Merge tag 'for-4.19-rc2-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs fixes from David Sterba:

 - fix for improper fsync after hardlink

 - fix for a corruption during file deduplication

 - use after free fixes

 - RCU warning fix

 - fix for buffered write to nodatacow file

* tag 'for-4.19-rc2-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
  btrfs: Fix suspicious RCU usage warning in btrfs_debug_in_rcu
  btrfs: use after free in btrfs_quota_enable
  btrfs: btrfs_shrink_device should call commit transaction at the end
  btrfs: fix qgroup_free wrong num_bytes in btrfs_subvolume_reserve_metadata
  Btrfs: fix data corruption when deduplicating between different files
  Btrfs: sync log after logging new name
  Btrfs: fix unexpected failure of nocow buffered writes after snapshotting when low on space

fs/btrfs/ctree.h

@@ -1280,6 +1280,7 @@ struct btrfs_root {
 	int send_in_progress;
 	struct btrfs_subvolume_writers *subv_writers;
 	atomic_t will_be_snapshotted;
+	atomic_t snapshot_force_cow;
 
 	/* For qgroup metadata reserved space */
 	spinlock_t qgroup_meta_rsv_lock;
@@ -3390,9 +3391,9 @@ do {									\
 #define btrfs_debug(fs_info, fmt, args...) \
 	btrfs_no_printk(fs_info, KERN_DEBUG fmt, ##args)
 #define btrfs_debug_in_rcu(fs_info, fmt, args...) \
-	btrfs_no_printk(fs_info, KERN_DEBUG fmt, ##args)
+	btrfs_no_printk_in_rcu(fs_info, KERN_DEBUG fmt, ##args)
 #define btrfs_debug_rl_in_rcu(fs_info, fmt, args...) \
-	btrfs_no_printk(fs_info, KERN_DEBUG fmt, ##args)
+	btrfs_no_printk_in_rcu(fs_info, KERN_DEBUG fmt, ##args)
 #define btrfs_debug_rl(fs_info, fmt, args...) \
 	btrfs_no_printk(fs_info, KERN_DEBUG fmt, ##args)
 #endif
@@ -3404,6 +3405,13 @@ do {							\
 	rcu_read_unlock();				\
 } while (0)
 
+#define btrfs_no_printk_in_rcu(fs_info, fmt, args...)	\
+do {							\
+	rcu_read_lock();				\
+	btrfs_no_printk(fs_info, fmt, ##args);		\
+	rcu_read_unlock();				\
+} while (0)
+
 #define btrfs_printk_ratelimited(fs_info, fmt, args...)		\
 do {								\
 	static DEFINE_RATELIMIT_STATE(_rs,			\

fs/btrfs/disk-io.c

@@ -1187,6 +1187,7 @@ static void __setup_root(struct btrfs_root *root, struct btrfs_fs_info *fs_info,
 	atomic_set(&root->log_batch, 0);
 	refcount_set(&root->refs, 1);
 	atomic_set(&root->will_be_snapshotted, 0);
+	atomic_set(&root->snapshot_force_cow, 0);
 	root->log_transid = 0;
 	root->log_transid_committed = -1;
 	root->last_log_commit = 0;

fs/btrfs/extent-tree.c

@@ -5800,7 +5800,7 @@ void btrfs_trans_release_chunk_metadata(struct btrfs_trans_handle *trans)
  * root: the root of the parent directory
  * rsv: block reservation
  * items: the number of items that we need do reservation
- * qgroup_reserved: used to return the reserved size in qgroup
+ * use_global_rsv: allow fallback to the global block reservation
  *
  * This function is used to reserve the space for snapshot/subvolume
  * creation and deletion. Those operations are different with the
@@ -5810,10 +5810,10 @@ void btrfs_trans_release_chunk_metadata(struct btrfs_trans_handle *trans)
  * the space reservation mechanism in start_transaction().
  */
 int btrfs_subvolume_reserve_metadata(struct btrfs_root *root,
-				     struct btrfs_block_rsv *rsv,
-				     int items,
+				     struct btrfs_block_rsv *rsv, int items,
 				     bool use_global_rsv)
 {
+	u64 qgroup_num_bytes = 0;
 	u64 num_bytes;
 	int ret;
 	struct btrfs_fs_info *fs_info = root->fs_info;
@@ -5821,12 +5821,11 @@ int btrfs_subvolume_reserve_metadata(struct btrfs_root *root,
 
