From: Miles Lane on
[ 43.478812] [ INFO: suspicious rcu_dereference_check() usage. ]
[ 43.478815] ---------------------------------------------------
[ 43.478820] net/mac80211/sta_info.c:125 invoked
rcu_dereference_check() without protection!
[ 43.478824]
[ 43.478824] other info that might help us debug this:
[ 43.478826]
[ 43.478829]
[ 43.478830] rcu_scheduler_active = 1, debug_locks = 1
[ 43.478834] no locks held by NetworkManager/4017.
[ 43.478836]
[ 43.478837] stack backtrace:
[ 43.478842] Pid: 4017, comm: NetworkManager Not tainted 2.6.35-rc2-git1 #8
[ 43.478846] Call Trace:
[ 43.478849] <IRQ> [<ffffffff81064e9c>] lockdep_rcu_dereference+0x9d/0xa5
[ 43.478876] [<ffffffffa010cb3c>] sta_info_get_bss+0x71/0x12d [mac80211]
[ 43.478889] [<ffffffffa010cc0d>] ieee80211_find_sta+0x15/0x2f [mac80211]
[ 43.478902] [<ffffffffa019ae16>] iwlagn_tx_queue_reclaim+0xe7/0x1bb [iwlagn]
[ 43.478909] [<ffffffff8106530b>] ? mark_lock+0x2d/0x262
[ 43.478920] [<ffffffffa019e270>] iwlagn_rx_reply_tx+0x4cd/0x58a [iwlagn]
[ 43.478928] [<ffffffff811ca313>] ? is_swiotlb_buffer+0x2e/0x3b
[ 43.478937] [<ffffffffa0192ec2>] iwl_rx_handle+0x161/0x2bf [iwlagn]
[ 43.478946] [<ffffffffa019370a>] iwl_irq_tasklet+0x2eb/0x408 [iwlagn]
[ 43.478953] [<ffffffff8103f892>] tasklet_action+0xa7/0x10f
[ 43.478960] [<ffffffff810404cf>] __do_softirq+0x148/0x25a
[ 43.478966] [<ffffffff8100314c>] call_softirq+0x1c/0x28
[ 43.478972] [<ffffffff81004d20>] do_softirq+0x38/0x80
[ 43.478977] [<ffffffff8103ffd0>] irq_exit+0x45/0x94
[ 43.478983] [<ffffffff81004465>] do_IRQ+0xad/0xc4
[ 43.478989] [<ffffffff813ca3d3>] ret_from_intr+0x0/0xf
[ 43.478993] <EOI>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
From: Miles Lane on
On Tue, Jun 8, 2010 at 3:26 AM, Johannes Berg <johannes(a)sipsolutions.net> wrote:
> On Mon, 2010-06-07 at 16:59 -0700, Paul E. McKenney wrote:
>> On Mon, Jun 07, 2010 at 02:25:44PM -0400, Miles Lane wrote:
>> > [ � 43.478812] [ INFO: suspicious rcu_dereference_check() usage. ]
>> > [ � 43.478815] ---------------------------------------------------
>> > [ � 43.478820] net/mac80211/sta_info.c:125 invoked
>> > rcu_dereference_check() without protection!
>> > [ � 43.478824]
>> > [ � 43.478824] other info that might help us debug this:
>> > [ � 43.478826]
>> > [ � 43.478829]
>> > [ � 43.478830] rcu_scheduler_active = 1, debug_locks = 1
>> > [ � 43.478834] no locks held by NetworkManager/4017.
>>
>> Hmmm... �Johannes's update has been merged, and it requires that callers
>> either be in an RCU read-side critical section or hold either the
>> ->sta_lock or the ->sta_mtx, and this thread does none of this.
>>
>> Johannes, any thoughts?
>>
>> � � � � � � � � � � � � � � � � � � � � � � � � � � � Thanx, Paul
>>
>> > [ � 43.478837] stack backtrace:
>> > [ � 43.478842] Pid: 4017, comm: NetworkManager Not tainted 2.6.35-rc2-git1 #8
>> > [ � 43.478846] Call Trace:
>> > [ � 43.478849] �<IRQ> �[<ffffffff81064e9c>] lockdep_rcu_dereference+0x9d/0xa5
>> > [ � 43.478876] �[<ffffffffa010cb3c>] sta_info_get_bss+0x71/0x12d [mac80211]
>> > [ � 43.478889] �[<ffffffffa010cc0d>] ieee80211_find_sta+0x15/0x2f [mac80211]
>> > [ � 43.478902] �[<ffffffffa019ae16>] iwlagn_tx_queue_reclaim+0xe7/0x1bb [iwlagn]
>
> iwlwifi wasn't using rcu protection here -- already sent a patch fixing
> it. My mistake, I �think. Thanks for checking :)
>
> johannes
>
>

Yes, your patch works for me.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/