linux/net/sched
Lion Ackermann 103406b38c net/sched: Always pass notifications when child class becomes empty
Certain classful qdiscs may invoke their classes' dequeue handler on an
enqueue operation. This may unexpectedly empty the child qdisc and thus
make an in-flight class passive via qlen_notify(). Most qdiscs do not
expect such behaviour at this point in time and may re-activate the
class eventually anyways which will lead to a use-after-free.

The referenced fix commit attempted to fix this behavior for the HFSC
case by moving the backlog accounting around, though this turned out to
be incomplete since the parent's parent may run into the issue too.
The following reproducer demonstrates this use-after-free:

    tc qdisc add dev lo root handle 1: drr
    tc filter add dev lo parent 1: basic classid 1:1
    tc class add dev lo parent 1: classid 1:1 drr
    tc qdisc add dev lo parent 1:1 handle 2: hfsc def 1
    tc class add dev lo parent 2: classid 2:1 hfsc rt m1 8 d 1 m2 0
    tc qdisc add dev lo parent 2:1 handle 3: netem
    tc qdisc add dev lo parent 3:1 handle 4: blackhole

    echo 1 | socat -u STDIN UDP4-DATAGRAM:127.0.0.1:8888
    tc class delete dev lo classid 1:1
    echo 1 | socat -u STDIN UDP4-DATAGRAM:127.0.0.1:8888

Since backlog accounting issues leading to a use-after-frees on stale
class pointers is a recurring pattern at this point, this patch takes
a different approach. Instead of trying to fix the accounting, the patch
ensures that qdisc_tree_reduce_backlog always calls qlen_notify when
the child qdisc is empty. This solves the problem because deletion of
qdiscs always involves a call to qdisc_reset() and / or
qdisc_purge_queue() which ultimately resets its qlen to 0 thus causing
the following qdisc_tree_reduce_backlog() to report to the parent. Note
that this may call qlen_notify on passive classes multiple times. This
is not a problem after the recent patch series that made all the
classful qdiscs qlen_notify() handlers idempotent.

Fixes: 3f98113810 ("sch_hfsc: Fix qlen accounting bug when using peek in hfsc_enqueue()")
Signed-off-by: Lion Ackermann <nnamrec@gmail.com>
Reviewed-by: Jamal Hadi Salim <jhs@mojatatu.com>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Link: https://patch.msgid.link/d912cbd7-193b-4269-9857-525bee8bbb6a@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-07-02 14:42:17 -07:00
..
Kconfig sctp: use skb_crc32c() instead of __skb_checksum() 2025-05-21 15:40:16 -07:00
Makefile bpf: net_sched: Support implementation of Qdisc_ops in bpf 2025-04-17 10:54:33 -07:00
act_api.c tc: Return an error if filters try to attach too many actions 2025-04-15 17:13:11 -07:00
act_bpf.c
act_connmark.c
act_csum.c
act_ct.c net: convert to nla_get_*_default() 2024-11-11 10:32:06 -08:00
act_ctinfo.c net: convert to nla_get_*_default() 2024-11-11 10:32:06 -08:00
act_gact.c
act_gate.c net/sched: Switch to use hrtimer_setup() 2025-02-18 10:35:44 +01:00
act_ife.c
act_meta_mark.c
act_meta_skbprio.c
act_meta_skbtcindex.c
act_mirred.c net/sched: act_mirred: Move the recursion counter struct netdev_xmit 2025-05-15 15:23:31 +02:00
act_mpls.c net: convert to nla_get_*_default() 2024-11-11 10:32:06 -08:00
act_nat.c
act_pedit.c
act_police.c net: convert to nla_get_*_default() 2024-11-11 10:32:06 -08:00
act_sample.c
act_simple.c
act_skbedit.c
act_skbmod.c
act_tunnel_key.c net: fix geneve_opt length integer overflow 2025-04-03 15:47:35 -07:00
act_vlan.c tc: adjust network header after 2nd vlan push 2024-08-27 11:37:42 +02:00
bpf_qdisc.c bpf: net_sched: Make some Qdisc_ops ops mandatory 2025-05-02 15:35:37 -07:00
cls_api.c tc: Ensure we have enough buffer space when sending filter netlink notifications 2025-04-08 13:57:49 +02:00
cls_basic.c
cls_bpf.c net: sched: refine software bypass handling in tc_run 2025-01-20 09:21:27 +00:00
cls_cgroup.c
cls_flow.c treewide, timers: Rename from_timer() to timer_container_of() 2025-06-08 09:07:37 +02:00
cls_flower.c net: fix geneve_opt length integer overflow 2025-04-03 15:47:35 -07:00
cls_fw.c
cls_matchall.c net: sched: refine software bypass handling in tc_run 2025-01-20 09:21:27 +00:00
cls_route.c
cls_u32.c net: sched: refine software bypass handling in tc_run 2025-01-20 09:21:27 +00:00
em_canid.c
em_cmp.c move asm/unaligned.h to linux/unaligned.h 2024-10-02 17:23:23 -04:00
em_ipset.c
em_ipt.c
em_meta.c net: dismiss sk_forward_alloc_get() 2025-02-19 19:05:28 -08:00
em_nbyte.c
em_text.c
em_u32.c
ematch.c
sch_api.c net/sched: Always pass notifications when child class becomes empty 2025-07-02 14:42:17 -07:00
sch_blackhole.c
sch_cake.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net 2025-01-09 16:11:47 -08:00
sch_cbs.c net/sched: cbs: Fix integer overflow in cbs_set_port_rate() 2024-10-15 18:25:47 -07:00
sch_choke.c net: sched: fix ordering of qlen adjustment 2024-12-04 12:54:22 +00:00
sch_codel.c net_sched: Flush gso_skb list too during ->change() 2025-05-09 12:34:38 +01:00
sch_drr.c net_sched: drr: Fix double list add in class with netem as child qdisc 2025-04-28 15:55:06 -07:00
sch_etf.c
sch_ets.c net_sched: ets: fix a race in ets_qdisc_change() 2025-06-12 08:05:50 -07:00
sch_fifo.c pfifo_tail_enqueue: Drop new packet when sch->limit == 0 2025-02-05 18:13:58 -08:00
sch_fq.c net_sched: Flush gso_skb list too during ->change() 2025-05-09 12:34:38 +01:00
sch_fq_codel.c net_sched: Flush gso_skb list too during ->change() 2025-05-09 12:34:38 +01:00
sch_fq_pie.c treewide, timers: Rename from_timer() to timer_container_of() 2025-06-08 09:07:37 +02:00
sch_frag.c net/sched: Use nested-BH locking for sch_frag_data_storage 2025-05-15 15:23:31 +02:00
sch_generic.c treewide, timers: Rename from_timer() to timer_container_of() 2025-06-08 09:07:37 +02:00
sch_gred.c sched: address a potential NULL pointer dereference in the GRED scheduler. 2025-03-06 16:35:14 -08:00
sch_hfsc.c net_sched: hfsc: Address reentrant enqueue adding class to eltree twice 2025-05-28 08:55:35 +02:00
sch_hhf.c net_sched: Flush gso_skb list too during ->change() 2025-05-09 12:34:38 +01:00
sch_htb.c sch_htb: make htb_deactivate() idempotent 2025-05-05 13:51:32 -07:00
sch_ingress.c
sch_mq.c
sch_mqprio.c
sch_mqprio_lib.c
sch_mqprio_lib.h
sch_multiq.c
sch_netem.c netem: Update sch->q.qlen before qdisc_tree_reduce_backlog() 2025-02-05 18:14:46 -08:00
sch_pie.c treewide, timers: Rename from_timer() to timer_container_of() 2025-06-08 09:07:37 +02:00
sch_plug.c
sch_prio.c net_sched: prio: fix a race in prio_tune() 2025-06-12 08:05:49 -07:00
sch_qfq.c net_sched: qfq: Fix double list add in class with netem as child qdisc 2025-04-28 15:55:07 -07:00
sch_red.c Including fixes from bluetooth and wireless. 2025-06-12 09:50:36 -07:00
sch_sfb.c net/sched: Add drop reasons for AQM-based qdiscs 2024-12-17 13:27:29 +01:00
sch_sfq.c Including fixes from bluetooth and wireless. 2025-06-12 09:50:36 -07:00
sch_skbprio.c net_sched: skbprio: Remove overly strict queue assertions 2025-04-02 16:03:32 -07:00
sch_taprio.c net/sched: fix use-after-free in taprio_dev_notifier 2025-06-17 16:14:04 -07:00
sch_tbf.c net_sched: tbf: fix a race in tbf_change() 2025-06-12 08:05:50 -07:00
sch_teql.c