Extract the pixman renderbuffer configuration from the add_buffer function into
a helper function to simplify the addition of the GL renderer.
Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
Use helper function for setting up the pixman renderer to simplify the addition
of the GL renderer.
Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
clip_quad() is a dedicated clipping function for quads that doesn't
depend on any GL renderer internal structures. It can be moved out to
the clipper to be called by both the renderer and the clipping test
client without having to duplicate code.
Signed-off-by: Loïc Molinari <loic.molinari@gmail.com>
Since both the surface rect and the transformed quad are axis-aligned
in the simple clipping path, non-zero area detection can more
efficiently be checked post-clipping by comparing opposite edges.
Signed-off-by: Loïc Molinari <loic.molinari@gmail.com>
Add a basic check to let the clipper take the simple axis-aligned path
when nodes are solely transformed with a translation and/or a scaling.
That makes some nodes like sub-surfaces (which always have their
transform enabled) take the fast path in the common case.
Signed-off-by: Loïc Molinari <loic.molinari@gmail.com>
Move vertex clipper back to single-precision floating point
intermediates. Since positions are sent down the graphics hardware as
single-precision values, this prevents useless conversions between
single and double precision values and lets compilers fit twice as
much data into vector registers. It also removes a copy by letting the
clipper store vertices directly into the vertex buffer.
This is mostly reverting the conversion to double-precision that
happened along with the switch to the weston_coord struct for vertex
coordinates (commit 10e70bf23c).
Signed-off-by: Loïc Molinari <loic.molinari@gmail.com>
Let the graphics hardware handle the transformation from surface
position to texture coordinates. Paint nodes now have a single vertex
position attribute from which texture coordinates are derived. A new
vertex shader variant handles the transformation.
Signed-off-by: Loïc Molinari <loic.molinari@gmail.com>
The clipper transforms dirty rects to surface space before clipping.
Each dirty rect is transformed by the same matrix for each surface
rect. This change decouples the transformation and the clipping code
to transform and compute the bounding box of dirty rects just once.
Signed-off-by: Loïc Molinari <loic.molinari@gmail.com>
Clip dirty rects to surface rects in surface coordinate space.
Dispatch vertices in surface coordinates and let the graphics hardware
handle the transformations. Clipping in global coordinate space
implies a useless roundtrip on the CPU to get the clipped polygons
back in surface coordinates for the buffer transformation. Clipping in
surface coordinate space prevents that.
This might seem counter-intuitive at first because in surface space
it's the dirty rects that are clipped to axis-aligned surface rects,
while it's the opposite in global space.
The projection matrix now combines the view and the output transforms.
Signed-off-by: Loïc Molinari <loic.molinari@gmail.com>
Let the GL renderer render to FBOs and read the pixels into the Neat VNC
frame buffers. This allows to run the VNC backend with the GL renderer.
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
For software backends like VNC, support downloading the FBO renderbuffer
contents via glReadPixels automatically at the end of repaint_output.
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Add support for creating surfaceless outputs and rendering to FBOs.
The backend has to create FBOs with the create_fbo API and pass the
resulting weston_renderbuffer handles to repaint_output.
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Since the pixman_region32_t damage is initialized unconditionally, also
finalize it unconditionally. Otherwise we leak rectangle memory when
sb->output->frame is NULL.
Reported-by: Daniel Stone <daniels@collabora.com>
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Add a to_pixman_renderbuffer() helper to consolidate the
container_of(renderbuffer, ...) macro calls.
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
The override-redirect window will not be assigned a shell_surface
object. If it is used as a parent window, it will cause a crash
when calling the set_parent function.
The EWMH specification does not describe the behavior of an
override-redirect window as a parent window, so we should ignore
this case.
Signed-off-by: Liu, Kai1 <kai1.liu@intel.com>
Remove the now unused previous/total_damage regions and the
buffer/border_damage arrays, as well as the output_get_damage and
output_rotate_damage functions.
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Create dummy renderbuffers to track surface buffer damage on
demand. The renderbuffer representing the surface buffer that
is currently rendered to is inferred using buffer age.
This aligns damage tracking with the Pixman renderer and will
simplify adding FBO rendering support.
The previous/total_damage regions and the buffer/border_damage
arrays are now unused except for validation.
