From 82ccf268c0fc98dfc282ff2e240656530618d49e Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 31 May 2011 15:15:34 -0400 Subject: [PATCH] NLM: Don't hang forever on NLM unlock requests commit 0b760113a3a155269a3fba93a409c640031dd68f upstream. 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 Signed-off-by: Trond Myklebust Signed-off-by: Greg Kroah-Hartman --- fs/lockd/clntproc.c | 8 +++++++- include/linux/sunrpc/sched.h | 4 ++-- net/sunrpc/clnt.c | 3 +++ net/sunrpc/sched.c | 1 + 4 files changed, 13 insertions(+), 3 deletions(-) diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c index c81249fef11..c325a834f8d 100644 --- a/fs/lockd/clntproc.c +++ b/fs/lockd/clntproc.c @@ -709,7 +709,13 @@ 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 401097781fc..67f63ddb2fc 100644 --- a/include/linux/sunrpc/sched.h +++ b/include/linux/sunrpc/sched.h @@ -84,8 +84,8 @@ struct rpc_task { long tk_rtt; /* round-trip time (jiffies) */ pid_t tk_owner; /* Process id for batching tasks */ - unsigned char tk_priority : 2;/* Task priority */ - + unsigned char tk_priority : 2,/* Task priority */ + tk_rebind_retry : 2; #ifdef RPC_DEBUG unsigned short tk_pid; /* debugging aid */ #endif diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index d838bea672a..b0c56461a19 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1052,6 +1052,9 @@ 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 570da303b01..ac94477eb49 100644 --- a/net/sunrpc/sched.c +++ b/net/sunrpc/sched.c @@ -784,6 +784,7 @@ 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; -- 2.34.1