Messages in this thread | | | Subject | Re: [tip:core/locking 2/3] lockdep:define preallocated "stack" for BFS as per cpu variable | From | Peter Zijlstra <> | Date | Wed, 22 Jul 2009 15:03:50 +0200 |
| |
On Wed, 2009-07-22 at 20:58 +0800, tom.leiming@gmail.com wrote: > From: Ming Lei <tom.leiming@gmail.com>
This patch can use a changelog.
Why is this needed, isn't all that serialized by the graph_lock anyway? Or are there a few paths where this isn't the case and we're now racy?
> Signed-off-by: Ming Lei <tom.leiming@gmail.com> > --- > kernel/lockdep.c | 5 +++-- > 1 files changed, 3 insertions(+), 2 deletions(-) > > diff --git a/kernel/lockdep.c b/kernel/lockdep.c > index 1b1796a..1583439 100644 > --- a/kernel/lockdep.c > +++ b/kernel/lockdep.c > @@ -847,7 +847,7 @@ struct circular_queue { > unsigned int front, rear; > }; > > -static struct circular_queue lock_cq; > +static DEFINE_PER_CPU(struct circular_queue, lock_cq); > > unsigned int max_bfs_queue_depth; > > @@ -937,7 +937,7 @@ static int __bfs(struct lock_list *source_entry, > { > struct lock_list *entry; > struct list_head *head; > - struct circular_queue *cq = &lock_cq; > + struct circular_queue *cq = &get_cpu_var(lock_cq); > int ret = 1; > > if (match(source_entry, data)) { > @@ -993,6 +993,7 @@ static int __bfs(struct lock_list *source_entry, > } > } > exit: > + put_cpu_var(lock_cq); > return ret; > } >
| |