Temporarily back out revision 1.77. This changed cd9660_getattr() and

cd9660_readdir() to return the address of the file's first data block as
the inode number instead of the address of the directory entry, but
neglected to update cd9660_vget_internal() for the new inode numbering
scheme.

Since the NFS server calls VFS_VGET (cd9660_vget()) with inode numbers
returned through VOP_READDIR (cd9660_readdir()) when servicing a READDIRPLUS
request, these two interfaces must agree on the numbering scheme; failure to
do so caused panics and/or bogus information about the entries to be returned
to clients using READDIRPLUS (Solaris, FreeBSD w/ mount -o rdirplus).

PR:		63446

1 files changed, 6 insertions, 14 deletions

diff --git a/sys/isofs/cd9660/cd9660_vnops.c b/sys/isofs/cd9660/cd9660_vnops.c
index a7b15585fc33..40eb1ddd7e13 100644
--- a/sys/isofs/cd9660/cd9660_vnops.c
+++ b/sys/isofs/cd9660/cd9660_vnops.c
@@ -167,14 +167,7 @@ cd9660_getattr(ap)
 	struct iso_node *ip = VTOI(vp);
 
 	vap->va_fsid	= dev2udev(ip->i_dev);
-
-	/*
-	 * Don't use ip->i_ino for this since it is wrong for hard links.
-	 * ip->i_ino should be the same as ip->iso_start (or not exist),
-	 * but this currently doesn't work since we abuse it to look up
-	 * parent directories from inodes.
-	 */
-	vap->va_fileid	= ip->iso_start;
+	vap->va_fileid	= ip->i_number;
 
 	vap->va_mode	= ip->inode.iso_mode;
 	vap->va_nlink	= ip->inode.iso_links;
@@ -528,12 +521,11 @@ cd9660_readdir(ap)
 			break;
 		}
 
-		/*
-		 * The "inode number" is iso_start, not i_ino, as in
-		 * cd9660_getattr().
-		 */
-		idp->current.d_fileno = isonum_711(ep->ext_attr_length) +
-		    isonum_733(ep->extent);
+		if (isonum_711(ep->flags)&2)
+			idp->current.d_fileno = isodirino(ep, imp);
+		else
+			idp->current.d_fileno = dbtob(bp->b_blkno) +
+				entryoffsetinblock;
 
 		idp->curroff += reclen;