Hi, Thank you for fixing this!
On Thu, 2019-02-28 at 12:55 -0800, Eric Dumazet wrote: > Later (net-next) we might revert this patch by tracking all these > qlen uses and replace them by a more efficient method (not having > to access a precise qlen, but an empty/non_empty status that might > be less expensive to maintain/track). We are working on this. The final goal is using this info to enable TCQ_F_CAN_BYPASS for nolock qdisc, too, as we see regressions in uncontended scenarios with pfifo_fast compared to pre TCQ_F_NOLOCK implementation. Please let us know if the "empty/non_empty status" refactor is in your radar, so we can avoid conflicts. Thanks! Paolo
