RFR: 8179559 Solaris MulticastSocket issues

[Chris Hegarty]

> On 5 May 2017, at 18:36, Michael McMahon <michael.x.mcmahon at oracle.com> wrote:
> 
> 
> 
> On 05/05/2017, 18:30, Chris Hegarty wrote:
>>> On 5 May 2017, at 18:29, Michael McMahon<michael.x.mcmahon at oracle.com>  wrote:
>>> ...
>>> 
>>> Do you mean if the Inet6Address passed in has zero for the scope_id that the
>>> search should ignore it, and do what happens now, ie. return the first address found?
>> Yes, that is what I wanted to say. Does it make sense?
>> 
>> -Chris.
>> 
> Yes, I think so. It's not unreasonable to have an address that might be unique
> on the local system and in some circumstances might not need a scope id.
> So, a scope id of zero would never match then.
> 
> I won't regenerate the webrev but just change the condition to:
> 
>            if (scopeid != 0 && scopeid != ((struct sockaddr_in6*)addrP->addr)->sin6_scope_id)
>                        break;

Looks good.

-Chris.