linux-stable.git/net/sched, branch linux-6.14.y

net_sched: hfsc: Address reentrant enqueue adding class to eltree twice

2025-06-04T12:45:08+00:00

commit ac9fe7dd8e730a103ae4481147395cc73492d786 upstream.

Savino says:
    "We are writing to report that this recent patch
    (141d34391abbb315d68556b7c67ad97885407547) [1]
    can be bypassed, and a UAF can still occur when HFSC is utilized with
    NETEM.

    The patch only checks the cl->cl_nactive field to determine whether
    it is the first insertion or not [2], but this field is only
    incremented by init_vf [3].

    By using HFSC_RSC (which uses init_ed) [4], it is possible to bypass the
    check and insert the class twice in the eltree.
    Under normal conditions, this would lead to an infinite loop in
    hfsc_dequeue for the reasons we already explained in this report [5].

    However, if TBF is added as root qdisc and it is configured with a
    very low rate,
    it can be utilized to prevent packets from being dequeued.
    This behavior can be exploited to perform subsequent insertions in the
    HFSC eltree and cause a UAF."

To fix both the UAF and the infinite loop, with netem as an hfsc child,
check explicitly in hfsc_enqueue whether the class is already in the eltree
whenever the HFSC_RSC flag is set.

[1] https://web.git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=141d34391abbb315d68556b7c67ad97885407547
[2] https://elixir.bootlin.com/linux/v6.15-rc5/source/net/sched/sch_hfsc.c#L1572
[3] https://elixir.bootlin.com/linux/v6.15-rc5/source/net/sched/sch_hfsc.c#L677
[4] https://elixir.bootlin.com/linux/v6.15-rc5/source/net/sched/sch_hfsc.c#L1574
[5] https://lore.kernel.org/netdev/8DuRWwfqjoRDLDmBMlIfbrsZg9Gx50DHJc1ilxsEBNe2D6NMoigR_eIRIG0LOjMc3r10nUUZtArXx4oZBIdUfZQrwjcQhdinnMis_0G7VEk=@willsroot.io/T/#u

Fixes: 37d9cf1a3ce3 ("sched: Fix detection of empty queues in child qdiscs")
Reported-by: Savino Dicanosa 
Reported-by: William Liu 
Acked-by: Jamal Hadi Salim 
Tested-by: Victor Nogueira 
Signed-off-by: Pedro Tammela 
Link: https://patch.msgid.link/20250522181448.1439717-2-pctammela@mojatatu.com
Signed-off-by: Paolo Abeni 
Signed-off-by: Greg Kroah-Hartman

sch_hfsc: Fix qlen accounting bug when using peek in hfsc_enqueue()

2025-05-29T09:13:59+00:00

[ Upstream commit 3f981138109f63232a5fb7165938d4c945cc1b9d ]

When enqueuing the first packet to an HFSC class, hfsc_enqueue() calls the
child qdisc's peek() operation before incrementing sch->q.qlen and
sch->qstats.backlog. If the child qdisc uses qdisc_peek_dequeued(), this may
trigger an immediate dequeue and potential packet drop. In such cases,
qdisc_tree_reduce_backlog() is called, but the HFSC qdisc's qlen and backlog
have not yet been updated, leading to inconsistent queue accounting. This
can leave an empty HFSC class in the active list, causing further
consequences like use-after-free.

This patch fixes the bug by moving the increment of sch->q.qlen and
sch->qstats.backlog before the call to the child qdisc's peek() operation.
This ensures that queue length and backlog are always accurate when packet
drops or dequeues are triggered during the peek.

Fixes: 12d0ad3be9c3 ("net/sched/sch_hfsc.c: handle corner cases where head may change invalidating calculated deadline")
Reported-by: Mingi Cho 
Signed-off-by: Cong Wang 
Reviewed-by: Simon Horman 
Link: https://patch.msgid.link/20250518222038.58538-2-xiyou.wangcong@gmail.com
Reviewed-by: Jamal Hadi Salim 
Signed-off-by: Paolo Abeni 
Signed-off-by: Sasha Levin

net_sched: Flush gso_skb list too during ->change()

2025-05-22T12:31:42+00:00

[ Upstream commit 2d3cbfd6d54a2c39ce3244f33f85c595844bd7b8 ]

Previously, when reducing a qdisc's limit via the ->change() operation, only
the main skb queue was trimmed, potentially leaving packets in the gso_skb
list. This could result in NULL pointer dereference when we only check
sch->limit against sch->q.qlen.

This patch introduces a new helper, qdisc_dequeue_internal(), which ensures
both the gso_skb list and the main queue are properly flushed when trimming
excess packets. All relevant qdiscs (codel, fq, fq_codel, fq_pie, hhf, pie)
are updated to use this helper in their ->change() routines.

Fixes: 76e3cc126bb2 ("codel: Controlled Delay AQM")
Fixes: 4b549a2ef4be ("fq_codel: Fair Queue Codel AQM")
Fixes: afe4fd062416 ("pkt_sched: fq: Fair Queue packet scheduler")
Fixes: ec97ecf1ebe4 ("net: sched: add Flow Queue PIE packet scheduler")
Fixes: 10239edf86f1 ("net-qdisc-hhf: Heavy-Hitter Filter (HHF) qdisc")
Fixes: d4b36210c2e6 ("net: pkt_sched: PIE AQM scheme")
Reported-by: Will 
Reported-by: Savy 
Signed-off-by: Cong Wang 
Signed-off-by: David S. Miller 
Signed-off-by: Sasha Levin

sch_htb: make htb_deactivate() idempotent

2025-05-18T06:25:57+00:00

[ Upstream commit 3769478610135e82b262640252d90f6efb05be71 ]

Alan reported a NULL pointer dereference in htb_next_rb_node()
after we made htb_qlen_notify() idempotent.

It turns out in the following case it introduced some regression:

htb_dequeue_tree():
  |-> fq_codel_dequeue()
    |-> qdisc_tree_reduce_backlog()
      |-> htb_qlen_notify()
        |-> htb_deactivate()
  |-> htb_next_rb_node()
  |-> htb_deactivate()

For htb_next_rb_node(), after calling the 1st htb_deactivate(), the
clprio[prio]->ptr could be already set to  NULL, which means
htb_next_rb_node() is vulnerable here.

For htb_deactivate(), although we checked qlen before calling it, in
case of qlen==0 after qdisc_tree_reduce_backlog(), we may call it again
which triggers the warning inside.

To fix the issues here, we need to:

1) Make htb_deactivate() idempotent, that is, simply return if we
   already call it before.
2) Make htb_next_rb_node() safe against ptr==NULL.

Many thanks to Alan for testing and for the reproducer.

Fixes: 5ba8b837b522 ("sch_htb: make htb_qlen_notify() idempotent")
Reported-by: Alan J. Wylie 
Signed-off-by: Cong Wang 
Link: https://patch.msgid.link/20250428232955.1740419-2-xiyou.wangcong@gmail.com
Signed-off-by: Jakub Kicinski 
Signed-off-by: Sasha Levin

sch_ets: make est_qlen_notify() idempotent

2025-05-09T07:56:07+00:00

commit a7a15f39c682ac4268624da2abdb9114bdde96d5 upstream.

est_qlen_notify() deletes its class from its active list with
list_del() when qlen is 0, therefore, it is not idempotent and
not friendly to its callers, like fq_codel_dequeue().

Let's make it idempotent to ease qdisc_tree_reduce_backlog() callers'
life. Also change other list_del()'s to list_del_init() just to be
extra safe.

Reported-by: Gerrard Tai 
Signed-off-by: Cong Wang 
Link: https://patch.msgid.link/20250403211033.166059-6-xiyou.wangcong@gmail.com
Acked-by: Jamal Hadi Salim 
Signed-off-by: Paolo Abeni 
Signed-off-by: Greg Kroah-Hartman

sch_qfq: make qfq_qlen_notify() idempotent

2025-05-09T07:56:07+00:00

commit 55f9eca4bfe30a15d8656f915922e8c98b7f0728 upstream.

qfq_qlen_notify() always deletes its class from its active list
with list_del_init() _and_ calls qfq_deactivate_agg() when the whole list
becomes empty.

To make it idempotent, just skip everything when it is not in the active
list.

Also change other list_del()'s to list_del_init() just to be extra safe.