 	if (test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) {
 		/* One for parent inode, two for dir entries */
-		num_bytes = 3 * fs_info->nodesize;
-		ret = btrfs_qgroup_reserve_meta_prealloc(root, num_bytes, true);
+		qgroup_num_bytes = 3 * fs_info->nodesize;
+		ret = btrfs_qgroup_reserve_meta_prealloc(root,
+				qgroup_num_bytes, true);
 		if (ret)
 			return ret;
-	} else {
-		num_bytes = 0;
 	}
 
 	num_bytes = btrfs_calc_trans_metadata_size(fs_info, items);
@@ -5838,8 +5837,8 @@ int btrfs_subvolume_reserve_metadata(struct btrfs_root *root,
 	if (ret == -ENOSPC && use_global_rsv)
 		ret = btrfs_block_rsv_migrate(global_rsv, rsv, num_bytes, 1);
 
-	if (ret && num_bytes)
-		btrfs_qgroup_free_meta_prealloc(root, num_bytes);
+	if (ret && qgroup_num_bytes)
+		btrfs_qgroup_free_meta_prealloc(root, qgroup_num_bytes);
 
 	return ret;
 }

fs/btrfs/inode.c

@@ -1271,7 +1271,7 @@ static noinline int run_delalloc_nocow(struct inode *inode,
 	u64 disk_num_bytes;
 	u64 ram_bytes;
 	int extent_type;
-	int ret, err;
+	int ret;
 	int type;
 	int nocow;
 	int check_prev = 1;
@@ -1403,11 +1403,8 @@ next_slot:
 			 * if there are pending snapshots for this root,
 			 * we fall into common COW way.
 			 */
-			if (!nolock) {
-				err = btrfs_start_write_no_snapshotting(root);
-				if (!err)
-					goto out_check;
-			}
+			if (!nolock && atomic_read(&root->snapshot_force_cow))
+				goto out_check;
 			/*
 			 * force cow if csum exists in the range.
 			 * this ensure that csum for a given extent are
@@ -1416,9 +1413,6 @@ next_slot:
 			ret = csum_exist_in_range(fs_info, disk_bytenr,
 						  num_bytes);
 			if (ret) {
-				if (!nolock)
-					btrfs_end_write_no_snapshotting(root);
-
 				/*
 				 * ret could be -EIO if the above fails to read
 				 * metadata.
@@ -1431,11 +1425,8 @@ next_slot:
 				WARN_ON_ONCE(nolock);
 				goto out_check;
 			}
-			if (!btrfs_inc_nocow_writers(fs_info, disk_bytenr)) {
-				if (!nolock)
-					btrfs_end_write_no_snapshotting(root);
+			if (!btrfs_inc_nocow_writers(fs_info, disk_bytenr))
 				goto out_check;
-			}
 			nocow = 1;
 		} else if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
 			extent_end = found_key.offset +
@@ -1448,8 +1439,6 @@ next_slot:
 out_check:
 		if (extent_end <= start) {
 			path->slots[0]++;
-			if (!nolock && nocow)
-				btrfs_end_write_no_snapshotting(root);
 			if (nocow)
 				btrfs_dec_nocow_writers(fs_info, disk_bytenr);
 			goto next_slot;
@@ -1471,8 +1460,6 @@ out_check:
 					     end, page_started, nr_written, 1,
 					     NULL);
 			if (ret) {
-				if (!nolock && nocow)
-					btrfs_end_write_no_snapshotting(root);
 				if (nocow)
 					btrfs_dec_nocow_writers(fs_info,
 								disk_bytenr);
@@ -1492,8 +1479,6 @@ out_check:
 					  ram_bytes, BTRFS_COMPRESS_NONE,
 					  BTRFS_ORDERED_PREALLOC);
 			if (IS_ERR(em)) {
-				if (!nolock && nocow)
-					btrfs_end_write_no_snapshotting(root);
 				if (nocow)
 					btrfs_dec_nocow_writers(fs_info,
 								disk_bytenr);
@@ -1532,8 +1517,6 @@ out_check:
 					     EXTENT_CLEAR_DATA_RESV,
 					     PAGE_UNLOCK | PAGE_SET_PRIVATE2);
 
