ceph: voluntarily drop Ax cap for requests that create new inode

MDS need to rdlock directory inode's authlock when handling these
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>

fs/ceph/caps.c

@@ -3939,11 +3939,20 @@ int ceph_encode_inode_release(void **p, struct inode *inode,
 
 	cap = __get_cap_for_mds(ci, mds);
 	if (cap && __cap_is_valid(cap)) {
-		if (force ||
-		    ((cap->issued & drop) &&
-		     (cap->issued & unless) == 0)) {
-			if ((cap->issued & drop) &&
-			    (cap->issued & unless) == 0) {
+		unless &= cap->issued;
+		if (unless) {
+			if (unless & CEPH_CAP_AUTH_EXCL)
+				drop &= ~CEPH_CAP_AUTH_SHARED;
+			if (unless & CEPH_CAP_LINK_EXCL)
+				drop &= ~CEPH_CAP_LINK_SHARED;
+			if (unless & CEPH_CAP_XATTR_EXCL)
+				drop &= ~CEPH_CAP_XATTR_SHARED;
+			if (unless & CEPH_CAP_FILE_EXCL)
+				drop &= ~CEPH_CAP_FILE_SHARED;
+		}
+
+		if (force || (cap->issued & drop)) {
+			if (cap->issued & drop) {
 				int wanted = __ceph_caps_wanted(ci);
 				if ((ci->i_ceph_flags & CEPH_I_NODELAY) == 0)
 					wanted |= cap->mds_wanted;
@@ -3975,7 +3984,7 @@ int ceph_encode_inode_release(void **p, struct inode *inode,
 			*p += sizeof(*rel);
 			ret = 1;
 		} else {
-			dout("encode_inode_release %p cap %p %s\n",
+			dout("encode_inode_release %p cap %p %s (noop)\n",
 			     inode, cap, ceph_cap_string(cap->issued));
 		}
 	}

fs/ceph/dir.c

@@ -835,7 +835,7 @@ static int ceph_mknod(struct inode *dir, struct dentry *dentry,
 	set_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags);
 	req->r_args.mknod.mode = cpu_to_le32(mode);
 	req->r_args.mknod.rdev = cpu_to_le32(rdev);
-	req->r_dentry_drop = CEPH_CAP_FILE_SHARED;
+	req->r_dentry_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_AUTH_EXCL;
 	req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
 	if (acls.pagelist) {
 		req->r_pagelist = acls.pagelist;
@@ -887,7 +887,7 @@ static int ceph_symlink(struct inode *dir, struct dentry *dentry,
 	set_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags);
 	req->r_dentry = dget(dentry);
 	req->r_num_caps = 2;
-	req->r_dentry_drop = CEPH_CAP_FILE_SHARED;
+	req->r_dentry_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_AUTH_EXCL;
 	req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
 	err = ceph_mdsc_do_request(mdsc, dir, req);
 	if (!err && !req->r_reply_info.head->is_dentry)
@@ -936,7 +936,7 @@ static int ceph_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
 	req->r_parent = dir;
 	set_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags);
 	req->r_args.mkdir.mode = cpu_to_le32(mode);
-	req->r_dentry_drop = CEPH_CAP_FILE_SHARED;
+	req->r_dentry_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_AUTH_EXCL;
 	req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
 	if (acls.pagelist) {
 		req->r_pagelist = acls.pagelist;

fs/ceph/file.c

@@ -396,7 +396,7 @@ int ceph_atomic_open(struct inode *dir, struct dentry *dentry,
 	req->r_dentry = dget(dentry);
 	req->r_num_caps = 2;
 	if (flags & O_CREAT) {
-		req->r_dentry_drop = CEPH_CAP_FILE_SHARED;
+		req->r_dentry_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_AUTH_EXCL;
 		req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
 		if (acls.pagelist) {
 			req->r_pagelist = acls.pagelist;