public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* [syzbot] [net?] unregister_netdevice: waiting for DEV to become free (8)
@ 2023-06-10  1:34 syzbot
       [not found] ` <20230621070710.380373-1-astrajoan@yahoo.com>
                   ` (3 more replies)
  0 siblings, 4 replies; 13+ messages in thread
From: syzbot @ 2023-06-10  1:34 UTC (permalink / raw)
  To: arnd, bridge, davem, edumazet, kuba, linux-kernel, netdev,
	nikolay, pabeni, roopa, syzkaller-bugs

Hello,

syzbot found the following issue on:

HEAD commit:    67faabbde36b selftests/bpf: Add missing prototypes for sev..
git tree:       bpf-next
console+strace: https://syzkaller.appspot.com/x/log.txt?x=1381363b280000
kernel config:  https://syzkaller.appspot.com/x/.config?x=5335204dcdecfda
dashboard link: https://syzkaller.appspot.com/bug?extid=881d65229ca4f9ae8c84
compiler:       gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2
syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=132faf93280000
C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=10532add280000

Downloadable assets:
disk image: https://storage.googleapis.com/syzbot-assets/751a0490d875/disk-67faabbd.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/2c5106cd9f1f/vmlinux-67faabbd.xz
kernel image: https://storage.googleapis.com/syzbot-assets/62c1154294e4/bzImage-67faabbd.xz

The issue was bisected to:

commit ad2f99aedf8fa77f3ae647153284fa63c43d3055
Author: Arnd Bergmann <arnd@arndb.de>
Date:   Tue Jul 27 13:45:16 2021 +0000

    net: bridge: move bridge ioctls out of .ndo_do_ioctl

bisection log:  https://syzkaller.appspot.com/x/bisect.txt?x=146de6f1280000
final oops:     https://syzkaller.appspot.com/x/report.txt?x=166de6f1280000
console output: https://syzkaller.appspot.com/x/log.txt?x=126de6f1280000

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+881d65229ca4f9ae8c84@syzkaller.appspotmail.com
Fixes: ad2f99aedf8f ("net: bridge: move bridge ioctls out of .ndo_do_ioctl")

unregister_netdevice: waiting for bridge0 to become free. Usage count = 2
leaked reference.
 __netdev_tracker_alloc include/linux/netdevice.h:4070 [inline]
 netdev_hold include/linux/netdevice.h:4099 [inline]
 dev_ifsioc+0xbc0/0xeb0 net/core/dev_ioctl.c:408
 dev_ioctl+0x250/0x1090 net/core/dev_ioctl.c:605
 sock_do_ioctl+0x15a/0x230 net/socket.c:1215
 sock_ioctl+0x1f8/0x680 net/socket.c:1318
 vfs_ioctl fs/ioctl.c:51 [inline]
 __do_sys_ioctl fs/ioctl.c:870 [inline]
 __se_sys_ioctl fs/ioctl.c:856 [inline]
 __x64_sys_ioctl+0x197/0x210 fs/ioctl.c:856
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x39/0xb0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x63/0xcd


---
This report is generated by a bot. It may contain errors.
See https://goo.gl/tpsmEJ for more information about syzbot.
syzbot engineers can be reached at syzkaller@googlegroups.com.

syzbot will keep track of this issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.
For information about bisection process see: https://goo.gl/tpsmEJ#bisection

If the bug is already fixed, let syzbot know by replying with:
#syz fix: exact-commit-title

If you want syzbot to run the reproducer, reply with:
#syz test: git://repo/address.git branch-or-commit-hash
If you attach or paste a git patch, syzbot will apply it before testing.

If you want to change bug's subsystems, reply with:
#syz set subsystems: new-subsystem
(See the list of subsystem names on the web dashboard)

If the bug is a duplicate of another bug, reply with:
#syz dup: exact-subject-of-another-report

If you want to undo deduplication, reply with:
#syz undup

^ permalink raw reply	[flat|nested] 13+ messages in thread
* [PATCH] can: j1939: implement NETDEV_UNREGISTER notification handler
@ 2025-08-25 11:01 Tetsuo Handa
  2025-08-25 13:35 ` [syzbot] [net?] unregister_netdevice: waiting for DEV to become free (8) syzbot
  0 siblings, 1 reply; 13+ messages in thread
From: Tetsuo Handa @ 2025-08-25 11:01 UTC (permalink / raw)
  To: syzbot+881d65229ca4f9ae8c84, LKML

syzbot is reporting

  unregister_netdevice: waiting for vcan0 to become free. Usage count = 2

problem, for j1939 protocol did not have NETDEV_UNREGISTER notification
handler for undoing changes made by j1939_sk_bind().

Commit 25fe97cb7620 ("can: j1939: move j1939_priv_put() into sk_destruct
callback") expects that a call to j1939_priv_put() can be unconditionally
delayed until j1939_sk_sock_destruct() is called. But we need to call
j1939_priv_put() against an extra ref held by j1939_sk_bind() call
(as a part of undoing changes made by j1939_sk_bind()) as soon as
NETDEV_UNREGISTER notification fires (i.e. before j1939_sk_sock_destruct()
is called via j1939_sk_release()). Otherwise, the extra ref on "struct
j1939_priv" held by j1939_sk_bind() call prevents "struct net_device" from
dropping the usage count to 1; making it impossible for
unregister_netdevice() to continue.

Reported-by: syzbot <syzbot+881d65229ca4f9ae8c84@syzkaller.appspotmail.com>
Closes: https://syzkaller.appspot.com/bug?extid=881d65229ca4f9ae8c84
Fixes: 9d71dd0c7009 ("can: add support of SAE J1939 protocol")
Fixes: 25fe97cb7620 ("can: j1939: move j1939_priv_put() into sk_destruct callback")
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
---
#syz test

 net/can/j1939/j1939-priv.h |  1 +
 net/can/j1939/main.c       |  3 +++
 net/can/j1939/socket.c     | 49 ++++++++++++++++++++++++++++++++++++++
 3 files changed, 53 insertions(+)

diff --git a/net/can/j1939/j1939-priv.h b/net/can/j1939/j1939-priv.h
index 31a93cae5111..81f58924b4ac 100644
--- a/net/can/j1939/j1939-priv.h
+++ b/net/can/j1939/j1939-priv.h
@@ -212,6 +212,7 @@ void j1939_priv_get(struct j1939_priv *priv);
 
 /* notify/alert all j1939 sockets bound to ifindex */
 void j1939_sk_netdev_event_netdown(struct j1939_priv *priv);
+void j1939_sk_netdev_event_unregister(struct j1939_priv *priv);
 int j1939_cancel_active_session(struct j1939_priv *priv, struct sock *sk);
 void j1939_tp_init(struct j1939_priv *priv);
 
diff --git a/net/can/j1939/main.c b/net/can/j1939/main.c
index 7e8a20f2fc42..3706a872ecaf 100644
--- a/net/can/j1939/main.c
+++ b/net/can/j1939/main.c
@@ -377,6 +377,9 @@ static int j1939_netdev_notify(struct notifier_block *nb,
 		j1939_sk_netdev_event_netdown(priv);
 		j1939_ecu_unmap_all(priv);
 		break;
+	case NETDEV_UNREGISTER:
+		j1939_sk_netdev_event_unregister(priv);
+		break;
 	}
 
 	j1939_priv_put(priv);
diff --git a/net/can/j1939/socket.c b/net/can/j1939/socket.c
index 493f49bfaf5d..72c649cec9e1 100644
--- a/net/can/j1939/socket.c
+++ b/net/can/j1939/socket.c
@@ -1303,6 +1303,55 @@ void j1939_sk_netdev_event_netdown(struct j1939_priv *priv)
 	read_unlock_bh(&priv->j1939_socks_lock);
 }
 
+void j1939_sk_netdev_event_unregister(struct j1939_priv *priv)
+{
+	struct sock *sk;
+	struct j1939_sock *jsk;
+	bool wait_rcu = false;
+
+ rescan: /* The caller is holding a ref on this "priv" via j1939_priv_get_by_ndev(). */
+	read_lock_bh(&priv->j1939_socks_lock);
+	list_for_each_entry(jsk, &priv->j1939_socks, list) {
+		/* Skip if j1939_jsk_add() is not called on this socket. */
+		if (!(jsk->state & J1939_SOCK_BOUND))
+			continue;
+		sk = &jsk->sk;
+		sock_hold(sk);
+		read_unlock_bh(&priv->j1939_socks_lock);
+		/* Check if j1939_jsk_del() is not yet called on this socket after holding
+		 * socket's lock, for both j1939_sk_bind() and j1939_sk_release() call
+		 * j1939_jsk_del() with socket's lock held.
+		 */
+		lock_sock(sk);
+		if (jsk->state & J1939_SOCK_BOUND) {
+			/* Neither j1939_sk_bind() nor j1939_sk_release() called j1939_jsk_del().
+			 * Make this socket no longer bound, by pretending as if j1939_sk_bind()
+			 * dropped old references but did not get new references.
+			 */
+			j1939_jsk_del(priv, jsk);
+			j1939_local_ecu_put(priv, jsk->addr.src_name, jsk->addr.sa);
+			j1939_netdev_stop(priv);
+			/* Call j1939_priv_put() now and prevent j1939_sk_sock_destruct() from
+			 * calling the corresponding j1939_priv_put().
+			 *
+			 * j1939_sk_sock_destruct() is supposed to call j1939_priv_put() after
+			 * an RCU grace period. But since the caller is holding a ref on this
+			 * "priv", we can defer synchronize_rcu() until immediately before
+			 * the caller calls j1939_priv_put().
+			 */
+			j1939_priv_put(priv);
+			jsk->priv = NULL;
+			wait_rcu = true;
+		}
+		release_sock(sk);
+		sock_put(sk);
+		goto rescan;
+	}
+	read_unlock_bh(&priv->j1939_socks_lock);
+	if (wait_rcu)
+		synchronize_rcu();
+}
+
 static int j1939_sk_no_ioctlcmd(struct socket *sock, unsigned int cmd,
 				unsigned long arg)
 {
-- 
2.51.0


^ permalink raw reply related	[flat|nested] 13+ messages in thread
* Re: unregister_netdevice: waiting for DEV to become free (8)
@ 2025-11-19 12:20 Tetsuo Handa
  2025-11-19 13:09 ` [syzbot] [net?] " syzbot
  0 siblings, 1 reply; 13+ messages in thread
From: Tetsuo Handa @ 2025-11-19 12:20 UTC (permalink / raw)
  To: syzbot; +Cc: linux-kernel

#syz test: git://git.code.sf.net/p/tomoyo/tomoyo.git master

^ permalink raw reply	[flat|nested] 13+ messages in thread
* Re: unregister_netdevice: waiting for DEV to become free (8)
  2025-11-19 12:20 Tetsuo Handa
@ 2025-11-19 13:13 Tetsuo Handa
  2025-11-19 13:57 ` [syzbot] [net?] " syzbot
  0 siblings, 1 reply; 13+ messages in thread
From: Tetsuo Handa @ 2025-11-19 13:13 UTC (permalink / raw)
  To: syzbot; +Cc: linux-kernel

Trying once more.

#syz test: git://git.code.sf.net/p/tomoyo/tomoyo.git master


^ permalink raw reply	[flat|nested] 13+ messages in thread
* Re: unregister_netdevice: waiting for DEV to become free (8)
  2025-11-19 13:13 Tetsuo Handa
@ 2025-11-19 14:00 Tetsuo Handa
  2025-11-19 14:47 ` [syzbot] [net?] " syzbot
  0 siblings, 1 reply; 13+ messages in thread
From: Tetsuo Handa @ 2025-11-19 14:00 UTC (permalink / raw)
  To: syzbot; +Cc: linux-kernel

Too timing-dependent to trigger using a reproducer?
Or, a reproducer for an already-fixed bug is used?

#syz test
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index d1a687444b27..798d60b3e2ad 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -2084,6 +2084,8 @@ enum netdev_reg_state {
  *
  *	FIXME: cleanup struct net_device such that network protocol info
  *	moves out.
+ *
+ *	@netdev_trace_buffer_list: Linked list for debugging refcount leak.
  */
 
 struct net_device {
@@ -2238,6 +2240,9 @@ struct net_device {
 #if IS_ENABLED(CONFIG_TLS_DEVICE)
 	const struct tlsdev_ops *tlsdev_ops;
 #endif
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+	struct list_head	netdev_trace_buffer_list;
+#endif
 
 	unsigned int		operstate;
 	unsigned char		link_mode;
@@ -3166,6 +3171,7 @@ enum netdev_cmd {
 	NETDEV_OFFLOAD_XSTATS_REPORT_USED,
 	NETDEV_OFFLOAD_XSTATS_REPORT_DELTA,
 	NETDEV_XDP_FEAT_CHANGE,
+	NETDEV_DEBUG_UNREGISTER,
 };
 const char *netdev_cmd_to_name(enum netdev_cmd cmd);
 
@@ -4345,9 +4351,15 @@ static inline bool dev_nit_active(const struct net_device *dev)
 
 void dev_queue_xmit_nit(struct sk_buff *skb, struct net_device *dev);
 
+void save_netdev_trace_buffer(struct net_device *dev, int delta);
+int trim_netdev_trace(unsigned long *entries, int nr_entries);
+
 static inline void __dev_put(struct net_device *dev)
 {
 	if (dev) {
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+		save_netdev_trace_buffer(dev, -1);
+#endif
 #ifdef CONFIG_PCPU_DEV_REFCNT
 		this_cpu_dec(*dev->pcpu_refcnt);
 #else
@@ -4359,6 +4371,9 @@ static inline void __dev_put(struct net_device *dev)
 static inline void __dev_hold(struct net_device *dev)
 {
 	if (dev) {
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+		save_netdev_trace_buffer(dev, 1);
+#endif
 #ifdef CONFIG_PCPU_DEV_REFCNT
 		this_cpu_inc(*dev->pcpu_refcnt);
 #else
diff --git a/kernel/softirq.c b/kernel/softirq.c
index 77198911b8dd..5f435c1e48d8 100644
--- a/kernel/softirq.c
+++ b/kernel/softirq.c
@@ -576,6 +576,10 @@ static inline bool lockdep_softirq_start(void) { return false; }
 static inline void lockdep_softirq_end(bool in_hardirq) { }
 #endif
 
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+static noinline void handle_softirqs(bool ksirqd);
+#endif
+
 static void handle_softirqs(bool ksirqd)
 {
 	unsigned long end = jiffies + MAX_SOFTIRQ_TIME;
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 45320e27a16c..e9c654a9d0bb 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -3145,6 +3145,10 @@ static bool manage_workers(struct worker *worker)
 	return true;
 }
 
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+static noinline void process_one_work(struct worker *worker, struct work_struct *work);
+#endif
+
 /**
  * process_one_work - process single work
  * @worker: self
diff --git a/net/can/j1939/main.c b/net/can/j1939/main.c
index a93af55df5fd..66e6624abfa3 100644
--- a/net/can/j1939/main.c
+++ b/net/can/j1939/main.c
@@ -124,6 +124,16 @@ static void j1939_can_recv(struct sk_buff *iskb, void *data)
 
 static DEFINE_MUTEX(j1939_netdev_lock);
 
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+static void dump_priv_trace_buffer(const struct net_device *ndev);
+static void erase_priv_trace_buffer(struct j1939_priv *priv);
+static noinline void save_priv_trace_buffer(struct j1939_priv *priv, int delta);
+#else
+static inline void dump_priv_trace_buffer(const struct net_device *ndev) { };
+static inline void erase_priv_trace_buffer(struct j1939_priv *priv) { };
+static inline void save_priv_trace_buffer(struct j1939_priv *priv, int delta) { };
+#endif
+
 static struct j1939_priv *j1939_priv_create(struct net_device *ndev)
 {
 	struct j1939_priv *priv;
@@ -137,6 +147,7 @@ static struct j1939_priv *j1939_priv_create(struct net_device *ndev)
 	priv->ndev = ndev;
 	kref_init(&priv->kref);
 	kref_init(&priv->rx_kref);
+	save_priv_trace_buffer(priv, 1);
 	dev_hold(ndev);
 
 	netdev_dbg(priv->ndev, "%s : 0x%p\n", __func__, priv);
@@ -164,17 +175,20 @@ static void __j1939_priv_release(struct kref *kref)
 	WARN_ON_ONCE(!list_empty(&priv->j1939_socks));
 
 	dev_put(ndev);
+	erase_priv_trace_buffer(priv);
 	kfree(priv);
 }
 
 void j1939_priv_put(struct j1939_priv *priv)
 {
+	save_priv_trace_buffer(priv, -1);
 	kref_put(&priv->kref, __j1939_priv_release);
 }
 
 void j1939_priv_get(struct j1939_priv *priv)
 {
 	kref_get(&priv->kref);
+	save_priv_trace_buffer(priv, 1);
 }
 
 static int j1939_can_rx_register(struct j1939_priv *priv)
@@ -282,6 +296,7 @@ struct j1939_priv *j1939_netdev_start(struct net_device *ndev)
 		kref_get(&priv_new->rx_kref);
 		mutex_unlock(&j1939_netdev_lock);
 		dev_put(ndev);
+		erase_priv_trace_buffer(priv);
 		kfree(priv);
 		return priv_new;
 	}
@@ -299,6 +314,7 @@ struct j1939_priv *j1939_netdev_start(struct net_device *ndev)
 	mutex_unlock(&j1939_netdev_lock);
 
 	dev_put(ndev);
+	erase_priv_trace_buffer(priv);
 	kfree(priv);
 
 	return ERR_PTR(ret);
@@ -364,6 +380,9 @@ static int j1939_netdev_notify(struct notifier_block *nb,
 	struct can_ml_priv *can_ml = can_get_ml_priv(ndev);
 	struct j1939_priv *priv;
 
+	if (msg == NETDEV_DEBUG_UNREGISTER)
+		dump_priv_trace_buffer(ndev);
+
 	if (!can_ml)
 		goto notify_done;
 
@@ -428,3 +447,79 @@ static __exit void j1939_module_exit(void)
 
 module_init(j1939_module_init);
 module_exit(j1939_module_exit);
+
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+
+#define PRIV_TRACE_BUFFER_SIZE 1024
+static struct priv_trace_buffer {
+	struct j1939_priv *priv; // no-ref
+	struct net_device *ndev; // no-ref
+	atomic_t count;
+	int nr_entries;
+	unsigned long entries[20];
+} priv_trace_buffer[PRIV_TRACE_BUFFER_SIZE];
+static bool priv_trace_buffer_exhausted;
+
+static void dump_priv_trace_buffer(const struct net_device *ndev)
+{
+	struct priv_trace_buffer *ptr;
+	int count, balance = 0;
+	int i;
+
+	for (i = 0; i < PRIV_TRACE_BUFFER_SIZE; i++) {
+		ptr = &priv_trace_buffer[i];
+		if (!ptr->priv || ptr->ndev != ndev)
+			continue;
+		count = atomic_read(&ptr->count);
+		balance += count;
+		pr_info("Call trace for %s@%p %+d at\n", ndev->name, ptr->priv, count);
+		stack_trace_print(ptr->entries, ptr->nr_entries, 4);
+	}
+	if (!priv_trace_buffer_exhausted)
+		pr_info("balance for %s@j1939_priv is %d\n", ndev->name, balance);
+	else
+		pr_info("balance for %s@j1939_priv is unknown\n", ndev->name);
+}
+
+static void erase_priv_trace_buffer(struct j1939_priv *priv)
+{
+	int i;
+
+	for (i = 0; i < PRIV_TRACE_BUFFER_SIZE; i++)
+		if (priv_trace_buffer[i].priv == priv)
+			priv_trace_buffer[i].priv = NULL;
+}
+
+static noinline void save_priv_trace_buffer(struct j1939_priv *priv, int delta)
+{
+	struct priv_trace_buffer *ptr;
+	unsigned long entries[ARRAY_SIZE(ptr->entries)];
+	unsigned long nr_entries;
+	int i;
+
+	if (in_nmi())
+		return;
+	nr_entries = stack_trace_save(entries, ARRAY_SIZE(ptr->entries), 1);
+	nr_entries = trim_netdev_trace(entries, nr_entries);
+	for (i = 0; i < PRIV_TRACE_BUFFER_SIZE; i++) {
+		ptr = &priv_trace_buffer[i];
+		if (ptr->priv == priv && ptr->nr_entries == nr_entries &&
+		    !memcmp(ptr->entries, entries, nr_entries * sizeof(unsigned long))) {
+			atomic_add(delta, &ptr->count);
+			return;
+		}
+	}
+	for (i = 0; i < PRIV_TRACE_BUFFER_SIZE; i++) {
+		ptr = &priv_trace_buffer[i];
+		if (!ptr->priv && !cmpxchg(&ptr->priv, NULL, priv)) {
+			ptr->ndev = priv->ndev;
+			atomic_set(&ptr->count, delta);
+			ptr->nr_entries = nr_entries;
+			memmove(ptr->entries, entries, nr_entries * sizeof(unsigned long));
+			return;
+		}
+	}
+	priv_trace_buffer_exhausted = true;
+}
+
+#endif
diff --git a/net/core/dev.c b/net/core/dev.c
index 2acfa44927da..c3a62c16fa15 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -1854,6 +1854,7 @@ const char *netdev_cmd_to_name(enum netdev_cmd cmd)
 	N(PRE_CHANGEADDR) N(OFFLOAD_XSTATS_ENABLE) N(OFFLOAD_XSTATS_DISABLE)
 	N(OFFLOAD_XSTATS_REPORT_USED) N(OFFLOAD_XSTATS_REPORT_DELTA)
 	N(XDP_FEAT_CHANGE)
+	N(DEBUG_UNREGISTER)
 	}
 #undef N
 	return "UNKNOWN_NETDEV_EVENT";
@@ -11429,6 +11430,14 @@ int netdev_refcnt_read(const struct net_device *dev)
 }
 EXPORT_SYMBOL(netdev_refcnt_read);
 
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+static void dump_netdev_trace_buffer(const struct net_device *dev);
+static void erase_netdev_trace_buffer(const struct net_device *dev);
+#else
+static inline void dump_netdev_trace_buffer(const struct net_device *dev) { }
+static inline void erase_netdev_trace_buffer(const struct net_device *dev) { }
+#endif
+
 int netdev_unregister_timeout_secs __read_mostly = 10;
 
 #define WAIT_REFS_MIN_MSECS 1
@@ -11502,11 +11511,16 @@ static struct net_device *netdev_wait_allrefs_any(struct list_head *list)
 
 		if (time_after(jiffies, warning_time +
 			       READ_ONCE(netdev_unregister_timeout_secs) * HZ)) {
+			rtnl_lock();
 			list_for_each_entry(dev, list, todo_list) {
 				pr_emerg("unregister_netdevice: waiting for %s to become free. Usage count = %d\n",
 					 dev->name, netdev_refcnt_read(dev));
 				ref_tracker_dir_print(&dev->refcnt_tracker, 10);
+				call_netdevice_notifiers(NETDEV_DEBUG_UNREGISTER, dev);
+				dump_netdev_trace_buffer(dev);
 			}
+			__rtnl_unlock();
+			rcu_barrier();
 
 			warning_time = jiffies;
 		}
@@ -11904,6 +11918,9 @@ struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name,
 
 	dev->priv_len = sizeof_priv;
 
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+	INIT_LIST_HEAD(&dev->netdev_trace_buffer_list);
+#endif
 	ref_tracker_dir_init(&dev->refcnt_tracker, 128, "netdev");
 #ifdef CONFIG_PCPU_DEV_REFCNT
 	dev->pcpu_refcnt = alloc_percpu(int);
@@ -12076,6 +12093,8 @@ void free_netdev(struct net_device *dev)
 
 	mutex_destroy(&dev->lock);
 
+	erase_netdev_trace_buffer(dev);
+
 	/*  Compatibility with error handling in drivers */
 	if (dev->reg_state == NETREG_UNINITIALIZED ||
 	    dev->reg_state == NETREG_DUMMY) {
@@ -13090,3 +13109,180 @@ static int __init net_dev_init(void)
 }
 
