diff options
| author | KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> | 2010-03-05 22:41:40 +0100 |
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2010-03-06 20:26:24 +0100 |
| commit | 34e55232e59f7b19050267a05ff1226e5cd122a5 (patch) | |
| tree | 6b94e776e87d2a2fe1ceca7c5606901575323900 /kernel/exit.c | |
| parent | mm: clean up mm_counter (diff) | |
| download | linux-34e55232e59f7b19050267a05ff1226e5cd122a5.tar.xz linux-34e55232e59f7b19050267a05ff1226e5cd122a5.zip | |
mm: avoid false sharing of mm_counter
Considering the nature of per mm stats, it's the shared object among
threads and can be a cache-miss point in the page fault path.
This patch adds per-thread cache for mm_counter. RSS value will be
counted into a struct in task_struct and synchronized with mm's one at
events.
Now, in this patch, the event is the number of calls to handle_mm_fault.
Per-thread value is added to mm at each 64 calls.
rough estimation with small benchmark on parallel thread (2threads) shows
[before]
4.5 cache-miss/faults
[after]
4.0 cache-miss/faults
Anyway, the most contended object is mmap_sem if the number of threads grows.
[akpm@linux-foundation.org: coding-style fixes]
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Minchan Kim <minchan.kim@gmail.com>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'kernel/exit.c')
| -rw-r--r-- | kernel/exit.c | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/kernel/exit.c b/kernel/exit.c index 45ed043b8bf5..10d3c5d5ae44 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -952,7 +952,8 @@ NORET_TYPE void do_exit(long code) preempt_count()); acct_update_integrals(tsk); - + /* sync mm's RSS info before statistics gathering */ + sync_mm_rss(tsk, tsk->mm); group_dead = atomic_dec_and_test(&tsk->signal->live); if (group_dead) { hrtimer_cancel(&tsk->signal->real_timer); |