Re: [PATCH] lockdep: Fix redundant_hardirqs_on incremented withirqs enabled

[Stephen Rothwell]

Hi Frederic,

On Thu, 15 Apr 2010 15:37:50 +0200 Frederic Weisbecker <fweisbec@xxxxxxxxx> wrote:
>
> When a path restore the flags while irqs are already enabled, we
> update the per cpu var redundant_hardirqs_on in a racy fashion
> and debug_atomic_inc() warns about this situation.
> 
> In this particular case, we need to explictly disable the irqs before
> updating this stat var in order to update it safely.
> 
> Reported-by: Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx>
> Signed-off-by: Frederic Weisbecker <fweisbec@xxxxxxxxx>
> Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> Cc: David Miller <davem@xxxxxxxxxxxxx>
> Cc: Benjamin Herrenschmidt <benh@xxxxxxxxxxxxxxxxxxx>
> ---
>  kernel/lockdep.c |    4 ++++
>  1 files changed, 4 insertions(+), 0 deletions(-)
> 
> diff --git a/kernel/lockdep.c b/kernel/lockdep.c
> index 78325f8..65d4336 100644
> --- a/kernel/lockdep.c
> +++ b/kernel/lockdep.c
> @@ -2298,7 +2298,11 @@ void trace_hardirqs_on_caller(unsigned long ip)
>  		return;
>  
>  	if (unlikely(curr->hardirqs_enabled)) {
> +		unsigned long flags;
> +
> +		raw_local_irq_save(flags);
>  		debug_atomic_inc(redundant_hardirqs_on);
> +		raw_local_irq_restore(flags);
>  		return;
>  	}
>  	/* we'll do an OFF -> ON transition: */

I applied this patch to linux-next today and it fixes my PowerPC boot
problems, so I will keep it in linux-next until it gets applied to the
tip tree or something better comes along.

Thanks.
-- 
Cheers,
Stephen Rothwell                    sfr@xxxxxxxxxxxxxxxx
http://www.canb.auug.org.au/~sfr/

Attachment: pgp00000.pgp
Description: PGP signature