 subsys_initcall(net_dev_init);
+
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+
+#define NETDEV_TRACE_BUFFER_SIZE 32768
+static struct netdev_trace_buffer {
+	struct list_head list;
+	int prev_count;
+	atomic_t count;
+	int nr_entries;
+	unsigned long entries[20];
+} netdev_trace_buffer[NETDEV_TRACE_BUFFER_SIZE];
+static LIST_HEAD(netdev_trace_buffer_list);
+static DEFINE_SPINLOCK(netdev_trace_buffer_lock);
+static bool netdev_trace_buffer_exhausted;
+
+static int netdev_trace_buffer_init(void)
+{
+	int i;
+
+	for (i = 0; i < NETDEV_TRACE_BUFFER_SIZE; i++)
+		list_add_tail(&netdev_trace_buffer[i].list, &netdev_trace_buffer_list);
+	return 0;
+}
+pure_initcall(netdev_trace_buffer_init);
+
+static void dump_netdev_trace_buffer(const struct net_device *dev)
+{
+	struct netdev_trace_buffer *ptr;
+	int count, balance = 0, pos = 0;
+
+	list_for_each_entry_rcu(ptr, &dev->netdev_trace_buffer_list, list,
+				/* list elements can't go away. */ 1) {
+		pos++;
+		count = atomic_read(&ptr->count);
+		balance += count;
+		if (ptr->prev_count == count)
+			continue;
+		ptr->prev_count = count;
+		pr_info("Call trace for %s[%d] %+d at\n", dev->name, pos, count);
+		stack_trace_print(ptr->entries, ptr->nr_entries, 4);
+		cond_resched();
+	}
+	if (!netdev_trace_buffer_exhausted)
+		pr_info("balance as of %s[%d] is %d\n", dev->name, pos, balance);
+}
+
+static void erase_netdev_trace_buffer(const struct net_device *dev)
+{
+	struct netdev_trace_buffer *ptr;
+	unsigned long flags;
+
+	spin_lock_irqsave(&netdev_trace_buffer_lock, flags);
+	while (!list_empty(&dev->netdev_trace_buffer_list)) {
+		ptr = list_first_entry(&dev->netdev_trace_buffer_list, typeof(*ptr), list);
+		list_del(&ptr->list);
+		list_add_tail(&ptr->list, &netdev_trace_buffer_list);
+	}
+	spin_unlock_irqrestore(&netdev_trace_buffer_lock, flags);
+}
+
+#ifdef CONFIG_KALLSYMS
+static noinline unsigned long __find_trim(unsigned long *entries, int nr_entries, const char *name)
+{
+	int i;
+	char buffer[KSYM_SYMBOL_LEN];
+	const int len = strlen(name);
+
+	for (i = 0; i < nr_entries; i++) {
+		snprintf(buffer, sizeof(buffer), "%pS", (void *)entries[i]);
+		if (!strncmp(buffer, name, len) && buffer[len] == '+')
+			return entries[i];
+	}
+	return 0;
+}
+
+static unsigned long caller_handle_softirqs;
+static unsigned long caller_process_one_work;
+static unsigned long caller_ksys_unshare;
+static unsigned long caller___sys_bind;
+static unsigned long caller___sock_sendmsg;
+
+static int __init net_check_symbols(void)
+{
+	if (!kallsyms_lookup_name("handle_softirqs"))
+		caller_handle_softirqs = -1;
+	if (!kallsyms_lookup_name("process_one_work"))
+		caller_process_one_work = -1;
+	if (!kallsyms_lookup_name("ksys_unshare"))
+		caller_ksys_unshare = -1;
+	if (!kallsyms_lookup_name("__sys_bind"))
+		caller___sys_bind = -1;
+	if (!kallsyms_lookup_name("sock_sendmsg_nosec") &&
+	    !kallsyms_lookup_name("__sock_sendmsg"))
+		caller___sock_sendmsg = -1;
+	return 0;
+}
+late_initcall(net_check_symbols);
+#endif
+
+int trim_netdev_trace(unsigned long *entries, int nr_entries)
+{
+#ifdef CONFIG_KALLSYMS
+	int i;
+
+	if (in_softirq()) {
+		if (unlikely(!caller_handle_softirqs))
+			caller_handle_softirqs = __find_trim(entries, nr_entries,
+							     "handle_softirqs");
+		for (i = 0; i < nr_entries; i++)
+			if (entries[i] == caller_handle_softirqs)
+				return i + 1;
+	} else if (current->flags & PF_WQ_WORKER) {
+		if (unlikely(!caller_process_one_work))
+			caller_process_one_work = __find_trim(entries, nr_entries,
+							      "process_one_work");
+		for (i = 0; i < nr_entries; i++)
+			if (entries[i] == caller_process_one_work)
+				return i + 1;
+	} else {
+		if (unlikely(!caller_ksys_unshare))
+			caller_ksys_unshare = __find_trim(entries, nr_entries, "ksys_unshare");
+		if (unlikely(!caller___sys_bind))
+			caller___sys_bind = __find_trim(entries, nr_entries, "__sys_bind");
+		if (unlikely(!caller___sock_sendmsg)) {
+			caller___sock_sendmsg = __find_trim(entries, nr_entries,
+							    "sock_sendmsg_nosec");
+			if (!caller___sock_sendmsg)
+				caller___sock_sendmsg = __find_trim(entries, nr_entries,
+								    "__sock_sendmsg");
+		}
+		for (i = 0; i < nr_entries; i++)
+			if (entries[i] == caller_ksys_unshare ||
+			    entries[i] == caller___sys_bind ||
+			    entries[i] == caller___sock_sendmsg)
+				return i + 1;
+	}
+#endif
+	return nr_entries;
+}
+EXPORT_SYMBOL(trim_netdev_trace);
+
+void save_netdev_trace_buffer(struct net_device *dev, int delta)
+{
+	struct netdev_trace_buffer *ptr;
+	unsigned long entries[ARRAY_SIZE(ptr->entries)];
+	unsigned long nr_entries;
+	unsigned long flags;
+
+	if (in_nmi())
+		return;
+	nr_entries = stack_trace_save(entries, ARRAY_SIZE(ptr->entries), 1);
+	nr_entries = trim_netdev_trace(entries, nr_entries);
+	list_for_each_entry_rcu(ptr, &dev->netdev_trace_buffer_list, list,
+				/* list elements can't go away. */ 1) {
+		if (ptr->nr_entries == nr_entries &&
+		    !memcmp(ptr->entries, entries, nr_entries * sizeof(unsigned long))) {
+			atomic_add(delta, &ptr->count);
+			return;
+		}
+	}
+	spin_lock_irqsave(&netdev_trace_buffer_lock, flags);
+	if (!list_empty(&netdev_trace_buffer_list)) {
+		ptr = list_first_entry(&netdev_trace_buffer_list, typeof(*ptr), list);
+		list_del(&ptr->list);
+		ptr->prev_count = 0;
+		atomic_set(&ptr->count, delta);
+		ptr->nr_entries = nr_entries;
+		memmove(ptr->entries, entries, nr_entries * sizeof(unsigned long));
+		list_add_tail_rcu(&ptr->list, &dev->netdev_trace_buffer_list);
+	} else {
+		netdev_trace_buffer_exhausted = true;
+	}
+	spin_unlock_irqrestore(&netdev_trace_buffer_lock, flags);
+}
+EXPORT_SYMBOL(save_netdev_trace_buffer);
+
+#endif
diff --git a/net/core/lock_debug.c b/net/core/lock_debug.c
index 9e9fb25314b9..78d611bb6d1c 100644
--- a/net/core/lock_debug.c
+++ b/net/core/lock_debug.c
@@ -29,6 +29,7 @@ int netdev_debug_event(struct notifier_block *nb, unsigned long event,
 	case NETDEV_DOWN:
 	case NETDEV_REBOOT:
 	case NETDEV_UNREGISTER:
+	case NETDEV_DEBUG_UNREGISTER:
 	case NETDEV_CHANGEMTU:
 	case NETDEV_CHANGEADDR:
 	case NETDEV_PRE_CHANGEADDR:
diff --git a/net/socket.c b/net/socket.c
index e8892b218708..fce536d2d8b9 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -734,6 +734,10 @@ static inline int sock_sendmsg_nosec(struct socket *sock, struct msghdr *msg)
 	return ret;
 }
 