-		if (!nolock && nocow)
-			btrfs_end_write_no_snapshotting(root);
 		cur_offset = extent_end;
 
 		/*
@@ -6639,6 +6622,8 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir,
 		drop_inode = 1;
 	} else {
 		struct dentry *parent = dentry->d_parent;
+		int ret;
+
 		err = btrfs_update_inode(trans, root, inode);
 		if (err)
 			goto fail;
@@ -6652,7 +6637,12 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir,
 				goto fail;
 		}
 		d_instantiate(dentry, inode);
-		btrfs_log_new_name(trans, BTRFS_I(inode), NULL, parent);
+		ret = btrfs_log_new_name(trans, BTRFS_I(inode), NULL, parent,
+					 true, NULL);
+		if (ret == BTRFS_NEED_TRANS_COMMIT) {
+			err = btrfs_commit_transaction(trans);
+			trans = NULL;
+		}
 	}
 
 fail:
@@ -9388,14 +9378,21 @@ static int btrfs_rename_exchange(struct inode *old_dir,
 	u64 new_idx = 0;
 	u64 root_objectid;
 	int ret;
-	int ret2;
 	bool root_log_pinned = false;
 	bool dest_log_pinned = false;
+	struct btrfs_log_ctx ctx_root;
+	struct btrfs_log_ctx ctx_dest;
+	bool sync_log_root = false;
+	bool sync_log_dest = false;
+	bool commit_transaction = false;
 
 	/* we only allow rename subvolume link between subvolumes */
 	if (old_ino != BTRFS_FIRST_FREE_OBJECTID && root != dest)
 		return -EXDEV;
 
+	btrfs_init_log_ctx(&ctx_root, old_inode);
+	btrfs_init_log_ctx(&ctx_dest, new_inode);
+
 	/* close the race window with snapshot create/destroy ioctl */
 	if (old_ino == BTRFS_FIRST_FREE_OBJECTID)
 		down_read(&fs_info->subvol_sem);
@@ -9542,15 +9539,29 @@ static int btrfs_rename_exchange(struct inode *old_dir,
 
