2020-12-05 15:54:19 +00:00
|
|
|
/*
|
|
|
|
KWin - the KDE window manager
|
|
|
|
This file is part of the KDE project.
|
|
|
|
|
|
|
|
SPDX-FileCopyrightText: 2006 Lubos Lunak <l.lunak@kde.org>
|
|
|
|
SPDX-FileCopyrightText: 2009 Fredrik Höglund <fredrik@kde.org>
|
|
|
|
SPDX-FileCopyrightText: 2013 Martin Gräßlin <mgraesslin@kde.org>
|
|
|
|
|
|
|
|
SPDX-License-Identifier: GPL-2.0-or-later
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include "x11xrenderbackend.h"
|
|
|
|
#include "main.h"
|
|
|
|
#include "platform.h"
|
|
|
|
#include "overlaywindow.h"
|
Introduce RenderLoop
At the moment, our frame scheduling infrastructure is still heavily
based on Xinerama-style rendering. Specifically, we assume that painting
is driven by a single timer, etc.
This change introduces a new type - RenderLoop. Its main purpose is to
drive compositing on a specific output, or in case of X11, on the
overlay window.
With RenderLoop, compositing is synchronized to vblank events. It
exposes the last and the next estimated presentation timestamp. The
expected presentation timestamp can be used by effects to ensure that
animations are synchronized with the upcoming vblank event.
On Wayland, every outputs has its own render loop. On X11, per screen
rendering is not possible, therefore the platform exposes the render
loop for the overlay window. Ideally, the Scene has to expose the
RenderLoop, but as the first step towards better compositing scheduling
it's good as is for the time being.
The RenderLoop tries to minimize the latency by delaying compositing as
close as possible to the next vblank event. One tricky thing about it is
that if compositing is too close to the next vblank event, animations
may become a little bit choppy. However, increasing the latency reduces
the choppiness.
Given that, there is no any "silver bullet" solution for the choppiness
issue, a new option has been added in the Compositing KCM to specify the
amount of latency. By default, it's "Medium," but if a user is not
satisfied with the upstream default, they can tweak it.
2020-11-19 08:52:29 +00:00
|
|
|
#include "renderloop_p.h"
|
2020-12-05 15:54:19 +00:00
|
|
|
#include "scene.h"
|
|
|
|
#include "screens.h"
|
Introduce RenderLoop
At the moment, our frame scheduling infrastructure is still heavily
based on Xinerama-style rendering. Specifically, we assume that painting
is driven by a single timer, etc.
This change introduces a new type - RenderLoop. Its main purpose is to
drive compositing on a specific output, or in case of X11, on the
overlay window.
With RenderLoop, compositing is synchronized to vblank events. It
exposes the last and the next estimated presentation timestamp. The
expected presentation timestamp can be used by effects to ensure that
animations are synchronized with the upcoming vblank event.
On Wayland, every outputs has its own render loop. On X11, per screen
rendering is not possible, therefore the platform exposes the render
loop for the overlay window. Ideally, the Scene has to expose the
RenderLoop, but as the first step towards better compositing scheduling
it's good as is for the time being.
The RenderLoop tries to minimize the latency by delaying compositing as
close as possible to the next vblank event. One tricky thing about it is
that if compositing is too close to the next vblank event, animations
may become a little bit choppy. However, increasing the latency reduces
the choppiness.
Given that, there is no any "silver bullet" solution for the choppiness
issue, a new option has been added in the Compositing KCM to specify the
amount of latency. By default, it's "Medium," but if a user is not
satisfied with the upstream default, they can tweak it.
2020-11-19 08:52:29 +00:00
|
|
|
#include "softwarevsyncmonitor.h"
|
2020-12-05 15:54:19 +00:00
|
|
|
#include "utils.h"
|
|
|
|
#include "x11_platform.h"
|
|
|
|
|
|
|
|
#include "kwinxrenderutils.h"
|
|
|
|
|
|
|
|
namespace KWin
|
|
|
|
{
|
|
|
|
|
|
|
|
X11XRenderBackend::X11XRenderBackend(X11StandalonePlatform *backend)
|
|
|
|
: XRenderBackend()
|
|
|
|
, m_backend(backend)
|
|
|
|
, m_overlayWindow(backend->createOverlayWindow())
|
|
|
|
, m_front(XCB_RENDER_PICTURE_NONE)
|
|
|
|
, m_format(0)
|
|
|
|
{
|
Introduce RenderLoop
At the moment, our frame scheduling infrastructure is still heavily
based on Xinerama-style rendering. Specifically, we assume that painting
is driven by a single timer, etc.
