NLM: Don't hang forever on NLM unlock requests

If the NLM daemon is killed on the NFS server, we can currently end up
hanging forever on an 'unlock' request, instead of aborting. Basically,
if the rpcbind request fails, or the server keeps returning garbage, we
really want to quit instead of retrying.

Tested-by: Vasily Averin <vvs@sw.ru>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@kernel.org

4 files changed, 13 insertions, 2 deletions

diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c
index adb45ec9038c..e374050a911c 100644
--- a/fs/lockd/clntproc.c
+++ b/fs/lockd/clntproc.c
@@ -708,7 +708,13 @@ static void nlmclnt_unlock_callback(struct rpc_task *task, void *data)
 
         if (task->tk_status < 0) {
                 dprintk("lockd: unlock failed (err = %d)\n", -task->tk_status);
-                goto retry_rebind;
+                switch (task->tk_status) {
+                case -EACCES:
+                case -EIO:
+                        goto die;
+                default:
+                        goto retry_rebind;
+                }
         }
         if (status == NLM_LCK_DENIED_GRACE_PERIOD) {
                 rpc_delay(task, NLMCLNT_GRACE_WAIT);
diff --git a/include/linux/sunrpc/sched.h b/include/linux/sunrpc/sched.h
index f73c482ec9c6..fe2d8e6b923b 100644
--- a/include/linux/sunrpc/sched.h
+++ b/include/linux/sunrpc/sched.h
@@ -84,7 +84,8 @@ struct rpc_task {
 #endif
         unsigned char           tk_priority : 2,/* Task priority */
                                 tk_garb_retry : 2,
-                                tk_cred_retry : 2;
+                                tk_cred_retry : 2,
+                                tk_rebind_retry : 2;
 };
 #define tk_xprt                 tk_client->cl_xprt
 
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index b84d7395535e..566bcfd067f6 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -1175,6 +1175,9 @@ call_bind_status(struct rpc_task *task)
                         status = -EOPNOTSUPP;
                         break;
                 }
+                if (task->tk_rebind_retry == 0)
+                        break;
+                task->tk_rebind_retry--;
                 rpc_delay(task, 3*HZ);
                 goto retry_timeout;
         case -ETIMEDOUT:
diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c
index 6b43ee7221d5..a27406b1654f 100644
--- a/net/sunrpc/sched.c
+++ b/net/sunrpc/sched.c
@@ -792,6 +792,7 @@ static void rpc_init_task(struct rpc_task *task, const struct rpc_task_setup *ta
         /* Initialize retry counters */
         task->tk_garb_retry = 2;
         task->tk_cred_retry = 2;
+        task->tk_rebind_retry = 2;
 
         task->tk_priority = task_setup_data->priority - RPC_PRIORITY_LOW;
         task->tk_owner = current->tgid;