On 21/11/2018 10:22, Simon Horman wrote:
>
> On Tue, 20 Nov 2018 at 14:09, Roi Dayan <r...@mellanox.com
> <mailto:r...@mellanox.com>> wrote:
>
>
>
> On 13/09/2018 17:58, Simon Horman wrote:
> > Thanks for the clarification below, I have applied this series to
> master.
> >
>
>
> Hi Simon,
>
> I worked with branch-2.10 today and noticed the following patch is there:
> d63ca5329ff9 dpctl: Properly reflect a rule's offloaded to HW state
>
> Since it's there type=offloaded not showing tc rules that failed to
> offload to hw
> and we should probably have the 2 patches in this series merged as well:
> a692410af0f7 dpctl: Expand the flow dump type filter
> 0d6b401cf60d dpif-netdev: Initialize dpif_flow attrs
>
> With them ovs will support type=tc for rules in tc but not offloaded to
> hw.
>
> Is it ok to merge them into branch-2.10 ?
> I tried it locally and didn't hit conflicts with cherry-pick.
>
>
> Hi Roi,
>
> I'm a little reluctant to add new features to branch-2.10.
> Perhaps it was an error to apply d63ca5329ff9 ("dpctl: Properly reflect a
> rule's offloaded to HW state")
> there and it should be reverted.
Hi Simon,
Sorry I didn't reply to you sooner. I took my time to think about revert or not
option.
At first I forgot that the commit also adds a "dp" flag when dumping
with -m to indicate if flow is offloaded or not and if its in tc or not.
So a user can still distinguish between rules in tc and offloaded or just in tc.
So I think its ok not to revert.
Thanks,
Roi
_______________________________________________
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev