Theodore Ts'o 42cf3452d5 jbd2: calculate statistics without holding j_state_lock and j_list_lock
The two hottest locks, and thus the biggest scalability bottlenecks,
in the jbd2 layer, are the j_list_lock and j_state_lock.  This has
inspired some people to do some truly unnatural things[1].

[1] https://www.usenix.org/system/files/conference/fast14/fast14-paper_kang.pdf

We don't need to be holding both j_state_lock and j_list_lock while
calculating the journal statistics, so move those calculations to the
very end of jbd2_journal_commit_transaction.

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
2014-03-08 19:51:16 -05:00
..
2013-10-24 23:43:27 -04:00
2014-01-25 03:13:02 -05:00
2014-01-25 03:13:02 -05:00
2014-01-23 18:54:14 +02:00
2013-06-29 12:56:53 +04:00
2014-01-26 11:51:09 +01:00
2014-02-02 16:24:07 -08:00
2013-12-04 12:27:46 +01:00
2014-01-28 18:56:37 -08:00
2014-02-08 10:50:58 -06:00
2013-09-16 18:20:25 -07:00
2014-01-25 03:14:05 -05:00
2013-06-29 12:56:32 +04:00
2014-01-25 03:13:03 -05:00
2013-06-29 12:56:39 +04:00
2013-12-22 11:03:49 -08:00
2013-12-05 16:36:21 -06:00
2013-06-29 12:57:04 +04:00
2013-10-24 23:34:54 -04:00
2013-09-13 23:06:40 -04:00
2013-06-29 12:57:05 +04:00
2014-01-26 12:37:55 -05:00
2013-11-23 22:33:47 -08:00
2013-09-10 18:56:31 -04:00
2013-10-24 23:34:54 -04:00
2013-11-09 00:16:20 -05:00
2013-10-24 23:34:54 -04:00
2014-01-26 08:26:40 -05:00
2013-11-23 22:33:47 -08:00
2013-10-24 23:35:00 -04:00
2013-10-24 23:34:54 -04:00
2014-01-22 19:36:57 +01:00
2013-11-09 00:16:31 -05:00