Hi Philipp,
Thanks for the review!
On 21 July 2017 at 13:30, Philipp Zabel <p.za...@pengutronix.de> wrote:
> On Tue, 2017-07-18 at 14:14 +0100, Daniel Stone wrote:
>> @@ -3804,7 +3840,8 @@ drm_output_enable(struct weston_output *base)
>> else
>> b->cursors_are_broken = 1;
>>
>> - weston_compositor_stack_plane(b->compositor, &output->scanout_plane,
>> + weston_compositor_stack_plane(b->compositor,
>> + &output->scanout_plane->base,
>> &b->compositor->primary_plane);
>
> If the scanout plane is still stacked during output enable ...
>
>> weston_log("Output %s, (connector %d, crtc %d)\n",
>> @@ -3831,20 +3868,11 @@ drm_output_deinit(struct weston_output *base)
>> [...]
>>
>> - weston_plane_release(&output->scanout_plane);
>> -
>
> ... but not released anymore during deinit, won't then unplugging and
> plugging in a monitor cause the second drm_output_enable to try to
> insert the scanout plane into ec->plane_list a second time?
Ugh, you're right; this is quite hairy.
> Also, will all view->plane references to the scanout plane in the
> compositor->view_list still be cleared with this removed?
I would imagine disabling the output would cause a repaint, which
would in turn cause the buffer to get moved out of the scanout plane.
That being said, I think a better plan would be to move the plane
assignment (finding output->{scanout,cursor}_plane) into output
creation, retain the stacking in output_enable, and then releasing the
plane in output_disable.
Cheers,
Daniel
_______________________________________________
wayland-devel mailing list
wayland-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/wayland-devel