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+static noinline int __sock_sendmsg(struct socket *sock, struct msghdr *msg);
+#endif
+
 static int __sock_sendmsg(struct socket *sock, struct msghdr *msg)
 {
 	int err = security_socket_sendmsg(sock, msg,


^ permalink raw reply related	[flat|nested] 13+ messages in thread
* Re: unregister_netdevice: waiting for DEV to become free (8)
  2025-11-19 14:00 Tetsuo Handa
@ 2026-03-02 10:56 Tetsuo Handa
  2026-03-02 11:21 ` [syzbot] [net?] " syzbot
  0 siblings, 1 reply; 13+ messages in thread
From: Tetsuo Handa @ 2026-03-02 10:56 UTC (permalink / raw)
  To: syzbot; +Cc: linux-kernel

#syz test: git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index d4e6e00bb90a..97862830f7d0 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -2102,6 +2102,8 @@ enum netdev_reg_state {
  *
  *	FIXME: cleanup struct net_device such that network protocol info
  *	moves out.
+ *
+ *	@netdev_trace_buffer_list: Linked list for debugging refcount leak.
  */
 
 struct net_device {
@@ -2257,6 +2259,9 @@ struct net_device {
 #if IS_ENABLED(CONFIG_TLS_DEVICE)
 	const struct tlsdev_ops *tlsdev_ops;
 #endif
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+	struct list_head	netdev_trace_buffer_list;
+#endif
 
 	unsigned int		operstate;
 	unsigned char		link_mode;
@@ -3185,6 +3190,7 @@ enum netdev_cmd {
 	NETDEV_OFFLOAD_XSTATS_REPORT_USED,
 	NETDEV_OFFLOAD_XSTATS_REPORT_DELTA,
 	NETDEV_XDP_FEAT_CHANGE,
+	NETDEV_DEBUG_UNREGISTER,
 };
 const char *netdev_cmd_to_name(enum netdev_cmd cmd);
 
@@ -4373,9 +4379,15 @@ static inline bool dev_nit_active(const struct net_device *dev)
 
 void dev_queue_xmit_nit(struct sk_buff *skb, struct net_device *dev);
 
+void save_netdev_trace_buffer(struct net_device *dev, int delta);
+int trim_netdev_trace(unsigned long *entries, int nr_entries);
+
 static inline void __dev_put(struct net_device *dev)
 {
 	if (dev) {
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+		save_netdev_trace_buffer(dev, -1);
+#endif
 #ifdef CONFIG_PCPU_DEV_REFCNT
 		this_cpu_dec(*dev->pcpu_refcnt);
 #else
@@ -4387,6 +4399,9 @@ static inline void __dev_put(struct net_device *dev)
 static inline void __dev_hold(struct net_device *dev)
 {
 	if (dev) {
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+		save_netdev_trace_buffer(dev, 1);
+#endif
 #ifdef CONFIG_PCPU_DEV_REFCNT
 		this_cpu_inc(*dev->pcpu_refcnt);
 #else
diff --git a/kernel/softirq.c b/kernel/softirq.c
index 77198911b8dd..5f435c1e48d8 100644
--- a/kernel/softirq.c
+++ b/kernel/softirq.c
@@ -576,6 +576,10 @@ static inline bool lockdep_softirq_start(void) { return false; }
 static inline void lockdep_softirq_end(bool in_hardirq) { }
 #endif
 
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+static noinline void handle_softirqs(bool ksirqd);
+#endif
+
 static void handle_softirqs(bool ksirqd)
 {
 	unsigned long end = jiffies + MAX_SOFTIRQ_TIME;
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index aeaec79bc09c..66cb4d8c00bb 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -3157,6 +3157,10 @@ static bool manage_workers(struct worker *worker)
 	return true;
 }
 
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+static noinline void process_one_work(struct worker *worker, struct work_struct *work);
+#endif
+
 /**
  * process_one_work - process single work
  * @worker: self
diff --git a/net/core/dev.c b/net/core/dev.c
index c1a9f7fdcffa..fbc205c82256 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -1874,6 +1874,7 @@ const char *netdev_cmd_to_name(enum netdev_cmd cmd)
 	N(PRE_CHANGEADDR) N(OFFLOAD_XSTATS_ENABLE) N(OFFLOAD_XSTATS_DISABLE)
 	N(OFFLOAD_XSTATS_REPORT_USED) N(OFFLOAD_XSTATS_REPORT_DELTA)
 	N(XDP_FEAT_CHANGE)
+	N(DEBUG_UNREGISTER)
 	}
 #undef N
 	return "UNKNOWN_NETDEV_EVENT";
@@ -11557,6 +11558,14 @@ int netdev_refcnt_read(const struct net_device *dev)
 }
 EXPORT_SYMBOL(netdev_refcnt_read);
 
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+static void dump_netdev_trace_buffer(const struct net_device *dev);
+static void erase_netdev_trace_buffer(const struct net_device *dev);
+#else
+static inline void dump_netdev_trace_buffer(const struct net_device *dev) { }
+static inline void erase_netdev_trace_buffer(const struct net_device *dev) { }
+#endif
+
 int netdev_unregister_timeout_secs __read_mostly = 10;
 
 #define WAIT_REFS_MIN_MSECS 1
@@ -11630,11 +11639,16 @@ static struct net_device *netdev_wait_allrefs_any(struct list_head *list)
 
 		if (time_after(jiffies, warning_time +
 			       READ_ONCE(netdev_unregister_timeout_secs) * HZ)) {
+			rtnl_lock();
 			list_for_each_entry(dev, list, todo_list) {
 				pr_emerg("unregister_netdevice: waiting for %s to become free. Usage count = %d\n",
 					 dev->name, netdev_refcnt_read(dev));
 				ref_tracker_dir_print(&dev->refcnt_tracker, 10);
+				call_netdevice_notifiers(NETDEV_DEBUG_UNREGISTER, dev);
+				dump_netdev_trace_buffer(dev);
 			}
+			__rtnl_unlock();
+			rcu_barrier();
 
 			warning_time = jiffies;
 		}
@@ -12032,6 +12046,9 @@ struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name,
 
 	dev->priv_len = sizeof_priv;
 
