Commit graph

13242 commits

Author SHA1 Message Date
Martin Gräßlin
4c9744fe8e Drop dependency on kdecoration.h 2014-12-02 13:50:26 +01:00
Martin Gräßlin
3d9a035312 Move MaximizeMode from KDecorationDefines to utils.h
Unfortunately it cannot be a enum defined in Client as client.h
depends on rules.h and with it in Client rules.h would depend on
client.h.
2014-12-02 13:49:08 +01:00
Martin Gräßlin
50019b4a31 Drop kdecoration.h include from placement.cpp 2014-12-02 11:48:51 +01:00
Martin Gräßlin
763a976c39 Drop kdecoration.h include from tabgroup.h 2014-12-02 11:48:13 +01:00
Martin Gräßlin
1747fde18b Toplevel does not inherit from KDecorationDefines any more
As an intermediate step Client inherits from KDecorationDefines.
2014-12-02 11:31:07 +01:00
Martin Gräßlin
6abf115d22 Drop kdecoration.h include from options.h 2014-12-02 11:19:40 +01:00
Martin Gräßlin
d4e2f19ef0 Workspace does not need to inherit from KDecorationDefines 2014-12-02 11:17:43 +01:00
Martin Gräßlin
fa4332754c Move Position from KDecorationDefines to Client 2014-12-02 11:15:54 +01:00
Martin Gräßlin
e38d56b0dd Move QuickTileMode from KDecorationDefines to Client 2014-12-02 10:52:16 +01:00
Martin Gräßlin
d4bf3602ce Move WindowOperation enum to Options
With this Options no longer needs to inherit from KDecorationDefines.
2014-12-02 10:41:47 +01:00
Martin Gräßlin
b9fb7df15a Do not use KDecorationDefines in Aurorae 2014-12-02 10:31:16 +01:00
Martin Gräßlin
f5c305bf64 Do not include kdecoration.h in plastik decoration plugin 2014-12-02 10:15:04 +01:00
Martin Gräßlin
4188736bc4 Stop building libkdecorations
Some defines are still used, so it's not yet deleted.
2014-12-02 09:58:29 +01:00
Martin Gräßlin
93f414a070 Options no longer inherits from KDecorationOptions 2014-12-02 09:51:52 +01:00
Martin Gräßlin
cb38028d98 Move maximize button click config to Options
In order to drop dependency on KDecorationOptions.
2014-12-02 09:43:24 +01:00
Martin Gräßlin
c2e53b9a2f Merge branch 'kdecorations2'
Conflicts:
	effects.cpp
	paintredirector.cpp
2014-12-02 08:38:10 +01:00
Martin Gräßlin
106e595709 DecorationSettings provides QVector for decorationButtons 2014-12-01 11:53:44 +01:00
Martin Gräßlin
40ce3e4ce3 [autotests] Give Xephyr more time to startup
The test is flaky during the wait for Xephyr. Giving it more time should
hopefully make it more reliable.

The better way would be to switch to the command line argument
-displayfd pipeFd

unfortunately the Xephyr on the CI system does not yet support this
command line argument.
2014-12-01 10:53:40 +01:00
Martin Gräßlin
027a7a1646 [kdecoration2] Decoration::shadow returns QSharedPointer
Adjust all components to use the QSharedPointer. Aurorae obviously has
to create a QSharedPointer, Shadow holds a QSharedPointer for the
DecorationShadow (advantage: is kept when the Decoration is destroyed),
and the KCM needs to add a property on PreviewItem to get access to the
Shadow. It's no longer a Q_PROPERTY on Decoration and we cannot re-add
it as a dynamic property (cannot be read from QML side).
2014-12-01 09:40:24 +01:00
Martin Gräßlin
6bdd0e96f6 [kdecoration] Decoration::client returns a QWeakPointer
In aurorae a wrapper property is added.
2014-11-28 16:28:39 +01:00
Martin Gräßlin
d1bf38e92c [kdecoration2] Detect double clicks on the decoration title bar
No longer handled by the Decoration.
2014-11-28 14:50:21 +01:00
Martin Gräßlin
0efa9d9c73 [kdecoration2] Decoration does not provide titleBarWheelEvent any more
Instead KWin needs to determine by itself whether it should act on the
wheel event. The Decoration will set it to accepted if it handled it.
2014-11-28 14:08:13 +01:00
Martin Gräßlin
0cc86656c9 DecorationButtonType::QuickHelp -> DecorationButtonType::ContextHelp 2014-11-28 11:27:31 +01:00
Martin Gräßlin
211834edcc [kdecoration] ::paint uses QRect instead of QRegion 2014-11-28 10:26:23 +01:00
Thomas Lübking
50cbe810f5 don't access array to check whether ptr is nullptr
Also unify with the below check whether array[0] is 0
(Was same condition and same branch)

