[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20130203183958.GO2632@linux.vnet.ibm.com>
Date: Sun, 3 Feb 2013 10:39:58 -0800
From: "Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>
To: Steven Whitehouse <swhiteho@...hat.com>
Cc: Dipankar Sarma <dipankar@...ibm.com>, linux-kernel@...r.kernel.org,
adas@...hat.com, hch@....de, npiggin@...il.com
Subject: Re: rcu: fix hlist_bl_set_first_rcu annotation
On Wed, Jan 30, 2013 at 07:07:57PM +0000, Steven Whitehouse wrote:
>
> Abhi noticed that we were getting a complaint from the RCU subsystem
> about access of an RCU protected list under the write side bit lock.
> This patch adds additional annotation to check both the RCU read
> lock and the write side bit lock before printing a message.
>
> Signed-off-by: Steven Whitehouse <swhiteho@...hat.com>
> Reported-by: Abhijith Das <adas@...hat.com>
> Tested-by: Abhijith Das <adas@...hat.com>
Looks plausible to me on first glance, copying Nick Piggin and Christoph
Hellwig. If they have no objections, I will queue this.
Thanx, Paul
> diff --git a/include/linux/list_bl.h b/include/linux/list_bl.h
> index 31f9d75..2eb8855 100644
> --- a/include/linux/list_bl.h
> +++ b/include/linux/list_bl.h
> @@ -125,6 +125,11 @@ static inline void hlist_bl_unlock(struct hlist_bl_head *b)
> __bit_spin_unlock(0, (unsigned long *)b);
> }
>
> +static inline bool hlist_bl_is_locked(struct hlist_bl_head *b)
> +{
> + return bit_spin_is_locked(0, (unsigned long *)b);
> +}
> +
> /**
> * hlist_bl_for_each_entry - iterate over list of given type
> * @tpos: the type * to use as a loop cursor.
> diff --git a/include/linux/rculist_bl.h b/include/linux/rculist_bl.h
> index cf1244f..4f216c5 100644
> --- a/include/linux/rculist_bl.h
> +++ b/include/linux/rculist_bl.h
> @@ -20,7 +20,7 @@ static inline void hlist_bl_set_first_rcu(struct hlist_bl_head *h,
> static inline struct hlist_bl_node *hlist_bl_first_rcu(struct hlist_bl_head *h)
> {
> return (struct hlist_bl_node *)
> - ((unsigned long)rcu_dereference(h->first) & ~LIST_BL_LOCKMASK);
> + ((unsigned long)rcu_dereference_check(h->first, hlist_bl_is_locked(h)) & ~LIST_BL_LOCKMASK);
> }
>
> /**
>
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists