ext4: do not ask jbd2 to write data for delalloc buffers

Currently we ask jbd2 to write all dirty allocated buffers before
committing a transaction when doing writeback of delay allocated blocks.
However this is unnecessary since we move all pages to writeback state
before dropping a transaction handle and then submit all the necessary
IO. We still need the transaction commit to wait for all the outstanding
writeback before flushing disk caches during transaction commit to avoid
data exposure issues though. Use the new jbd2 capability and ask it to
only wait for outstanding writeback during transaction commit when
writing back data in ext4_writepages().

Tested-by: "HUANG Weller (CM/ESW12-CN)" <[e-mail]>
Signed-off-by: Jan Kara <[e-mail]>
Signed-off-by: Theodore Ts'o <[e-mail]>

fs/ext4/ext4.h

···
 #define EXT4_GET_BLOCKS_ZERO			0x0200
 #define EXT4_GET_BLOCKS_CREATE_ZERO		(EXT4_GET_BLOCKS_CREATE |\
 					EXT4_GET_BLOCKS_ZERO)
+	/* Caller will submit data before dropping transaction handle. This
+	 * allows jbd2 to avoid submitting data before commit. */
+#define EXT4_GET_BLOCKS_IO_SUBMIT		0x0400
 
 /*
  * The bit position of these flags must not overlap with any of the

fs/ext4/ext4_jbd2.h

···
 	return 0;
 }
 
-static inline int ext4_jbd2_file_inode(handle_t *handle, struct inode *inode)
+static inline int ext4_jbd2_inode_add_write(handle_t *handle,
+					    struct inode *inode)
 {
 	if (ext4_handle_valid(handle))
 		return jbd2_journal_inode_add_write(handle,
 						    EXT4_I(inode)->jinode);
+	return 0;
+}
+
+static inline int ext4_jbd2_inode_add_wait(handle_t *handle,
+					   struct inode *inode)
+{
+	if (ext4_handle_valid(handle))
+		return jbd2_journal_inode_add_wait(handle,
+						   EXT4_I(inode)->jinode);
 	return 0;
 }
 

fs/ext4/inode.c

···
 		    !(flags & EXT4_GET_BLOCKS_ZERO) &&
 		    !IS_NOQUOTA(inode) &&
 		    ext4_should_order_data(inode)) {
-			ret = ext4_jbd2_file_inode(handle, inode);
+			if (flags & EXT4_GET_BLOCKS_IO_SUBMIT)
+				ret = ext4_jbd2_inode_add_wait(handle, inode);
+			else
+				ret = ext4_jbd2_inode_add_write(handle, inode);
 			if (ret)
 				return ret;
 		}
···
 	 * the data was copied into the page cache.
 	 */
 	get_blocks_flags = EXT4_GET_BLOCKS_CREATE |
-			   EXT4_GET_BLOCKS_METADATA_NOFAIL;
+			   EXT4_GET_BLOCKS_METADATA_NOFAIL |
+			   EXT4_GET_BLOCKS_IO_SUBMIT;
 	dioread_nolock = ext4_should_dioread_nolock(inode);
 	if (dioread_nolock)
 		get_blocks_flags |= EXT4_GET_BLOCKS_IO_CREATE_EXT;
···
 		err = 0;
 		mark_buffer_dirty(bh);
 		if (ext4_should_order_data(inode))
-			err = ext4_jbd2_file_inode(handle, inode);
+			err = ext4_jbd2_inode_add_write(handle, inode);
 	}
 
 unlock:

fs/ext4/move_extent.c

···
 
 	/* Even in case of data=writeback it is reasonable to pin
 	 * inode to transaction, to prevent unexpected data loss */
-	*err = ext4_jbd2_file_inode(handle, orig_inode);
+	*err = ext4_jbd2_inode_add_write(handle, orig_inode);
 
 unlock_pages:
 	unlock_page(pagep[0]);