* [PATCH -next] sched/numa: document CLONE_VM behavior
@ 2014-01-15 19:01 Davidlohr Bueso
2014-01-17 1:59 ` Rik van Riel
0 siblings, 1 reply; 2+ messages in thread
From: Davidlohr Bueso @ 2014-01-15 19:01 UTC (permalink / raw)
To: Ingo Molnar
Cc: Peter Zijlstra, Rik van Riel, Mel Gorman, Andrea Arcangeli,
Srikar Dronamraju, Johannes Weiner, linux-kernel
From: Davidlohr Bueso <davidlohr@hp.com>
The changelog for Rik's "sched/numa: Stay on the same node if CLONE_VM" patch
serves well documented in the code.
Signed-off-by: Davidlohr Bueso <davidlohr@hp.com>
---
kernel/sched/core.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 7f45fd5..a2d690c 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -1743,10 +1743,16 @@ static void __sched_fork(unsigned long clone_flags, struct task_struct *p)
#ifdef CONFIG_NUMA_BALANCING
if (p->mm && atomic_read(&p->mm->mm_users) == 1) {
- p->mm->numa_next_scan = jiffies + msecs_to_jiffies(sysctl_numa_balancing_scan_delay);
p->mm->numa_scan_seq = 0;
+ p->mm->numa_next_scan = jiffies +
+ msecs_to_jiffies(sysctl_numa_balancing_scan_delay);
}
+ /*
+ * A newly spawned thread should stay on the same NUMA node as its
+ * parent. This prevents processes from being "torn" across multiple
+ * NUMA nodes every time they spawn a new thread.
+ */
if (clone_flags & CLONE_VM)
p->numa_preferred_nid = current->numa_preferred_nid;
else
--
1.8.1.4
^ permalink raw reply related [flat|nested] 2+ messages in thread
end of thread, other threads:[~2014-01-17 1:59 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-01-15 19:01 [PATCH -next] sched/numa: document CLONE_VM behavior Davidlohr Bueso
2014-01-17 1:59 ` Rik van Riel
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.