1 files changed, 124 insertions, 116 deletions

diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index 5e659d054b40..220dfd87cbfa 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -499,7 +499,6 @@ struct inode *ceph_alloc_inode(struct super_block *sb)
 	ci->i_rdcache_gen = 0;
 	ci->i_rdcache_revoking = 0;
 
-	INIT_LIST_HEAD(&ci->i_unsafe_writes);
 	INIT_LIST_HEAD(&ci->i_unsafe_dirops);
 	INIT_LIST_HEAD(&ci->i_unsafe_iops);
 	spin_lock_init(&ci->i_unsafe_lock);
@@ -583,14 +582,6 @@ int ceph_drop_inode(struct inode *inode)
 	return 1;
 }
 
-void ceph_evict_inode(struct inode *inode)
-{
-	/* wait unsafe sync writes */
-	ceph_sync_write_wait(inode);
-	truncate_inode_pages_final(&inode->i_data);
-	clear_inode(inode);
-}
-
 static inline blkcnt_t calc_inode_blocks(u64 size)
 {
 	return (size + (1<<9) - 1) >> 9;
@@ -1016,21 +1007,38 @@ out:
 static void update_dentry_lease(struct dentry *dentry,
 				struct ceph_mds_reply_lease *lease,
 				struct ceph_mds_session *session,
-				unsigned long from_time)
+				unsigned long from_time,
+				struct ceph_vino *tgt_vino,
+				struct ceph_vino *dir_vino)
 {
 	struct ceph_dentry_info *di = ceph_dentry(dentry);
 	long unsigned duration = le32_to_cpu(lease->duration_ms);
 	long unsigned ttl = from_time + (duration * HZ) / 1000;
 	long unsigned half_ttl = from_time + (duration * HZ / 2) / 1000;
 	struct inode *dir;
+	struct ceph_mds_session *old_lease_session = NULL;
+
+	/*
+	 * Make sure dentry's inode matches tgt_vino. NULL tgt_vino means that
+	 * we expect a negative dentry.
+	 */
+	if (!tgt_vino && d_really_is_positive(dentry))
+		return;
+
+	if (tgt_vino && (d_really_is_negative(dentry) ||
+			!ceph_ino_compare(d_inode(dentry), tgt_vino)))
+		return;
 
 	spin_lock(&dentry->d_lock);
 	dout("update_dentry_lease %p duration %lu ms ttl %lu\n",
 	     dentry, duration, ttl);
 
-	/* make lease_rdcache_gen match directory */
 	dir = d_inode(dentry->d_parent);
 
+	/* make sure parent matches dir_vino */
+	if (!ceph_ino_compare(dir, dir_vino))
+		goto out_unlock;
+
 	/* only track leases on regular dentries */
 	if (ceph_snap(dir) != CEPH_NOSNAP)
 		goto out_unlock;
@@ -1044,8 +1052,10 @@ static void update_dentry_lease(struct dentry *dentry,
 	    time_before(ttl, di->time))
 		goto out_unlock;  /* we already have a newer lease. */
 
-	if (di->lease_session && di->lease_session != session)
-		goto out_unlock;
+	if (di->lease_session && di->lease_session != session) {
+		old_lease_session = di->lease_session;
+		di->lease_session = NULL;
+	}
 
 	ceph_dentry_lru_touch(dentry);
 
@@ -1058,6 +1068,8 @@ static void update_dentry_lease(struct dentry *dentry,
 	di->time = ttl;
 out_unlock:
 	spin_unlock(&dentry->d_lock);
+	if (old_lease_session)
+		ceph_put_mds_session(old_lease_session);
 	return;
 }
 
