NFSv4: The state manager must ignore EKEYEXPIRED.

Otherwise, we cannot recover state correctly.

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

2 files changed, 40 insertions, 10 deletions

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 88a9d93ec1bd..aa771c1af115 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -1186,7 +1186,7 @@ static int nfs4_do_open_reclaim(struct nfs_open_context *ctx, struct nfs4_state
         int err;
         do {
                 err = _nfs4_do_open_reclaim(ctx, state);
-                if (err != -NFS4ERR_DELAY && err != -EKEYEXPIRED)
+                if (err != -NFS4ERR_DELAY)
                         break;
                 nfs4_handle_exception(server, err, &exception);
         } while (exception.retry);
@@ -1256,6 +1256,13 @@ int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state
                         case -NFS4ERR_ADMIN_REVOKED:
                         case -NFS4ERR_BAD_STATEID:
                                 nfs4_state_mark_reclaim_nograce(server->nfs_client, state);
+                        case -EKEYEXPIRED:
+                                /*
+                                 * User RPCSEC_GSS context has expired.
+                                 * We cannot recover this stateid now, so
+                                 * skip it and allow recovery thread to
+                                 * proceed.
+                                 */
                         case -ENOMEM:
                                 err = 0;
                                 goto out;
@@ -1603,7 +1610,6 @@ static int nfs4_do_open_expired(struct nfs_open_context *ctx, struct nfs4_state
                         goto out;
                 case -NFS4ERR_GRACE:
                 case -NFS4ERR_DELAY:
-                case -EKEYEXPIRED:
                         nfs4_handle_exception(server, err, &exception);
                         err = 0;
                 }
@@ -3544,7 +3550,6 @@ int nfs4_proc_setclientid_confirm(struct nfs_client *clp,
                         case -NFS4ERR_RESOURCE:
                                 /* The IBM lawyers misread another document! */
                         case -NFS4ERR_DELAY:
-                        case -EKEYEXPIRED:
                                 err = nfs4_delay(clp->cl_rpcclient, &timeout);
                 }
         } while (err == 0);
@@ -4156,7 +4161,7 @@ static int nfs4_lock_reclaim(struct nfs4_state *state, struct file_lock *request
                 if (test_bit(NFS_DELEGATED_STATE, &state->flags) != 0)
                         return 0;
                 err = _nfs4_do_setlk(state, F_SETLK, request, NFS_LOCK_RECLAIM);
-                if (err != -NFS4ERR_DELAY && err != -EKEYEXPIRED)
+                if (err != -NFS4ERR_DELAY)
                         break;
                 nfs4_handle_exception(server, err, &exception);
         } while (exception.retry);
@@ -4181,7 +4186,6 @@ static int nfs4_lock_expired(struct nfs4_state *state, struct file_lock *request
                         goto out;
                 case -NFS4ERR_GRACE:
                 case -NFS4ERR_DELAY:
-                case -EKEYEXPIRED:
                         nfs4_handle_exception(server, err, &exception);
                         err = 0;
                 }
@@ -4327,13 +4331,21 @@ int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl)
                                 nfs4_state_mark_reclaim_nograce(server->nfs_client, state);
                                 err = 0;
                                 goto out;
+                        case -EKEYEXPIRED:
+                                /*
+                                 * User RPCSEC_GSS context has expired.
+                                 * We cannot recover this stateid now, so
+                                 * skip it and allow recovery thread to
+                                 * proceed.
+                                 */
+                                err = 0;
+                                goto out;
                         case -ENOMEM:
                         case -NFS4ERR_DENIED:
                                 /* kill_proc(fl->fl_pid, SIGLOST, 1); */
                                 err = 0;
                                 goto out;
                         case -NFS4ERR_DELAY:
-                        case -EKEYEXPIRED:
                                 break;
                 }
                 err = nfs4_handle_exception(server, err, &exception);
@@ -4562,7 +4574,6 @@ static void nfs4_get_lease_time_done(struct rpc_task *task, void *calldata)
         switch (task->tk_status) {
         case -NFS4ERR_DELAY:
         case -NFS4ERR_GRACE:
-        case -EKEYEXPIRED:
                 dprintk("%s Retry: tk_status %d\n", __func__, task->tk_status);
                 rpc_delay(task, NFS4_POLL_RETRY_MIN);
                 task->tk_status = 0;
@@ -5025,7 +5036,6 @@ static int nfs41_sequence_handle_errors(struct rpc_task *task, struct nfs_client
 {
         switch(task->tk_status) {
         case -NFS4ERR_DELAY:
-        case -EKEYEXPIRED:
                 rpc_delay(task, NFS4_POLL_RETRY_MAX);
                 return -EAGAIN;
         default:
@@ -5167,7 +5177,6 @@ static int nfs41_reclaim_complete_handle_errors(struct rpc_task *task, struct nf
         case -NFS4ERR_WRONG_CRED: /* What to do here? */
                 break;
         case -NFS4ERR_DELAY:
-        case -EKEYEXPIRED:
                 rpc_delay(task, NFS4_POLL_RETRY_MAX);
                 return -EAGAIN;
         default:
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index 940cf7c070af..40028ac2b69c 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -1063,6 +1063,14 @@ restart:
                                 /* Mark the file as being 'closed' */
                                 state->state = 0;
                                 break;
+                        case -EKEYEXPIRED:
+                                /*
+                                 * User RPCSEC_GSS context has expired.
+                                 * We cannot recover this stateid now, so
+                                 * skip it and allow recovery thread to
+                                 * proceed.
+                                 */
+                                break;
                         case -NFS4ERR_ADMIN_REVOKED:
                         case -NFS4ERR_STALE_STATEID:
                         case -NFS4ERR_BAD_STATEID:
@@ -1181,6 +1189,14 @@ static void nfs4_state_start_reclaim_nograce(struct nfs_client *clp)
         nfs4_state_mark_reclaim_helper(clp, nfs4_state_mark_reclaim_nograce);
 }
 
+static void nfs4_warn_keyexpired(const char *s)
+{
+        printk_ratelimited(KERN_WARNING "Error: state manager"
+                        " encountered RPCSEC_GSS session"
+                        " expired against NFSv4 server %s.\n",
+                        s);
+}
+
 static int nfs4_recovery_handle_error(struct nfs_client *clp, int error)
 {
         switch (error) {
@@ -1210,6 +1226,10 @@ static int nfs4_recovery_handle_error(struct nfs_client *clp, int error)
                         set_bit(NFS4CLNT_SESSION_RESET, &clp->cl_state);
                         /* Zero session reset errors */
                         return 0;
+                case -EKEYEXPIRED:
+                        /* Nothing we can do */
+                        nfs4_warn_keyexpired(clp->cl_hostname);
+                        return 0;
         }
         return error;
 }
@@ -1420,9 +1440,10 @@ static void nfs4_set_lease_expired(struct nfs_client *clp, int status)
                 case -NFS4ERR_DELAY:
                 case -NFS4ERR_CLID_INUSE:
                 case -EAGAIN:
-                case -EKEYEXPIRED:
                         break;
 
+                case -EKEYEXPIRED:
+                        nfs4_warn_keyexpired(clp->cl_hostname);
                 case -NFS4ERR_NOT_SAME: /* FixMe: implement recovery
                                          * in nfs4_exchange_id */
                 default: