GitLab

Summary:
The .wait() in this case isn't really time depdenent, we just need to
process the event loop.

Given we know this tests fails, there's no point me waiting 30s.

Test Plan: Test failed, but faster

Reviewers: #kwin, zzag

Reviewed By: #kwin, zzag

Subscribers: kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D19351

Summary:
Everything on the wl_surface is double buffered.

When we create an XdgShell toplevel or popup we shouldn't treat it as
attached until it's committed to the surface.

A client should commit the surface after it's sent it's initial state of
the Xdg topLevel; minimumSize, title, app_id, etc.

By blocking sending configure events we will have flushed the correct
initial state before sending a single atomic correct event to the
client. It also adds a hook to re-evaluate rules now that all properties
are set.

Arguably this applies to WlShellSurface too, but I've left it unchanged
as it's deprecated and hard to verify real client behaviour.

Test Plan: Ran all unit tests

Reviewers: #kwin, zzag

Reviewed By: #kwin, zzag

Subscribers: zzag, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D18583

Summary:
We currently check the final position kwin knows about.
This patch also checks the position the popup is told that
it's in as well as acking configure requests correctly.

Reviewers: #kwin, zzag

Reviewed By: #kwin, zzag

Subscribers: zzag, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D18955

Summary:
We can handle the case of a null size inside
ShellClient::requestGeometry and that makes that method more robust.

With that we don't need to handle separate code paths in
the handler of RequestGeometryBlocker.

It's not an exact identical code path, but everything still works out.

If the geometry is unchanged, we'll save the positionAfterResize  but
when we apply it doSetGeometry will no-op.

There's also an assumption that toggling maximised will always send a
valid size, but that's true for the current state.

Test Plan: Ran unit tests

Reviewers: #kwin, zzag

Reviewed By: #kwin, zzag

Subscribers: kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D19331

Summary:
Currently, each frame callback sent by KWin has the current time in
nanoseconds, but the protocol spec states that we have to send the time
in milliseconds. This is the reason why animations that are driven by
frame callbacks are too fast.

In addition to that, m_timeSinceStart isn't actually "time since start,"
it's rather accumulated duration of all painting cycles. If there is
something to draw and it takes quite a while to compose the scene, maybe
m_timeSinceStart will be close enough to the current time. So, it has
been replaced with QElapsedTimer, this makes the current time correct
and also simplifies code a little bit.

Test Plan: The triangle in weston-subsurfaces no longer spins very fast.

Reviewers: #kwin, romangg

Reviewed By: #kwin, romangg

Subscribers: romangg, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D18656

Summary:
We need AniMap only for test purposes so it would be better to make the
typedef protected (and mark as internal as well).

Test Plan: Compiles.

Reviewers: #kwin, davidedmundson

Reviewed By: #kwin, davidedmundson

Subscribers: davidedmundson, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D17833

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

Summary:
In a recent patch the newly added xwayland class ended up being
responsible for continuing the startup, calling back into the main app
to spawn the workspace.

It moves the flow of startup about so it's not very readable or
following class structure.

This patch moves the code back into main_wayland and removes the
duplication between xwayland and non-xwayland modes.

There was also a misnaming of methods.

Previously:
continueStartupWithScreens was called after platform screens are created
continueStartupWithScene was called after the scene was created
continueStartupWithXwayland was called before xwayland is created

This was confusing, so the names have been shuffled around to follow a
consistent pattern of what has been done so far.

Test Plan:
Started kwin_wayland in normal and xwayland mode
Ran unit tests (though some failed due to a local unrelated and as yet unindentified bug)

Reviewers: #kwin, romangg

Reviewed By: #kwin, romangg

Subscribers: romangg, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D19232

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

Summary: When creating a surface and setting it as fullscreen before
attaching its buffer KWin does not know the original dimensions of the
surface and tries to use an invalid value when unsetting the fullscreen
flag. This patch fixes this by sending a configure with the size of 0,0
- which according to xdg-shell spec means that the client is requested
to set its size by itself.

Test Plan:
1. Create a fullscreen Wayland window.
2. Toggle fullscreen off.
3. KWin should send a configure event with size 0,0 instead of 1,1.

Reviewers: #kwin, davidedmundson

Reviewed By: #kwin, davidedmundson

Subscribers: zzag, graesslin, davidedmundson, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D16710

Summary:
On Wayland we cannot switch from OpenGL to QPainter compositor as this
would break any running OpenGL application. KWin registers it's
EGLDisplay to Wayland and without OpenGL this doesn't make sense any
more. We are not able to render OpenGL buffers in the QPainter
compositor.

