Vladimir Oltean
2021-Jul-20 14:12 UTC
[Bridge] [PATCH v5 net-next 00/10] Let switchdev drivers offload and unoffload bridge ports at their own convenience
On Tue, Jul 20, 2021 at 05:01:48PM +0300, Ido Schimmel wrote:> > The patches were split from a larger series for easier review: > > This is not what I meant. I specifically suggested to get the TX > forwarding offload first and then extending the API with an argument to > opt-in for the replay / cleanup:Yeah, ok, I did not get that and I had already reposted by the time you clarified, sorry. Anyway, is it so bad that we cannot look at the patches in the order that they are in right now (even if this means that maybe a few more days would pass)? To me it makes a bit more sense anyway to first consolidate the code that is already in the tree right now, before adding new logic. And I don't really want to rebase the patches again to change the ordering and risk a build breakage without a good reason.
Ido Schimmel
2021-Jul-20 14:25 UTC
[Bridge] [PATCH v5 net-next 00/10] Let switchdev drivers offload and unoffload bridge ports at their own convenience
On Tue, Jul 20, 2021 at 02:12:01PM +0000, Vladimir Oltean wrote:> On Tue, Jul 20, 2021 at 05:01:48PM +0300, Ido Schimmel wrote: > > > The patches were split from a larger series for easier review: > > > > This is not what I meant. I specifically suggested to get the TX > > forwarding offload first and then extending the API with an argument to > > opt-in for the replay / cleanup: > > Yeah, ok, I did not get that and I had already reposted by the time you > clarified, sorry. > > Anyway, is it so bad that we cannot look at the patches in the order > that they are in right now (even if this means that maybe a few more > days would pass)? To me it makes a bit more sense anyway to first > consolidate the code that is already in the tree right now, before > adding new logic. And I don't really want to rebase the patches again to > change the ordering and risk a build breakage without a good reason.If you don't want to change the order, then at least make the replay/cleanup optional and set it to 'false' for mlxsw. This should mean that the only change in mlxsw should be adding calls to switchdev_bridge_port_offload() / switchdev_bridge_port_unoffload() in mlxsw_sp_bridge_port_create() / mlxsw_sp_bridge_port_destroy(), respectively.