[ovs-dev] [PATCH 1/2] netdev-vport: Always implement get_ifindex for netdev-vport

[Simon Horman]

On Tue, Aug 08, 2017 at 01:35:41PM +0300, Roi Dayan wrote:
> 
> 
> On 08/08/2017 11:04, Simon Horman wrote:
> >On Tue, Aug 08, 2017 at 07:36:25AM +0300, Roi Dayan wrote:
> >>
> >>
> >>On 07/08/2017 20:05, Ben Pfaff wrote:
> >>>On Mon, Aug 07, 2017 at 07:00:31AM -0700, Ben Pfaff wrote:
> >>>>On Mon, Aug 07, 2017 at 07:32:02AM +0300, Roi Dayan wrote:
> >>>>>From: Paul Blakey <paulb at mellanox.com>
> >>>>>
> >>>>>Always implement get_ifindex without checking if offload is
> >>>>>enabled or not as this should not be related. From ovs-dpctl
> >>>>>we cannot tell if offload is enabled or not as other_config is
> >>>>>not being read.
> >>>>>
> >>>>>Signed-off-by: Paul Blakey <paulb at mellanox.com>
> >>>>>Reviewed-by: Roi Dayan <roid at mellanox.com>
> >>>>
> >>>>Applied to master and branch-2.8, thanks!
> >>>
> >>>Sorry, I had to revert this because it caused several unit test
> >>>failures: 770 781 783 787 788 791 2189 2378.
> >>>
> >>
> >>This is because of the warnings from get_ifindex which resolved in
> >>the second patch but was missing the ratelimiting you mentioned.
> >>I submitted V2 of it to add back the ratelimiting
> >>"netdev-linux: Reduce log level for ENODEV errors getting ifindex"
> >
> >In that case shouldn't the patch order be reversed to avoid
> >the (temporary) regression Ben pointed out?
> >
> 
> right.
> should i post V3 for changing the order or is it something that
> can be done when merged?

I guess its up to Ben, but surely posting v3 would not hurt.