[RFC]mm: batch activate_page() to reduce lock contention

* [RFC]mm: batch activate_page() to reduce lock contention
@ 2010-07-20  7:18 Shaohua Li
  2010-07-21 16:06 ` Minchan Kim
                   ` (2 more replies)
  0 siblings, 3 replies; 20+ messages in thread
From: Shaohua Li @ 2010-07-20  7:18 UTC (permalink / raw)
  To: linux-mm; +Cc: Andrew Morton, Andi Kleen, Wu, Fengguang

The zone->lru_lock is heavily contented in workload where activate_page()
is frequently used. We could do batch activate_page() to reduce the lock
contention. The batched pages will be added into zone list when the pool
is full or page reclaim is trying to drain them.

For example, in a 4 socket 64 CPU system, create a sparse file and 64 processes,
processes shared map to the file. Each process read access the whole file and
then exit. The process exit will do unmap_vmas() and cause a lot of
activate_page() call. In such workload, we saw about 58% total time reduction
with below patch.

But we did see some strange regression. The regression is small (usually < 2%)
and most are from multithread test and none heavily use activate_page(). For
example, in the same system, we create 64 threads. Each thread creates a private
mmap region and does read access. We measure the total time and saw about 2%
regression. But in such workload, 99% time is on page fault and activate_page()
takes no time. Very strange, we haven't a good explanation for this so far,
hopefully somebody can share a hint.

Signed-off-by: Shaohua Li <shaohua.li@intel.com>

diff --git a/mm/swap.c b/mm/swap.c
index 3ce7bc3..4a3fd7f 100644
--- a/mm/swap.c
+++ b/mm/swap.c
@@ -39,6 +39,7 @@ int page_cluster;
 
 static DEFINE_PER_CPU(struct pagevec[NR_LRU_LISTS], lru_add_pvecs);
 static DEFINE_PER_CPU(struct pagevec, lru_rotate_pvecs);
+static DEFINE_PER_CPU(struct pagevec, activate_page_pvecs);
 
 /*
  * This path almost never happens for VM activity - pages are normally
@@ -175,11 +176,10 @@ static void update_page_reclaim_stat(struct zone *zone, struct page *page,
 /*
  * FIXME: speed this up?
  */
-void activate_page(struct page *page)
+static void __activate_page(struct page *page)
 {
 	struct zone *zone = page_zone(page);
 
-	spin_lock_irq(&zone->lru_lock);
 	if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) {
 		int file = page_is_file_cache(page);
 		int lru = page_lru_base_type(page);
@@ -192,7 +192,46 @@ void activate_page(struct page *page)
 
 		update_page_reclaim_stat(zone, page, file, 1);
 	}
-	spin_unlock_irq(&zone->lru_lock);
+}
+
+static void activate_page_drain_cpu(int cpu)
+{
+	struct pagevec *pvec = &per_cpu(activate_page_pvecs, cpu);
+	struct zone *last_zone = NULL, *zone;
+	int i, j;
+
+	for (i = 0; i < pagevec_count(pvec); i++) {
+		zone = page_zone(pvec->pages[i]);
+		if (zone == last_zone)
+			continue;
+
+		if (last_zone)
+			spin_unlock_irq(&last_zone->lru_lock);
+		last_zone = zone;
+		spin_lock_irq(&last_zone->lru_lock);
+
+		for (j = i; j < pagevec_count(pvec); j++) {
+			struct page *page = pvec->pages[j];
+
+			if (last_zone != page_zone(page))
+				continue;
+			__activate_page(page);
+		}
+	}
+	if (last_zone)
+		spin_unlock_irq(&last_zone->lru_lock);
+	release_pages(pvec->pages, pagevec_count(pvec), pvec->cold);
+	pagevec_reinit(pvec);
+}
+
+void activate_page(struct page *page)
+{
+	struct pagevec *pvec = &get_cpu_var(activate_page_pvecs);
+
+	page_cache_get(page);
+	if (!pagevec_add(pvec, page))
+		activate_page_drain_cpu(smp_processor_id());
+	put_cpu_var(activate_page_pvecs);
 }
 
 /*
@@ -297,6 +336,7 @@ static void drain_cpu_pagevecs(int cpu)
 void lru_add_drain(void)
 {
 	drain_cpu_pagevecs(get_cpu());
+	activate_page_drain_cpu(smp_processor_id());
 	put_cpu();
 }
 


--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

^ permalink raw reply related	[flat|nested] 20+ messages in thread