Merge tag 'ceph-for-4.16-rc1' of git://github.com/ceph/ceph-client

Pull ceph updates from Ilya Dryomov:
 "Things have been very quiet on the rbd side, as work continues on the
  big ticket items slated for the next merge window.

  On the CephFS side we have a large number of cap handling
  improvements, a fix for our long-standing abuse of ->journal_info in
  ceph_readpages() and yet another dentry pointer management patch"

* tag 'ceph-for-4.16-rc1' of git://github.com/ceph/ceph-client:
  ceph: improving efficiency of syncfs
  libceph: check kstrndup() return value
  ceph: try to allocate enough memory for reserved caps
  ceph: fix race of queuing delayed caps
  ceph: delete unreachable code in ceph_check_caps()
  ceph: limit rate of cap import/export error messages
  ceph: fix incorrect snaprealm when adding caps
  ceph: fix un-balanced fsc->writeback_count update
  ceph: track read contexts in ceph_file_info
  ceph: avoid dereferencing invalid pointer during cached readdir
  ceph: use atomic_t for ceph_inode_info::i_shared_gen
  ceph: cleanup traceless reply handling for rename
  ceph: voluntarily drop Fx cap for readdir request
  ceph: properly drop caps for setattr request
  ceph: voluntarily drop Lx cap for link/rename requests
  ceph: voluntarily drop Ax cap for requests that create new inode
  rbd: whitelist RBD_FEATURE_OPERATIONS feature bit
  rbd: don't NULL out ->obj_request in rbd_img_obj_parent_read_full()
  rbd: use kmem_cache_zalloc() in rbd_img_request_create()
  rbd: obj_request->completion is unused

1 files changed, 42 insertions, 14 deletions

diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index ab81652198c4..c6ec5aa46100 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -494,7 +494,7 @@ struct inode *ceph_alloc_inode(struct super_block *sb)
 	ci->i_wrbuffer_ref = 0;
 	ci->i_wrbuffer_ref_head = 0;
 	atomic_set(&ci->i_filelock_ref, 0);
-	ci->i_shared_gen = 0;
+	atomic_set(&ci->i_shared_gen, 0);
 	ci->i_rdcache_gen = 0;
 	ci->i_rdcache_revoking = 0;
 
@@ -1041,7 +1041,7 @@ static void update_dentry_lease(struct dentry *dentry,
 	if (ceph_snap(dir) != CEPH_NOSNAP)
 		goto out_unlock;
 
-	di->lease_shared_gen = ceph_inode(dir)->i_shared_gen;
+	di->lease_shared_gen = atomic_read(&ceph_inode(dir)->i_shared_gen);
 
 	if (duration == 0)
 		goto out_unlock;
@@ -1080,6 +1080,27 @@ static struct dentry *splice_dentry(struct dentry *dn, struct inode *in)
 
 	BUG_ON(d_inode(dn));
 
+	if (S_ISDIR(in->i_mode)) {
+		/* If inode is directory, d_splice_alias() below will remove
+		 * 'realdn' from its origin parent. We need to ensure that
+		 * origin parent's readdir cache will not reference 'realdn'
+		 */
+		realdn = d_find_any_alias(in);
+		if (realdn) {
+			struct ceph_dentry_info *di = ceph_dentry(realdn);
+			spin_lock(&realdn->d_lock);
+
+			realdn->d_op->d_prune(realdn);
+
+			di->time = jiffies;
+			di->lease_shared_gen = 0;
+			di->offset = 0;
+
+			spin_unlock(&realdn->d_lock);
+			dput(realdn);
+		}
+	}
+
 	/* dn must be unhashed */
 	if (!d_unhashed(dn))
 		d_drop(dn);
@@ -1295,8 +1316,8 @@ retry_lookup:
 		if (!rinfo->head->is_target) {
 			dout("fill_trace null dentry\n");
 			if (d_really_is_positive(dn)) {
-				ceph_dir_clear_ordered(dir);
 				dout("d_delete %p\n", dn);
+				ceph_dir_clear_ordered(dir);
 				d_delete(dn);
 			} else if (have_lease) {
 				if (d_unhashed(dn))
@@ -1323,7 +1344,6 @@ retry_lookup:
 			dout(" %p links to %p %llx.%llx, not %llx.%llx\n",
 			     dn, d_inode(dn), ceph_vinop(d_inode(dn)),
 			     ceph_vinop(in));
-			ceph_dir_clear_ordered(dir);
 			d_invalidate(dn);
 			have_lease = false;
 		}
@@ -1573,9 +1593,19 @@ retry_lookup:
 		} else if (d_really_is_positive(dn) &&
 			   (ceph_ino(d_inode(dn)) != tvino.ino ||
 			    ceph_snap(d_inode(dn)) != tvino.snap)) {
+			struct ceph_dentry_info *di = ceph_dentry(dn);
 			dout(" dn %p points to wrong inode %p\n",
 			     dn, d_inode(dn));
-			__ceph_dir_clear_ordered(ci);
+
+			spin_lock(&dn->d_lock);
+			if (di->offset > 0 &&
+			    di->lease_shared_gen ==
+			    atomic_read(&ci->i_shared_gen)) {
+				__ceph_dir_clear_ordered(ci);
+				di->offset = 0;
+			}
+			spin_unlock(&dn->d_lock);
+
 			d_delete(dn);
 			dput(dn);
 			goto retry_lookup;
@@ -1600,9 +1630,7 @@ retry_lookup:
 				 &req->r_caps_reservation);
 		if (ret < 0) {
 			pr_err("fill_inode badness on %p\n", in);
-			if (d_really_is_positive(dn))
-				__ceph_dir_clear_ordered(ci);
-			else
+			if (d_really_is_negative(dn))
 				iput(in);
 			d_drop(dn);
 			err = ret;
@@ -2000,8 +2028,8 @@ int __ceph_setattr(struct inode *inode, struct iattr *attr)
 			ceph_encode_timespec(&req->r_args.setattr.atime,
 					     &attr->ia_atime);
 			mask |= CEPH_SETATTR_ATIME;
-			release |= CEPH_CAP_FILE_CACHE | CEPH_CAP_FILE_RD |
-				CEPH_CAP_FILE_WR;
+			release |= CEPH_CAP_FILE_SHARED |
+				   CEPH_CAP_FILE_RD | CEPH_CAP_FILE_WR;
 		}
 	}
 	if (ia_valid & ATTR_MTIME) {
@@ -2022,8 +2050,8 @@ int __ceph_setattr(struct inode *inode, struct iattr *attr)
 			ceph_encode_timespec(&req->r_args.setattr.mtime,
 					     &attr->ia_mtime);
 			mask |= CEPH_SETATTR_MTIME;
-			release |= CEPH_CAP_FILE_SHARED | CEPH_CAP_FILE_RD |
-				CEPH_CAP_FILE_WR;
+			release |= CEPH_CAP_FILE_SHARED |
+				   CEPH_CAP_FILE_RD | CEPH_CAP_FILE_WR;
 		}
 	}
 	if (ia_valid & ATTR_SIZE) {
@@ -2041,8 +2069,8 @@ int __ceph_setattr(struct inode *inode, struct iattr *attr)
 			req->r_args.setattr.old_size =
 				cpu_to_le64(inode->i_size);
 			mask |= CEPH_SETATTR_SIZE;
-			release |= CEPH_CAP_FILE_SHARED | CEPH_CAP_FILE_RD |
-				CEPH_CAP_FILE_WR;
+			release |= CEPH_CAP_FILE_SHARED | CEPH_CAP_FILE_EXCL |
+				   CEPH_CAP_FILE_RD | CEPH_CAP_FILE_WR;
 		}
 	}