Commit graph

19 commits

Author SHA1 Message Date
Vlad Zahorodnii
ab9a9c8428 [libkwineffects] Detect AMD Navi GPUs
Reviewers: #kwin, davidedmundson

Reviewed By: #kwin, davidedmundson

Subscribers: davidedmundson, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D27168
2020-02-05 22:27:15 +02:00
Vlad Zahorodnii
55b4912004 Update my email address 2020-01-14 18:17:18 +02:00
Martin Flöser
92197242ce Add support for virgl in GLPlatform
Summary:
This change adds detection support for virgl (Mesa gallium virtio guest driver).
Results in proper detection in supportInformation and debug console.

Test Plan: Modified test passes, run KWin_Wayland in kvm with virgl and verified supportInformation

Subscribers: kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D25056
2019-10-29 20:57:02 +01:00
Vlad Zahorodnii
f93875535c Update my last name 2019-09-29 17:03:25 +03:00
Vlad Zagorodniy
9b9d97db2b [libkwineffects] Detect modern Radeon GPUs
Summary:
Currently, KWin is not able to detect Radeon GPUs with GCN architecture.
This patch tries to address that.

Reviewers: #kwin, graesslin

Reviewed By: #kwin, graesslin

Subscribers: graesslin, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D17715
2018-12-21 13:43:13 +02:00
Vlad Zagorodniy
e02573263f [libkwineffects] Introduce timeline redirect modes
Summary:
The redirect modes control behavior of the timeline when its direction
is changed at the start or target position. For example, consider the
following piece of code:

    TimeLine timeLine(1000ms, TimeLine::Forward);
    timeLine.setDirection(TimeLine::Backward);

What should happen when the direction of the timeline was changed to go
backward? Should the current value of the timeline go from 1 to 0, or
should the timeline stop its "execution"?

In the relaxed mode, the timeline will go from 1 to 0.
In the strict mode, the timeline will stop its execution.

Different effects may prefer different modes for source and target
positions. For example, most C++ effect would prefer relaxed mode for
source position, and strict mode for target position. On the other side,
scripted effects(AnimationEffect) would prefer strict mode for source
position, and relaxed mode for target position(because of set).

Reviewers: #kwin, graesslin

Reviewed By: #kwin, graesslin

Subscribers: kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D16447
2018-11-17 13:43:39 +02:00
Vlad Zagorodniy
ee88951b17 [libkwineffects] Add TimeLine helper
Summary:
Most effects use QTimeLine in the following manner

```lang=cpp
if (...) {
    m_timeline->setCurrentTime(m_timeline->currentTime() + time);
} else {
    m_timeline->setCurrentTime(m_timeline->currentTime() - time);
}
```

Because effects do not rely on a timer that QTimeLine has, they can't
toggle direction of the QTimeLine, which makes somewhat harder to write
effects. In some cases that's obvious what condition to use to figure
out whether to add or subtract `time`, but there are cases when it's
not. In addition to that, setCurrentTime allows to have negative
currentTime, which in some cases causes bugs.

And overall, the way effects use QTimeLine is really hack-ish. It makes
more sense just to use an integer accumulator(like the Fall Apart
effect is doing) than to use QTimeLine.

Another problem with QTimeLine is that it's a QObject and some effects
do

```lang=cpp
class WindowInfo
{
public:
    ~WindowInfo();

    QTimeLine *timeLine;
};

WindowInfo::~WindowInfo()
{
    delete timeLine;
}

// ...

QHash<EffectWindow*, WindowInfo> m_windows;
```

which is unsafe.

This change adds the TimeLine class. The TimeLine class is a timeline
helper that designed specifically for needs of effects.

Demo

```lang=cpp
TimeLine timeLine(1000, TimeLine::Forward);
timeLine.setEasingCurve(QEasingCurve::Linear);

timeLine.value(); // 0.0
timeLine.running(); // false
timeLine.done(); // false

timeLine.update(420);
timeLine.value(); // 0.42
timeLine.running(); // true
timeLine.done(); // false

timeLine.toggleDirection();
timeLine.value(); // 0.42
timeLine.running(); // true
timeLine.done(); // false

timeLine.update(100);
timeLine.value(); // 0.32
timeLine.running(); // true
timeLine.done(); // false

timeLine.update(1000);
timeLine.value(); // 0.0
timeLine.running(); // false
timeLine.done(); // true
```

Test Plan: Ran tests.

