From: "Yan, Zheng" <zyan@redhat.com>
Date: Thu, 23 Nov 2017 17:59:13 +0800
Subject: ceph: voluntarily drop Lx cap for link/rename requests
Git-commit: d19a0b540182a742338d88f6501e4f29a30d7541
Patch-mainline: v4.16-rc1
References: FATE#324714

MDS need to xlock inode's linklock when handling link/rename requests.
Voluntarily dropping CEPH_CAP_AUTH_EXCL avoids a cap revoke message.

Signed-off-by: "Yan, Zheng" <zyan@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Acked-by: Luis Henriques <lhenriques@suse.com>
---
 fs/ceph/dir.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -982,7 +982,7 @@ static int ceph_link(struct dentry *old_
 	req->r_dentry_drop = CEPH_CAP_FILE_SHARED;
 	req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
 	/* release LINK_SHARED on source inode (mds will lock it) */
-	req->r_old_inode_drop = CEPH_CAP_LINK_SHARED;
+	req->r_old_inode_drop = CEPH_CAP_LINK_SHARED | CEPH_CAP_LINK_EXCL;
 	err = ceph_mdsc_do_request(mdsc, dir, req);
 	if (err) {
 		d_drop(dentry);
@@ -1095,7 +1095,7 @@ static int ceph_rename(struct inode *old
 	req->r_dentry_drop = CEPH_CAP_FILE_SHARED;
 	req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
 	/* release LINK_RDCACHE on source inode (mds will lock it) */
-	req->r_old_inode_drop = CEPH_CAP_LINK_SHARED;
+	req->r_old_inode_drop = CEPH_CAP_LINK_SHARED | CEPH_CAP_LINK_EXCL;
 	if (d_really_is_positive(new_dentry))
 		req->r_inode_drop = drop_caps_for_unlink(d_inode(new_dentry));
 	err = ceph_mdsc_do_request(mdsc, old_dir, req);