While it's theoretically possible to switch from QPainter to OpenGL it
doesn't make any sense for the same reason. Any running OpenGL
application would be using llvmpipe and could not be switched to proper
OpenGL.

This change stores the selected compositing type in Platform and the
implementations can use it to restrict the supported compositors. On X11
we don't need this, all other Platforms implement the restriction. Thus
it's no longer possible to switch the backends at runtime.

Test Plan:
Adjusted tests run, no runtime test as gui doesn't support
switching to QPainter anyway.

Reviewers: kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D19084

Summary:
Using the Wayland protocol for OpenGL is no longer used or useful. KWin
internal windows only use OpenGL through QtQuick. We either have the
sharing context (KWin uses OpenGL for compositing) or we have the
QPainter compositor which also turns QtQuick to use software renderer.

Thus a situation where the Wayland platform context is useful doesn't
exist any more. Removing it helps getting the QPA plugin Wayland free.

Test Plan:
Run nested KWin triggering Outline once with OpenGL and once
with QPainter compositor.

Reviewers: #kwin

Subscribers: kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D19012

Summary:
In 403038 the user expected the DBus call to run to finish before
processing the next request. For a singleshot script that makes a lot of
sense, otherwise you have no idea when it finishes.

This also allows us to return errors.

CCBUG: 403038

Test Plan:
NUM=$(qdbus org.kde.KWin /Scripting org.kde.kwin.Scripting.loadScript /noFile.js Minimize)
qdbus org.kde.KWin /$NUM org.kde.kwin.Scripting.run
  Error: org.kde.kwin.Scripting.FileError
  Could not open /noFile.js

Running a real script behaved effectively the same as before.

Reviewers: #kwin, graesslin

Reviewed By: #kwin, graesslin

Subscribers: zzag, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D18127

Summary:
placeIn did not handle the case for a popup not having had the size
already set and only being available via
m_xdgShellPopup->initialSize(). This is needed if we want to call placeIn
at the correct time, before the window is mapped.

There was also a logic bug when sliding popups. We called the
confusingly named setX thinking it would be move the popup keeping the
width the same. In practice it moves the left edge keeping the right
position the same. This wasn't an issue as the size was
discarded.

Handling the resize constraint is not yet done, but it should now be
even more trivial.

Reviewers: #kwin, zzag

Reviewed By: #kwin, zzag

Subscribers: zzag, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D18956

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

`client = new ShellClient(surface)` is most likely a leftover after a
merge conflict.

Summary:
Despite plasma frameworks doing it's own scaling with fonts, it's been
requested to use kwin/wayland scaling.

Like DRM, when kscreen is not used, scale value is loaded from a config
file.

Config format is
[HWComposerOutputs][0]
Scale=N

The 0 is to map similarly to DRM and support multi-screen, but with a
screen index
rather than a UUID based on EDID.

Because we don't support multi screen this is always 0 for now.

Test Plan: Ran with the config value unset and with the config value at
Scale=3.

Reviewers: #kwin, romangg

Reviewed By: #kwin, romangg

Subscribers: bshah, romangg, nicolasfella, zzag, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D18810

Summary: Shm textures via Gl must be reloaded in case their size changes.

Test Plan: testShmBufferSizeChangeOnSubSurface passes.

Reviewers: #kwin, graesslin

Reviewed By: #kwin, graesslin

Subscribers: davidedmundson, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D18453

Summary:
Adds an autotest to show that KWin fails an assertion when a client tries to
resize a sub-surface.

Since it is the first autotest dealing with sub-surfaces explicitly additional
autotest helpers are introduced to allow that.

We also add a new signal in Compositor to spy on to know when the buffer swap
has been completed.

Test Plan:
Test fails as expected:
```
QFATAL : KWin::BufferSizeChangeTest::testShmBufferSizeChangeOnSubSurface() ASSERT: "image.size() == m_size" in file /home/roman/dev/kde/src/kde/workspace/kwin/platformsupport/scenes/opengl/abstract_egl_backend.cpp, line 394
FAIL!  : KWin::BufferSizeChangeTest::testShmBufferSizeChangeOnSubSurface() Received a fatal error.
   Loc: [Unknown file(0)]
Totals: 4 passed, 1 failed, 0 skipped, 0 blacklisted, 367ms
********* Finished testing of KWin::BufferSizeChangeTest *********
```

Reviewers: #kwin, zzag

Subscribers: zzag, graesslin, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D18452

Summary:
The virtual platform did not actually present content and spin the
compositor. Let's do this from now on , first for the EGL GBM backend.