@@ -1108,61 +1120,27 @@ out:
  *
  * Called with snap_rwsem (read).
  */
-int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req,
-		    struct ceph_mds_session *session)
+int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req)
 {
+	struct ceph_mds_session *session = req->r_session;
 	struct ceph_mds_reply_info_parsed *rinfo = &req->r_reply_info;
 	struct inode *in = NULL;
-	struct ceph_vino vino;
+	struct ceph_vino tvino, dvino;
 	struct ceph_fs_client *fsc = ceph_sb_to_client(sb);
 	int err = 0;
 
 	dout("fill_trace %p is_dentry %d is_target %d\n", req,
 	     rinfo->head->is_dentry, rinfo->head->is_target);
 
-#if 0
-	/*
-	 * Debugging hook:
-	 *
-	 * If we resend completed ops to a recovering mds, we get no
-	 * trace.  Since that is very rare, pretend this is the case
-	 * to ensure the 'no trace' handlers in the callers behave.
-	 *
-	 * Fill in inodes unconditionally to avoid breaking cap
-	 * invariants.
-	 */
-	if (rinfo->head->op & CEPH_MDS_OP_WRITE) {
-		pr_info("fill_trace faking empty trace on %lld %s\n",
-			req->r_tid, ceph_mds_op_name(rinfo->head->op));
-		if (rinfo->head->is_dentry) {
-			rinfo->head->is_dentry = 0;
-			err = fill_inode(req->r_locked_dir,
-					 &rinfo->diri, rinfo->dirfrag,
-					 session, req->r_request_started, -1);
-		}
-		if (rinfo->head->is_target) {
-			rinfo->head->is_target = 0;
-			ininfo = rinfo->targeti.in;
-			vino.ino = le64_to_cpu(ininfo->ino);
-			vino.snap = le64_to_cpu(ininfo->snapid);
-			in = ceph_get_inode(sb, vino);
-			err = fill_inode(in, &rinfo->targeti, NULL,
-					 session, req->r_request_started,
-					 req->r_fmode);
-			iput(in);
-		}
-	}
-#endif
-
 	if (!rinfo->head->is_target && !rinfo->head->is_dentry) {
 		dout("fill_trace reply is empty!\n");
-		if (rinfo->head->result == 0 && req->r_locked_dir)
+		if (rinfo->head->result == 0 && req->r_parent)
 			ceph_invalidate_dir_request(req);
 		return 0;
 	}
 
 	if (rinfo->head->is_dentry) {
-		struct inode *dir = req->r_locked_dir;
+		struct inode *dir = req->r_parent;
 
 		if (dir) {
 			err = fill_inode(dir, NULL,
@@ -1188,8 +1166,8 @@ int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req,
 			dname.name = rinfo->dname;
 			dname.len = rinfo->dname_len;
 			dname.hash = full_name_hash(parent, dname.name, dname.len);
-			vino.ino = le64_to_cpu(rinfo->targeti.in->ino);
-			vino.snap = le64_to_cpu(rinfo->targeti.in->snapid);
+			tvino.ino = le64_to_cpu(rinfo->targeti.in->ino);
+			tvino.snap = le64_to_cpu(rinfo->targeti.in->snapid);
 retry_lookup:
 			dn = d_lookup(parent, &dname);
 			dout("d_lookup on parent=%p name=%.*s got %p\n",
@@ -1206,8 +1184,8 @@ retry_lookup:
 				}
 				err = 0;
 			} else if (d_really_is_positive(dn) &&
-				   (ceph_ino(d_inode(dn)) != vino.ino ||
-				    ceph_snap(d_inode(dn)) != vino.snap)) {
+				   (ceph_ino(d_inode(dn)) != tvino.ino ||
+				    ceph_snap(d_inode(dn)) != tvino.snap)) {
 				dout(" dn %p points to wrong inode %p\n",
 				     dn, d_inode(dn));
 				d_delete(dn);
@@ -1221,10 +1199,10 @@ retry_lookup:
 	}
 
 	if (rinfo->head->is_target) {
-		vino.ino = le64_to_cpu(rinfo->targeti.in->ino);
-		vino.snap = le64_to_cpu(rinfo->targeti.in->snapid);
+		tvino.ino = le64_to_cpu(rinfo->targeti.in->ino);
+		tvino.snap = le64_to_cpu(rinfo->targeti.in->snapid);
 
