[PATCH 6/6] writeback: redirty a fully scanned inode

[Wu Fengguang]

Redirty inode when its writeback_index wrapped around.
This is mainly to avoid unnecessary IOs for fast dirtier
that do lots of overwrites.

CC: Jan Kara <jack@xxxxxxx>
CC: Theodore Ts'o <tytso@xxxxxxx>
CC: Dave Chinner <david@xxxxxxxxxxxxx>
CC: Jens Axboe <jens.axboe@xxxxxxxxxx>
CC: Chris Mason <chris.mason@xxxxxxxxxx>
CC: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Signed-off-by: Wu Fengguang <fengguang.wu@xxxxxxxxx>
---
 fs/fs-writeback.c |    6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

--- linux.orig/fs/fs-writeback.c	2009-09-23 18:44:56.000000000 +0800
+++ linux/fs/fs-writeback.c	2009-09-23 18:52:22.000000000 +0800
@@ -397,6 +397,7 @@ writeback_single_inode(struct inode *ino
 {
 	struct address_space *mapping = inode->i_mapping;
 	int wait = wbc->sync_mode == WB_SYNC_ALL;
+	pgoff_t start_index;
 	unsigned dirty;
 	int ret;
 
@@ -434,6 +435,7 @@ writeback_single_inode(struct inode *ino
 
 	spin_unlock(&inode_lock);
 
+	start_index = mapping->writeback_index;
 	ret = do_writepages(mapping, wbc);
 
 	/* Don't write the inode if only I_DIRTY_PAGES was set */
@@ -485,7 +487,9 @@ writeback_single_inode(struct inode *ino
 				 * soon as the queue becomes uncongested.
 				 */
 				inode->i_state |= I_DIRTY_PAGES;
-				if (wbc->nr_to_write <= 0) {
+select_queue:
+				if (wbc->nr_to_write <= 0 &&
+				    start_index < mapping->writeback_index) {
 					/*
 					 * slice used up: queue for next turn
 					 */


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/