Moving a window is a compositor action and happens immediately. Waiting
here is pointless. Make sure instead that the action happens immediately
by asserting the position.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3842>
wait and wait_configure after a single resize is useless and a race.
Resize is a client side action which doesn't result in a configure and a
wait doesn't sync for a resize as well.
Sometimes the resize is paired with another action, such as maximize,
fullscreen or show. In those cases a configure will be generated and a
previous resize is accounted for.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3842>
Commit e775052429 changed the code such that resetting the actor is done
when a surface role is assigned. The dnd surface assigned vfunc doesn't
chain up which means the code to reset the actor is never hit and the
dnd surface never shows up.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3540
Fixes: e775052429 ("wayland/actor-surface: Reset the actor on role-assignment")
In these checks, it's an important detail to preserve subpixel information
in order to correctly determine whether the coordinates are inside or
outside a view.
Otherwise, small enough motion towards the left/top might get rounded
to 0, be seen as "inside the view", and the pointer coordinates be allowed
to escape the viewport constraints.
This was figured out by Pascal Nowack before me, with a difference of
minutes. Credit where credit is due.
Closes: https://gitlab.gnome.org/GNOME/mutter/-/issues/3530
Fixes: 6c972546f1 ("mtk: Add Rectangle.contains_point")
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3820>
In meta_seat_impl_remove_virtual_input_device(), the 'device'
variable is first removed from MetaSeatImpl, then a "device
removed" event is generated with it.
The problem here is that, if this is the last reference of
'device', the removal from MetaSeatImpl will destroy it. Then
the freed variable will be used to create the "device removed"
event, which is a use-after-free situation.
Fix that by owning an extra ref to 'device' as long as the
function is executing. Do this by declaring a g_autoptr
variable with the extra ref. This g_autoptr variable is cleaned
up by the end of the function, which achieves the desired effect.
Spotted by Coverity.
CID: #1594046
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3835>
Right now the unmapped signal doesn't always fire which means we didn't
see a surface that's being unmapped in these code paths before. In
particular the resource, window and role can be gone. Handle those
cases.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3783>
This allows us one less level of indentation of all the tests. It is not
entirely true, the X11 backend test case script can run without it, but
it isn't valuable enough to run without the native backend being enabled
to complicate building.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3812>
Instead of having get_seat_id() handle most cases, and then special
casing another case outside of it, let it handle them all, making all
users just able to call get_seat_id().
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3812>
Except the X11 backend test, it still needs an X server. Eventually we
can replace it with the equivalent that uses Xwayland, but that needs a
"scaled down" mutter that runs as the host compositor for Xwayland, that
doesn't expose anything on the session bus.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3812>
It's constructed as a native backend headless backend, but uses a custom
monitor manager (based on MetaMonitorManagerNative) that creates a fake
monitor. Rendering is unconditionally done with the surfaceless
renderer.
The test devices used now use virtual devices, meaning some changes to
the tests to e.g. not set names, and not dealing with input devices
directly.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3812>
This is where the viewports are updated, and so lets move connecting the
signal together with it. It also helps a future change to the test
backend where it creates a custom monitor manager.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3812>
Test touch devices are autocleaned up, but if a test case explicitly
needs it to be removed at a certain point, make sure to also clear the
pointer so that it doesn't get removed twice.
Right now it's harmless, but in a later commit it'll expected to only
remove a device once.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3812>
`update_secondary_gpu_state_post_swap_buffers` decides what our front
buffer object will be. There is only one answer. So return it as the
function result instead of making the caller figure it out.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3830>
It's always equal to `onscreen_native->next_frame` and we can't eliminate
that copy so easily. Removing the parameter removes all ambiguity about
where the next frame will come from.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3829>
Inside the "if (clutter_actor_has_accessible (actor))" condition,
the 'atk_child' variable is set and a signal is emitted on it.
There is a classic ref/unref dance around the signal to guarantee
that 'atk_child' won't be destroyed.
However, this ref/unref dance doesn't work, because the unref is
done *before* the 'atk_child' variable is used again. So if this
was the last reference to it, it would have been destroyed in the
unref call, then used for another signal emission a few lines down.
That's a use-after-free.
Fix that by declaring the 'atk_child' variable with g_autoptr. This
delays the unref until the very end of the function, and is NULL safe.
Also add a sneaky assertion, just for extra safety.
Part-of: <https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3828>