Test Plan: All tests pass according to Vlad

Reviewers: #kwin, zzag

Reviewed By: #kwin, zzag

Subscribers: graesslin, zzag, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D18436

Summary:
Most of the functionality which is special to internal clients is moved
from ShellClient to InternalClient. As KWin's qpa is still bound to the
Wayland protocol InternalClient inherits from ShellClient. Due to that
some aspects in ShellClient are "weird". ShellClient still detects
whether it's an internal client and uses the variable m_internal to
capture the state. This is required as we cannot use the isInternal
method. Most of m_internal usage is in init which is called from
constructor of ShellClient. Thus it's not possible to call into virtual
methods of InternalClient.

Also some of the code is duplicated and some methods are temporarily
marked as virtual.

The next step will be to remove ShmBuffer for internal windows which
should decouple the two implementations further with the long term goal
of having InternalClient inherit AbstractClient directly.

Test Plan:
Run nested KWin, triggered outline (OpenGL case) and debug console (shm case).
InternalWindow unit test still passes.

Reviewers: #kwin

Subscribers: kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D18569

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

Summary:
This gives us XdgOutput support a requirement now that wayland server
reports that we support it and Qt5.12 then expects it to be sent.

It also provides an easy starting point for supporting wayland scaling
and a slight step towards multi-output.

Test Plan:
Ran on phone. Got output. Turning screen on and off works.
The double tap doesn't work great, but that was true before
(via bshah's hack to build on the current state)

Reviewers: #kwin, romangg

Reviewed By: #kwin, romangg

Subscribers: romangg, nicolasfella, zzag, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D18808

Test Plan: Compile + grep

Reviewers: #kwin

Subscribers: kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D19177

Summary:
Currently, KWin/Wayland crashes when the compositor is reinitialized.
The reason for that is ShellClient's DecorationRenderer gets destroyed
when the scene is already gone, thus there is no current OpenGL context.
Client works around that issue by destroying scene-specific DecorationRender
in finishCompositing. Such a workaround could be applied to ShellClient
as well, but it would make code more confusing because DecoratedClientImpl
also tries to destroy DecorationRenderer.

A better approach would be to notify DecoratedClientImpl that
compositing is about to be finished, so it can destroy the decoration
renderer when the scene is still alive. This not only fixes the
previously mentioned issue in ShellClient, but also makes code a little
bit tidier.

Test Plan:
Start Plasma on Wayland session, change any compositor settings (e.g.
animation speed).

Reviewers: #kwin, davidedmundson

Reviewed By: #kwin, davidedmundson

Subscribers: davidedmundson, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D18921

Summary:
Some X clients acting as drag sources might try to optimize finding the current
target window by checking if a window manager, that sets the root window
_NET_CLIENT_LIST_STACKING property, is present. An example for this is Chromium
and since KWin sets the property the drag proxy windows must be added to this
list. Otherwise the origin client will not detect the proxy window and not send
an XdndEnter message.

Test Plan: Manually with Chromium.

Reviewers: #kwin

Subscribers: zzag, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D15628

Summary:
On X several target atoms are established to provide data similar to the
text/uri-list target format (the respective MIME on Wayland is called
text/x-uri).

Firefox can send link data in the NETSCAPE_URL format ones. Chromium on the
other side sends link data in the text/x-moz-url format, which transports
UTF-16 text.

For both these peculiarities this patch provides converter functions, that
translate these formats into the Wayland native text/x-uri format. In the
other direction no translation is necessary. Both browsers supports the
text/uri-list format when receiving link lists.

Test Plan: Manually with Firefox and Chromium.

Reviewers: #kwin

Subscribers: zzag, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D15629

Summary:
Building upon the generic X Selection support this patch establishes another
selection class representing the XDND selection and provides interfaces
to communicate drags originating from Xwayland windows to the Wayland
server KWin and drags originating from Wayland native drags to Xwayland.

For Wayland native drags KWin will claim the XDND selection as owner and
will simply translate all relevant events to the XDND protocol and receive
alike messages by X clients.

