Re: [REGRESSION] 6.14.3 panic - kernel NULL pointer dereference in htb_dequeue

[Cong Wang <xiyou.wangcong@gmail.com>]

On Sun, Apr 27, 2025 at 09:35:48PM +0100, Alan J. Wylie wrote:
> On Sun, 27 Apr 2025 20:42:54 +0100
> "Alan J. Wylie" <alan@wylie.me.uk> wrote:
> 
> > That would be https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net.git/ ?
> > 
> > I've just cloned that. I'll do a build and a test.
> 
> $ uname -r
> 6.15.0-rc3-00109-gf73f05c6f711
> 
> It's crashed. Same place as usual. I tried again, same thing.
> 
>  htb_dequeue+0x42e/0x610 [sch_htb]
> 
> Rather than a ping flood, I was running a Speedtest. Both times it
> crashed during the upload test, not the download.
> 
> https://www.speedtest.net/
> 
> Could running an iptables firewall perhaps have anything to do with it?

I doubt it is related to iptables. I will try some TCP traffic on my
side later, but I suspect this is related to the type of packets.

Meanwhile, since I still can't reproduce it here, do you mind applying
both of my patches on top of -net and test again?

For your convenience, below is the combined patch of the previous two
patches, which can be applied on -net.

Thanks!

----->

diff --git a/net/sched/sch_htb.c b/net/sched/sch_htb.c
index 4b9a639b642e..9d88fff120bc 100644
--- a/net/sched/sch_htb.c
+++ b/net/sched/sch_htb.c
@@ -348,7 +348,8 @@ static void htb_add_to_wait_tree(struct htb_sched *q,
  */
 static inline void htb_next_rb_node(struct rb_node **n)
 {
-	*n = rb_next(*n);
+	if (*n)
+		*n = rb_next(*n);
 }
 
 /**
@@ -1487,7 +1488,8 @@ static void htb_qlen_notify(struct Qdisc *sch, unsigned long arg)
 
 	if (!cl->prio_activity)
 		return;
-	htb_deactivate(qdisc_priv(sch), cl);
+	if (!cl->leaf.q->q.qlen)
+		htb_deactivate(qdisc_priv(sch), cl);
 }
 
 static inline int htb_parent_last_child(struct htb_class *cl)