Re: [BUG] general protection fault in qfq_qlen_notify

[Jamal Hadi Salim]

Hi,

On Sun, Feb 22, 2026 at 8:27 PM GangMin Kim <km.kim1503@xxxxxxxxx> wrote:
>
> Dear Linux kernel developers and maintainers,
> Using a modified version of syzkaller, I identified a new bug and
> refined the PoC, and the bug-related information is attached
> below.Please let me know if you need any further information.
>
> Summary
>
> A NULL Pointer Dereference occurs because there is no proper check for
> whether cl is NULL when a class is found via cops->find() and then
> used in cops->qlen_notify().
> This description provides an explanation and KASAN report for
> qfq_qlen_notify, but the vulnerability can occur in various other
> places such as drr_qlen_notify, hfsc_qlen_notify, htb_qlen_notify,
> etc.
>

This is a duplicate of the issue reported by Manas Ghandat
<ghandatmanas@xxxxxxxxx> on the 17th.
We are still looking at best way to resolve this issue.

cheers,
jamal

> Keywords
> - net/sched
>
> Kernel Info
> Version: (Output of /proc/version)
> - Linux version 7.0.0-rc1
> Commit: (Git hash if applicable)
> - 6de23f81a5e08be8fbf5e8d7e9febc72a5b5f27f
>
> Description(Root Cause)
>
> ```c
> void qdisc_tree_reduce_backlog(struct Qdisc *sch, int n, int len)
> {
>     const struct Qdisc_class_ops *cops;
>     unsigned long cl;
>     u32 parentid;
>     bool notify;
>     int drops;
>
>     drops = max_t(int, n, 0);
>     rcu_read_lock();
>     while ((parentid = sch->parent)) {
>         if (parentid == TC_H_ROOT)
>             break;
>
>         if (sch->flags & TCQ_F_NOPARENT)
>             break;
>         /* Notify parent qdisc only if child qdisc becomes empty. */
>         notify = !sch->q.qlen;
>         /* TODO: perform the search on a per txq basis */
>         sch = qdisc_lookup_rcu(qdisc_dev(sch), TC_H_MAJ(parentid));
>         if (sch == NULL) {
>             WARN_ON_ONCE(parentid != TC_H_ROOT);
>             break;
>         }
>         cops = sch->ops->cl_ops;
>         if (notify && cops->qlen_notify) { // [1]
>             /* Note that qlen_notify must be idempotent as it may get called
>              * multiple times.
>              */
>             cl = cops->find(sch, parentid); // [2]
>             cops->qlen_notify(sch, cl); // [3]
>         }
>         sch->q.qlen -= n;
>         sch->qstats.backlog -= len;
>         __qdisc_qstats_drop(sch, drops);
>     }
>     rcu_read_unlock();
> }
> ```
>
> When condition [1] is satisfied, cl is retrieved at [2] and then
> passed to [3]. Here, [2] can return NULL, and using this value at [3]
> can cause a NULL Pointer Dereference.
>
> ```c
> static unsigned long qfq_search_class(struct Qdisc *sch, u32 classid)
> {
>     return (unsigned long)qfq_find_class(sch, classid);
> }
>
> static struct qfq_class *qfq_find_class(struct Qdisc *sch, u32 classid)
> {
>     struct qfq_sched *q = qdisc_priv(sch);
>     struct Qdisc_class_common *clc;
>
>     clc = qdisc_class_find(&q->clhash, classid);
>     if (clc == NULL)
>         return NULL; // [4]
>     return container_of(clc, struct qfq_class, common);
> }
> ```
>
> When QFQ is configured, the above function is executed at [2], and
> NULL can be returned at [4].
>
> ```c
> static void qfq_qlen_notify(struct Qdisc *sch, unsigned long arg)
> {
