From 91876b13b8b3cbff5cca8476d4b4eebe5f6038cc Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Thu, 28 Mar 2013 14:01:33 -0400
Subject: NFSv4: Fix another reboot recovery race

If the open_context for the file is not yet fully initialised,
then open recovery cannot succeed, and since nfs4_state_find_open_context
returns an ENOENT, we end up treating the file as being irrecoverable.

What we really want to do, is just defer the recovery until later.

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

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 3e7d42f..8134394 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -1373,7 +1373,7 @@ static int nfs4_open_reclaim(struct nfs4_state_owner *sp, struct nfs4_state *sta
 
 	ctx = nfs4_state_find_open_context(state);
 	if (IS_ERR(ctx))
-		return PTR_ERR(ctx);
+		return -EAGAIN;
 	ret = nfs4_do_open_reclaim(ctx, state);
 	put_nfs_open_context(ctx);
 	return ret;
@@ -1814,7 +1814,7 @@ static int nfs4_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *sta
 
 	ctx = nfs4_state_find_open_context(state);
 	if (IS_ERR(ctx))
-		return PTR_ERR(ctx);
+		return -EAGAIN;
 	ret = nfs4_do_open_expired(ctx, state);
 	put_nfs_open_context(ctx);
 	return ret;
@@ -1936,10 +1936,8 @@ static int _nfs4_open_and_get_state(struct nfs4_opendata *opendata,
 	if (ret != 0)
 		goto out;
 
-	if (read_seqcount_retry(&sp->so_reclaim_seqcount, seq)) {
+	if (read_seqcount_retry(&sp->so_reclaim_seqcount, seq))
 		nfs4_schedule_stateid_recovery(server, state);
-		nfs4_wait_clnt_recover(server->nfs_client);
-	}
 	*res = state;
 out:
 	return ret;
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index 685b1e9..b924bdd 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -1476,6 +1476,8 @@ restart:
 				 */
 				nfs4_state_mark_recovery_failed(state, status);
 				break;
+			case -EAGAIN:
+				ssleep(1);
 			case -NFS4ERR_ADMIN_REVOKED:
 			case -NFS4ERR_STALE_STATEID:
 			case -NFS4ERR_BAD_STATEID:
-- 
cgit v0.10.2