This change introduces a new type - RenderLoop. Its main purpose is to
drive compositing on a specific output, or in case of X11, on the
overlay window.
With RenderLoop, compositing is synchronized to vblank events. It
exposes the last and the next estimated presentation timestamp. The
expected presentation timestamp can be used by effects to ensure that
animations are synchronized with the upcoming vblank event.
On Wayland, every outputs has its own render loop. On X11, per screen
rendering is not possible, therefore the platform exposes the render
loop for the overlay window. Ideally, the Scene has to expose the
RenderLoop, but as the first step towards better compositing scheduling
it's good as is for the time being.
The RenderLoop tries to minimize the latency by delaying compositing as
close as possible to the next vblank event. One tricky thing about it is
that if compositing is too close to the next vblank event, animations
may become a little bit choppy. However, increasing the latency reduces
the choppiness.
Given that, there is no any "silver bullet" solution for the choppiness
issue, a new option has been added in the Compositing KCM to specify the
amount of latency. By default, it's "Medium," but if a user is not
satisfied with the upstream default, they can tweak it.
2020-11-19 08:52:29 +00:00
|
|
|
// Fallback to software vblank events for now. Maybe use the Present extension or
|
|
|
|
// something to get notified when the overlay window is actually presented?
|
|
|
|
m_vsyncMonitor = SoftwareVsyncMonitor::create(this);
|
|
|
|
connect(backend->renderLoop(), &RenderLoop::refreshRateChanged, this, [this, backend]() {
|
|
|
|
m_vsyncMonitor->setRefreshRate(backend->renderLoop()->refreshRate());
|
|
|
|
});
|
|
|
|
m_vsyncMonitor->setRefreshRate(backend->renderLoop()->refreshRate());
|
|
|
|
|
|
|
|
connect(m_vsyncMonitor, &VsyncMonitor::vblankOccurred, this, &X11XRenderBackend::vblank);
|
|
|
|
|
2020-12-05 15:54:19 +00:00
|
|
|
init(true);
|
|
|
|
}
|
|
|
|
|
|
|
|
X11XRenderBackend::~X11XRenderBackend()
|
|
|
|
{
|
Introduce RenderLoop
At the moment, our frame scheduling infrastructure is still heavily
based on Xinerama-style rendering. Specifically, we assume that painting
is driven by a single timer, etc.
This change introduces a new type - RenderLoop. Its main purpose is to
drive compositing on a specific output, or in case of X11, on the
overlay window.
With RenderLoop, compositing is synchronized to vblank events. It
exposes the last and the next estimated presentation timestamp. The
expected presentation timestamp can be used by effects to ensure that
animations are synchronized with the upcoming vblank event.
On Wayland, every outputs has its own render loop. On X11, per screen
rendering is not possible, therefore the platform exposes the render
loop for the overlay window. Ideally, the Scene has to expose the
RenderLoop, but as the first step towards better compositing scheduling
it's good as is for the time being.
The RenderLoop tries to minimize the latency by delaying compositing as
close as possible to the next vblank event. One tricky thing about it is
that if compositing is too close to the next vblank event, animations
may become a little bit choppy. However, increasing the latency reduces
the choppiness.
Given that, there is no any "silver bullet" solution for the choppiness
issue, a new option has been added in the Compositing KCM to specify the
amount of latency. By default, it's "Medium," but if a user is not
satisfied with the upstream default, they can tweak it.