BUG: 341010
2014-11-27 20:01:10 +01:00
Thomas Lübking
7747a38189 step into the focus chain at the beginning
if the index is 0, we close the circle, but if the
index is -1 (invalid) we're "somewhere" and step
into the focus chain - reasonably at the beginning

BUG: 340594
REVIEW: 121024
2014-11-27 20:01:10 +01:00
Thomas Lübking
b26ef9302a don't restore unrestorable window sizes on Q'unmax
BUG: 340600
REVIEW: 121022
2014-11-27 20:01:10 +01:00
Thomas Lübking
ec8f7e1fac ensure disjunct screenedges on overlapping screens
If two or more screens overlap (partially or fully)
we must not create double edges since kwin would
trigger both - since they've the same actions bound
(for sure) the action will called twice (so either
eg. toggle or skip one VD etc.)

BUG: 338381
REVIEW: 120060
2014-11-27 20:01:10 +01:00
Martin Gräßlin
66df3eefdf Cleanup: reduce usage of QDesktopWidget 2014-11-27 09:32:43 +01:00
Martin Gräßlin
bf4ba54647 Remove displayWidth and displayHeight from the compositor
Instead of getting size from displayWidth() and displayHeight() use
the information we have from Screens. This means there is only one
place to have the information and by that we can ensure that all
components use the same data to rely on. displayWidth/displayHeight
seem to provide the wrong information when unplugging an output
without disabling the output. This results in rendering artefacts.
But KWin::Screens has the correct information available.
2014-11-27 09:00:19 +01:00
Martin Gräßlin
cc98c5d9c7 Use screens()->geometry() in Workspace::desktopResized
It used to get the individual screen geometries from QDesktopWidget.
We don't need to go through QDesktopWidget any more as Screens
contains all the information we need in a more reliable way.
2014-11-27 09:00:19 +01:00
Martin Gräßlin
2eb876743c [screens] Replace DesktopWidgetScreens by XRandRScreens
A new implementation of the Screens interface is added which uses XRandR
directly instead of relying on QDesktopWidget. The implementation is
provided in a new implementation file screens_xrandr.cpp.

XRandRScreens comes with a unit test. Unfortunately it's rather difficult
to provide a proper unit test against XRandR. Xvfb (which is obviously
used on the CI system) doesn't provide the XRandR extension. Also on a
"normal" developer system one would not want to just execute the test as
the results are not predictable (number of available outputs?) and the
test would mess up the setup resulting in nobody wanting to execute the
test.

As a solution to both problems the unit test starts Xephyr as a nested
X server. This allows to have at least some limited tests against XRandR.
Nevertheless there are a few things which I was not able to test:
* multiple outputs
* no output at all

The nested X Server approach makes the interaction rather complex. Qt
opens it's connection against the main X Server thus QX11Info provides
a wrong connection and also KWin::connection() which is heavily used by
xcbutils and thus all the RandR wrappers have the wrong connection. To
circumvent this problem the test is GUILESS. In case it would call into
any code using QX11Info, it would probably either runtime fail or crash.

REVIEW: 117614
2014-11-27 09:00:19 +01:00
Martin Gräßlin
6d64113ed4 Do not include utils.h in xcbutils.h
Only needed for ScopedCPointer which can be defined in both utils and
xcbutils.h.

Not having utils.h in xcbutils.h makes it easier to have unit tests use
xcbutils.h as it removes the dependencies.
2014-11-27 09:00:19 +01:00
Fredrik Höglund
9d01ed16be Remove GLTexturePrivate::bind()/unbind()
These virtual methods are not reimplemented anywhere.
2014-11-26 22:01:54 +01:00
Fredrik Höglund
59b5746351 Mark SceneOpenGL::Texture::discard() as override final 2014-11-26 22:01:54 +01:00
Fredrik Höglund
e47d8671e9 De-virtualize SceneOpenGL::Texture::load()
SceneOpenGL::Texture is not subclassed anywhere so there is no reason
for this function to be virtual.
2014-11-26 22:01:54 +01:00
Fredrik Höglund
4b9448db5e Inline GLTexture::load(QImage) in the constructor
This is the only place where it is called.

