[PATCH] mm: fix calculation accounting dirtyable highmem

From: Minchan Kim
Date: Tue Jul 12 2016 - 22:22:49 EST


When I tested vmscale in mmtest in 32bit, I found the benchmark
was slow down 0.5 times.

base node
1 global-1
User 12.98 16.04
System 147.61 166.42
Elapsed 26.48 38.08

With vmstat, I found IO wait avg is much increased compared to
base.

The reason was highmem_dirtyable_memory accumulates free pages
and highmem_file_pages from HIGHMEM to MOVABLE zones which was
wrong. With that, dirth_thresh in throtlle_vm_write is always
0 so that it calls congestion_wait frequently if writeback
starts.

With this patch, it is much recovered.

base node fi
1 global-1 fix
User 12.98 16.04 13.78
System 147.61 166.42 143.92
Elapsed 26.48 38.08 29.64

Signed-off-by: Minchan Kim <minchan@xxxxxxxxxx>
---
mm/page-writeback.c | 16 ++++++++++------
1 file changed, 10 insertions(+), 6 deletions(-)

diff --git a/mm/page-writeback.c b/mm/page-writeback.c
index 8db1db2..bf27594 100644
--- a/mm/page-writeback.c
+++ b/mm/page-writeback.c
@@ -307,27 +307,31 @@ static unsigned long highmem_dirtyable_memory(unsigned long total)
{
#ifdef CONFIG_HIGHMEM
int node;
- unsigned long x = 0;
+ unsigned long x;
int i;
- unsigned long dirtyable = highmem_file_pages;
+ unsigned long dirtyable = 0;

for_each_node_state(node, N_HIGH_MEMORY) {
for (i = ZONE_NORMAL + 1; i < MAX_NR_ZONES; i++) {
struct zone *z;
+ unsigned long nr_pages;

if (!is_highmem_idx(i))
continue;

z = &NODE_DATA(node)->node_zones[i];
- dirtyable += zone_page_state(z, NR_FREE_PAGES);
+ if (!populated_zone(z))
+ continue;

+ nr_pages = zone_page_state(z, NR_FREE_PAGES);
/* watch for underflows */
- dirtyable -= min(dirtyable, high_wmark_pages(z));
-
- x += dirtyable;
+ nr_pages -= min(nr_pages, high_wmark_pages(z));
+ dirtyable += nr_pages;
}
}

+ x = dirtyable + highmem_file_pages;
+
/*
* Unreclaimable memory (kernel memory or anonymous memory
* without swap) can bring down the dirtyable pages below
--
1.9.1