vboxsf: don't allow to change the inode type

vboxsf_init_inode() is used both for initial setup of inode and for metadata
updates.  Tell it whether we are updating a live inode or setting up a new
instance and have it refuse to change type in the former case.

[fixed the braino caught by Hans de Goede <hdegoede@redhat.com>]

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
diff --git a/fs/vboxsf/super.c b/fs/vboxsf/super.c
index d7816c0..4f5e59f 100644
--- a/fs/vboxsf/super.c
+++ b/fs/vboxsf/super.c
@@ -207,7 +207,7 @@ static int vboxsf_fill_super(struct super_block *sb, struct fs_context *fc)
 		err = -ENOMEM;
 		goto fail_unmap;
 	}
-	vboxsf_init_inode(sbi, iroot, &sbi->root_info);
+	vboxsf_init_inode(sbi, iroot, &sbi->root_info, false);
 	unlock_new_inode(iroot);
 
 	droot = d_make_root(iroot);
@@ -418,7 +418,7 @@ static int vboxsf_reconfigure(struct fs_context *fc)
 
 	/* Apply changed options to the root inode */
 	sbi->o = ctx->o;
-	vboxsf_init_inode(sbi, iroot, &sbi->root_info);
+	vboxsf_init_inode(sbi, iroot, &sbi->root_info, true);
 
 	return 0;
 }