2020-11-19 08:52:29 +00:00
|
|
|
// No completion events will be received for in-flight frames, this may lock the
|
|
|
|
// render loop. We need to ensure that the render loop is back to its initial state
|
|
|
|
// if the render backend is about to be destroyed.
|
|
|
|
RenderLoopPrivate::get(kwinApp()->platform()->renderLoop())->invalidate();
|
|
|
|
|
2020-12-05 15:54:19 +00:00
|
|
|
if (m_front) {
|
|
|
|
xcb_render_free_picture(connection(), m_front);
|
|
|
|
}
|
|
|
|
m_overlayWindow->destroy();
|
|
|
|
}
|
|
|
|
|
|
|
|
OverlayWindow *X11XRenderBackend::overlayWindow()
|
|
|
|
{
|
|
|
|
return m_overlayWindow.data();
|
|
|
|
}
|
|
|
|
|
|
|
|
void X11XRenderBackend::showOverlay()
|
|
|
|
{
|
|
|
|
if (m_overlayWindow->window()) { // show the window only after the first pass, since
|
|
|
|
m_overlayWindow->show(); // that pass may take long
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void X11XRenderBackend::init(bool createOverlay)
|
|
|
|
{
|
|
|
|
if (m_front != XCB_RENDER_PICTURE_NONE)
|
|
|
|
xcb_render_free_picture(connection(), m_front);
|
|
|
|
bool haveOverlay = createOverlay ? m_overlayWindow->create() : (m_overlayWindow->window() != XCB_WINDOW_NONE);
|
|
|
|
if (haveOverlay) {
|
|
|
|
m_overlayWindow->setup(XCB_WINDOW_NONE);
|
|
|
|
ScopedCPointer<xcb_get_window_attributes_reply_t> attribs(xcb_get_window_attributes_reply(connection(),
|
|
|
|
xcb_get_window_attributes_unchecked(connection(), m_overlayWindow->window()), nullptr));
|
|
|
|
if (!attribs) {
|
|
|
|
setFailed("Failed getting window attributes for overlay window");
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
m_format = XRenderUtils::findPictFormat(attribs->visual);
|
|
|
|
if (m_format == 0) {
|
|
|
|
setFailed("Failed to find XRender format for overlay window");
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
m_front = xcb_generate_id(connection());
|
|
|
|
xcb_render_create_picture(connection(), m_front, m_overlayWindow->window(), m_format, 0, nullptr);
|
|
|
|
} else {
|
|
|
|
// create XRender picture for the root window
|
|
|
|
m_format = XRenderUtils::findPictFormat(kwinApp()->x11DefaultScreen()->root_visual);
|
|
|
|
if (m_format == 0) {
|
|
|
|
setFailed("Failed to find XRender format for root window");
|
|
|
|
return; // error
|
|
|
|
}
|
|
|
|
m_front = xcb_generate_id(connection());
|
|
|
|
const uint32_t values[] = {XCB_SUBWINDOW_MODE_INCLUDE_INFERIORS};
|
|
|
|
xcb_render_create_picture(connection(), m_front, rootWindow(), m_format, XCB_RENDER_CP_SUBWINDOW_MODE, values);
|
|
|
|
}
|
|
|
|
createBuffer();
|
|
|
|
}
|
|
|
|
|
|
|
|
void X11XRenderBackend::createBuffer()
|
|
|
|
{
|
|
|
|
xcb_pixmap_t pixmap = xcb_generate_id(connection());
|
|
|
|
const auto displaySize = screens()->displaySize();
|
|
|
|
xcb_create_pixmap(connection(), Xcb::defaultDepth(), pixmap, rootWindow(), displaySize.width(), displaySize.height());
|
|
|
|
xcb_render_picture_t b = xcb_generate_id(connection());
|
|
|
|
xcb_render_create_picture(connection(), b, pixmap, m_format, 0, nullptr);
|
|
|
|
xcb_free_pixmap(connection(), pixmap); // The picture owns the pixmap now
|
|
|
|
setBuffer(b);
|
|
|
|
}
|
|
|
|
|
|
|
|
void X11XRenderBackend::present(int mask, const QRegion &damage)
|
|
|
|
{
|
Introduce RenderLoop
At the moment, our frame scheduling infrastructure is still heavily
based on Xinerama-style rendering. Specifically, we assume that painting
is driven by a single timer, etc.
This change introduces a new type - RenderLoop. Its main purpose is to
drive compositing on a specific output, or in case of X11, on the
overlay window.
With RenderLoop, compositing is synchronized to vblank events. It
exposes the last and the next estimated presentation timestamp. The
expected presentation timestamp can be used by effects to ensure that
animations are synchronized with the upcoming vblank event.
On Wayland, every outputs has its own render loop. On X11, per screen
rendering is not possible, therefore the platform exposes the render
loop for the overlay window. Ideally, the Scene has to expose the
RenderLoop, but as the first step towards better compositing scheduling
it's good as is for the time being.