When an X client claims the XDND selection KWin is notified via the X protocol
and it decides if it allows the X drag to transcend into the Wayland protocol.
If this is the case the mouse position is tracked and on entering a Wayland
native window a proxy X Window is mapped to the top of the window stack. This
proxy window acts as a drag destination for the drag origin window and again
X messages will be translated into respective Wayland protocol calls. If the
cursor leaves the Wayland window geometry before a drop is registered, the
proxy window is unmapped, what triggers a subsequent drag leave event.

In both directions the necessary core integration is minimal. There is a single
call to be done in the drag and drop event filter through the Xwayland
interface class.

From my tests this patch facilitates drags between any Qt/KDE apps. What needs
extra care are the browsers, which use target formats, that are not directly
compatible with the Wayland protocol's MIME representation. For Chromium an
additional integration step must be done in order to provide it with a net
window stack containing the proxy window.

Test Plan: Manually. Auto tests planned.

Reviewers: #kwin

Subscribers: zzag, kwin, alexde

Tags: #kwin

Maniphest Tasks: T4611

Differential Revision: https://phabricator.kde.org/D15627

Summary: Adds an interface class to access Xwayland members from within KWin core.

Reviewers: #kwin, davidedmundson, zzag

Subscribers: zzag, davidedmundson, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D15419

Summary:
With the gneric X selections infrastructure and clipboard support the X
clipboard sync helper utility can be removed. Also rename its autotest as it
tests the inner workings of the new mechanism since this mechanism was
introduced.

Test Plan: Autotest still passes under new name.

Reviewers: #kwin, davidedmundson

Reviewed By: #kwin, davidedmundson

Subscribers: graesslin, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D15063

Summary:
In this patch an infrastructure is created to represent generic X selections
in a Wayland session and use them for data transfers between Xwayland windows
and Wayland native clients.

The central manager is the DataBridge class, in which Selection objects can be
created. This is hard-coded and such a Selection object persists until the end
of the session, so no arbitrary selections can be created on the fly. For now
the X Clipboard selection is supported, whose corresponding mechanism in the
Wayland protocol is just called Selection.

A Selection object listens for selection owner changes on the X side and for
similar events into the Wayland server interfaces. If a data provider is
available a selection source object is created by the Selection object. In case
data is requested on the other side, a data transfer is initialized by creating
a Transfer object. A Selection keeps track of all transfers and makes sure that
they are destroyed when they are finished or in case they idle because of
misbehaving clients.

The Clipboard class translates the X Clipboard via a proxy window. Selection
changes on the Wayland side are listened to through a new signal on the active
KWayland seat interface.

The previously used X clipboard syncer helper is disabled. The clipboard sync
autotest is changed to the new mechanism.

BUG: 394765
BUG: 395313

Test Plan: Manually and clipboard sync autotest.

Reviewers: #kwin

Subscribers: zzag, graesslin, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D15061

Summary:
By using the new Xwayland class we can reduce code duplication.

An abstract parent class is introduced, that allows interfacing
with the Xwayland class from the test binaries.

Test Plan: Autotests still pass.

Reviewers: #kwin, davidedmundson

Reviewed By: #kwin, davidedmundson

Subscribers: kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D15022

Summary:
The Xwayland code path is moved from ApplicationWayland to a dedicated class
Xwayland in a new top-level directory xwl.

This is a direct preparation step for generic support of Xwayland Selections.

On a longer timescale this should also allow us to further separate Wayland
native functionality from Xwayland to allow us at one point to build KWin's
Wayland binary optionally without X dependencies. Another long term goal, that
becomes possible through this separation is to recover from Xwayland crashes.

Test Plan: Manually and auto tests

Reviewers: #kwin, davidedmundson

Subscribers: davidedmundson, zzag, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D15020

GIT_SILENT

Summary:
Currently, the Desktop Grid effect tries to copy a client when user
moves it. But on X11 such behavior doesn't make sense because
a client can't be on multiple desktops (except on all).

BUG: 404442
FIXED-IN: 5.15.1

Reviewers: #kwin, davidedmundson

Reviewed By: #kwin, davidedmundson

Subscribers: kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D19116

Summary: Required for kwin_wayland to find it's own plugin.

Test Plan: My kwin_wayland was crashing and gdb showed wrong plugin being loaded

Reviewers: #kwin

Subscribers: kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D19010

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"