From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-it0-f70.google.com (mail-it0-f70.google.com [209.85.214.70]) by kanga.kvack.org (Postfix) with ESMTP id B3F3D6B025E for ; Tue, 12 Jul 2016 22:22:01 -0400 (EDT) Received: by mail-it0-f70.google.com with SMTP id j8so69500051itb.1 for ; Tue, 12 Jul 2016 19:22:01 -0700 (PDT) Received: from lgeamrelo13.lge.com (LGEAMRELO13.lge.com. [156.147.23.53]) by mx.google.com with ESMTP id t1si839529itb.59.2016.07.12.19.22.00 for ; Tue, 12 Jul 2016 19:22:01 -0700 (PDT) From: Minchan Kim Subject: [PATCH] mm: fix calculation accounting dirtyable highmem Date: Wed, 13 Jul 2016 11:23:13 +0900 Message-Id: <1468376593-26444-1-git-send-email-minchan@kernel.org> Sender: owner-linux-mm@kvack.org List-ID: To: Andrew Morton Cc: Mel Gorman , linux-kernel@vger.kernel.org, linux-mm@kvack.org, Minchan Kim 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 --- 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 -- 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: email@kvack.org