On Mon, May 16, 2016 at 3:25 PM, Alexander Duyck <alexander.du...@gmail.com> wrote: > On Mon, May 16, 2016 at 2:33 PM, Tom Herbert <t...@herbertland.com> wrote: >> Add encap_hlen and ip_tunnel_encap structure to ip6_tnl. Add functions >> for getting encap hlen, setting up encap on a tunnel, performing >> encapsulation operation. >> >> Signed-off-by: Tom Herbert <t...@herbertland.com> >> --- >> include/net/ip6_tunnel.h | 58 ++++++++++++++++++++++++++++++++ >> net/ipv4/ip_tunnel_core.c | 5 +++ >> net/ipv6/ip6_tunnel.c | 85 >> ++++++++++++++++++++++++++++++++++++++++++----- >> 3 files changed, 139 insertions(+), 9 deletions(-) >> > > So it looks like you completely dropped the two spots that were > updating mtu and max_headroom with the t->hlen. I thought you needed > to at least have a check that used t->encap_hlen here in order to > avoid overflowing the buffer or exceeding skb_headroom, or am I > missing something? > Sorry, you're probably right. max_headroom seems to be an absolute value. mtu being calculated seems relative to what is in skbuff already.
Tom > - Alex
