On Mon, Mar 27, 2006 at 12:01:41PM +0100, Hugo Santos wrote: > > One small issue with the patch is tunnel6_rcv's return of 0 when no > handler handles the packet which will increase IPSTATS_MIB_INDELIVERS. > I guess -1 should be returned on error. I was checking ip_input and the > expected behaviour there is different, any value != 0 will trigger a > re-submit.
Incrementing the counter makes sense because we do have a protocol handler that looked at the packet and decided that it couldn't be handled. You can compare this with the case where UDP couldn't find a socket for a packet. > Also, there's a bigger problem which is the ICMP error generation. I > acknowledge you mentioned that this is work for another patch but i > would like to comment it now. Right now you kept the icmpv6_send in > ip6_tunnel's discard path, which doesn't work that well if you have > another tunnel handler following like xfrm6_tunnel. I would suggest > having the icmpv6_send in tunnel6_rcv's error path, so if no handler > was able to process the packet, the source receives the DEST_UNREACH > error. I'm not sure if this is desirable in IPSec but i'm sure you'll > be able to comment on that. Yes it will be moved out in a subsequent patch. Cheers, -- Visit Openswan at http://www.openswan.org/ Email: Herbert Xu ~{PmV>HI~} <[EMAIL PROTECTED]> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt - To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
