NFSv4.1 handle DS stateid errors

Handle DS READ and WRITE stateid errors by recovering the stateid on the MDS.

NFS4ERR_OLD_STATEID is ignored as the client always sends a
state sequenceid of zero for DS READ and WRITE stateids.

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

3 files changed, 31 insertions, 1 deletions

diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c
index 87f7544f3dce..97d53574bf53 100644
--- a/fs/nfs/delegation.c
+++ b/fs/nfs/delegation.c
@@ -474,6 +474,7 @@ void nfs_remove_bad_delegation(struct inode *inode)
                 nfs_free_delegation(delegation);
         }
 }
+EXPORT_SYMBOL_GPL(nfs_remove_bad_delegation);
 
 /**
  * nfs_expire_all_delegation_types
diff --git a/fs/nfs/nfs4filelayout.c b/fs/nfs/nfs4filelayout.c
index 47e8f3435d38..b2d3bb5971bb 100644
--- a/fs/nfs/nfs4filelayout.c
+++ b/fs/nfs/nfs4filelayout.c
@@ -36,6 +36,7 @@
 #include <linux/sunrpc/metrics.h>
 
 #include "internal.h"
+#include "delegation.h"
 #include "nfs4filelayout.h"
 
 #define NFSDBG_FACILITY         NFSDBG_PNFS_LD
@@ -86,12 +87,31 @@ static int filelayout_async_handle_error(struct rpc_task *task,
                                          struct nfs_client *clp,
                                          int *reset)
 {
+        struct nfs_server *mds_server = NFS_SERVER(state->inode);
+        struct nfs_client *mds_client = mds_server->nfs_client;
+
         if (task->tk_status >= 0)
                 return 0;
-
         *reset = 0;
 
         switch (task->tk_status) {
+        /* MDS state errors */
+        case -NFS4ERR_DELEG_REVOKED:
+        case -NFS4ERR_ADMIN_REVOKED:
+        case -NFS4ERR_BAD_STATEID:
+                if (state != NULL)
+                        nfs_remove_bad_delegation(state->inode);
+        case -NFS4ERR_OPENMODE:
+                if (state == NULL)
+                        break;
+                nfs4_schedule_stateid_recovery(mds_server, state);
+                goto wait_on_recovery;
+        case -NFS4ERR_EXPIRED:
+                if (state != NULL)
+                        nfs4_schedule_stateid_recovery(mds_server, state);
+                nfs4_schedule_lease_recovery(mds_client);
+                goto wait_on_recovery;
+        /* DS session errors */
         case -NFS4ERR_BADSESSION:
         case -NFS4ERR_BADSLOT:
         case -NFS4ERR_BAD_HIGH_SLOT:
@@ -117,8 +137,15 @@ static int filelayout_async_handle_error(struct rpc_task *task,
                 *reset = 1;
                 break;
         }
+out:
         task->tk_status = 0;
         return -EAGAIN;
+wait_on_recovery:
+        rpc_sleep_on(&mds_client->cl_rpcwaitq, task, NULL);
+        if (test_bit(NFS4CLNT_MANAGER_RUNNING, &mds_client->cl_state) == 0)
+                rpc_wake_up_queued_task(&mds_client->cl_rpcwaitq, task);
+        goto out;
+
 }
 
 /* NFS_PROTO call done callback routines */
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index 1dad5c53c7fa..a58d02a0c27d 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -1072,6 +1072,7 @@ void nfs4_schedule_lease_recovery(struct nfs_client *clp)
                 set_bit(NFS4CLNT_CHECK_LEASE, &clp->cl_state);
         nfs4_schedule_state_manager(clp);
 }
+EXPORT_SYMBOL_GPL(nfs4_schedule_lease_recovery);
 
 void nfs4_schedule_path_down_recovery(struct nfs_client *clp)
 {
@@ -1109,6 +1110,7 @@ void nfs4_schedule_stateid_recovery(const struct nfs_server *server, struct nfs4
         nfs4_state_mark_reclaim_nograce(clp, state);
         nfs4_schedule_state_manager(clp);
 }
+EXPORT_SYMBOL_GPL(nfs4_schedule_stateid_recovery);
 
 void nfs_inode_find_state_and_recover(struct inode *inode,
                 const nfs4_stateid *stateid)