md/raid5: close recently introduced race in stripe_head management.

As release_stripe and __release_stripe decrement ->count and then
manipulate ->lru both under ->device_lock, it is important that
get_active_stripe() increments ->count and clears ->lru also under
->device_lock.

However we currently list_del_init ->lru under the lock, but increment
the ->count outside the lock.  This can lead to races and list
corruption.

So move the atomic_inc(&sh->count) up inside the ->device_lock
protected region.

Note that we still increment ->count without device lock in the case
where get_free_stripe() was called, and in fact don't take
->device_lock at all in that path.
This is safe because if the stripe_head can be found by
get_free_stripe, then the hash lock assures us the no-one else could
possibly be calling release_stripe() at the same time.

Fixes: 566c09c534
Cc: stable@vger.kernel.org (3.13)
Reported-and-tested-by: Ian Kumlien <ian.kumlien@gmail.com>
Signed-off-by: NeilBrown <neilb@suse.de>
This commit is contained in:
NeilBrown 2014-01-22 11:45:03 +11:00
parent 9f97e4b128
commit 7da9d450ab

View file

@ -675,8 +675,10 @@ get_active_stripe(struct r5conf *conf, sector_t sector,
|| !conf->inactive_blocked), || !conf->inactive_blocked),
*(conf->hash_locks + hash)); *(conf->hash_locks + hash));
conf->inactive_blocked = 0; conf->inactive_blocked = 0;
} else } else {
init_stripe(sh, sector, previous); init_stripe(sh, sector, previous);
atomic_inc(&sh->count);
}
} else { } else {
spin_lock(&conf->device_lock); spin_lock(&conf->device_lock);
if (atomic_read(&sh->count)) { if (atomic_read(&sh->count)) {
@ -695,13 +697,11 @@ get_active_stripe(struct r5conf *conf, sector_t sector,
sh->group = NULL; sh->group = NULL;
} }
} }
atomic_inc(&sh->count);
spin_unlock(&conf->device_lock); spin_unlock(&conf->device_lock);
} }
} while (sh == NULL); } while (sh == NULL);
if (sh)
atomic_inc(&sh->count);
spin_unlock_irq(conf->hash_locks + hash); spin_unlock_irq(conf->hash_locks + hash);
return sh; return sh;
} }