NFSv4.1: Don't loop forever in nfs4_proc_create_session

If a server for some reason keeps sending NFS4ERR_DELAY errors, we can end
up looping forever inside nfs4_proc_create_session, and so the usual
mechanisms for detecting if the nfs_client is dead don't work.

Fix this by ensuring that we loop inside the nfs4_state_manager thread
instead.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>

diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h
index e1c261d..c4a6983 100644 (file)
--- a/fs/nfs/nfs4_fs.h
+++ b/fs/nfs/nfs4_fs.h
@@ -47,6 +47,7 @@ enum nfs4_client_state {
        NFS4CLNT_LAYOUTRECALL,
        NFS4CLNT_SESSION_RESET,
        NFS4CLNT_RECALL_SLOT,
+       NFS4CLNT_LEASE_CONFIRM,
 };
 
 enum nfs4_session_state {

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 628e35f..50c8148 100644 (file)
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -3754,18 +3754,17 @@ int nfs4_proc_setclientid(struct nfs_client *clp, u32 program,
                status = rpc_call_sync(clp->cl_rpcclient, &msg, 0);
                if (status != -NFS4ERR_CLID_INUSE)
                        break;
-               if (signalled())
+               if (loop != 0) {
+                       ++clp->cl_id_uniquifier;
                        break;
-               if (loop++ & 1)
-                       ssleep(clp->cl_lease_time / HZ + 1);
-               else
-                       if (++clp->cl_id_uniquifier == 0)
-                               break;
+               }
+               ++loop;
+               ssleep(clp->cl_lease_time / HZ + 1);
        }
        return status;
 }
 
-static int _nfs4_proc_setclientid_confirm(struct nfs_client *clp,
+int nfs4_proc_setclientid_confirm(struct nfs_client *clp,
                struct nfs4_setclientid_res *arg,
                struct rpc_cred *cred)
 {
@@ -3790,26 +3789,6 @@ static int _nfs4_proc_setclientid_confirm(struct nfs_client *clp,
        return status;
 }
 
-int nfs4_proc_setclientid_confirm(struct nfs_client *clp,
-               struct nfs4_setclientid_res *arg,
-               struct rpc_cred *cred)
-{
-       long timeout = 0;
-       int err;
-       do {
-               err = _nfs4_proc_setclientid_confirm(clp, arg, cred);
-               switch (err) {
-                       case 0:
-                               return err;
-                       case -NFS4ERR_RESOURCE:
-                               /* The IBM lawyers misread another document! */
-                       case -NFS4ERR_DELAY:
-                               err = nfs4_delay(clp->cl_rpcclient, &timeout);
-               }
-       } while (err == 0);
-       return err;
-}
-
 struct nfs4_delegreturndata {
        struct nfs4_delegreturnargs args;
        struct nfs4_delegreturnres res;
@@ -5222,20 +5201,10 @@ int nfs4_proc_create_session(struct nfs_client *clp)
        int status;
        unsigned *ptr;
        struct nfs4_session *session = clp->cl_session;
-       long timeout = 0;
-       int err;
 
        dprintk("--> %s clp=%p session=%p\n", __func__, clp, session);
 
-       do {
-               status = _nfs4_proc_create_session(clp);
-               if (status == -NFS4ERR_DELAY) {
-                       err = nfs4_delay(clp->cl_rpcclient, &timeout);
-                       if (err)
-                               status = err;
-               }
-       } while (status == -NFS4ERR_DELAY);
-
+       status = _nfs4_proc_create_session(clp);
        if (status)
                goto out;
 

diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index 4dfb34b..4a810c8 100644 (file)
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -64,10 +64,15 @@ static LIST_HEAD(nfs4_clientid_list);
 
 int nfs4_init_clientid(struct nfs_client *clp, struct rpc_cred *cred)
 {
-       struct nfs4_setclientid_res clid;
+       struct nfs4_setclientid_res clid = {
+               .clientid = clp->cl_clientid,
+               .confirm = clp->cl_confirm,
+       };
        unsigned short port;
        int status;
 
+       if (test_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state))
+               goto do_confirm;
        port = nfs_callback_tcpport;
        if (clp->cl_addr.ss_family == AF_INET6)
                port = nfs_callback_tcpport6;
@@ -75,10 +80,14 @@ int nfs4_init_clientid(struct nfs_client *clp, struct rpc_cred *cred)
        status = nfs4_proc_setclientid(clp, NFS4_CALLBACK, port, cred, &clid);
        if (status != 0)
                goto out;
+       clp->cl_clientid = clid.clientid;
+       clp->cl_confirm = clid.confirm;
+       set_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state);
+do_confirm:
        status = nfs4_proc_setclientid_confirm(clp, &clid, cred);
        if (status != 0)
                goto out;
-       clp->cl_clientid = clid.clientid;
+       clear_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state);
        nfs4_schedule_state_renewal(clp);
 out:
        return status;
@@ -230,13 +239,18 @@ int nfs41_init_clientid(struct nfs_client *clp, struct rpc_cred *cred)
 {
        int status;
 
+       if (test_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state))
+               goto do_confirm;
        nfs4_begin_drain_session(clp);
        status = nfs4_proc_exchange_id(clp, cred);
        if (status != 0)
                goto out;
+       set_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state);
+do_confirm:
        status = nfs4_proc_create_session(clp);
        if (status != 0)
                goto out;
+       clear_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state);
        nfs41_setup_state_renewal(clp);
        nfs_mark_client_ready(clp, NFS_CS_READY);
 out:
@@ -1584,20 +1598,22 @@ static int nfs4_recall_slot(struct nfs_client *clp) { return 0; }
  */
 static void nfs4_set_lease_expired(struct nfs_client *clp, int status)
 {
-       if (nfs4_has_session(clp)) {
-               switch (status) {
-               case -NFS4ERR_DELAY:
-               case -NFS4ERR_CLID_INUSE:
-               case -EAGAIN:
-                       break;
+       switch (status) {
+       case -NFS4ERR_CLID_INUSE:
+       case -NFS4ERR_STALE_CLIENTID:
+               clear_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state);
+               break;
+       case -NFS4ERR_DELAY:
+       case -EAGAIN:
+               ssleep(1);
+               break;
 
-               case -EKEYEXPIRED:
-                       nfs4_warn_keyexpired(clp->cl_hostname);
-               case -NFS4ERR_NOT_SAME: /* FixMe: implement recovery
-                                        * in nfs4_exchange_id */
-               default:
-                       return;
-               }
+       case -EKEYEXPIRED:
+               nfs4_warn_keyexpired(clp->cl_hostname);
+       case -NFS4ERR_NOT_SAME: /* FixMe: implement recovery
+                                * in nfs4_exchange_id */
+       default:
+               return;
        }
        set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state);
 }

diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index 216cea5..87694ca 100644 (file)
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -47,6 +47,7 @@ struct nfs_client {
 
 #ifdef CONFIG_NFS_V4
        u64                     cl_clientid;    /* constant */
+       nfs4_verifier           cl_confirm;     /* Clientid verifier */
        unsigned long           cl_state;
 
        spinlock_t              cl_lock;