On Sun, Dec 21, 2014 at 6:17 PM, Jesse Gross <je...@nicira.com> wrote: > On Fri, Dec 19, 2014 at 7:25 PM, Pravin B Shelar <pshe...@nicira.com> wrote: >> Today vport-send has complex error handling because it involves >> freeing skb and updating stats depending on return value from >> vport send implementation. >> This can be simplified by delegating responsibility of freeing >> skb to the vport implementation for all cases. So that >> vport-send needs just update stats. >> >> Signed-off-by: Pravin B Shelar <pshe...@nicira.com> > > It's somewhat non-obvious to me that handle_offloads should free the > skb in the event of an error. It seems like this introduces more > complexity to this patch as well. > > As an example of the problem, I think that the previous patch > introduces a double free in the non-compat case. This is because it > updates the skb freeing for the compat code in handle_offloads() but > the equivalent update to the non-compat code isn't made until this > patch.
I now see that the upstream iptunnel_handle_offloads() also frees the skb on error but it seems like this has introduced some errors as well. It looks like several tunnel protocols continue using the skb after calling handle_offloads without any kind of check. _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
