On Wednesday 16 January 2008 11:06:21 Herbert Xu wrote: > Rusty Russell <rusty@xxxxxxxxxxxxxxx> wrote: > > It's far easier to deal with GSO if we don't have to parse the packet > > to figure out the header length. Add the field to the virtio_net_hdr > > struct (and fix the spaces that somehow crept in there). > > > > Signed-off-by: Rusty Russell <rusty@xxxxxxxxxxxxxxx> > > --- > > drivers/net/virtio_net.c | 4 +++- > > include/linux/virtio_net.h | 11 ++++++----- > > 2 files changed, 9 insertions(+), 6 deletions(-) > > > > diff -r 24ef33a4ab14 drivers/net/virtio_net.c > > --- a/drivers/net/virtio_net.c Tue Jan 15 16:59:58 2008 +1100 > > +++ b/drivers/net/virtio_net.c Tue Jan 15 21:21:40 2008 +1100 > > @@ -126,6 +126,7 @@ static void receive_skb(struct net_devic > > /* Header must be checked, and gso_segs computed. */ > > skb_shinfo(skb)->gso_type |= SKB_GSO_DODGY; > > skb_shinfo(skb)->gso_segs = 0; > > + skb_set_transport_header(skb, hdr->gso_hdr_len); > > Why do we need this? When receiving GSO packets from an untrusted > source the network stack will fill in the transport header offset > after verifying that the headers are sane. Thanks for clarifying; it simplifies things. I'll re-test and resend. Rusty. _______________________________________________ Virtualization mailing list Virtualization@xxxxxxxxxxxxxxxxxxxxxxxxxx https://lists.linux-foundation.org/mailman/listinfo/virtualization
