svn commit: r186955 - in head/sys: conf netinet

[Robert Watson]

On Sat, 10 Jan 2009, Julian Elischer wrote:

>>> I'm happy to (eventually) also implement the BSDI API once I actually 
>>> spend time looking at what the difference in behaviours are. If we're 
>>> lucky, the only difference is where the socket option hooks in and the 
>>> actual network behaviour is the same.
>>> 
>>> (Meanwhile, I think I have to go off and implement this particular 
>>> behaviour in Squid, and see if the OpenBSD support indeed does function as 
>>> advertised.)
>> 
>> If the API turns out to be effectly semantically the same, or better, then 
>> I think the suggestion is to entirely replace, rather than supplement, the 
>> socket option you just added with it.  There's no point in having 
>> pointlessly divergent APIs where it can be avoided.
>
> I think just making the name the same should be enough..

Well, I think that depends.  If it's a SOL_SOCKET-layer option, we still need 
some way for the protocol layer to either accept or veto setting the option, 
depending on whether it supports it.  For example, I think SPX sockets should 
reject the option being set if they don't support it, so we'll need to figure 
out something there to either pass down the SOL_SOCKET option explicitly, or 
check with the protocol somehow as to whether or not to accept it.

Robert N M Watson
Computer Laboratory
University of Cambridge