-		in = ceph_get_inode(sb, vino);
+		in = ceph_get_inode(sb, tvino);
 		if (IS_ERR(in)) {
 			err = PTR_ERR(in);
 			goto done;
@@ -1233,8 +1211,8 @@ retry_lookup:
 
 		err = fill_inode(in, req->r_locked_page, &rinfo->targeti, NULL,
 				session, req->r_request_started,
-				(!req->r_aborted && rinfo->head->result == 0) ?
-				req->r_fmode : -1,
+				(!test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags) &&
+				rinfo->head->result == 0) ?  req->r_fmode : -1,
 				&req->r_caps_reservation);
 		if (err < 0) {
 			pr_err("fill_inode badness %p %llx.%llx\n",
@@ -1247,8 +1225,9 @@ retry_lookup:
 	 * ignore null lease/binding on snapdir ENOENT, or else we
 	 * will have trouble splicing in the virtual snapdir later
 	 */
-	if (rinfo->head->is_dentry && !req->r_aborted &&
-	    req->r_locked_dir &&
+	if (rinfo->head->is_dentry &&
+            !test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags) &&
+	    test_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags) &&
 	    (rinfo->head->is_target || strncmp(req->r_dentry->d_name.name,
 					       fsc->mount_options->snapdir_name,
 					       req->r_dentry->d_name.len))) {
@@ -1257,17 +1236,19 @@ retry_lookup:
 		 * mknod symlink mkdir  : null -> new inode
 		 * unlink               : linked -> null
 		 */
-		struct inode *dir = req->r_locked_dir;
+		struct inode *dir = req->r_parent;
 		struct dentry *dn = req->r_dentry;
 		bool have_dir_cap, have_lease;
 
 		BUG_ON(!dn);
 		BUG_ON(!dir);
 		BUG_ON(d_inode(dn->d_parent) != dir);
-		BUG_ON(ceph_ino(dir) !=
-		       le64_to_cpu(rinfo->diri.in->ino));
-		BUG_ON(ceph_snap(dir) !=
-		       le64_to_cpu(rinfo->diri.in->snapid));
+
+		dvino.ino = le64_to_cpu(rinfo->diri.in->ino);
+		dvino.snap = le64_to_cpu(rinfo->diri.in->snapid);
+
+		BUG_ON(ceph_ino(dir) != dvino.ino);
+		BUG_ON(ceph_snap(dir) != dvino.snap);
 
 		/* do we have a lease on the whole dir? */
 		have_dir_cap =
@@ -1319,12 +1300,13 @@ retry_lookup:
 				ceph_dir_clear_ordered(dir);
 				dout("d_delete %p\n", dn);
 				d_delete(dn);
-			} else {
-				if (have_lease && d_unhashed(dn))
+			} else if (have_lease) {
+				if (d_unhashed(dn))
 					d_add(dn, NULL);
 				update_dentry_lease(dn, rinfo->dlease,
 						    session,
-						    req->r_request_started);
+						    req->r_request_started,
+						    NULL, &dvino);
 			}
 			goto done;
 		}
@@ -1347,15 +1329,19 @@ retry_lookup:
 			have_lease = false;
 		}
 
-		if (have_lease)
+		if (have_lease) {
+			tvino.ino = le64_to_cpu(rinfo->targeti.in->ino);
+			tvino.snap = le64_to_cpu(rinfo->targeti.in->snapid);
 			update_dentry_lease(dn, rinfo->dlease, session,
-					    req->r_request_started);
+					    req->r_request_started,
+					    &tvino, &dvino);
+		}
 		dout(" final dn %p\n", dn);
-	} else if (!req->r_aborted &&
-		   (req->r_op == CEPH_MDS_OP_LOOKUPSNAP ||
-		    req->r_op == CEPH_MDS_OP_MKSNAP)) {
+	} else if ((req->r_op == CEPH_MDS_OP_LOOKUPSNAP ||
+		    req->r_op == CEPH_MDS_OP_MKSNAP) &&
+		   !test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags)) {
 		struct dentry *dn = req->r_dentry;
-		struct inode *dir = req->r_locked_dir;
+		struct inode *dir = req->r_parent;
 
