> On Nov 25, 2015, at 10:52 AM, Joe Stringer <j...@ovn.org> wrote:
> 
> On 25 November 2015 at 10:31, Jarno Rajahalme <ja...@ovn.org 
> <mailto:ja...@ovn.org>> wrote:
>> 
>>> On Nov 24, 2015, at 5:02 PM, Joe Stringer <j...@ovn.org> wrote:
>>> 
>>> On 24 November 2015 at 13:41, Jarno Rajahalme <ja...@ovn.org> wrote:
>>>> Sometimes xlate_actions() fails due to too deep recursion, too many
>>>> MPLS labels, or missing recirculation context.  Make xlate_actions()
>>>> clear out the produced odp actions in these cases to make it easy for
>>>> the caller to install a drop flow (instead or installing a flow with
>>>> partially translated actions).  Also, return a specific error code, so
>>>> that the error can be properly propagated where meaningful.
>>>> 
>>>> Before this patch it was possible that the revalidation installed a
>>>> flow with a recirculation ID with an invalid recirc ID (== 0), due to
>>>> the introduction of in-place modification in commit 43b2f131a229
>>>> (ofproto: Allow in-place modifications of datapath flows).
>>>> 
>>>> Signed-off-by: Jarno Rajahalme <ja...@ovn.org>
>>> 
>>> Should this also set the error when receiving packets on a mirror port
>>> in xlate_actions()? Or when receiving tagged VLAN traffic that doesn't
>>> correspond to the port's vlan tag? Or when a group has no live bucket?
>>> Are there any other cases that should also be covered? (I just scanned
>>> across ofproto/ofproto-dpif-xlate.c looking for cases where we're
>>> already logging that we drop the packet, but maybe there's a reasoning
>>> behind not including these - if so, please enlighten me)
>> 
>> No reasoning for missing those, I just did not notice them. Thanks for 
>> pointing them out.
> 
> OK, I thought it may have been something like "expected errors" vs.
> "unexpected errors".

Looking into these I noticed this to be the case. Must discern whether to fail 
just the individual action v.s. the whole pipeline.

  Jarno

_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev

Reply via email to