Previously renderers were not told when the output (framebuffer they
need to draw) size changed. Renderers just pulled that information out
from weston_output::current_mode when they happened to need it. This
makes some things awkward, like resizing the shadow or intermediate
buffers. In fact, Pixman-renderer does not even support resizing its
shadow buffer, nor does GL-renderer. DRM-backend has to destroy and
re-create the renderer output state anyway, but rdp, x11 and wayland
backends would be natural users of resizing API.
This commit adds an API for resizing with empty implementations. Actual
implementations will be added in following patches for each renderer
while moving parts of resizing code from backends into the renderers.
No-op renderer needs no implementation.
Only wayland-backend has actual resizing code already, and that is made
to call the new API. Unfortunately, Pixman and GL renderers differ: one
does not blit them while the other does. In order to assert the
functionality of each renderer to keep the API consistent,
wayland-backend needs to lie to pixman-renderer. That's not new, it
already does so in wayland_output_get_shm_buffer() where the 'pm_image'
addresses only the interior area instead of the whole buffer.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>
Except the module dir path, they're one and the same. This change
warrants a libweston version bump, if it hasn't been done already.
Signed-off-by: Marius Vlad <marius.vlad@collabora.com>
Replace all uses of weston_transform_region with
weston_matrix_transform_region, then remove the function completely.
Signed-off-by: Derek Foreman <derek.foreman@collabora.com>
As wayland-backend is blitting the output decorations into the output
buffer itself, it pretends towards the pixman-renderer that there is no
decorations area. The pixman_image_create_bits() call wraps the
previously allocated buffer with an offset so that pixman-renderer will
paint in the right position.
The bug is that this pixman image was using the original buffer width
and height, instead of the composited area width and height. So the
pixman image looks too big to pixman-renderer, but the renderer didn't
care. The image being too big does risk access out of bounds in
pixman-renderer.
I found this when I was making renderers explicitly aware of the
frambuffer size and resizing, added asserts, and they surprisingly
failed. This fixes that.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>
A following patch is going to need the introduced 'area' and 'fb_size'
variables. Until then though, a little hack is needed to avoid no-gl
builds failing with error: variable 'fb_size' set but not used.
While starting to use struct weston_geometry, convert also the input and
opaque regions to use it. This shortens and simplifies the code, as we
can drop the roughly duplicate code of doing stuff for with vs. without
a frame.
No change in behavior, this is pure refactoring.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>
We were only destroying these when the parent display removed the output
global. Do it on shutdown too, so we can avoid leaking it.
Signed-off-by: Daniel Stone <daniels@collabora.com>
As a first step towards heterogeneous outputs, ignore other backends'
heads and outputs. This is done by checking the destroy callbacks for
heads and outputs.
See: https://gitlab.freedesktop.org/wayland/weston/-/issues/268
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
The repaint_data is entirely backend specific. Moreover, it is only used by the
drm backend, while other backends ignore the repaint data.
There will always be only one repaint active, thus, there is no need to pass the
repaint data from the outside.
The repaint_data breaks with the multi-backend series, which calls repaint begin
for all backends to get the repaint_data. The repaint_data of the last backend
will then be passed to all other backend. At the moment, this works, because the
drm backend is the only backend that implements the begin_repaint call.
Another option would be to track the repaint data per backend in the compositor,
but actually, it the backend needs to track state across the calls, it's its own
responsibility.
Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
weston_touch_destroy(), which is called from weston_seat_release(),
asserts that all its touch devices have been destroyed. The Wayland
backend currently destroys the touch devices ... immediately after
calling weston_seat_release().
Invert the ordering so that touch devices are destroyed first and we
don't trip over the assert.
Signed-off-by: Daniel Stone <daniels@collabora.com>
The struct wayland_input objects tracking the outer compositor's
wl_seats are now properly destroyed when the wayland backend is.
Signed-off-by: Manuel Stoeckl <code@mstoeckl.com>
The wl_display_roundtrip call was originally introduced to let the
display_add_seat function wait until a wl_seat.name event was received.
This change replaces the wl_display_roundtrip call with an
asynchronous, nonrecursive equivalent. Now a wl_display.sync callback
is used to delay the final steps of adding a seat until one protocol
roundtrip has occured/the name has been received.
Signed-off-by: Manuel Stoeckl <code@mstoeckl.com>
The wayland.c actually include 'xdg-shell-client-protocol.h' instead of
the server one, so fix it. Otherwise, it's possible to get build failure
due to race condition.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Daniel Stone <daniels@collabora.com>
[daniels: Found in OpenEmbedded/Yocto source.]
Use the real name of the seat instead of calling each seat 'default'. This makes
it easier to identify the current seat in a multi-seat environment.
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
As observed in #420 (Running Weston under Weston's kiosk shell with
multiple outputs causes the scrollback to go nuts), not
being able to cope with (a correct) resize of the parent surface would
cause the client weston instance to spin forever. If dispatching
failed, just exit.
Fixes#420
Signed-off-by: Marius Vlad <marius.vlad@collabora.com>
As an in-flight resize call might cause a call to
wayland_output_destroy_shm_buffers() to go over a list of free_buffers
list. Just initialize the lists before attempting to create the parent
surface to avoid it.
Signed-off-by: Marius Vlad <marius.vlad@collabora.com>
This header is for sharing fallback definitions for drm_fourcc.h. A new
test in tests/yuv-buffer-test.c is going to be needing XYUV8888 format,
and more new formats will be expected with HDR supports.
