[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Xen-devel] Re: [1/4] [NET] back: Fix maximum fragment check

On Sat, Jul 01, 2006 at 01:17:31PM +0100, Keir Fraser wrote:
> Ah, okay, can that not be handled properly by some TSO-supporting NICs 
> then? You need to do something smarter than simply duplicate the ECN 
> bits in every segment header? What would happen if netfront sent 
> through a packet with ECE/CWR and didn't set the GSO_ECN flag -- is it 
> safe but stupid, will it break something, should dodgy packets have the 
> GSO_ECN flag recomputed?

Getting the ECN bit wrong is OK.  The worst result is that the packets
generated will have incorrect CWR markings which will only penalise the
connection belonging to the guest.

If the guest wanted to achieve that, it can always do the same thing by
segmenting the packets itself anyway.

Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@xxxxxxxxxxxxxxxxxxx>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.