Huang Ying 9a01b56b1a f2fs: avoid node page to be written twice in gc_node_segment
In gc_node_segment, if node page gc is run concurrently with node page
writeback, and check_valid_map and get_node_page run after page locked
and before cur_valid_map is updated as below, it is possible for the
page to be written twice unnecessarily.

			sync_node_pages
			  try_lock_page
			  ...
check_valid_map		  f2fs_write_node_page
			    ...
			    write_node_page
			      do_write_page
			        allocate_data_block
				  ...
				  refresh_sit_entry /* update cur_valid_map */
				  ...
			    ...
			    unlock_page
get_node_page
...
set_page_dirty
...
f2fs_put_page
  unlock_page

This can be solved via calling check_valid_map after get_node_page again.

Signed-off-by: Huang, Ying <ying.huang@intel.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
2014-09-09 13:15:07 -07:00
..
2014-08-08 15:57:24 -07:00
2014-08-08 15:57:20 -07:00
2014-06-04 16:53:53 -07:00
2014-08-06 18:01:12 -07:00
2014-08-07 14:40:09 -04:00
2014-08-08 18:13:21 -07:00
2014-07-23 15:01:44 -07:00
2014-08-26 16:17:48 -04:00
2014-01-25 03:14:05 -05:00
2014-08-06 18:01:12 -07:00
2014-07-15 22:40:23 +02:00
2014-07-31 15:52:51 +03:00
2014-08-24 15:47:27 -07:00
2014-08-07 14:40:09 -04:00
2014-08-07 14:40:11 -04:00
2014-08-08 15:57:23 -07:00
2014-08-08 15:57:31 -07:00
2014-05-06 17:31:10 -04:00
2014-02-07 16:08:57 -08:00
2014-08-07 14:40:08 -04:00
2014-08-07 14:40:09 -04:00
2014-05-06 17:39:42 -04:00
2014-04-01 23:19:08 -04:00
2014-04-01 23:19:08 -04:00
2014-07-23 10:18:02 -07:00