Allow multiple values per directory entry

When using zfs to back a Lustre filesystem it's advantageous to
to store a fid with the object id in the directory zap.  The only
technical impediment to doing this is that the zpl code expects
a single value in the zap per directory entry.

This change relaxes that requirement such that multiple entries
are allowed provided the first one is the object id.  The zpl
code will just ignore additional entries.  This allows the ZoL
count to mount datasets which are being used as Lustre server
backends.

Once the upstream feature flags support is merged in this change
should be updated to a read-only feature.  Until this occurs
other zfs implementations will not be able to read the zfs
filesystems created by Lustre.

Signed-off-by: Brian Behlendorf <[email protected]>

2 files changed, 28 insertions, 6 deletions

diff --git a/module/zfs/zfs_dir.c b/module/zfs/zfs_dir.c
index 47550ec63..8f1a0c2cc 100644
--- a/module/zfs/zfs_dir.c
+++ b/module/zfs/zfs_dir.c
@@ -64,11 +64,11 @@ static int
 zfs_match_find(zfs_sb_t *zsb, znode_t *dzp, char *name, boolean_t exact,
     boolean_t update, int *deflags, pathname_t *rpnp, uint64_t *zoid)
 {
+	boolean_t conflict = B_FALSE;
 	int error;
 
 	if (zsb->z_norm) {
 		matchtype_t mt = MT_FIRST;
-		boolean_t conflict = B_FALSE;
 		size_t bufsz = 0;
 		char *buf = NULL;
 
@@ -84,11 +84,23 @@ zfs_match_find(zfs_sb_t *zsb, znode_t *dzp, char *name, boolean_t exact,
 		 */
 		error = zap_lookup_norm(zsb->z_os, dzp->z_id, name, 8, 1,
 		    zoid, mt, buf, bufsz, &conflict);
-		if (!error && deflags)
-			*deflags = conflict ? ED_CASE_CONFLICT : 0;
 	} else {
 		error = zap_lookup(zsb->z_os, dzp->z_id, name, 8, 1, zoid);
 	}
+
+	/*
+	 * Allow multiple entries provided the first entry is
+	 * the object id.  Non-zpl consumers may safely make
+	 * use of the additional space.
+	 *
+	 * XXX: This should be a feature flag for compatibility
+	 */
+	if (error == EOVERFLOW)
+		error = 0;
+
+	if (zsb->z_norm && !error && deflags)
+		*deflags = conflict ? ED_CASE_CONFLICT : 0;
+
 	*zoid = ZFS_DIRENT_OBJ(*zoid);
 
 #ifdef HAVE_DNLC
diff --git a/module/zfs/zfs_vnops.c b/module/zfs/zfs_vnops.c
index b7f5daaaf..c1391984a 100644
--- a/module/zfs/zfs_vnops.c
+++ b/module/zfs/zfs_vnops.c
@@ -2012,12 +2012,22 @@ zfs_readdir(struct inode *ip, void *dirent, filldir_t filldir,
 					goto update;
 			}
 
+			/*
+			 * Allow multiple entries provided the first entry is
+			 * the object id.  Non-zpl consumers may safely make
+			 * use of the additional space.
+			 *
+			 * XXX: This should be a feature flag for compatibility
+			 */
 			if (zap.za_integer_length != 8 ||
-			    zap.za_num_integers != 1) {
+			    zap.za_num_integers == 0) {
 				cmn_err(CE_WARN, "zap_readdir: bad directory "
-				    "entry, obj = %lld, offset = %lld\n",
+				    "entry, obj = %lld, offset = %lld, "
+				    "length = %d, num = %lld\n",
 				    (u_longlong_t)zp->z_id,
-				    (u_longlong_t)*pos);
+				    (u_longlong_t)*pos,
+				    zap.za_integer_length,
+				    (u_longlong_t)zap.za_num_integers);
 				error = ENXIO;
 				goto update;
 			}