 	if (root_log_pinned) {
 		parent = new_dentry->d_parent;
-		btrfs_log_new_name(trans, BTRFS_I(old_inode), BTRFS_I(old_dir),
-				parent);
+		ret = btrfs_log_new_name(trans, BTRFS_I(old_inode),
+					 BTRFS_I(old_dir), parent,
+					 false, &ctx_root);
+		if (ret == BTRFS_NEED_LOG_SYNC)
+			sync_log_root = true;
+		else if (ret == BTRFS_NEED_TRANS_COMMIT)
+			commit_transaction = true;
+		ret = 0;
 		btrfs_end_log_trans(root);
 		root_log_pinned = false;
 	}
 	if (dest_log_pinned) {
-		parent = old_dentry->d_parent;
-		btrfs_log_new_name(trans, BTRFS_I(new_inode), BTRFS_I(new_dir),
-				parent);
+		if (!commit_transaction) {
+			parent = old_dentry->d_parent;
+			ret = btrfs_log_new_name(trans, BTRFS_I(new_inode),
+						 BTRFS_I(new_dir), parent,
+						 false, &ctx_dest);
+			if (ret == BTRFS_NEED_LOG_SYNC)
+				sync_log_dest = true;
+			else if (ret == BTRFS_NEED_TRANS_COMMIT)
+				commit_transaction = true;
+			ret = 0;
+		}
 		btrfs_end_log_trans(dest);
 		dest_log_pinned = false;
 	}
@@ -9583,8 +9594,26 @@ out_fail:
 			dest_log_pinned = false;
 		}
 	}
-	ret2 = btrfs_end_transaction(trans);
-	ret = ret ? ret : ret2;
+	if (!ret && sync_log_root && !commit_transaction) {
+		ret = btrfs_sync_log(trans, BTRFS_I(old_inode)->root,
+				     &ctx_root);
+		if (ret)
+			commit_transaction = true;
+	}
+	if (!ret && sync_log_dest && !commit_transaction) {
+		ret = btrfs_sync_log(trans, BTRFS_I(new_inode)->root,
+				     &ctx_dest);
+		if (ret)
+			commit_transaction = true;
+	}
+	if (commit_transaction) {
+		ret = btrfs_commit_transaction(trans);
+	} else {
+		int ret2;
+
+		ret2 = btrfs_end_transaction(trans);
+		ret = ret ? ret : ret2;
+	}
 out_notrans:
 	if (new_ino == BTRFS_FIRST_FREE_OBJECTID)
 		up_read(&fs_info->subvol_sem);
@@ -9661,6 +9690,9 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	int ret;
 	u64 old_ino = btrfs_ino(BTRFS_I(old_inode));
 	bool log_pinned = false;
+	struct btrfs_log_ctx ctx;
+	bool sync_log = false;
+	bool commit_transaction = false;
 
 	if (btrfs_ino(BTRFS_I(new_dir)) == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)
 		return -EPERM;
@@ -9818,8 +9850,15 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	if (log_pinned) {
 		struct dentry *parent = new_dentry->d_parent;
 
-		btrfs_log_new_name(trans, BTRFS_I(old_inode), BTRFS_I(old_dir),
-				parent);
+		btrfs_init_log_ctx(&ctx, old_inode);
+		ret = btrfs_log_new_name(trans, BTRFS_I(old_inode),
+					 BTRFS_I(old_dir), parent,
+					 false, &ctx);
+		if (ret == BTRFS_NEED_LOG_SYNC)
+			sync_log = true;
+		else if (ret == BTRFS_NEED_TRANS_COMMIT)
+			commit_transaction = true;
+		ret = 0;
 		btrfs_end_log_trans(root);
 		log_pinned = false;
 	}
@@ -9856,7 +9895,19 @@ out_fail:
 		btrfs_end_log_trans(root);
 		log_pinned = false;
 	}
-	btrfs_end_transaction(trans);
+	if (!ret && sync_log) {
+		ret = btrfs_sync_log(trans, BTRFS_I(old_inode)->root, &ctx);
+		if (ret)
+			commit_transaction = true;
+	}
+	if (commit_transaction) {
+		ret = btrfs_commit_transaction(trans);
+	} else {
+		int ret2;
+
+		ret2 = btrfs_end_transaction(trans);
+		ret = ret ? ret : ret2;
+	}
 out_notrans:
 	if (old_ino == BTRFS_FIRST_FREE_OBJECTID)
 		up_read(&fs_info->subvol_sem);

fs/btrfs/ioctl.c

@@ -747,6 +747,7 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir,
 	struct btrfs_pending_snapshot *pending_snapshot;
 	struct btrfs_trans_handle *trans;
 	int ret;
+	bool snapshot_force_cow = false;
 
 	if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state))
 		return -EINVAL;
@@ -763,6 +764,11 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir,
 		goto free_pending;
 	}
 