There was also no need for this function to be virtual since it's not
reimplemented anywhere.

This also fixes the GLTexturePrivate being created only to be
immediately deleted and replaced by a new GLTexturePrivate when the
GLTexture(QImage) constructor was used.

With this change the structure of the GLTexture effectively becomes
immutable to the outside world; the contents of the texture images can
change, but the texture images themselves cannot be reallocated.
2014-11-26 22:01:54 +01:00
Fredrik Höglund
9c7720abca Remove GLTexture::load(QPixmap) and load(QString)
These methods are only called from the GLTexture constructors,
so inline them there.
2014-11-26 22:01:54 +01:00
Fredrik Höglund
fa52af4765 Remove SceneOpenGL::createTexture(QPixmap)
...and associated code.

It is no longer used.
2014-11-26 22:01:54 +01:00
Fredrik Höglund
34f1408134 Make the EffectFrame textures regular GLTextures
There is valid reason for these textures to be Scene::Texture's.
2014-11-26 22:01:54 +01:00
Fredrik Höglund
383a5ac319 Remove GLTexturePrivate::convertToGLFormat()
QImage::Format_RGBA8888_Premultiplied is the equivalent of
GL_RGBA/GL_UNSIGNED_BYTE, so use QImage::convertToFormat()
instead.

This commit also makes the texture upload code work correctly
on big-endian systems.
2014-11-26 22:01:23 +01:00
Fredrik Höglund
6ac7371113 Remove most calls to checkGLError()
...and rely on KHR_debug/GL_ARB_debug_output instead.
2014-11-25 21:57:57 +01:00
Fredrik Höglund
c0deadfaad Add support for KHR_debug
Register a callback function for debug output when KHR_debug or
ARB_debug_output is supported.

Only error messages and warnings about undefined behavior are
enabled in release builds, while all debug messages are enabled
in debug builds.
2014-11-25 21:57:15 +01:00
Martin Gräßlin
7c62cd19ef Verify Workspace pointer in X11EventFilter::~X11EventFilter
In case the inheriting class of X11EventFilter is a child of
Workspace and deleted by the QObject dtor of Workspace the
Workspace::self() pointer is no longer valid and already set to
nullptr.
2014-11-25 12:57:37 +01:00
Martin Gräßlin
b7a8bb4f52 Drop build option KWIN_BUILD_SCREENEDGES
The build option got introduced for Plasma Active back in a time
when we did not properly aim for convergence. In a Plasma 5 world
we want to have only one shell and one window manager which adjust
itself. This means we don't want a differently compiled kwin for
plasma active, but the same one. Thus the build option doesn't
make much sense any more. A KWin for touch interface needs to support
screenedges for the case that mouse is plugged in.

CCBUG: 340960
REVIEW: 121200
2014-11-24 10:48:49 +01:00
Martin Gräßlin
482f89b91f Fix build without KF5Activities
BUG: 340961
FIXED-IN: 5.2.0
REVIEW: 121202
2014-11-24 10:47:08 +01:00
Martin Gräßlin
26cf3195be [kcmkwin/tabbox] Find all layouts in lnf package
First try to locate all look-and-feel packages and check whether
they provide a WindowSwitcher. If they do they are inserted in
the list with the Package name being the user visible name.
So the good old sidebar is now called "Breeze".

REVIEW: 120914
2014-11-24 10:39:26 +01:00
Martin Gräßlin
0f7a990316 Use Protocol support in NETWinInfo instead of fetching them ourselves
Client does no longer need to track the protocols as this is also
provided by NETWinInfo as of version 5.3.

REVIEW: 120255
2014-11-24 10:37:18 +01:00
Fredrik Höglund
cb39eb0cc5 Use consistent names for the static members in GLTexturePrivate 2014-11-23 20:20:16 +01:00
Fredrik Höglund
854f66c164 Make s_supportsUnpack a static member of GLTexturePrivate 2014-11-23 20:16:02 +01:00