Hello,
I have these in my logs :
bad: scheduling while atomic!
Call Trace:
[<c011aa54>] schedule+0x59e/0x5a4
[<c0118575>] pte_alloc_one+0x55/0x68
[<c011ad6d>] wait_for_completion+0x91/0xe6
[<c011aaaa>] default_wake_function+0x0/0x12
[<c011aaaa>] default_wake_function+0x0/0x12
[<c012dc05>] synchronize_kernel+0x31/0x3c
[<c012dbca>] wakeme_after_rcu+0x0/0xa
[<c04131ef>] unregister_netdevice+0x141/0x27a
[<c0419f8b>] rtmsg_ifinfo+0x91/0xda
[<c03316c8>] lapbeth_device_event+0xa6/0xf0
[<c0129dc5>] notifier_call_chain+0x27/0x3e
[<c041320f>] unregister_netdevice+0x161/0x27a
[<c0330cf2>] tun_chr_close+0xe8/0xf2
[<c0330c0a>] tun_chr_close+0x0/0xf2
[<c0153340>] __fput+0x120/0x132
[<c015199b>] filp_close+0x59/0x86
[<c0151a2c>] sys_close+0x64/0x98
[<c0108b17>] syscall_call+0x7/0xb
Any other info that could help someone ?
Regards,
Paul
Bring up/down network devices with lapbether causes scheduling while
atomic (if preempt enabled).
The calls to rcu_read_lock are unnecessary since lapb_device_event
is called from notifier with the rtnetlink semaphore held, it is
already protected from the labp_devices list changing.
Patch against 2.6.6-rc1
diff -Nru a/drivers/net/wan/lapbether.c b/drivers/net/wan/lapbether.c
--- a/drivers/net/wan/lapbether.c Fri Apr 16 11:00:35 2004
+++ b/drivers/net/wan/lapbether.c Fri Apr 16 11:00:35 2004
@@ -392,6 +392,8 @@
/*
* Handle device status changes.
+ *
+ * Called from notifier with RTNL held.
*/
static int lapbeth_device_event(struct notifier_block *this,
unsigned long event, void *ptr)
@@ -402,7 +404,6 @@
if (!dev_is_ethdev(dev))
return NOTIFY_DONE;
- rcu_read_lock();
switch (event) {
case NETDEV_UP:
/* New ethernet device -> new LAPB interface */
@@ -422,7 +423,6 @@
lapbeth_free_device(lapbeth);
break;
}
- rcu_read_unlock();
return NOTIFY_DONE;
}
On Fri, 16 Apr 2004 13:16:33 -0700
Stephen Hemminger <[email protected]> wrote:
> Bring up/down network devices with lapbether causes scheduling while
> atomic (if preempt enabled).
>
> The calls to rcu_read_lock are unnecessary since lapb_device_event
> is called from notifier with the rtnetlink semaphore held, it is
> already protected from the labp_devices list changing.
Applied, thanks Stephen.
Thanks Stephen for being so fast !
Regards,
Paul
Paul Rolland, rol(at)as2917.net
ex-AS2917 Network administrator and Peering Coordinator
--
Please no HTML, I'm not a browser - Pas d'HTML, je ne suis pas un navigateur
"Some people dream of success... while others wake up and work hard at it"
> -----Message d'origine-----
> De : [email protected]
> [mailto:[email protected]] De la part de
> Stephen Hemminger
> Envoy? : vendredi 16 avril 2004 22:17
> ? : David S. Miller
> Cc : Paul Rolland; [email protected]; [email protected]
> Objet : Re: [2.6.5] Bad scheduling while atomic
>
> Bring up/down network devices with lapbether causes scheduling while
> atomic (if preempt enabled).
>
> The calls to rcu_read_lock are unnecessary since lapb_device_event
> is called from notifier with the rtnetlink semaphore held, it is
> already protected from the labp_devices list changing.
>
> Patch against 2.6.6-rc1
>
> diff -Nru a/drivers/net/wan/lapbether.c b/drivers/net/wan/lapbether.c
> --- a/drivers/net/wan/lapbether.c Fri Apr 16 11:00:35 2004
> +++ b/drivers/net/wan/lapbether.c Fri Apr 16 11:00:35 2004
> @@ -392,6 +392,8 @@
>
> /*
> * Handle device status changes.
> + *
> + * Called from notifier with RTNL held.
> */
> static int lapbeth_device_event(struct notifier_block *this,
> unsigned long event, void *ptr)
> @@ -402,7 +404,6 @@
> if (!dev_is_ethdev(dev))
> return NOTIFY_DONE;
>
> - rcu_read_lock();
> switch (event) {
> case NETDEV_UP:
> /* New ethernet device -> new LAPB interface */
> @@ -422,7 +423,6 @@
> lapbeth_free_device(lapbeth);
> break;
> }
> - rcu_read_unlock();
>
> return NOTIFY_DONE;
> }
> -
> To unsubscribe from this list: send the line "unsubscribe
> linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
>