They can be removed next.
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Split the buffer age query out into a separate function.
The following patches will replace the remainder of the
output_get_damage function.
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
In preparation of turning mem leak detection on in CI, this patch adds
LSAN_OPTIONS to be able to pass the suppression file.
This *does not* turn the mem leak detection on yet, as we still
need some additional mem leak fixes.
This removes fast unwind set to zero as well, as that will add a massive
runtime penalty to some of our tests, and we'll no longer need it as
we've included the entire libfontconfig as a mem leak source in the
suppression file.
Signed-off-by: Marius Vlad <marius.vlad@collabora.com>
As we can't really use `cleanup_after_cairo()` in our headless backend,
add more leak suppression entry, found while running the entire
test suite.
FcConfigSubstituteWithPat was already added (which seems to be main
source) but found a couple of more while running the entire suite.
The issue is that we need to set fast unwind to zero in order track
those leaks but that would result into some massive runtime penalty and
we get into the issue of timing out some of the tests.
So rather than doing that, just add the entire libconfig library and be
done with it. With it, this removes any pango/cairo former symbols as
that seems to catch all other leaks as well.
Signed-off-by: Marius Vlad <marius.vlad@collabora.com>
In some circumstances (like system load) we seem to be racing with the
threads fontconfig creates and the resource release happening inside
cleanup_after_cairo(), and we would still find a cached entry holding a
font map reference while we are on the compositor exit path --
which happens in cleanup_after_cairo() when calling
cairo_debug_reset_static_data().
This was introduced with commit 823580e070, 'backend-headless: fully
release pango and fontconfig', as a way to have a clean memory leak
report, but due to the fact we can't influence how libraries manage
their threads, the solution (for now at least) would be to just remove
it entirely.
Running it at the end in the test itself, or before calling exit(2),
while it does narrow the window, it still exhibits the cairo crash assert
related to having cached entries in that font map hash table.
Reference: https://gitlab.freedesktop.org/wayland/weston/-/issues/756
Signed-off-by: Marius Vlad <marius.vlad@collabora.com>
If the source is not supported, we won't receive the capture
information. So the capture info (size/format) will be zeroed, and we
fail while trying to create a buffer for the screenshot with size/format
zeroed.
With this patch we fail if we don't receive the capture info, what makes
the failure reason more explicit.
Signed-off-by: Leandro Ribeiro <leandro.ribeiro@collabora.com>
Since 2d70bdfdcd "drm-backend: add support
to output capture writeback source", the DRM-backend was broken for KMS
devices that do not support the atomic API. This fixes that.
We don't support writeback screenshots without atomic modeset support.
So for such devices, we never update the output capture info
(weston_output_update_capture_info()) for the writeback source.
The function that we use to pull writeback tasks
(weston_output_pull_capture_task()) asserts that the capture providers
(renderers, DRM-backend) did not forget to update the capture info
(size/format) if something changed. But as we've never updated the
capture info for such devices, it is zeroed, leading to an assert hit.
With this patch we only pull the capture task for KMS devices that
support the atomic API.
Signed-off-by: Leandro Ribeiro <leandro.ribeiro@collabora.com>
In 2d70bdfdcd "drm-backend: add support to
output capture writeback source" we've ensured that disable_planes
should be false in order to support writeback capture tasks.
But this was wrong; disable_planes is transient (it is true when
there's some sort of content recording happening), and we enable/disable
that during compositor's lifetime.
This is dangerous and may result in a crash. Imagine the following
sequence:
1. screen recording starts, disable_planes is set to true.
2. for whatever reason the output size changes, and we end up
not updating capture info because we think that writeback is not
supported by the device.
3. screen recording stops, disable_planes is set to false.
4. user tries to take a writeback screenshot, and the
DRM-backend will pull a writeback capture task with
weston_output_pull_capture_task().
5. this function has an assert to ensure that the DRM-backend
did not forget to update the capture info, and we hit that
assert.
With this patch we drop disable_planes being false as a condition to
support writeback. So now we keep the capture info up-to-date even when
screen recording is happening, and we gracefully fail writeback tasks.
