Re: [RFC net-next 2/2] net: sched: fix tp destroy race conditions in flower

2019-10-03 Thread John Hurley
On Thu, Oct 3, 2019 at 5:18 PM Vlad Buslov wrote: > > > On Thu 03 Oct 2019 at 02:14, John Hurley wrote: > > Flower has rule HW offload functions available that drivers can choose to > > register for. For the deletion case, these are triggered after filters > > have been removed from lookup tables

Re: [RFC net-next 2/2] net: sched: fix tp destroy race conditions in flower

2019-10-03 Thread Vlad Buslov
On Thu 03 Oct 2019 at 02:14, John Hurley wrote: > Flower has rule HW offload functions available that drivers can choose to > register for. For the deletion case, these are triggered after filters > have been removed from lookup tables both at the flower level, and the > higher cls_api level. Wi

[RFC net-next 2/2] net: sched: fix tp destroy race conditions in flower

2019-10-02 Thread John Hurley
Flower has rule HW offload functions available that drivers can choose to register for. For the deletion case, these are triggered after filters have been removed from lookup tables both at the flower level, and the higher cls_api level. With flower running without RTNL locking, this can lead to ra