>     struct qfq_sched *q = qdisc_priv(sch);
>     struct qfq_class *cl = (struct qfq_class *)arg; // [5]
>
>     if (list_empty(&cl->alist)) // [6]
>         return;
>     qfq_deactivate_class(q, cl);
> }
> ```
>
> If NULL is returned at [2], that value is passed directly to [3],
> which becomes cl through [5]. Subsequently, a NULL Pointer Dereference
> occurs at [6].
>
> Kasan Report
> Oops: general protection fault, probably for non-canonical address
> 0xdffffc000000000b: 0000 [#1] SMP KASAN NOPTI
> KASAN: null-ptr-deref in range [0x0000000000000058-0x000000000000005f]
> CPU: 0 UID: 0 PID: 377 Comm: test Not tainted 7.0.0-rc1 #54 PREEMPT(full)
> Hardware name: QEMU Ubuntu 24.04 PC v2 (i440FX + PIIX, arch_caps fix,
> 1996), BIOS 1.16.3-debian-1.16.3-2 04/01/2014
> RIP: 0010:qfq_qlen_notify+0x2e/0x180 net/sched/sch_qfq.c:1434
> Code: 41 57 41 56 41 55 41 54 55 48 89 fd 53 48 89 f3 4c 8d 7b 58 e8
> e3 b3 8a fd 4c 89 fa 48 b8 00 00 00 00 00 fc ff df 48 c1 ea 03 <80> 3c
> 02 00 0f 85 f8 00 00 00 48 8b 43 58 49 39 c7 0f 84 b9 00 00
> RSP: 0018:ffff88811a447258 EFLAGS: 00010212
> RAX: dffffc0000000000 RBX: 0000000000000000 RCX: ffffffff83e2af6d
> RDX: 000000000000000b RSI: 0000000000000000 RDI: ffff88811b524000
> RBP: ffff88811b524000 R08: 0000000000000000 R09: fffffbfff0d3595c
> R10: ffffffff869acae7 R11: 00000000ac3123a9 R12: ffffffff859b0b60
> R13: ffffffff859b0b40 R14: 0000000000000000 R15: 0000000000000058
> FS:  00007ff24e5da700(0000) GS:ffff88834ad0d000(0000) knlGS:0000000000000000
> CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 00007ff24e6e95f0 CR3: 000000011a5ff000 CR4: 0000000000750ef0
> PKRU: 55555554
> Call Trace:
>  <TASK>
>  __qdisc_qstats_drop include/net/sch_generic.h:958 [inline]
>  qdisc_tree_reduce_backlog+0x128/0x400 net/sched/sch_api.c:812
>  tbf_offload_graft net/sched/sch_tbf.c:191 [inline]
>  tbf_graft+0x1cc/0x590 net/sched/sch_tbf.c:573
>  qdisc_refcount_dec_if_one include/net/sch_generic.h:152 [inline]
>  qdisc_graft+0x2ef/0x1460 net/sched/sch_api.c:1117
>  __tc_modify_qdisc net/sched/sch_api.c:1631 [inline]
>  tc_modify_qdisc+0xf4f/0x1e40 net/sched/sch_api.c:1819
>  rcu_read_unlock include/linux/rcupdate.h:883 [inline]
>  rtnetlink_rcv_msg+0x3b9/0xa90 net/core/rtnetlink.c:6913
>  netlink_rcv_skb+0x12e/0x390 net/netlink/af_netlink.c:2539
>  kfree_skb_reason include/linux/skbuff.h:1322 [inline]
>  kfree_skb include/linux/skbuff.h:1331 [inline]
>  netlink_unicast_kernel net/netlink/af_netlink.c:1321 [inline]
>  netlink_unicast+0x6c1/0x970 net/netlink/af_netlink.c:1344
>  netlink_sendmsg+0x79c/0xc50 net/netlink/af_netlink.c:2465
>  __sock_release net/socket.c:673 [inline]
>  ____sys_sendmsg+0x8b2/0xa50 net/socket.c:690
>  sendmsg_copy_msghdr net/socket.c:2621 [inline]
>  ___sys_sendmsg+0x120/0x1c0 net/socket.c:2642
>  __sys_sendmsg+0x147/0x1f0 net/socket.c:2681
>  arch_atomic64_read arch/x86/include/asm/atomic64_64.h:15 [inline]
>  raw_atomic64_read include/linux/atomic/atomic-arch-fallback.h:2583 [inline]
>  raw_atomic_long_read include/linux/atomic/atomic-long.h:38 [inline]
>  atomic_long_read include/linux/atomic/atomic-instrumented.h:3189 [inline]
>  unwind_reset_info include/linux/unwind_deferred.h:37 [inline]
>  exit_to_user_mode include/linux/irq-entry-common.h:296 [inline]
