On Sat, Aug 16, 2003 at 10:39:56PM -0700, David S. Miller wrote: > > Wait a second, I think this fix is wrong. BTW this bug was definitely there before I moved the tunnel processing code since my move only affected IPCOMP tunnels :) > First of all, you cannot mangle the packet contents. > Other taps might be looking at the original copy, > ie. we might have a clone. But this is xfrm_rcv_encap where we're doing all sorts of nasty things to the packet including decrypting/decompressing it. > Second of all, you can achieve the affect by just > nudging up skb->mac.raw to the correct spot. Hmm skb->mac.raw is at the correct spot already. The problem is that there is a heap of crap between the ll header and the real IP header so the packet socket's rcv function breaks down. > Have a look at what the ipip.c driver does. I'm afraid I can't see the correct solution here. Could you enlighten me please? Cheers, -- Debian GNU/Linux 3.0 is out! ( http://www.debian.org/ ) Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt - : send the line "unsubscribe linux-net" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
