aboutsummaryrefslogtreecommitdiffstats
path: root/fs
diff options
context:
space:
mode:
authorJeff Layton <jlayton@redhat.com>2008-02-06 11:34:13 -0500
committerJ. Bruce Fields <bfields@citi.umich.edu>2008-02-10 18:09:36 -0500
commitc64e80d55db81df22a7f25b75ab4ba4c55db4749 (patch)
treebc844c9fbc4a19887b108d457f0d8fdc651630e2 /fs
parent9706501e43a80ce48b319214a0a9e562deded35b (diff)
NLM: don't requeue block if it was invalidated while GRANT_MSG was in flight
It's possible for lockd to catch a SIGKILL while a GRANT_MSG callback is in flight. If this happens we don't want lockd to insert the block back into the nlm_blocked list. This helps that situation, but there's still a possible race. Fixing that will mean adding real locking for nlm_blocked. Signed-off-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Diffstat (limited to 'fs')
-rw-r--r--fs/lockd/svclock.c11
1 files changed, 11 insertions, 0 deletions
diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c
index 82db7b323b83..fe9bdb4a220c 100644
--- a/fs/lockd/svclock.c
+++ b/fs/lockd/svclock.c
@@ -795,6 +795,17 @@ static void nlmsvc_grant_callback(struct rpc_task *task, void *data)
795 795
796 dprintk("lockd: GRANT_MSG RPC callback\n"); 796 dprintk("lockd: GRANT_MSG RPC callback\n");
797 797
798 /* if the block is not on a list at this point then it has
799 * been invalidated. Don't try to requeue it.
800 *
801 * FIXME: it's possible that the block is removed from the list
802 * after this check but before the nlmsvc_insert_block. In that
803 * case it will be added back. Perhaps we need better locking
804 * for nlm_blocked?
805 */
806 if (list_empty(&block->b_list))
807 return;
808
798 /* Technically, we should down the file semaphore here. Since we 809 /* Technically, we should down the file semaphore here. Since we
799 * move the block towards the head of the queue only, no harm 810 * move the block towards the head of the queue only, no harm
800 * can be done, though. */ 811 * can be done, though. */