+	/*
+	 * Force new buffered writes to reserve space even when NOCOW is
+	 * possible. This is to avoid later writeback (running dealloc) to
+	 * fallback to COW mode and unexpectedly fail with ENOSPC.
+	 */
 	atomic_inc(&root->will_be_snapshotted);
 	smp_mb__after_atomic();
 	/* wait for no snapshot writes */
@@ -773,6 +779,14 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir,
 	if (ret)
 		goto dec_and_free;
 
+	/*
+	 * All previous writes have started writeback in NOCOW mode, so now
+	 * we force future writes to fallback to COW mode during snapshot
+	 * creation.
+	 */
+	atomic_inc(&root->snapshot_force_cow);
+	snapshot_force_cow = true;
+
 	btrfs_wait_ordered_extents(root, U64_MAX, 0, (u64)-1);
 
 	btrfs_init_block_rsv(&pending_snapshot->block_rsv,
@@ -837,6 +851,8 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir,
 fail:
 	btrfs_subvolume_release_metadata(fs_info, &pending_snapshot->block_rsv);
 dec_and_free:
+	if (snapshot_force_cow)
+		atomic_dec(&root->snapshot_force_cow);
 	if (atomic_dec_and_test(&root->will_be_snapshotted))
 		wake_up_var(&root->will_be_snapshotted);
 free_pending:
@@ -3453,6 +3469,25 @@ static int btrfs_extent_same_range(struct inode *src, u64 loff, u64 olen,
 
 		same_lock_start = min_t(u64, loff, dst_loff);
 		same_lock_len = max_t(u64, loff, dst_loff) + len - same_lock_start;
+	} else {
+		/*
+		 * If the source and destination inodes are different, the
+		 * source's range end offset matches the source's i_size, that
+		 * i_size is not a multiple of the sector size, and the
+		 * destination range does not go past the destination's i_size,
+		 * we must round down the length to the nearest sector size
+		 * multiple. If we don't do this adjustment we end replacing
+		 * with zeroes the bytes in the range that starts at the
+		 * deduplication range's end offset and ends at the next sector
+		 * size multiple.
+		 */
+		if (loff + olen == i_size_read(src) &&
+		    dst_loff + len < i_size_read(dst)) {
+			const u64 sz = BTRFS_I(src)->root->fs_info->sectorsize;
+
+			len = round_down(i_size_read(src), sz) - loff;
+			olen = len;
+		}
 	}
 
 again:

fs/btrfs/qgroup.c

@@ -1019,10 +1019,9 @@ out_add_root:
 	spin_unlock(&fs_info->qgroup_lock);
 
 	ret = btrfs_commit_transaction(trans);
-	if (ret) {
-		trans = NULL;
+	trans = NULL;
+	if (ret)
 		goto out_free_path;
-	}
 
 	ret = qgroup_rescan_init(fs_info, 0, 1);
 	if (!ret) {

fs/btrfs/tree-log.c

@@ -6025,14 +6025,25 @@ void btrfs_record_snapshot_destroy(struct btrfs_trans_handle *trans,
  * Call this after adding a new name for a file and it will properly
  * update the log to reflect the new name.
  *
- * It will return zero if all goes well, and it will return 1 if a
- * full transaction commit is required.
+ * @ctx can not be NULL when @sync_log is false, and should be NULL when it's
+ * true (because it's not used).
+ *
+ * Return value depends on whether @sync_log is true or false.
+ * When true: returns BTRFS_NEED_TRANS_COMMIT if the transaction needs to be
+ *            committed by the caller, and BTRFS_DONT_NEED_TRANS_COMMIT
+ *            otherwise.
+ * When false: returns BTRFS_DONT_NEED_LOG_SYNC if the caller does not need to
+ *             to sync the log, BTRFS_NEED_LOG_SYNC if it needs to sync the log,
+ *             or BTRFS_NEED_TRANS_COMMIT if the transaction needs to be
+ *             committed (without attempting to sync the log).
  */
 int btrfs_log_new_name(struct btrfs_trans_handle *trans,
 			struct btrfs_inode *inode, struct btrfs_inode *old_dir,
-			struct dentry *parent)
+			struct dentry *parent,
+			bool sync_log, struct btrfs_log_ctx *ctx)
 {
 	struct btrfs_fs_info *fs_info = trans->fs_info;
+	int ret;
 