Reviewers: #kwin, davidedmundson, graesslin

Reviewed By: #kwin, davidedmundson, graesslin

Subscribers: romangg, graesslin, anthonyfieroni, davidedmundson, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D13740
2018-06-30 09:58:11 +03:00
David Edmundson
a3cff85e7a Remove Qt module declarations in includes
Summary:
Test Plan: Compiles

Subscribers: kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D13359
2018-06-05 18:07:23 +01:00
Fabian Vogt
e302f87598 Properly detect Gallium drivers with newer Mesa
Summary:
The renderer string does not contain "Gallium 0.4 on" anymore,
instead it directly contains the gallium driver's name.
So assume that every unknown renderer is a gallium driver.

Test Plan: Added a testcase, it succeeds only with this patch.

Reviewers: #plasma, graesslin

Subscribers: kwin, plasma-devel, #kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D9495
2018-01-12 09:05:03 +01:00
Martin Flöser
1c61e61119 [autotests] Support new way to run the tests without need to install
Summary:
KWin was quite good in ensuring that you don't need to install by
passing paths to the tests. The new way is much nicer, so code is
adjusted for the new way. Also if we require a newer ECM in future we
need to support the new way.

No guarantee that the tests don't pick something up from the system env,
that needs more testing.

References: https://community.kde.org/Guidelines_and_HOWTOs/Making_apps_run_uninstalled

Test Plan: The tests which loaded helpers pass

Reviewers: #kwin, #plasma

Subscribers: plasma-devel, kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D7543
2017-11-21 20:34:52 +01:00
Martin Gräßlin
356510b0aa [libkwineffects] Port getXServerVersion from X11 to xcb
Removes one of the last pure XLib usages and also means that in theory
we can detect the Xwayland version number. In practice that only works
when restarting the compositor as detect is invoked before the XWayland
connection is created.
2016-08-23 11:45:59 +02:00
Martin Gräßlin
83d8181675 [libkwinglutils] Revert version hack for Qualcomm Adreno on libhybris
The idea to set proper version is good, but it results in epoxy
thinking it does not have the entry points and terminating KWin.
2016-08-12 15:46:55 +02:00
Martin Gräßlin
a5c8356d40 [libkwineffects] Add detection code for Qualcomm Adreno to GLPlatform
Summary:
The Qualcom Adreno classes are recognized and a version detection
workaround is added for libhybris which only announces GLES version 2
although GLES version 3 is supported. KWin at least used to work with
GLES version 3 which gives us e.g. framebuffer blit.

Reviewers: #kwin, bshah

Subscribers: kwin

Tags: #kwin

Differential Revision: https://phabricator.kde.org/D2415
2016-08-12 09:59:38 +02:00
Martin Gräßlin
79641df928 [autotests/libkwineffects] Add test data for Adreno 330
This is the device used in Nexus 5 and reference for Plasma Mobile.
The output is with using libhybris on top of hwcomposer backend.
2016-08-12 09:16:28 +02:00
Martin Gräßlin
4cdb078376 [autotest/libkwineffects] Add more test data for glplatformtest 2016-08-09 08:40:37 +02:00
Martin Gräßlin
1e13c5ad8f [autotests/libkwineffects] Test GLPlatform::driverToString and chipClassToString 2016-08-09 08:30:09 +02:00
Martin Gräßlin
0778cab42d [autotest/libkwineffects] Add more test data for the glplatform test
Example data harvested from bugs.kde.org: open bugs against kwin with
a comment containing supportInformation.

We can see that especially detecting modern radeon gpus is not working.
2016-08-08 17:40:04 +02:00
Martin Gräßlin
31790dc00c [autotest] Introduce a test infrastructure for GLPlatform
The new test can load "profiles" from kconfig files in the test data.
Based on that the glGetString return values are mocked and GLPlatform
can perform detect without having to interact with a real GL library.

That way we can verify that the detect code works correctly. As a first
test the settings of one Intel/IvyBridge is included. More tests can be
added easily (e.g. looking at various supportInformation output in
bugs.kde.org). Also this allows to more easily add detect code for GPUs
we do not know yet. And to simulate conditions where the detect code
failed resulting in no compositing at all.
2016-08-05 08:47:32 +02:00
Martin Gräßlin
fb8434671f [autotests] Move libkwineffects/autotests to autotests/libkwineffects
All autotests shall be in autotests subdirectory.
2016-07-18 08:47:45 +02:00