 		/* fill out a snapdir LOOKUPSNAP dentry */
 		BUG_ON(!dn);
@@ -1370,6 +1356,26 @@ retry_lookup:
 			goto done;
 		}
 		req->r_dentry = dn;  /* may have spliced */
+	} else if (rinfo->head->is_dentry) {
+		struct ceph_vino *ptvino = NULL;
+
+		if ((le32_to_cpu(rinfo->diri.in->cap.caps) & CEPH_CAP_FILE_SHARED) ||
+		    le32_to_cpu(rinfo->dlease->duration_ms)) {
+			dvino.ino = le64_to_cpu(rinfo->diri.in->ino);
+			dvino.snap = le64_to_cpu(rinfo->diri.in->snapid);
+
+			if (rinfo->head->is_target) {
+				tvino.ino = le64_to_cpu(rinfo->targeti.in->ino);
+				tvino.snap = le64_to_cpu(rinfo->targeti.in->snapid);
+				ptvino = &tvino;
+			}
+
+			update_dentry_lease(req->r_dentry, rinfo->dlease,
+				session, req->r_request_started, ptvino,
+				&dvino);
+		} else {
+			dout("%s: no dentry lease or dir cap\n", __func__);
+		}
 	}
 done:
 	dout("fill_trace done err=%d\n", err);
@@ -1478,13 +1484,20 @@ int ceph_readdir_prepopulate(struct ceph_mds_request *req,
 	u32 fpos_offset;
 	struct ceph_readdir_cache_control cache_ctl = {};
 
-	if (req->r_aborted)
+	if (test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags))
 		return readdir_prepopulate_inodes_only(req, session);
 
-	if (rinfo->hash_order && req->r_path2) {
-		last_hash = ceph_str_hash(ci->i_dir_layout.dl_dir_hash,
-					  req->r_path2, strlen(req->r_path2));
-		last_hash = ceph_frag_value(last_hash);
+	if (rinfo->hash_order) {
+		if (req->r_path2) {
+			last_hash = ceph_str_hash(ci->i_dir_layout.dl_dir_hash,
+						  req->r_path2,
+						  strlen(req->r_path2));
+			last_hash = ceph_frag_value(last_hash);
+		} else if (rinfo->offset_hash) {
+			/* mds understands offset_hash */
+			WARN_ON_ONCE(req->r_readdir_offset != 2);
+			last_hash = le32_to_cpu(rhead->args.readdir.offset_hash);
+		}
 	}
 
 	if (rinfo->dir_dir &&
@@ -1509,7 +1522,7 @@ int ceph_readdir_prepopulate(struct ceph_mds_request *req,
 	}
 
 	if (ceph_frag_is_leftmost(frag) && req->r_readdir_offset == 2 &&
-	    !(rinfo->hash_order && req->r_path2)) {
+	    !(rinfo->hash_order && last_hash)) {
 		/* note dir version at start of readdir so we can tell
 		 * if any dentries get dropped */
 		req->r_dir_release_cnt = atomic64_read(&ci->i_release_count);
@@ -1523,14 +1536,14 @@ int ceph_readdir_prepopulate(struct ceph_mds_request *req,
 	/* FIXME: release caps/leases if error occurs */
 	for (i = 0; i < rinfo->dir_nr; i++) {
 		struct ceph_mds_reply_dir_entry *rde = rinfo->dir_entries + i;
-		struct ceph_vino vino;
+		struct ceph_vino tvino, dvino;
 
 		dname.name = rde->name;
 		dname.len = rde->name_len;
 		dname.hash = full_name_hash(parent, dname.name, dname.len);
 
