Just did a cherry-pick.
There was a trivial conflict which was due to an extra init line added in sch_drr.c, traced back to 67c9e6270f30.
sch_ets.c doesn't exist so didn't (rather can't) put that commit.
I don't have the vulnerability reproducer so I cannot test, but it looks okay to me as the original logic applied cleanly.
Thanks, Siddh
Cong Wang (4): sch_drr: make drr_qlen_notify() idempotent sch_hfsc: make hfsc_qlen_notify() idempotent sch_qfq: make qfq_qlen_notify() idempotent codel: remove sch->q.qlen check before qdisc_tree_reduce_backlog()
net/sched/sch_codel.c | 5 +---- net/sched/sch_drr.c | 7 ++++--- net/sched/sch_fq_codel.c | 6 ++---- net/sched/sch_hfsc.c | 8 ++++++-- net/sched/sch_qfq.c | 7 +++++-- 5 files changed, 18 insertions(+), 15 deletions(-)