Fix rpc shutdown event condition bug
authorLinus Torvalds <torvalds@g5.osdl.org>
Tue, 10 Jan 2006 16:56:39 +0000 (08:56 -0800)
committerLinus Torvalds <torvalds@g5.osdl.org>
Tue, 10 Jan 2006 16:56:39 +0000 (08:56 -0800)
We want to wait for the cl_users to go down to zero, not for it to stay
positive.  Quoth Trond (who wasn't even the author, but acked the wrong
version): "Argh! I need to increase my daily caffeine dosages."

Signed-off-by: Linus Torvalds <torvalds@osdl.org>
net/sunrpc/clnt.c

index a44da8b3d24076383c97574f4bb27cb957f22eb8..4cef7fa2b740d838093614037a78053a7c462f3e 100644 (file)
@@ -269,7 +269,7 @@ rpc_shutdown_client(struct rpc_clnt *clnt)
                clnt->cl_dead = 0;
                rpc_killall_tasks(clnt);
                wait_event_timeout(destroy_wait,
-                       atomic_read(&clnt->cl_users) > 0, 1*HZ);
+                       !atomic_read(&clnt->cl_users), 1*HZ);
        }
 
        if (atomic_read(&clnt->cl_users) < 0) {