The RenderLoop tries to minimize the latency by delaying compositing as
close as possible to the next vblank event. One tricky thing about it is
that if compositing is too close to the next vblank event, animations
may become a little bit choppy. However, increasing the latency reduces
the choppiness.
Given that, there is no any "silver bullet" solution for the choppiness
issue, a new option has been added in the Compositing KCM to specify the
amount of latency. By default, it's "Medium," but if a user is not
satisfied with the upstream default, they can tweak it.
2020-11-19 08:52:29 +00:00
|
|
|
m_vsyncMonitor->arm();
|
|
|
|
|
2020-12-05 15:54:19 +00:00
|
|
|
const auto displaySize = screens()->displaySize();
|
|
|
|
if (mask & Scene::PAINT_SCREEN_REGION) {
|
|
|
|
// Use the damage region as the clip region for the root window
|
|
|
|
XFixesRegion frontRegion(damage);
|
|
|
|
xcb_xfixes_set_picture_clip_region(connection(), m_front, frontRegion, 0, 0);
|
|
|
|
// copy composed buffer to the root window
|
|
|
|
xcb_xfixes_set_picture_clip_region(connection(), buffer(), XCB_XFIXES_REGION_NONE, 0, 0);
|
|
|
|
xcb_render_composite(connection(), XCB_RENDER_PICT_OP_SRC, buffer(), XCB_RENDER_PICTURE_NONE,
|
|
|
|
m_front, 0, 0, 0, 0, 0, 0, displaySize.width(), displaySize.height());
|
|
|
|
xcb_xfixes_set_picture_clip_region(connection(), m_front, XCB_XFIXES_REGION_NONE, 0, 0);
|
|
|
|
} else {
|
|
|
|
// copy composed buffer to the root window
|
|
|
|
xcb_render_composite(connection(), XCB_RENDER_PICT_OP_SRC, buffer(), XCB_RENDER_PICTURE_NONE,
|
|
|
|
m_front, 0, 0, 0, 0, 0, 0, displaySize.width(), displaySize.height());
|
|
|
|
}
|
|
|
|
|
|
|
|
xcb_flush(connection());
|
|
|
|
}
|
|
|
|
|
Introduce RenderLoop
At the moment, our frame scheduling infrastructure is still heavily
based on Xinerama-style rendering. Specifically, we assume that painting
is driven by a single timer, etc.
This change introduces a new type - RenderLoop. Its main purpose is to
drive compositing on a specific output, or in case of X11, on the
overlay window.
With RenderLoop, compositing is synchronized to vblank events. It
exposes the last and the next estimated presentation timestamp. The
expected presentation timestamp can be used by effects to ensure that
animations are synchronized with the upcoming vblank event.
On Wayland, every outputs has its own render loop. On X11, per screen
rendering is not possible, therefore the platform exposes the render
loop for the overlay window. Ideally, the Scene has to expose the
RenderLoop, but as the first step towards better compositing scheduling
it's good as is for the time being.
The RenderLoop tries to minimize the latency by delaying compositing as
close as possible to the next vblank event. One tricky thing about it is
that if compositing is too close to the next vblank event, animations
may become a little bit choppy. However, increasing the latency reduces
the choppiness.
Given that, there is no any "silver bullet" solution for the choppiness
issue, a new option has been added in the Compositing KCM to specify the
amount of latency. By default, it's "Medium," but if a user is not
satisfied with the upstream default, they can tweak it.
2020-11-19 08:52:29 +00:00
|
|
|
void X11XRenderBackend::vblank(std::chrono::nanoseconds timestamp)
|
|
|
|
{
|
|
|
|
RenderLoopPrivate *renderLoopPrivate = RenderLoopPrivate::get(m_backend->renderLoop());
|
|
|
|
renderLoopPrivate->notifyFrameCompleted(timestamp);
|
|
|
|
}
|
|
|
|
|
2020-12-05 15:54:19 +00:00
|
|
|
void X11XRenderBackend::screenGeometryChanged(const QSize &size)
|
|
|
|
{
|
|
|
|
Q_UNUSED(size)
|
|
|
|
init(false);
|
|
|
|
}
|
|
|
|
|
|
|
|
bool X11XRenderBackend::usesOverlayWindow() const
|
|
|
|
{
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
} // namespace KWin
|