[PATCH] libmbim-glib, proxy: Filter indications based on MBIM_CID_BASIC_CONNECT_DEVICE_SERVICE_SUBSCRIBE_LIST
Greg Suarez
gpsuarez2512 at gmail.com
Mon Jun 23 11:52:29 PDT 2014
On Mon, Jun 23, 2014 at 11:31 AM, Aleksander Morgado <
aleksander at aleksander.es> wrote:
> On Mon, Jun 23, 2014 at 8:25 PM, Greg Suarez <gpsuarez2512 at gmail.com>
> wrote:
> >> There's one key thing missing still in the behavior we're supposed to
> >> see w.r.t. indications. In the current code with your latest patches,
> >> there is no difference between standard and non-standard services. If
> >> I'm not mistaken, by default all indications of "standard" services
> >> should be treated as enabled; while all indications of non-standard
> >> services shouldn't.
> >
> >
> > Just so I understand what you're saying...
> > So Client-A enables indications for non-standard services and then
> Client-B
> > connects to the proxy and expects the non-standard indications disabled.
> > So the proxy shouldn't forward the non-standard indications to Client-B
> > until
> > Client-B enables the indications, correct?
>
> Yes.
>
> Another flow:
> * Client A starts, and by default expects standard indications.
> * Client B starts, and by default expects standard indications.
> * At this point, all standard indications should go to both clients A and
> B.
> * Client B unregisters itself from all standard indications, and
> registers some non-standard ones.
> * At this point, all standard indications should go to client A, and
> client B should only get those non-standard indications it was
> registered for.
>
The current patch already handles this flow. Since the client provides
only the
services and cids that it expects to receive in the
MBIM_CID_DEVICE_SERVICE_SUBSCRIBE_LIST,
the patch saves the MBIM_DEVICES_SERVICE_SUBSCRIBE_LIST and uses it to
check if an
indication should be passed on to the client. So if Client-B unregisters
from a standard service it will
not receive those indications. And since the proxy will never actually
unregister from standard services
with the actual device, Client A will continue to receive indications from
standard services.
> Could you re-check the standard to confirm this logic, and if so,
> provide in the commit message the section in which this is explained?
>
> Sure
Greg
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/libmbim-devel/attachments/20140623/c2b4fbb8/attachment.html>
More information about the libmbim-devel
mailing list