Loïc Molinari authored 1 year ago and Pekka Paalanen committed 1 year ago

Add a few tests ensuring the box32 quad clipping wrapper works as
expected.

Signed-off-by: Loïc Molinari <loic.molinari@collabora.com>

Loïc Molinari authored 1 year ago and Pekka Paalanen committed 1 year ago

The non-zero area check of clipper_quad_clip() is incorrect for quads
initialized with a polygon starting with a vertical edge. In order to
handle polygons starting with an horizontal edge and polygons starting
with a vertical one, it must check opposite vertices for equality.

The test previously described as "Box intersects entire smaller
aligned quad" is now described as "Clockwise winding and top/left
initial vertex". This test keeps the same values as before but all
combinations of winding order and first edge orientations are
also tested. The QUAD() macro isn't used anymore to do so.

Signed-off-by: Loïc Molinari <loic.molinari@collabora.com>

Loïc Molinari authored 1 year ago and Pekka Paalanen committed 1 year ago

Add quad clipping tests checking intersections at all edges and
corners of axis-aligned and unaligned quads with negative and positive
values.

Signed-off-by: Loïc Molinari <loic.molinari@collabora.com>

Loïc Molinari authored 1 year ago and Pekka Paalanen committed 1 year ago

Simplify box and quad declarations using dedicated macros.

Hard-coded literals are used instead of preprocessor constants for the
sake of brevity and because, as several new tests will be added, it
appears to be easier to understand those when values are inlined.

Descriptions have been revised to better understand the intent of each
test. Previous descriptions used a coord system with Y pointing up,
new descriptions use Y pointing down in order to improve consistency
with the rest of the code base and to have a common ground when
talking about winding order.

All the tests keep the same input values with the exception of the
last 2 tests with (new) descriptions: "Rotated quad with edges
adjacent to box corners" and "Rotated quad with edges cutting out box
corners". These tests had a different winding order and have been
modified so that all tests use a clockwise order (new axis
convention). The last test also gets a counter-clockwise version in
order to ensure the opposite winding order is correctly supported too.

Signed-off-by: Loïc Molinari <loic.molinari@collabora.com>

Loïc Molinari authored 1 year ago and Pekka Paalanen committed 1 year ago

When the first vertex passed to the clipper is clipped into two
vertices, the second vertex can sometimes be emitted as the first
clipped vertices while the first vertex is emitted as the last one. A
new utility function assert_vertices() is added to handle that
case. The function also checks the number of clipped vertices and the
clipped vertices in one go.

Signed-off-by: Loïc Molinari <loic.molinari@collabora.com>

The fullscreen should take precedence if the client requests both
window states at the same time.

Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>

Marius Vlad authored 1 year ago and Pekka Paalanen committed 1 year ago

A possible solution to warn out users that messages will be dropped
otherwise.

Fixes: #858

Signed-off-by: Marius Vlad <marius.vlad@collabora.com>

Wismill authored 1 year ago and Daniel Stone committed 1 year ago

There is a typo in the call to `zwp_input_method_context_v1_modifiers`:
the `mods_latched` and `mods_locked` arguments have wrong values.

Fixed by forwarding the correct masks.

Signed-off-by: Pierre Le Marre <dev@wismill.eu>

I want re-use this variable for printing the colorimetry mode list as
well in a future patch, so a generic name is less confusing.

Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

Performance-wise this is moot, but since we are detecting if the raw
EDID data changed, might as well use it.

Now we have a path where the head's device_changed is almost guaranteed
since EDID changed, and that's useful for the next patch.

We can only do this, because the core initializes a head with values
that we would be setting anyway when EDID is missing, e.g. disconnected
head on compositor start-up.

Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

Store the EDID data as-is, so that we can tell when the EDID blob has
changed. This is not too useful yet, because all the weston_head_set_*()
API raises the device_changed flag only if the information actually
changes. However, I want to expose the libdisplay-info di_info structure
through weston_head, and those cannot be (as) easily compared.

We need to know when the EDID blob changes, so we can call
weston_head_set_device_changed() appropriately when updating di_info.

Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

One of the three fixture setups of alpha-blending test requires
color-lcms.so. If color-lcms.so is not built, that fixture fails.

Make it skip as necessary, making the test suite pass with
color-management-lcms=false build option.

Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

This is very trivial test, which does not skip correctly when built with
color-management-lcms=false. It fails instead.

Since alpha-blending test already covers everything that color-manager
test did, remove color-manager test.

My editor eliminated a stray whitespace, too.

Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

Nicholas Niro authored 1 year ago and Pekka Paalanen committed 1 year ago

Added an entry with examples into the man page and also
into frontend/main.c for the '--help' interface.

Signed-off-by: Nicholas Niro <blowfist@xroutine.net>

Nicholas Niro authored 1 year ago and Pekka Paalanen committed 1 year ago

This command is being executed in parallel with the westen instance,
just like the autolaunch config.

I recently came across the kiosk-shell and found out I could start a program exclusive
weston instance using it and that opened my eyes to new possibilities.

With the desktop-shell, it is necessary to set up quite a few options like the panel launchers,
the background image/color and a few other things and that indeed make the configuration file mandatory.

With the kiosk-shell all you really care about is the underlying program,
the vast majority of the configuration file options are not relevant for that shell.
That made me wonder how convenient it would be to forego the configuration file
and implement the autolaunch option directly in the weston program. That indeed worked pretty
well and that is why I decided to propose this merge request.

I think this avenue opens up a different set of uses cases for weston where rather than just have one
"big" desktop-shell instance, we could have multiple smaller and potentially specific usage instances.
Yes, it can be done with configuration files currently. But what it boils down to is convenience.
Maybe this convenience will enable other people to start more than just one weston instance
in the future. For instance, to quickly watch an image or opening up a pdf file.

This patch was made in a matter that is meant to be consistent with the intuitive
way other programs accept a command input, like so :

weston <some options> -- some_command option1 ... optionN

Further work would be necessary to remove the requirement for the '--'
option. To do that, we would need to check if the option is a valid command
and not just a mistyped option.

There may be some conflict with the current autolaunch implementation.
I'm not sure if both 'command' and 'autolaunch' could be used at the
same time using this implementation. I think it would be necessary to
have a distinct watch and pid variable in the 'wet' context variable
for the command to support this.

Signed-off-by: Nicholas Niro <blowfist@xroutine.net>

Nicholas Niro authored 1 year ago and Pekka Paalanen committed 1 year ago

This toggles parse_options to ignore the rest of the remaining
options.

Signed-off-by: Nicholas Niro <blowfist@xroutine.net>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Signed-off-by: Dylan Aïssi <dylan.aissi@collabora.com>

Just like we already do for planes with proper zpos. Otherwise we'll
often end up choosing the primary plane instead of an overlay one
in `drm_output_find_plane_for_view()`.

Signed-off-by: Robert Mader <robert.mader@collabora.com>

Ray Smith authored 1 year ago and Pekka Paalanen committed 1 year ago

This seems to think formats needs to be NULL terminated, but it doesn't
and gl_renderer_get_egl_config asserts that all formats_count elements
are not NULL.

This happens when EGL_KHR_no_config_context is not supported.

Signed-off-by: Ray Smith <rsmith@brightsign.biz>

It adds the followig paragraph:
```
Starting from version 5, the invalid_format protocol error is sent if
all planes don't use the same modifier.
```

We already assumed this in some places and, most importantly, it's
required by the kernel. Thus alter `dmabuf_attributes.modifier` to make
it clear that different modifiers for multi-planar dmabufs were never
supported.

Signed-off-by: Robert Mader <robert.mader@collabora.com>

Signed-off-by: Robert Mader <robert.mader@collabora.com>

We should use the actual stride to create tmp pixmap.

Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>

The gl_renderer_do_read_pixels() is expecting stride in bytes.

Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>

Derek Foreman authored 1 year ago and Marius Vlad committed 1 year ago

commit 79212ba9 fixed a bug by
introducing a new one.

Before that point we could clip paint node damage to stale
visibility data. After that point we post damage for occluded
views, leading to large amounts of pointless drawing.

Add back the clip to visible region, in
weston_output_flush_damage_for_plane(), where we have up to date
visibility region information.

fixes 79212ba9

Signed-off-by: Derek Foreman <derek.foreman@collabora.com>

Colin Kinloch authored 1 year ago and Marius Vlad committed 1 year ago

This used to not be a problem as both `struct stacking` and `struct
window` start with a pointer to `struct display`.

Signed-off-by: Colin Kinloch <colin.kinloch@collabora.com>

Chao Guo authored 1 year ago and Daniel Stone committed 1 year ago

If client changes the buffer transform or size and re-commit the surface
before surface dirtys buffer (weston_surface_attach), the viewport
source validity check may fail. Because the width_from_buffer and
height_from_buffer cannot be updated based on new buffer transform.

Therefore, add the situation of size dirty (WESTON_SURFACE_DIRTY_SIZE) to
source validity check, width_from_buffer and height_from_buffer can be
updated in time when the buffer transform and scale change.

Signed-off-by: Chao Guo <chao.guo@nxp.com>

zhou liang authored 1 year ago and zhou liang committed 1 year ago

The `drm_output_deinit` should use the drm device passed by the
function, should not use the `b->drm` device.

Signed-off-by: Zhou Liang <174381115@qq.com>

zhou liang authored 1 year ago and zhou liang committed 1 year ago

If both the head and writeback are not found, then we should add
connectors to the drm device passed by the function, not the b->drm
device.

Signed-off-by: Zhou Liang <174381115@qq.com>

zhou liang authored 1 year ago and zhou liang committed 1 year ago

In a multi-GPU environment, different cards may contain connectors with the
same ID, and drm_head_find_by_connector just use the connector_id to find
the connector, it may find the wrong connector.
Fix this by find the connector based on the drm device and connector id.

Signed-off-by: Zhou Liang <174381115@qq.com>

If a view is non-opaque - such as an overlay over a video - we shouldn't
force it to be on the primary plane, as that's where the underlying
content should be placed, such as the video view.

dc0de9ee already mentioned: "This check should be changed in future to
only filter for opaque views, but that's for another time."

Adding "Fixes" at this is arguably a bug fix:
Fixes: dc0de9ee (backend-drm: Move overlay vs. primary plane check earlier)
Fixes: 2538aacc (backend-drm: Construct a zpos candidate list of planes)

Signed-off-by: Robert Mader <robert.mader@collabora.com>

Currently we flush damage for the "primary plane" every repaint, but this
is folly.

The drm backend may skip rendering entirely if using an all-planes
composition. This could leave the renderer plane in a messy state if a
surface on an overlay plane disappears.

Instead, let the backends flush the primary plane damage when they know
they need to render.

Fixes #864

Signed-off-by: Derek Foreman <derek.foreman@collabora.com>

And with it, bump libweston to next major version, 14. We seems like
we never used that argument so better just removed it.

Signed-off-by: Marius Vlad <marius.vlad@collabora.com>

The idle_animation_destroy task should be removed when destroying
animations by the other callers such as handle_animation_view_destroy().

Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>

Michael Olbrich authored 1 year ago and Marius Vlad committed 1 year ago

Otherwise shseat->focused_ivisurf can point to deleted memory.
This does not happen with the hmi-controller because it explicitly assigns a new
focused surface. But the ivi-shell should not relay on the controller here.

Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de>

Khem Raj authored 1 year ago and Marius Vlad committed 1 year ago

Latest musl has removed the declaration from string.h [1] as it only
implements POSIX version alone and string.h in glibc implements GNU
version of basename. This now results in compile errors on musl.

This might be a warning with older compilers but it is error with
Clang-17+ as it treats -Wimplicit-function-declaration as error

Switch the use in backlight_init function to use POSIX version

[1] https://git.musl-libc.org/cgit/musl/commit/?id=725e17ed6dff4d0cd22487bb64470881e86a92e7



Signed-off-by: Khem Raj <raj.khem@gmail.com>

Pekka Paalanen authored 1 year ago and Daniel Stone committed 1 year ago

"Frontend" is a much more descriptive name for the code that is in the
directory called "compositor".

Fixes: #633



Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

Pekka Paalanen authored 1 year ago and Daniel Stone committed 1 year ago

Looks like these don't actually need it.

Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>

This was added in 4def21c1.

0c1ab2ad removed all uses of NULL
weston_compositor, making the workaround unnecessary.

Drop the workaround, it's dead code.

Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.com>