net_sched: prio: fix a race in prio_tune()

Gerrard Tai reported a race condition in PRIO, whenever SFQ perturb timer
fires at the wrong time.

The race is as follows:

CPU 0 CPU 1
[1]: lock root
[2]: qdisc_tree_flush_backlog()
[3]: unlock root
|
| [5]: lock root
| [6]: rehash
| [7]: qdisc_tree_reduce_backlog()
|
[4]: qdisc_put()

This can be abused to underflow a parent's qlen.

Calling qdisc_purge_queue() instead of qdisc_tree_flush_backlog()
should fix the race, because all packets will be purged from the qdisc
before releasing the lock.

Fixes: 7b8e0b6e6599 ("net: sched: prio: delay destroying child qdiscs on change")
Reported-by: Gerrard Tai <[e-mail]>
Suggested-by: Gerrard Tai <[e-mail]>
Signed-off-by: Eric Dumazet <[e-mail]>
Link: https://patch.msgid.link/[e-mail]
Signed-off-by: Jakub Kicinski <[e-mail]>

net/sched/sch_prio.c

···
 	memcpy(q->prio2band, qopt->priomap, TC_PRIO_MAX+1);
 
 	for (i = q->bands; i < oldbands; i++)
-		qdisc_tree_flush_backlog(q->queues[i]);
+		qdisc_purge_queue(q->queues[i]);
 
 	for (i = oldbands; i < q->bands; i++) {
 		q->queues[i] = queues[i];