Files
kernel-tenderloin-3.0/include/linux
KAMEZAWA Hiroyuki 08e552c69c memcg: synchronized LRU
A big patch for changing memcg's LRU semantics.

Now,
  - page_cgroup is linked to mem_cgroup's its own LRU (per zone).

  - LRU of page_cgroup is not synchronous with global LRU.

  - page and page_cgroup is one-to-one and statically allocated.

  - To find page_cgroup is on what LRU, you have to check pc->mem_cgroup as
    - lru = page_cgroup_zoneinfo(pc, nid_of_pc, zid_of_pc);

  - SwapCache is handled.

And, when we handle LRU list of page_cgroup, we do following.

	pc = lookup_page_cgroup(page);
	lock_page_cgroup(pc); .....................(1)
	mz = page_cgroup_zoneinfo(pc);
	spin_lock(&mz->lru_lock);
	.....add to LRU
	spin_unlock(&mz->lru_lock);
	unlock_page_cgroup(pc);

But (1) is spin_lock and we have to be afraid of dead-lock with zone->lru_lock.
So, trylock() is used at (1), now. Without (1), we can't trust "mz" is correct.

This is a trial to remove this dirty nesting of locks.
This patch changes mz->lru_lock to be zone->lru_lock.
Then, above sequence will be written as

        spin_lock(&zone->lru_lock); # in vmscan.c or swap.c via global LRU
	mem_cgroup_add/remove/etc_lru() {
		pc = lookup_page_cgroup(page);
		mz = page_cgroup_zoneinfo(pc);
		if (PageCgroupUsed(pc)) {
			....add to LRU
		}
        spin_lock(&zone->lru_lock); # in vmscan.c or swap.c via global LRU

This is much simpler.
(*) We're safe even if we don't take lock_page_cgroup(pc). Because..
    1. When pc->mem_cgroup can be modified.
       - at charge.
       - at account_move().
    2. at charge
       the PCG_USED bit is not set before pc->mem_cgroup is fixed.
    3. at account_move()
       the page is isolated and not on LRU.

Pros.
  - easy for maintenance.
  - memcg can make use of laziness of pagevec.
  - we don't have to duplicated LRU/Active/Unevictable bit in page_cgroup.
  - LRU status of memcg will be synchronized with global LRU's one.
  - # of locks are reduced.
  - account_move() is simplified very much.
Cons.
  - may increase cost of LRU rotation.
    (no impact if memcg is not configured.)

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Li Zefan <lizf@cn.fujitsu.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-01-08 08:31:05 -08:00
..
2009-01-04 12:17:42 +01:00
2009-01-07 17:38:32 -05:00
2008-11-11 21:15:50 -05:00
2008-12-29 08:29:50 +01:00
2008-12-03 22:12:38 -08:00
2008-12-03 22:12:38 -08:00
2009-01-06 15:59:29 -08:00
2009-01-01 10:12:19 +10:30
2008-12-29 08:29:50 +01:00
2008-11-26 12:13:34 +01:00
2009-01-04 13:33:20 -08:00
2008-11-12 17:17:18 -08:00
2009-01-08 08:31:03 -08:00
2008-11-30 11:00:15 -08:00
2009-01-02 12:19:34 -08:00
2008-12-29 17:47:23 +10:00
2008-12-25 11:01:43 +11:00
2008-11-24 18:57:41 -05:00
2008-12-31 18:07:38 -05:00
2008-12-29 07:39:34 -05:00
2008-12-15 23:44:31 -08:00
2008-12-29 11:27:46 +02:00
2008-12-31 18:07:42 -05:00
2008-12-01 19:14:02 +01:00
2009-01-06 15:59:01 -08:00
2008-12-08 01:14:16 -08:00
2009-01-07 14:29:17 +01:00
2008-12-31 18:07:42 -05:00
2008-10-29 22:02:09 +01:00
2008-11-03 18:21:45 +01:00
2009-01-05 08:40:30 -08:00
2009-01-08 08:31:00 -08:00
2009-01-05 08:40:30 -08:00
2008-12-31 15:11:46 +01:00
2009-01-03 14:11:07 +01:00
2008-12-28 22:43:21 -05:00
2009-01-08 08:31:05 -08:00
2009-01-08 08:31:05 -08:00
2009-01-06 15:58:58 -08:00
2008-11-06 15:41:21 -08:00
2008-12-08 14:31:59 +01:00
2008-11-19 18:49:57 -08:00
2009-01-04 16:13:40 -08:00
2009-01-07 17:38:31 -05:00
2008-12-23 15:21:45 -05:00
2008-12-19 15:22:54 -05:00
2008-12-21 14:21:14 +11:00
2009-01-08 08:31:05 -08:00
2008-11-28 16:24:56 -08:00
2008-12-04 09:09:37 +01:00
2008-11-07 22:56:00 -08:00
2008-11-20 04:10:00 -08:00
2009-01-06 10:44:30 -08:00
2008-10-23 00:11:07 -04:00
2009-01-06 15:59:12 -08:00
2008-10-23 18:54:05 +04:00
2008-12-20 09:15:46 +01:00
2009-01-05 18:31:12 -08:00
2008-11-12 17:17:17 -08:00
2008-11-16 19:39:21 -08:00
2009-01-04 13:33:20 -08:00
2009-01-07 09:58:22 +11:00
2008-12-15 23:42:33 -08:00
2008-12-30 09:05:16 +10:30
2008-11-24 21:27:22 -08:00
2008-11-02 10:15:07 -08:00
2009-01-08 08:31:05 -08:00
2009-01-01 10:12:25 +10:30
2008-12-12 17:01:38 +01:00
2009-01-02 10:19:36 -08:00
2009-01-07 10:00:14 -08:00
2009-01-06 15:59:09 -08:00