-		vino.ino = le64_to_cpu(rde->inode.in->ino);
-		vino.snap = le64_to_cpu(rde->inode.in->snapid);
+		tvino.ino = le64_to_cpu(rde->inode.in->ino);
+		tvino.snap = le64_to_cpu(rde->inode.in->snapid);
 
 		if (rinfo->hash_order) {
 			u32 hash = ceph_str_hash(ci->i_dir_layout.dl_dir_hash,
@@ -1559,8 +1572,8 @@ retry_lookup:
 				goto out;
 			}
 		} else if (d_really_is_positive(dn) &&
-			   (ceph_ino(d_inode(dn)) != vino.ino ||
-			    ceph_snap(d_inode(dn)) != vino.snap)) {
+			   (ceph_ino(d_inode(dn)) != tvino.ino ||
+			    ceph_snap(d_inode(dn)) != tvino.snap)) {
 			dout(" dn %p points to wrong inode %p\n",
 			     dn, d_inode(dn));
 			d_delete(dn);
@@ -1572,7 +1585,7 @@ retry_lookup:
 		if (d_really_is_positive(dn)) {
 			in = d_inode(dn);
 		} else {
-			in = ceph_get_inode(parent->d_sb, vino);
+			in = ceph_get_inode(parent->d_sb, tvino);
 			if (IS_ERR(in)) {
 				dout("new_inode badness\n");
 				d_drop(dn);
@@ -1617,8 +1630,9 @@ retry_lookup:
 
 		ceph_dentry(dn)->offset = rde->offset;
 
+		dvino = ceph_vino(d_inode(parent));
 		update_dentry_lease(dn, rde->lease, req->r_session,
-				    req->r_request_started);
+				    req->r_request_started, &tvino, &dvino);
 
 		if (err == 0 && skipped == 0 && cache_ctl.index >= 0) {
 			ret = fill_readdir_cache(d_inode(parent), dn,
@@ -1632,7 +1646,7 @@ next_item:
 	}
 out:
 	if (err == 0 && skipped == 0) {
-		req->r_did_prepopulate = true;
+		set_bit(CEPH_MDS_R_DID_PREPOPULATE, &req->r_req_flags);
 		req->r_readdir_cache_idx = cache_ctl.index;
 	}
 	ceph_readdir_cache_release(&cache_ctl);
@@ -1644,20 +1658,17 @@ out:
 	return err;
 }
 
-int ceph_inode_set_size(struct inode *inode, loff_t size)
+bool ceph_inode_set_size(struct inode *inode, loff_t size)
 {
 	struct ceph_inode_info *ci = ceph_inode(inode);
-	int ret = 0;
+	bool ret;
 
 	spin_lock(&ci->i_ceph_lock);
 	dout("set_size %p %llu -> %llu\n", inode, inode->i_size, size);
 	i_size_write(inode, size);
 	inode->i_blocks = calc_inode_blocks(size);
 
-	/* tell the MDS if we are approaching max_size */
-	if ((size << 1) >= ci->i_max_size &&
-	    (ci->i_reported_size << 1) < ci->i_max_size)
-		ret = 1;
+	ret = __ceph_should_report_size(ci);
 
 	spin_unlock(&ci->i_ceph_lock);
 	return ret;
@@ -1720,7 +1731,7 @@ static void ceph_invalidate_work(struct work_struct *work)
 
 	mutex_lock(&ci->i_truncate_mutex);
 
