Re: [PATCH tip/core/rcu 3/9] drivers/md: Use rcu_dereference() for accessing rcu pointer

[Pranith Kumar]

On Fri, Nov 21, 2014 at 9:58 AM, Kirill A. Shutemov
<kirill@xxxxxxxxxxxxx> wrote:
> On Fri, Nov 21, 2014 at 09:30:36AM -0500, Pranith Kumar wrote:
>> On 11/21/2014 08:31 AM, Kirill A. Shutemov wrote:
>> > On Tue, Oct 28, 2014 at 03:09:56PM -0700, Paul E. McKenney wrote:
>> >> From: Pranith Kumar <bobby.prani@xxxxxxxxx>
>> >>
>> >> Got Paul's email wrong the first time.
>> >>
>> >> The map field in 'struct mapped_device' is an rcu pointer. Use rcu_dereference()
>> >> while accessing it.
>> >>
>> >> Signed-off-by: Pranith Kumar <bobby.prani@xxxxxxxxx>
>> >> Signed-off-by: Paul E. McKenney <paulmck@xxxxxxxxxxxxxxxxxx>
>> >
>> > On current -next I see this:
>> >
>> > [    6.388264] ===============================
>> > [    6.389571] [ INFO: suspicious RCU usage. ]
>> > [    6.390869] 3.18.0-rc5-next-20141121-08303-g44cae4530372 #2 Not tainted
>> > [    6.392185] -------------------------------
>> > [    6.393479] /home/kas/git/public/linux/drivers/md/dm.c:2853 suspicious rcu_dereference_check() usage!
>> > [    6.394801]
>> > other info that might help us debug this:
>> >
>>
>> Hi Kirill,
>>
>> We are dereferencing an RCU pointer with the suspend_lock held which is causing this warning.
>>
>> Can you please check if the following patch helps? Thanks!
>
> Nope. The same issue.
>
> IIUC, the problem is that you dereference pointer outside rcu_read_lock()
> section, not that suspend_lock is held.
>

I am not sure we should be taking rcu_read_lock() there as I am not
sure how long that critical section might last. Can someone who is
more familiar with the code take a look?

I will try to look for a solution too in the mean time.

Thanks!
-- 
Pranith
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/