Reported-by: Gerrard Tai 
Signed-off-by: Cong Wang 
Reviewed-by: Simon Horman 
Link: https://patch.msgid.link/20250403211033.166059-5-xiyou.wangcong@gmail.com
Acked-by: Jamal Hadi Salim 
Signed-off-by: Paolo Abeni 
Signed-off-by: Greg Kroah-Hartman

sch_hfsc: make hfsc_qlen_notify() idempotent

2025-05-09T07:56:07+00:00

commit 51eb3b65544c9efd6a1026889ee5fb5aa62da3bb upstream.

hfsc_qlen_notify() is not idempotent either and not friendly
to its callers, like fq_codel_dequeue(). Let's make it idempotent
to ease qdisc_tree_reduce_backlog() callers' life:

1. update_vf() decreases cl->cl_nactive, so we can check whether it is
non-zero before calling it.

2. eltree_remove() always removes RB node cl->el_node, but we can use
   RB_EMPTY_NODE() + RB_CLEAR_NODE() to make it safe.

Reported-by: Gerrard Tai 
Signed-off-by: Cong Wang 
Reviewed-by: Simon Horman 
Link: https://patch.msgid.link/20250403211033.166059-4-xiyou.wangcong@gmail.com
Acked-by: Jamal Hadi Salim 
Signed-off-by: Paolo Abeni 
Signed-off-by: Greg Kroah-Hartman

sch_drr: make drr_qlen_notify() idempotent

2025-05-09T07:56:07+00:00

commit df008598b3a00be02a8051fde89ca0fbc416bd55 upstream.

drr_qlen_notify() always deletes the DRR class from its active list
with list_del(), therefore, it is not idempotent and not friendly
to its callers, like fq_codel_dequeue().

Let's make it idempotent to ease qdisc_tree_reduce_backlog() callers'
life. Also change other list_del()'s to list_del_init() just to be
extra safe.

Reported-by: Gerrard Tai 
Signed-off-by: Cong Wang 
Reviewed-by: Simon Horman 
Link: https://patch.msgid.link/20250403211033.166059-3-xiyou.wangcong@gmail.com
Acked-by: Jamal Hadi Salim 
Signed-off-by: Paolo Abeni 
Signed-off-by: Greg Kroah-Hartman

sch_htb: make htb_qlen_notify() idempotent

2025-05-09T07:56:07+00:00

commit 5ba8b837b522d7051ef81bacf3d95383ff8edce5 upstream.

htb_qlen_notify() always deactivates the HTB class and in fact could
trigger a warning if it is already deactivated. Therefore, it is not
idempotent and not friendly to its callers, like fq_codel_dequeue().

Let's make it idempotent to ease qdisc_tree_reduce_backlog() callers'
life.

Reported-by: Gerrard Tai 
Signed-off-by: Cong Wang 
Reviewed-by: Simon Horman 
Link: https://patch.msgid.link/20250403211033.166059-2-xiyou.wangcong@gmail.com
Acked-by: Jamal Hadi Salim 
Signed-off-by: Paolo Abeni 
Signed-off-by: Greg Kroah-Hartman

net_sched: qfq: Fix double list add in class with netem as child qdisc

2025-05-09T07:55:57+00:00

[ Upstream commit f139f37dcdf34b67f5bf92bc8e0f7f6b3ac63aa4 ]

As described in Gerrard's report [1], there are use cases where a netem
child qdisc will make the parent qdisc's enqueue callback reentrant.
In the case of qfq, there won't be a UAF, but the code will add the same
classifier to the list twice, which will cause memory corruption.

This patch checks whether the class was already added to the agg->active
list (cl_is_active) before doing the addition to cater for the reentrant
case.

[1] https://lore.kernel.org/netdev/CAHcdcOm+03OD2j6R0=YHKqmy=VgJ8xEOKuP6c7mSgnp-TEJJbw@mail.gmail.com/

Fixes: 37d9cf1a3ce3 ("sched: Fix detection of empty queues in child qdiscs")
Acked-by: Jamal Hadi Salim 
Signed-off-by: Victor Nogueira 
Link: https://patch.msgid.link/20250425220710.3964791-5-victor@mojatatu.com
Signed-off-by: Jakub Kicinski 
Signed-off-by: Sasha Levin