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+	INIT_LIST_HEAD(&dev->netdev_trace_buffer_list);
+#endif
 	ref_tracker_dir_init(&dev->refcnt_tracker, 128, "netdev");
 #ifdef CONFIG_PCPU_DEV_REFCNT
 	dev->pcpu_refcnt = alloc_percpu(int);
@@ -12204,6 +12221,8 @@ void free_netdev(struct net_device *dev)
 
 	mutex_destroy(&dev->lock);
 
+	erase_netdev_trace_buffer(dev);
+
 	/*  Compatibility with error handling in drivers */
 	if (dev->reg_state == NETREG_UNINITIALIZED ||
 	    dev->reg_state == NETREG_DUMMY) {
@@ -13306,3 +13325,171 @@ static int __init net_dev_init(void)
 }
 
 subsys_initcall(net_dev_init);
+
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+
+#define NETDEV_TRACE_BUFFER_SIZE 32768
+static struct netdev_trace_buffer {
+	struct list_head list;
+	int prev_count;
+	atomic_t count;
+	int nr_entries;
+	unsigned long entries[20];
+} netdev_trace_buffer[NETDEV_TRACE_BUFFER_SIZE];
+static LIST_HEAD(netdev_trace_buffer_list);
+static DEFINE_SPINLOCK(netdev_trace_buffer_lock);
+static bool netdev_trace_buffer_exhausted;
+
+static int netdev_trace_buffer_init(void)
+{
+	int i;
+
+	for (i = 0; i < NETDEV_TRACE_BUFFER_SIZE; i++)
+		list_add_tail(&netdev_trace_buffer[i].list, &netdev_trace_buffer_list);
+	return 0;
+}
+pure_initcall(netdev_trace_buffer_init);
+
+static void dump_netdev_trace_buffer(const struct net_device *dev)
+{
+	struct netdev_trace_buffer *ptr;
+	int count, balance = 0, pos = 0;
+
+	list_for_each_entry_rcu(ptr, &dev->netdev_trace_buffer_list, list,
+				/* list elements can't go away. */ 1) {
+		pos++;
+		count = atomic_read(&ptr->count);
+		balance += count;
+		if (ptr->prev_count == count)
+			continue;
+		ptr->prev_count = count;
+		pr_info("Call trace for %s[%d] %+d at\n", dev->name, pos, count);
+		stack_trace_print(ptr->entries, ptr->nr_entries, 4);
+		cond_resched();
+	}
+	if (!netdev_trace_buffer_exhausted)
+		pr_info("balance as of %s[%d] is %d\n", dev->name, pos, balance);
+}
+
+static void erase_netdev_trace_buffer(const struct net_device *dev)
+{
+	struct netdev_trace_buffer *ptr;
+	unsigned long flags;
+
+	spin_lock_irqsave(&netdev_trace_buffer_lock, flags);
+	while (!list_empty(&dev->netdev_trace_buffer_list)) {
+		ptr = list_first_entry(&dev->netdev_trace_buffer_list, typeof(*ptr), list);
+		list_del(&ptr->list);
+		list_add_tail(&ptr->list, &netdev_trace_buffer_list);
+	}
+	spin_unlock_irqrestore(&netdev_trace_buffer_lock, flags);
+}
+
+int trim_netdev_trace(unsigned long *entries, int nr_entries)
+{
+#ifdef CONFIG_KALLSYMS
+	char buffer[32] = { };
+	char *cp;
+	int i;
+
+	if (in_softirq()) {
+		static unsigned long __data_racy caller;
+
+		if (!caller) {
+			for (i = 0; i < nr_entries; i++) {
+				snprintf(buffer, sizeof(buffer) - 1, "%ps", (void *)entries[i]);
+				cp = strchr(buffer, ' ');
+				if (cp)
+					*cp = '\0';
+				if (!strcmp(buffer, "handle_softirqs")) {
+					caller = entries[i];
+					break;
+				}
+			}
+		}
+		for (i = 0; i < nr_entries; i++)
+			if (entries[i] == caller)
+				return i + 1;
+	} else if (current->flags & PF_WQ_WORKER) {
+		static unsigned long __data_racy caller;
+
+		if (!caller) {
+			for (i = 0; i < nr_entries; i++) {
+				snprintf(buffer, sizeof(buffer) - 1, "%ps", (void *)entries[i]);
+				cp = strchr(buffer, ' ');
+				if (cp)
+					*cp = '\0';
+				if (!strcmp(buffer, "process_one_work")) {
+					caller = entries[i];
+					break;
+				}
+			}
+		}
+		for (i = 0; i < nr_entries; i++)
+			if (entries[i] == caller)
+				return i + 1;
+	} else {
+		for (i = 0; i < nr_entries; i++) {
+			snprintf(buffer, sizeof(buffer) - 1, "%ps", (void *)entries[i]);
+			cp = strchr(buffer, ' ');
+			if (cp)
+				*cp = '\0';
+			if (buffer[0] == 'k') {
+				if (!strcmp(buffer, "ksys_unshare"))
+					return i + 1;
+			} else if (buffer[0] == 's') {
+				if (!strcmp(buffer, "sock_sendmsg_nosec") ||
+				    !strcmp(buffer, "sock_recvmsg_nosec"))
+					return i + 1;
+			} else if (buffer[0] == '_') {
+				if (!strcmp(buffer, "__sys_bind") ||
+				    !strcmp(buffer, "__sock_release") ||
+				    !strcmp(buffer, "__sys_bpf"))
+					return i + 1;
+			} else {
+				if (!strcmp(buffer, "do_sock_setsockopt"))
+					return i + 1;
+			}
+		}
+	}
+#endif
+	return nr_entries;
+}
+EXPORT_SYMBOL(trim_netdev_trace);
+
+void save_netdev_trace_buffer(struct net_device *dev, int delta)
+{
+	struct netdev_trace_buffer *ptr;
+	unsigned long entries[ARRAY_SIZE(ptr->entries)];
+	unsigned long nr_entries;
+	unsigned long flags;
+
+	if (in_nmi())
+		return;
+	nr_entries = stack_trace_save(entries, ARRAY_SIZE(ptr->entries), 1);
+	nr_entries = trim_netdev_trace(entries, nr_entries);
+	list_for_each_entry_rcu(ptr, &dev->netdev_trace_buffer_list, list,
+				/* list elements can't go away. */ 1) {
+		if (ptr->nr_entries == nr_entries &&
+		    !memcmp(ptr->entries, entries, nr_entries * sizeof(unsigned long))) {
+			atomic_add(delta, &ptr->count);
+			return;
+		}
+	}
+	spin_lock_irqsave(&netdev_trace_buffer_lock, flags);
+	if (!list_empty(&netdev_trace_buffer_list)) {
+		ptr = list_first_entry(&netdev_trace_buffer_list, typeof(*ptr), list);
+		list_del(&ptr->list);
+		ptr->prev_count = 0;
+		atomic_set(&ptr->count, delta);
+		ptr->nr_entries = nr_entries;
+		memmove(ptr->entries, entries, nr_entries * sizeof(unsigned long));
+		list_add_tail_rcu(&ptr->list, &dev->netdev_trace_buffer_list);
+	} else {
+		netdev_trace_buffer_exhausted = true;
+	}
+	spin_unlock_irqrestore(&netdev_trace_buffer_lock, flags);
+}
+EXPORT_SYMBOL(save_netdev_trace_buffer);
+
+#endif
diff --git a/net/core/lock_debug.c b/net/core/lock_debug.c
index 9e9fb25314b9..78d611bb6d1c 100644
--- a/net/core/lock_debug.c
+++ b/net/core/lock_debug.c
@@ -29,6 +29,7 @@ int netdev_debug_event(struct notifier_block *nb, unsigned long event,
 	case NETDEV_DOWN:
 	case NETDEV_REBOOT:
 	case NETDEV_UNREGISTER:
+	case NETDEV_DEBUG_UNREGISTER:
 	case NETDEV_CHANGEMTU:
 	case NETDEV_CHANGEADDR:
 	case NETDEV_PRE_CHANGEADDR:
diff --git a/net/socket.c b/net/socket.c
index 05952188127f..53c4b1fd3ef7 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -650,7 +650,11 @@ struct socket *sock_alloc(void)
 }
 EXPORT_SYMBOL(sock_alloc);
 
