[Spice-devel] [PATCH] display: factor out draw_until
Frediano Ziglio
fziglio at redhat.com
Wed Nov 25 05:50:32 PST 2015
>
> From: Marc-André Lureau <marcandre.lureau at gmail.com>
>
> Acked-by: Fabiano FidĂȘncio <fidencio at redhat.com>
> ---
> server/display-channel.c | 62
> +++++++++++++++++++++++-------------------------
> 1 file changed, 30 insertions(+), 32 deletions(-)
>
> diff --git a/server/display-channel.c b/server/display-channel.c
> index d54f725..1837c9b 100644
> --- a/server/display-channel.c
> +++ b/server/display-channel.c
> @@ -1282,6 +1282,24 @@ static void surface_update_dest(RedSurface *surface,
> const SpiceRect *area)
> canvas->ops->read_bits(canvas, dest, -stride, area);
> }
>
> +static void draw_until(DisplayChannel *display, RedSurface *surface,
> Drawable *last)
> +{
> + RingItem *ring_item;
> + Container *container;
> + Drawable *now;
> +
> + do {
> + ring_item = ring_get_tail(&surface->current_list);
> + now = SPICE_CONTAINEROF(ring_item, Drawable, surface_list_link);
> + now->refs++;
> + container = now->tree_item.base.container;
> + current_remove_drawable(display, now);
> + container_cleanup(container);
> + drawable_draw(display, now);
> + display_channel_drawable_unref(display, now);
> + } while (now != last);
> +}
> +
> /*
> * Renders drawables for updating the requested area, but only drawables
> that are older
> * than 'last' (exclusive).
> @@ -1345,23 +1363,17 @@ void display_channel_draw_till(DisplayChannel
> *display, const SpiceRect *area, i
> return;
> }
>
> - do {
> - Container *container;
> + /* drawable_draw may call display_channel_draw for the surfaces
> + * 'now' depends on. Notice, that it is valid to call
> + * display_channel_draw in this case and not
> + * display_channel_draw_till: It is impossible that there was
> + * newer item then 'last' in one of the surfaces that
> + * display_channel_draw is called for, Otherwise, 'now' would have
> + * already been rendered. See the call for
> + * red_handle_depends_on_target_surface in red_process_draw
> + */
I don't understand now the sense of this comment here.
Should not be inside the new function?
It refer to drawable_draw and now but in this function there are neither
of them.
> + draw_until(display, surface, surface_last);
>
> - ring_item = ring_get_tail(&surface->current_list);
> - now = SPICE_CONTAINEROF(ring_item, Drawable, surface_list_link);
> - now->refs++;
> - container = now->tree_item.base.container;
> - current_remove_drawable(display, now);
> - container_cleanup(container);
> - /* drawable_draw may call display_channel_draw for the surfaces
> 'now' depends on. Notice,
> - that it is valid to call display_channel_draw in this case and
> not display_channel_draw_till:
> - It is impossible that there was newer item then 'last' in one of
> the surfaces
> - that display_channel_draw is called for, Otherwise, 'now' would
> have already been rendered.
> - See the call for red_handle_depends_on_target_surface in
> red_process_draw */
> - drawable_draw(display, now);
> - display_channel_drawable_unref(display, now);
> - } while (now != surface_last);
> surface_update_dest(surface, area);
> }
>
> @@ -1398,22 +1410,8 @@ void display_channel_draw(DisplayChannel *display,
> const SpiceRect *area, int su
> }
> region_destroy(&rgn);
>
> - if (!last) {
> - surface_update_dest(surface, area);
> - return;
> - }
> -
> - do {
> - Container *container;
> + if (last)
> + draw_until(display, surface, last);
>
> - ring_item = ring_get_tail(&surface->current_list);
> - now = SPICE_CONTAINEROF(ring_item, Drawable, surface_list_link);
> - now->refs++;
> - container = now->tree_item.base.container;
> - current_remove_drawable(display, now);
> - container_cleanup(container);
> - drawable_draw(display, now);
> - display_channel_drawable_unref(display, now);
> - } while (now != last);
> surface_update_dest(surface, area);
> }
Frediano
More information about the Spice-devel
mailing list