NFSv4: Fix a nfs4_state_manager() race

[ Upstream commit ed1cc05 ]

If the NFS4CLNT_RUN_MANAGER flag got set just before we cleared
NFS4CLNT_MANAGER_RUNNING, then we might have won the race against
nfs4_schedule_state_manager(), and are responsible for handling the
recovery situation.

Fixes: aeabb3c ("NFSv4: Fix a NFSv4 state manager deadlock")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

fs/nfs/nfs4state.c

@@ -2710,6 +2710,13 @@ static void nfs4_state_manager(struct nfs_client *clp)
 		nfs4_end_drain_session(clp);
 		nfs4_clear_state_manager_bit(clp);
 
+		if (test_bit(NFS4CLNT_RUN_MANAGER, &clp->cl_state) &&
+		    !test_and_set_bit(NFS4CLNT_MANAGER_RUNNING,
+				      &clp->cl_state)) {
+			memflags = memalloc_nofs_save();
+			continue;
+		}
+
 		if (!test_and_set_bit(NFS4CLNT_RECALL_RUNNING, &clp->cl_state)) {
 			if (test_and_clear_bit(NFS4CLNT_DELEGRETURN, &clp->cl_state)) {
 				nfs_client_return_marked_delegations(clp);