one tries to
get a ClientConnection for a wl_client* and it doesn't exist yet a
new one will be created and a clientConnected signal will be emitted.
Also there is a clientDisconnected signal.
ClientConnection provides access to pid, uid and gid. The idea is
to extend it to provide access to all the resources created for the
client.
Required information is passed through the ctor of Private.
::create is still virtual as ShellSurfaceInterface is retrieving
client information. This can be fixed once we have a better wl_client
encapsulation class.
The Resource base class is supposed to be used by all interface
classes which get created for a wl_resource.
Most interface classes are adjusted, but there are some exceptions:
* BufferInterface: is different as the wl_resource is already created
* PointerInterface and KeyboardInterface: those two need changes, the
implementation differs from all other interface implementations.
Version and interface get passed to the ctor allowing Global::Private
to implement ::create instead of providing a pure virtual method.
Also the static bind method is added to the Global::Private which
delegates into a pure virtual method.
New base class KWayland::Server::Global which all Interface classes
for a wl_global inherit. Furthermore there is a shared base class
for all the Private classes of that type.
The BufferInterface used nested calls of wl_shm_buffer_begin_access
and allowed multiple different BufferInterfaces to be in a block
between wl_shm_buffer_begin_access and wl_shm_buffer_end_access.
But this is not allowed: nesting is only possible if it accesses
the same buffer!
This resulted in an abort when we accessed two BufferInterfaces
at the same time. The added test case aborts in the previous
version.
The fix now changes the semantic of the method. The BufferInterface
doesn't hold the QImage any more, but creates a new one every time
::data is invoked. The QImage is created with a custom cleanup
function which calls to wl_shm_buffer_end_access. It ensures that
we don't call into wl_shm_buffer_begin_access as long as there is
a valid QImage for any other BufferInterface still around and
instead returns a null QImage.
Thus a user of a BufferInterface::data should destroy the returned
QImage as soon as it's no longer needed or create a deep copy if
it needs to be kept around.
Counter part for the copyClient test app: the pasteClient test app
creates and maps a shell surface and waits for selection announce,
initiates the copy/paste mechanismn and prints out what got pasted.
After success it closes itself.
The test application copyClient creates a ShellSurface, maps it as
fullscreen, renders a green buffer and when it gets pointer focus it
sets "foo" in mimetype "text/plain" into the selection.
This test application "testRenderingServer" is able to start a
Wayland server and uses a QWidget to render the "scene". It manages
ShellSurfaces and renders them on the QWidget in a very static way.
Furthermore input events (key, pointer, wheel) are forwarded to the
top most Surface.
This test server is able to support kwin_wayland as a client, but
weston's test application in latest release fail due to missing
xdg_shell support.
Qt::MouseButton is mapped to the linux buttons. The mapping does
not match the mapping used in QtWayland module [1] as that seems
to be incorrect to me. E.g. Qt::BackButton is not mapped to BTN_BACK.
[1] qtwayland/src/client/qwaylandinputdevice.cpp
Proof-of-concept code illustrating how to start a Wayland server
before creating the QGuiApplication and also starting an Xwayland
server before creating the QGuiApplication. Thus the QGuiApplication
is able to connect to the Xwayland server started in the same
application.
Event dispatching in Display requires a QSocketNotifier which
requires the QCoreApplication to be created. This requirement renders
it impossible to create the server before the QCoreApplication is
created. But there are use cases which need this:
Let's assume we want to integrate a wayland server into an existing
X11 dependent application which uses QApplication and enforces the
xcb plugin. This means said application requires that an X Server is
up and running before the QApplication is created. If we want said
application to connect to an Xwayland server we need to be able to
create the Wayland server AND the Xwayland server before creating
the QApplication.
The solution is to only create the socket notifier if the
QCoreApplication exists. For the case that it doesn't exist an
additional method is added which needs to be called once the
QCoreApplication got created. In addition a manual event dispatch
method is added which allows to block for events when we don't
have a QSocketNotifier set up yet.
Only selection part is implemented, drag'n'drop still needs to be
implemented.
Unit test is not properly testing whether the data can be transferred.
This needs some better architecture to have multiple processes which
perform the source and target part.
This implements the subcompositor and subsurface protocol on both
Client and Server side.
Client:
New classes SubCompositor and SubSurface. The SubCompositor can be
created through the Registry and creates the SubSurface which is
bound to a Surface and has a parent Surface. The SubSurface class
provides convenient wrappers for all calls exposed in the
wl_subsurface interface.
Server:
New classes SubCompositorInterface and SubSurfaceInterface. Support
for all commands is added, though the API probably still can need
some fine tuning. The synchronized vs. desynchronized behavior is
not yet exposed in the API. This could also be delegated towards
the user of the library.
This makes it possible to install and then use it. Installation is still
commented since we can't give enough stability guarantees for now.
In detail:
- do not actually install headers
- generate the export header into Wayland/Server
- include it from there
REVIEW:120579
Emitted when the Wayland display is done flushing the initial interface
callbacks, announcing wl_display properties. This can be used to compress
events. Note that this signal is emitted only after announcing interfaces,
such as outputs, but not after receiving callbacks of interface properties,
such as the output's geometry, modes, etc..
This signal is emitted from the wl_display_sync callback.
REVIEW:120471
The test started to fail on my system without changing anything. I
assume my system got an update to latest wayland version and it
looks like the registry doesn't announce anything if nothing is
registered. So now our display at least creates the Shm interface.
In addition the test case is slightly improved with a few smaller
changes added during the debug session.
The EventQueue is a wrapper for wl_event_queue. The usage is for the
case that the connection is in a different thread.
The EventQueue needs to be added to the Registry before Registry::create
is invoked. This ensures that the Registry gets added to the EventQueue
and all objects created by the Registry (and further down the tree)
will be added to the same EventQueue.
The auto-tests which already used a wl_event_queue are transited to the
EventQueue class. That's also the reason why there is no dedicated
unit test: the new code is sufficiently tested by the existing tests.
This makes the API more consistent and allows to properly cleanup
Keyboard and Pointer if the connection dies. Like with Shell and
ShellSurface signals are emitted from Seat when the interface is
going to be released or destroyed. These are connected to the
methods of the created Pointer and Keyboard.