-static void __sock_release(struct socket *sock, struct inode *inode)
+static
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+noinline
+#endif
+void __sock_release(struct socket *sock, struct inode *inode)
 {
 	const struct proto_ops *ops = READ_ONCE(sock->ops);
 
@@ -722,7 +726,13 @@ static noinline void call_trace_sock_send_length(struct sock *sk, int ret,
 	trace_sock_send_length(sk, ret, 0);
 }
 
-static inline int sock_sendmsg_nosec(struct socket *sock, struct msghdr *msg)
+static
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+noinline
+#else
+inline
+#endif
+int sock_sendmsg_nosec(struct socket *sock, struct msghdr *msg)
 {
 	int ret = INDIRECT_CALL_INET(READ_ONCE(sock->ops)->sendmsg, inet6_sendmsg,
 				     inet_sendmsg, sock, msg,
@@ -1072,8 +1082,13 @@ static noinline void call_trace_sock_recv_length(struct sock *sk, int ret, int f
 	trace_sock_recv_length(sk, ret, flags);
 }
 
-static inline int sock_recvmsg_nosec(struct socket *sock, struct msghdr *msg,
-				     int flags)
+static
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+noinline
+#else
+inline
+#endif
+int sock_recvmsg_nosec(struct socket *sock, struct msghdr *msg, int flags)
 {
 	int ret = INDIRECT_CALL_INET(READ_ONCE(sock->ops)->recvmsg,
 				     inet6_recvmsg,
@@ -2532,9 +2547,12 @@ static int copy_msghdr_from_user(struct msghdr *kmsg,
 	return err < 0 ? err : 0;
 }
 
-static int ____sys_sendmsg(struct socket *sock, struct msghdr *msg_sys,
-			   unsigned int flags, struct used_address *used_address,
-			   unsigned int allowed_msghdr_flags)
+static
+#ifdef CONFIG_NET_DEV_REFCNT_TRACKER
+noinline
+#endif
+int ____sys_sendmsg(struct socket *sock, struct msghdr *msg_sys, unsigned int flags,
+		    struct used_address *used_address, unsigned int allowed_msghdr_flags)
 {
 	unsigned char ctl[sizeof(struct cmsghdr) + 20]
 				__aligned(sizeof(__kernel_size_t));


^ permalink raw reply related	[flat|nested] 13+ messages in thread
end of thread, other threads:[~2026-03-02 11:21 UTC | newest]

Thread overview: 13+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2023-06-10  1:34 [syzbot] [net?] unregister_netdevice: waiting for DEV to become free (8) syzbot
     [not found] ` <20230621070710.380373-1-astrajoan@yahoo.com>
2023-06-21  8:46   ` Dongliang Mu
     [not found] ` <20230819081057.330728-1-astrajoan@yahoo.com>
2023-08-19  9:25   ` [PATCH] net: bridge: Fix refcnt issues in dev_ioctl Nikolay Aleksandrov
     [not found]     ` <20230819225048.dxxzv47fo64g24qx@Astras-Ubuntu>
2023-08-22 10:40       ` Nikolay Aleksandrov
2025-08-25 12:35 ` [syzbot] [net?] unregister_netdevice: waiting for DEV to become free (8) Hillf Danton
2025-08-25 13:51   ` syzbot
2025-08-26  1:50 ` Hillf Danton
2025-08-26  2:48   ` syzbot
  -- strict thread matches above, loose matches on Subject: below --
2025-08-25 11:01 [PATCH] can: j1939: implement NETDEV_UNREGISTER notification handler Tetsuo Handa
2025-08-25 13:35 ` [syzbot] [net?] unregister_netdevice: waiting for DEV to become free (8) syzbot
2025-11-19 12:20 Tetsuo Handa
2025-11-19 13:09 ` [syzbot] [net?] " syzbot
2025-11-19 13:13 Tetsuo Handa
2025-11-19 13:57 ` [syzbot] [net?] " syzbot
2025-11-19 14:00 Tetsuo Handa
2025-11-19 14:47 ` [syzbot] [net?] " syzbot
2026-03-02 10:56 Tetsuo Handa
2026-03-02 11:21 ` [syzbot] [net?] " syzbot

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox