[Spice-devel] [RFC PATCH spice-server v2 17/19] WIP cursor-channel: Allow to not handle a QXL device

[Jonathon Jongsma]

On Wed, 2017-06-14 at 16:40 +0100, Frediano Ziglio wrote:
> TODO: not a constant ID.
> Christophe F was preparing a patch to get rid of the
> qxl parameter to the CursorChannel.

It's fine that this part is split out to a separate commit, but it
really needs additional explanation about why we're doing it: in what
(future) scenario where qxl will be NULL, etc?


> 
> Signed-off-by: Frediano Ziglio <fziglio at redhat.com>
> ---
>  server/cursor-channel.c | 6 ++++--
>  1 file changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/server/cursor-channel.c b/server/cursor-channel.c
> index a876113..4abcd53 100644
> --- a/server/cursor-channel.c
> +++ b/server/cursor-channel.c
> @@ -94,7 +94,9 @@ static void cursor_item_unref(CursorItem *item)
>          return;
>  
>      cursor_cmd = item->red_cursor;
> -    red_qxl_release_resource(item->qxl, cursor_cmd-
> >release_info_ext);
> +    if (item->qxl) {
> +        red_qxl_release_resource(item->qxl, cursor_cmd-
> >release_info_ext);
> +    }
>      red_put_cursor_cmd(cursor_cmd);
>      free(cursor_cmd);
>  
> @@ -297,7 +299,7 @@ CursorChannel* cursor_channel_new(RedsState
> *server, QXLInstance *qxl,
>                          "spice-server", server,
>                          "core-interface", core,
>                          "channel-type", SPICE_CHANNEL_CURSOR,
> -                        "id", qxl->id,
> +                        "id", qxl ? qxl->id : 1,
>                          "migration-flags", 0,
>                          "qxl", qxl,
>                          "handle-acks", TRUE,