 	/*
 	 * this will force the logging code to walk the dentry chain
@@ -6047,9 +6058,34 @@ int btrfs_log_new_name(struct btrfs_trans_handle *trans,
 	 */
 	if (inode->logged_trans <= fs_info->last_trans_committed &&
 	    (!old_dir || old_dir->logged_trans <= fs_info->last_trans_committed))
-		return 0;
+		return sync_log ? BTRFS_DONT_NEED_TRANS_COMMIT :
+			BTRFS_DONT_NEED_LOG_SYNC;
+
+	if (sync_log) {
+		struct btrfs_log_ctx ctx2;
+
+		btrfs_init_log_ctx(&ctx2, &inode->vfs_inode);
+		ret = btrfs_log_inode_parent(trans, inode, parent, 0, LLONG_MAX,
+					     LOG_INODE_EXISTS, &ctx2);
+		if (ret == BTRFS_NO_LOG_SYNC)
+			return BTRFS_DONT_NEED_TRANS_COMMIT;
+		else if (ret)
+			return BTRFS_NEED_TRANS_COMMIT;
+
+		ret = btrfs_sync_log(trans, inode->root, &ctx2);
+		if (ret)
+			return BTRFS_NEED_TRANS_COMMIT;
+		return BTRFS_DONT_NEED_TRANS_COMMIT;
+	}
+
+	ASSERT(ctx);
+	ret = btrfs_log_inode_parent(trans, inode, parent, 0, LLONG_MAX,
+				     LOG_INODE_EXISTS, ctx);
+	if (ret == BTRFS_NO_LOG_SYNC)
+		return BTRFS_DONT_NEED_LOG_SYNC;
+	else if (ret)
+		return BTRFS_NEED_TRANS_COMMIT;
 
-	return btrfs_log_inode_parent(trans, inode, parent, 0, LLONG_MAX,
-				      LOG_INODE_EXISTS, NULL);
+	return BTRFS_NEED_LOG_SYNC;
 }
 

fs/btrfs/tree-log.h

@@ -71,8 +71,16 @@ void btrfs_record_unlink_dir(struct btrfs_trans_handle *trans,
 			     int for_rename);
 void btrfs_record_snapshot_destroy(struct btrfs_trans_handle *trans,
 				   struct btrfs_inode *dir);
+/* Return values for btrfs_log_new_name() */
+enum {
+	BTRFS_DONT_NEED_TRANS_COMMIT,
+	BTRFS_NEED_TRANS_COMMIT,
+	BTRFS_DONT_NEED_LOG_SYNC,
+	BTRFS_NEED_LOG_SYNC,
+};
 int btrfs_log_new_name(struct btrfs_trans_handle *trans,
 			struct btrfs_inode *inode, struct btrfs_inode *old_dir,
-			struct dentry *parent);
+			struct dentry *parent,
+			bool sync_log, struct btrfs_log_ctx *ctx);
 
 #endif

fs/btrfs/volumes.c

@@ -4491,7 +4491,12 @@ again:
 
 	/* Now btrfs_update_device() will change the on-disk size. */
 	ret = btrfs_update_device(trans, device);
-	btrfs_end_transaction(trans);
+	if (ret < 0) {
+		btrfs_abort_transaction(trans, ret);
+		btrfs_end_transaction(trans);
+	} else {
+		ret = btrfs_commit_transaction(trans);
+	}
 done:
 	btrfs_free_path(path);
 	if (ret) {