| Message ID | 20170406181050.12137-2-bart.vanassche@sandisk.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Thu, Apr 06, 2017 at 11:10:46AM -0700, Bart Van Assche wrote: > Since the next patch in this series will use RCU to iterate over > tag_list, make this safe. Add lockdep_assert_held() statements > in functions that iterate over tag_list to make clear that using > list_for_each_entry() instead of list_for_each_entry_rcu() is > fine in these functions. > > Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com> > Cc: Christoph Hellwig <hch@lst.de> > Cc: Hannes Reinecke <hare@suse.com> > --- > block/blk-mq.c | 6 ++++++ > 1 file changed, 6 insertions(+) > > diff --git a/block/blk-mq.c b/block/blk-mq.c > index f7cd3208bcdf..b5580b09b4a5 100644 > --- a/block/blk-mq.c > +++ b/block/blk-mq.c > @@ -2076,6 +2076,8 @@ static void blk_mq_update_tag_set_depth(struct blk_mq_tag_set *set, bool shared) > { > struct request_queue *q; > > + lockdep_assert_held(&set->tag_list_lock); > + > list_for_each_entry(q, &set->tag_list, tag_set_list) { > blk_mq_freeze_queue(q); > queue_set_hctx_shared(q, shared); > @@ -2096,6 +2098,8 @@ static void blk_mq_del_queue_tag_set(struct request_queue *q) > blk_mq_update_tag_set_depth(set, false); > } > mutex_unlock(&set->tag_list_lock); > + > + synchronize_rcu(); Looks synchronize_rcu() is only needed in deletion path, so it can be moved to blk_mq_del_queue_tag_set(). Also list_del_init/list_add_tail() need to be replaced with RCU safe version in functions operating &set->tag_list. > } > > static void blk_mq_add_queue_tag_set(struct blk_mq_tag_set *set, > @@ -2601,6 +2605,8 @@ void blk_mq_update_nr_hw_queues(struct blk_mq_tag_set *set, int nr_hw_queues) > { > struct request_queue *q; > > + lockdep_assert_held(&set->tag_list_lock); > + > if (nr_hw_queues > nr_cpu_ids) > nr_hw_queues = nr_cpu_ids; > if (nr_hw_queues < 1 || nr_hw_queues == set->nr_hw_queues) > -- > 2.12.0 >
On Fri, 2017-04-07 at 17:46 +0800, Ming Lei wrote: > On Thu, Apr 06, 2017 at 11:10:46AM -0700, Bart Van Assche wrote: > > Since the next patch in this series will use RCU to iterate over > > tag_list, make this safe. Add lockdep_assert_held() statements > > in functions that iterate over tag_list to make clear that using > > list_for_each_entry() instead of list_for_each_entry_rcu() is > > fine in these functions. > > > > Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com> > > Cc: Christoph Hellwig <hch@lst.de> > > Cc: Hannes Reinecke <hare@suse.com> > > --- > > block/blk-mq.c | 6 ++++++ > > 1 file changed, 6 insertions(+) > > > > diff --git a/block/blk-mq.c b/block/blk-mq.c > > index f7cd3208bcdf..b5580b09b4a5 100644 > > --- a/block/blk-mq.c > > +++ b/block/blk-mq.c > > @@ -2076,6 +2076,8 @@ static void blk_mq_update_tag_set_depth(struct blk_mq_tag_set *set, bool shared) > > { > > struct request_queue *q; > > > > + lockdep_assert_held(&set->tag_list_lock); > > + > > list_for_each_entry(q, &set->tag_list, tag_set_list) { > > blk_mq_freeze_queue(q); > > queue_set_hctx_shared(q, shared); > > @@ -2096,6 +2098,8 @@ static void blk_mq_del_queue_tag_set(struct request_queue *q) > > blk_mq_update_tag_set_depth(set, false); > > } > > mutex_unlock(&set->tag_list_lock); > > + > > + synchronize_rcu(); > > Looks synchronize_rcu() is only needed in deletion path, so it can > be moved to blk_mq_del_queue_tag_set(). > > Also list_del_init/list_add_tail() need to be replaced with RCU > safe version in functions operating &set->tag_list. Hello Ming, I will replace list_del_init() / list_add_tail() by their RCU equivalents. Regarding synchronize_rcu(): have you noticed that that call has been added to blk_mq_del_queue_tag_set(), the function you requested to move that call to? Bart.
diff --git a/block/blk-mq.c b/block/blk-mq.c index f7cd3208bcdf..b5580b09b4a5 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -2076,6 +2076,8 @@ static void blk_mq_update_tag_set_depth(struct blk_mq_tag_set *set, bool shared) { struct request_queue *q; + lockdep_assert_held(&set->tag_list_lock); + list_for_each_entry(q, &set->tag_list, tag_set_list) { blk_mq_freeze_queue(q); queue_set_hctx_shared(q, shared); @@ -2096,6 +2098,8 @@ static void blk_mq_del_queue_tag_set(struct request_queue *q) blk_mq_update_tag_set_depth(set, false); } mutex_unlock(&set->tag_list_lock); + + synchronize_rcu(); } static void blk_mq_add_queue_tag_set(struct blk_mq_tag_set *set, @@ -2601,6 +2605,8 @@ void blk_mq_update_nr_hw_queues(struct blk_mq_tag_set *set, int nr_hw_queues) { struct request_queue *q; + lockdep_assert_held(&set->tag_list_lock); + if (nr_hw_queues > nr_cpu_ids) nr_hw_queues = nr_cpu_ids; if (nr_hw_queues < 1 || nr_hw_queues == set->nr_hw_queues)
Since the next patch in this series will use RCU to iterate over tag_list, make this safe. Add lockdep_assert_held() statements in functions that iterate over tag_list to make clear that using list_for_each_entry() instead of list_for_each_entry_rcu() is fine in these functions. Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Hannes Reinecke <hare@suse.com> --- block/blk-mq.c | 6 ++++++ 1 file changed, 6 insertions(+)