Age | Commit message (Collapse) | Author |
|
|
|
This avoids relying on the implicit wlr_output.pending state.
|
|
References: https://gitlab.freedesktop.org/wlroots/wlroots/-/merge_requests/4046
|
|
This allows for layer shell surfaces to receive focus while the surface is explicitly focused, i.e allowing
text fields to receive keyboard input just like a regular surface.
|
|
https://gitlab.freedesktop.org/wlroots/wlroots/-/merge_requests/4159
> ../sway/desktop/output.c:618:47: error: too few arguments to function 'wlr_renderer_begin_buffer_pass'
> 618 | struct wlr_render_pass *render_pass = wlr_renderer_begin_buffer_pass(
> | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
|
|
|
|
|
This pattern is being slowly removed from wlroots.
|
|
|
|
We need to provide an unclipped dst_box.
Fixes: https://github.com/swaywm/sway/issues/7573
Regressed by: https://github.com/swaywm/sway/pull/7552
|
|
Clear was done using sway_output's logical dimensions, instead of the
wlr_output physical dimensions. This meant that when output scaling was
applied, only a part of the screen would be cleared.
Use the wlr_output dimensions instead.
Regressed by: https://github.com/swaywm/sway/pull/7552
|
|
The new wlr_render_pass API provides src_box, dst_box and clip
parameters for texture rendition. Rather than clipping the dst_box,
which control the projection matrix and leads to compression, intersect
the damage and clip box and pass these as a clip parameter.
Fixes: https://github.com/swaywm/sway/issues/7579
Regressed by: https://github.com/swaywm/sway/pull/7552
|
|
|
|
|
|
|
|
This lets us easily add rendering state that we need in the future
|
|
|
|
|
|
Reduces code duplication.
|
|
Signed-off-by: Elyes Haouas <ehaouas@noos.fr>
|
|
During direct scan-out, pass the damaged region to the wlroots
backend.
|
|
When there is no damage, no need to perform an output commit, even
when direct scan-out is used.
|
|
|
|
We already do this in handle_commit().
|
|
#7524 was a partial fix. Seems like this is still an issue when
unplugging and plugging the monitor back in.
Closes: https://github.com/swaywm/sway/issues/7528
|
|
Otherwise the initial bounds would be `INT_MAX` until `handle_mode` or `handle_commit` is called :)
|
|
|
|
Check whether output->damage_ring.current is empty before calling
wlr_output_attach_render(). Saves us from having to un-do that
via wlr_output_rollback().
|
|
That way output_render() only does what it says on the tin.
|
|
|
|
This sends fractional-scale-v1 events before the first configure
event. That way clients have all of the metadata they need to render
the first frame.
|
|
Closes: https://github.com/swaywm/sway/issues/7464
|
|
References: https://gitlab.freedesktop.org/wlroots/wlroots/-/merge_requests/3511
|
|
We can centralize all output-related surface events from there.
|
|
|
|
This way we can move focus on the same seat an activation token
originates from.
|
|
For compatibility with i3 [1].
[1]: https://i3wm.org/docs/ipc.html#_output_event
|
|
fixes #7394
Test cases:
* zwlr_output_configuration_head_v1_set_adaptive_sync 0->0, no change
* 0->1, enabled
* 1->0, disabled
* 1->1, no change
Similar tests with an incapable display resulted in `"Adaptive sync
failed, ignoring"` messages as expected.
|
|
Since [1], wlr_renderer_begin() can fail. Check its return value
and bail.
This fixes an assertion error (when begin() fails and then we try
to render something) after a GPU reset.
[1]: https://gitlab.freedesktop.org/wlroots/wlroots/-/commit/a541c9510a4cf544313bc9b0503d75820b42444e
|
|
|
|
|
|
|
|
|
|
|
|
|
|
We want to create a context before knowing the pid it will match with.
|
|
|
|
Views now maintain a reference to a launch context which, as a last
resort, is populated at map time with a context associated with its pid.
This opens the possibility of populating it before map via another
source, e.g. xdga-tokens or configuration.
|
|
Soon we will match views with more than just a pid.
|
|
This reuses wlroots token tracking for workspace matching. It doesn't
export any xdga tokens for clients yet.
|