Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 18 Jul 2001 18:20:00 -0700 (PDT)
From:      Scott Hazen Mueller <scott@zorch.sf-bay.org>
To:        freebsd-bugs@FreeBSD.org
Subject:   Re: bin/29071: a little hack to rwhod 
Message-ID:  <200107190120.f6J1K0D79273@freefall.freebsd.org>

next in thread | raw e-mail | index | archive | help
The following reply was made to PR bin/29071; it has been noted by GNATS.

From: Scott Hazen Mueller <scott@zorch.sf-bay.org>
To: brian@Awfulhak.org, FreeBSD-gnats-submit@FreeBSD.ORG
Cc:  
Subject: Re: bin/29071: a little hack to rwhod 
Date: Wed, 18 Jul 2001 18:14:05 -0700 (PDT)

 >> >Number:         29071
 >> >Category:       bin
 >> >Synopsis:       relay patch for rwhod
 
 >However, I'm not sure I like the idea of sending the packet back out with a
 >special X marker.  I'd prefer if the target network(s) could be specified:
 
 That part of it is inelegant, to say the least.  I looked at trying to
 identify the source address and using that for loop control...however, that's
 harder than it appears.  The send socket is (of course) bound with INADDR_ANY
 (if anything, the structure is actually bzero'ed), so while we get an address
 in the received packet, it's moderately hard(*) to correlate it back to our-
 selves.  The Right Thing, of course, is to maintain a list of our interface
 addresses, and compare all of those against the address in the received
 packet.
 
 (*) Moderately hard here means well beyond my exceedingly rusty C programming
 skills.
 
 >  rwhod -r 172.17.10.255 -r 10.0.0.255 etc
 
 Assuming connected interfaces, this probably would work.  rwhod builds a list
 of "neighbors" when it starts up, so these -r addresses could just be inserted
 into the neighbor list...  Question, here, though - syntactically, is the
 intent to relay packets from anywhere to 172.17.10.255 & 10.0.0.255, or
 packets from 172.17.10.255 & 10.0.0.255 to all interfaces?  Probably the
 former, but just checking.  My hack just does all to all...
 
 >If rwhod kept a note of the payload it sends out, and simply makes sure it
 >doesn't send the same thing out twice in a row within (say) 15 seconds, it
 >may be better ?  This could be implemented fairly easily by just maintaining
 >a list of timestamped outbound packets and expiring them when it notices
 >they're out of date....
 
 I think looking for it's own interface address(es) as the source would be the
 cleaner solution, but I'm not 100% confident it's easily implementable.  This
 would be the next best thing.
 
                                 \scott
 

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-bugs" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200107190120.f6J1K0D79273>