Signed-off-by: Leandro Ribeiro <leandro.ribeiro@collabora.com>
The surface buffer must be committed before a ivi-controller adds ivi
surface to a ivi layer. This constraint is necessary information for the
ivi-controller.
Signed-off-by: Tomohito Esaki <etom@igel.co.jp>
One .TP was lost, causing --wait-for-debugger section to be merged in
the previous section. Fix it.
Fixes 5ffda17e21
Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>
This adds three new helpers: one to iterate over all debug scopes
created/added and other two are for simpler getters for the scope name
and the description.
Included with this change is also a simple test to retrieve them.
This is an alternative to using the debug scope list advertised when
using the weston-debug private extension. libweston users can use this
directly to know which scopes they can subscribe to, and there's no need
to have a client implementation for the weston-debug protocol.
Signed-off-by: Marius Vlad <marius.vlad@collabora.com>
Don't forget to transfer the fullscreen setting from the config to the backend.
Without this, weston tries to resize the window with the windowed output API but
that is not registered with new_config.fullscreen == true.
This code was accidentally lost in 0a5bb7acff
("backend-wayland: Use renderer enum type for config selection"). So just
restore it.
Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>
We were printing only the matrices (cmsSigMatrixElemType) up to now.
Start printing the curve sets (cmsSigCurveSetElemType) as well.
Signed-off-by: Leandro Ribeiro <leandro.ribeiro@collabora.com>
Function matrix_print() is called only by pipeline_print(), which
already checks if the log scope is enabled. So remove the repeated
check from matrix_print().
Signed-off-by: Leandro Ribeiro <leandro.ribeiro@collabora.com>
Which should have been with commit ed012ee505, 'libweston:
Store view instead of surface, and add flags, to activation data'.
Reported-by: Pekka Paalanen <pekka.paalanen@collabora.com>
Signed-off-by: Marius Vlad <marius.vlad@collabora.com>
The addition in a former commit of the flags field in the activation data will
let us pass the reason for activation to the constraint logic. We use that
reason here to unconditionally enable constraints in the recently 'fullscreened'
surface.
Signed-off-by: Sergio Gómez <sergio.g.delreal@gmail.com>
When a toplevel xdg surface is changing from non-fullscreen to fullscreen upon a
client request, activate its corresponding shell surface.
This will let us use the activation mechanism to enforce the enabling of
pointer confinement.
Signed-off-by: Sergio Gómez <sergio.g.delreal@gmail.com>
When the position or size of a fullscreen surface changes (e.g., due to change
in the output), the surface commit logic in desktop_surface_committed() resets
the fullscreen by first calling unset_fullscreen() and later on calling
shell_configure_fullscreen(). One part of this reset is the recreation of the
black view curtain (destroyed in unset_fullscreen() and created again in
shell_configure_fullscreen()).
In the upcoming commit we will replace the call to shell_configure_fullscreen()
with a call to activate() (since we want to activate this shell surface, and
activate() already has a call to shell_configure_fullscreen() in it).
The code in activate() by default lowers the fullscreen layer of the shell
surface through lower_fullscreen_layer(), which is called before
shell_configure_fullscreen(). This lowering function assumes that the fullscreen
shell surface has a valid black view curtain, which as said before was removed
in the call to unset_fullscreen(). Add the check to guard against this case.
Signed-off-by: Sergio Gómez <sergio.g.delreal@gmail.com>
This will be used to let know the constraints code that the reason for
activation is that the client has requested to set the surface to fullscreen.
Signed-off-by: Sergio Gómez <sergio.g.delreal@gmail.com>
Since we want to pass the view to the surface activation listener inside the
constraints code, and the surface is reachable from the view anyway.
The flags field will let us pass the reason for activation to the constraints
code, which will then handle especially the fullscreen case.
Signed-off-by: Sergio Gómez <sergio.g.delreal@gmail.com>
Currently, and for legacy reasons, weston_seat_set_keyboard_focus() contains
logic related to surface activation. Since this function is always called from
weston_view_activate_input(), move that code there where it seems more
appropriate.
This will help us in subsequent commits by avoiding to have to change the
signature of weston_seat_set_keyboard_focus(), which would make that function
even more awkward than it currently is.
Signed-off-by: Sergio Gómez <sergio.g.delreal@gmail.com>