Share these fallback definitions in one place instead of copying them
all over.
All users of drm_fourcc.h are converted to include weston-drm-fourcc.h
instead for consistency: have the same definitions available everywhere.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>
This prevents a segfault in libwayland-cursor when the parent compositor
doesn’t have any of the needed protocols implemented.
Signed-off-by: Emmanuel Gil Peyrot <linkmauve@linkmauve.fr>
pixman_renderer_output_create currently takes a flags enum bitmask for
its options. Switch this to using a structure, so we can introduce other
non-boolean options.
Signed-off-by: Daniel Stone <daniels@collabora.com>
gl_rendererer's output_window_create has a lot of arguments now. Add a
structure for the options to make it more clear what is what.
This is in preparation for adding bare-integer arguments which are ripe
for confusion when passing positional arguments.
Signed-off-by: Daniel Stone <daniels@collabora.com>
gl_rendererer's output_create has a lot of arguments now. Add a
structure for the options to make it more clear what is what.
This is in preparation for adding bare-integer arguments which are ripe
for confusion when passing positional arguments.
Signed-off-by: Daniel Stone <daniels@collabora.com>
In order to start the repaint loop, the Wayland backend tries to damage
the full SHM buffer, but doesn't actually damage the full area if we
have a frame.
Store the buffer's width and height alongside the buffer itself, so we
can damage the full area when required.
Signed-off-by: Daniel Stone <daniels@collabora.com>
We have two kinds of libweston users: internal and external. Weston, the
frontend, counts as an external user, and should not have access to libweston
private headers. The shell plugins are external users as well, because we
intend people to be able to write them. Renderers, backends, and some plugins
are internal users who will need access to private headers.
Create two different Meson dependency objects, one for each kind.
This makes it less likely to accidentally use a private header.
Screen-share is a Weston plugin and therefore counts as an external user, but
it needs the backend API to deliver input. Until we are comfortable exposing
public API for that purpose, let it use internal headers.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>
In case the base EGLConfig is needed, gl_renderer_display_create() needs to
know it should use EGL_WINDOW_BIT or EGL_PBUFFER_BIT.
The PBUFFER case is added for when the headless backend will grow GL-renderer
support.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>
Define common_inc which includes both public_inc and the project root directory.
The project root directory will allow access to config.h and all the shared/
headers.
Replacing all custom '.', '..', '../..', '../shared' etc. include paths with
common_inc reduces clutter in the target definitions and enforces the common
#include directive style, as e.g. including shared/ headers without the
subdirectory name no longer works.
Unfortunately this does not prevent one from using private libweston headers
with the usual include pattern for public headers.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>
When all shared/ headers are included in the same way, we can drop unnecessary
include seach paths from the compiler.
This include style was chosen because it is prevalent in the code base. Doing
anything different would have been a bigger patch.
This also means that we need to keep the project root directory in the include
search path, which means that one could accidentally include private headers
with
#include "libweston/dbus.h"
or even
#include <libweston/dbus.h>
IMO such problem is smaller than the churn caused by any of the alternatives,
and we should be able to catch those in review. We might even be able to catch
those with grep in CI if necessary.
The "bad" include style was found with:
$ for h in shared/*.h; do git grep -F $(basename $h); done | grep -vF '"shared/'
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>
Now that all backends pass in a list of acceptable DRM formats, that is used to
determine if the EGLConfig has an alpha channel or not. Therefore the
opaque_attribs and alpha_attribs are now useless, and we can remove the whole
config_attribs argument from the API.
gl_renderer_get_egl_config() uses an internal attrib list that matches at least
the union of the opaque_attribs and alpha_attribs matches.
Overall, behaviour should remain unchanged.
The new attribute array becomes variable in the future, so it is left
non-const.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>
Define a specific DRM format for the GL-renderer to render in. It goes through
fuzzy matching in egl-glue.c which ensures we get exactly the number of bits
for each channel, but does not require an exact format match.
This ensures we get the bit depth we expect instead of the first arbitrary
EGLConfig.
This should not change the current behaviour, because Mesa EGL takes care to
order the configs as apps expect.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>
No caller ever used anything but NULL here, so just use NULL to simplify code.
In fact, no EGL platform defined today even defines any platform attributes
except the X11 platform for choosing a non-default SCREEN.
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>
Depending on system loading, weston-launcher could drop the drm
master access before the compositor and all the clients receive
the notification. In this case, some commit could be sent to the
drm driver too late and get refused with error EACCES.
This error condition is not properly managed and causes weston to
hang.
Change the return type of start_repaint_loop() and repaint_flush()
from void to int, and return 0 on success or -1 if the repaint has
to be cancelled.
In the callers of start_repaint_loop() and repaint_flush() handle
the return value and cancel the repaint when needed.
In backend-drm detect the error EACCES and return -1.
Note: to keep the code cleaner, this change inverts the execution
order between weston_output_schedule_repaint_reset() and
repaint_cancel().
No need to wait for suspend or for any notification; in case the
weston reschedules a repaint, it will get EACCES again.
At resume, damage-all guarantees a complete repaint.
This fix is for atomic modeset only.
Legacy modeset suffers from similar problems, but it is not fixed
by this change. Since drm_pending_state_apply() never returns
error for legacy modeset, this change has no impact on legacy
modeset.
Signed-off-by: Antonio Borneo <antonio.borneo@st.com>
Fixes: https://gitlab.freedesktop.org/wayland/weston/issues/117