-	if (ACCESS_ONCE(fsc->mount_state) == CEPH_MOUNT_SHUTDOWN) {
+	if (READ_ONCE(fsc->mount_state) == CEPH_MOUNT_SHUTDOWN) {
 		pr_warn_ratelimited("invalidate_pages %p %lld forced umount\n",
 				    inode, ceph_ino(inode));
 		mapping_set_error(inode->i_mapping, -EIO);
@@ -2013,7 +2024,6 @@ int __ceph_setattr(struct inode *inode, struct iattr *attr)
 		    attr->ia_size > inode->i_size) {
 			i_size_write(inode, attr->ia_size);
 			inode->i_blocks = calc_inode_blocks(attr->ia_size);
-			inode->i_ctime = attr->ia_ctime;
 			ci->i_reported_size = attr->ia_size;
 			dirtied |= CEPH_CAP_FILE_EXCL;
 		} else if ((issued & CEPH_CAP_FILE_SHARED) == 0 ||
@@ -2035,7 +2045,6 @@ int __ceph_setattr(struct inode *inode, struct iattr *attr)
 		     inode->i_ctime.tv_sec, inode->i_ctime.tv_nsec,
 		     attr->ia_ctime.tv_sec, attr->ia_ctime.tv_nsec,
 		     only ? "ctime only" : "ignored");
-		inode->i_ctime = attr->ia_ctime;
 		if (only) {
 			/*
 			 * if kernel wants to dirty ctime but nothing else,
@@ -2058,7 +2067,7 @@ int __ceph_setattr(struct inode *inode, struct iattr *attr)
 	if (dirtied) {
 		inode_dirty_flags = __ceph_mark_dirty_caps(ci, dirtied,
 							   &prealloc_cf);
-		inode->i_ctime = current_time(inode);
+		inode->i_ctime = attr->ia_ctime;
 	}
 
 	release &= issued;
@@ -2069,11 +2078,6 @@ int __ceph_setattr(struct inode *inode, struct iattr *attr)
 	if (inode_dirty_flags)
 		__mark_inode_dirty(inode, inode_dirty_flags);
 
-	if (ia_valid & ATTR_MODE) {
-		err = posix_acl_chmod(inode, attr->ia_mode);
-		if (err)
-			goto out_put;
-	}
 
 	if (mask) {
 		req->r_inode = inode;
@@ -2081,19 +2085,18 @@ int __ceph_setattr(struct inode *inode, struct iattr *attr)
 		req->r_inode_drop = release;
 		req->r_args.setattr.mask = cpu_to_le32(mask);
 		req->r_num_caps = 1;
+		req->r_stamp = attr->ia_ctime;
 		err = ceph_mdsc_do_request(mdsc, NULL, req);
 	}
 	dout("setattr %p result=%d (%s locally, %d remote)\n", inode, err,
 	     ceph_cap_string(dirtied), mask);
 
 	ceph_mdsc_put_request(req);
-	if (mask & CEPH_SETATTR_SIZE)
-		__ceph_do_pending_vmtruncate(inode);
-	ceph_free_cap_flush(prealloc_cf);
-	return err;
-out_put:
-	ceph_mdsc_put_request(req);
 	ceph_free_cap_flush(prealloc_cf);
+
+	if (err >= 0 && (mask & CEPH_SETATTR_SIZE))
+		__ceph_do_pending_vmtruncate(inode);
+
 	return err;
 }
 
@@ -2112,7 +2115,12 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr)
 	if (err != 0)
 		return err;
 
-	return __ceph_setattr(inode, attr);
+	err = __ceph_setattr(inode, attr);
+
+	if (err >= 0 && (attr->ia_valid & ATTR_MODE))
+		err = posix_acl_chmod(inode, attr->ia_mode);
+
+	return err;
 }
 
 /*
@@ -2185,10 +2193,10 @@ int ceph_permission(struct inode *inode, int mask)
  * Get all attributes.  Hopefully somedata we'll have a statlite()
  * and can limit the fields we require to be accurate.
  */
-int ceph_getattr(struct vfsmount *mnt, struct dentry *dentry,
-		 struct kstat *stat)
+int ceph_getattr(const struct path *path, struct kstat *stat,
+		 u32 request_mask, unsigned int flags)
 {
-	struct inode *inode = d_inode(dentry);
+	struct inode *inode = d_inode(path->dentry);
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	int err;