>  syscall_exit_to_user_mode include/linux/entry-common.h:327 [inline]
>  do_syscall_64+0xf1/0x530 arch/x86/entry/syscall_64.c:100
>  entry_SYSCALL_64_after_hwframe+0x77/0x7f
> RIP: 0033:0x7ff24e7c4e4d
> Code: 28 89 54 24 1c 48 89 74 24 10 89 7c 24 08 e8 ca ee ff ff 8b 54
> 24 1c 48 8b 74 24 10 41 89 c0 8b 7c 24 08 b8 2e 00 00 00 0f 05 <48> 3d
> 00 f0 ff ff 77 33 44 89 c7 48 89 44 24 08 e8 fe ee ff ff 48
> RSP: 002b:00007ff24e5d9d90 EFLAGS: 00000293 ORIG_RAX: 000000000000002e
> RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007ff24e7c4e4d
> RDX: 0000000000000000 RSI: 00007ff24e5d9dd0 RDI: 0000000000000003
> RBP: 00007ff24e5d9e30 R08: 0000000000000000 R09: 0000000000050000
> R10: 0000000000000000 R11: 0000000000000293 R12: 00007ffcdc50ddde
> R13: 00007ffcdc50dddf R14: 00007ff24e5d9fc0 R15: 0000000000022000
>  </TASK>
> Modules linked in:
> ---[ end trace 0000000000000000 ]---
> RIP: 0010:qfq_qlen_notify+0x2e/0x180 net/sched/sch_qfq.c:1434
> Code: 41 57 41 56 41 55 41 54 55 48 89 fd 53 48 89 f3 4c 8d 7b 58 e8
> e3 b3 8a fd 4c 89 fa 48 b8 00 00 00 00 00 fc ff df 48 c1 ea 03 <80> 3c
> 02 00 0f 85 f8 00 00 00 48 8b 43 58 49 39 c7 0f 84 b9 00 00
> RSP: 0018:ffff88811a447258 EFLAGS: 00010212
> RAX: dffffc0000000000 RBX: 0000000000000000 RCX: ffffffff83e2af6d
> RDX: 000000000000000b RSI: 0000000000000000 RDI: ffff88811b524000
> RBP: ffff88811b524000 R08: 0000000000000000 R09: fffffbfff0d3595c
> R10: ffffffff869acae7 R11: 00000000ac3123a9 R12: ffffffff859b0b60
> R13: ffffffff859b0b40 R14: 0000000000000000 R15: 0000000000000058
> FS:  00007ff24e5da700(0000) GS:ffff88834ad0d000(0000) knlGS:0000000000000000
> CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 00007ff24e6e95f0 CR3: 000000011a5ff000 CR4: 0000000000750ef0
> PKRU: 55555554
> ----------------
> Code disassembly (best guess):
>    0:   41 57                   push   %r15
>    2:   41 56                   push   %r14
>    4:   41 55                   push   %r13
>    6:   41 54                   push   %r12
>    8:   55                      push   %rbp
>    9:   48 89 fd                mov    %rdi,%rbp
>    c:   53                      push   %rbx
>    d:   48 89 f3                mov    %rsi,%rbx
>   10:   4c 8d 7b 58             lea    0x58(%rbx),%r15
>   14:   e8 e3 b3 8a fd          call   0xfd8ab3fc
>   19:   4c 89 fa                mov    %r15,%rdx
>   1c:   48 b8 00 00 00 00 00    movabs $0xdffffc0000000000,%rax
>   23:   fc ff df
>   26:   48 c1 ea 03             shr    $0x3,%rdx
> * 2a:   80 3c 02 00             cmpb   $0x0,(%rdx,%rax,1) <-- trapping
> instruction
>   2e:   0f 85 f8 00 00 00       jne    0x12c
>   34:   48 8b 43 58             mov    0x58(%rbx),%rax
>   38:   49 39 c7                cmp    %rax,%r15
>   3b:   0f                      .byte 0xf
>   3c:   84                      .byte 0x84
>   3d:   b9                      .byte 0xb9
> R13: ffffffff859b0b40 R14: 0000000000000000 R15: 0000000000000058
> FS:  00007ff24e5da700(0000) GS:ffff88834ad0d000(0000) knlGS:0000000000000000
> CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 00007ff24e6e95f0 CR3: 000000011a5ff000 CR4: 0000000000750ef0
> PKRU: 55555554
> Kernel panic - not syncing: Fatal exception in interrupt
> Kernel Offset: disabled
> ---[ end Kernel panic - not syncing: Fatal exception in interrupt ]---