Good catch, Erich.  Only discovery packets are looped back, and we should know 
our own device ID.

John

-----Original Message-----
From: iotivity-dev-bounces at lists.iotivity.org 
[mailto:[email protected]] On Behalf Of Keane, Erich
Sent: Wednesday, January 27, 2016 8:57 AM
To: iotivity-dev at lists.iotivity.org; Macieira, Thiago
Subject: Re: [dev] OCDevAddr - Determine if sender isSelf in case of 
CLIENT_SERVER mode.

isSelf is easy enough on some of the messages, since DeviceId (server
id?) is included in discovery packets. ?That filtering would be possible with a 
simple function.

At one point there was discussion about getting a stable per-machine ID, but 
I'm not sure if that was ever implemented. ?That could be used to filter as 
well without having to bother with IP addresses.

On Tue, 2016-01-26 at 20:41 -0800, Thiago Macieira wrote:
> On Wednesday 27 January 2016 00:33:34 Morrow, Joseph L wrote:
> > Hi All,
> > 
> > I am wondering if anyone has already written any code to determine 
> > if the OCDevAddr of the sender (in the EH) is the local machine or 
> > not.
> > Also,
> > would it be difficult to ask that an "isSelf" flag be added to the 
> > OCDevAddr struct? I think checking for "isSelf" would be much easier 
> > at the transport layer than in the application layer.
> > 
> > I am looking to ignore any requests from my own CLIENT_SERVER 
> > application.
> 
> Hi Joey
> 
> Local machine does not imply self. It's possible that there are other 
> stacks running on the same machine and you should handle them as if 
> they came from someone else.
> 
> To do that, we need to enable IPV6_MULTICAST_LOOP.
> 
> However, you're right: it would be highly useful to know if a looped 
> back packet is one of ours. In fact, we should probably discard them 
> and not pass them on to the upper layers and the application. The 
> problem is that this is not an easy task in user space: each machine 
> has probably quite a few IP addresses.
> 
> Interesting challenge.
_______________________________________________
iotivity-dev mailing list
iotivity-dev at lists.iotivity.org
https://lists.iotivity.org/mailman/listinfo/iotivity-dev

Reply via email to