mm/page-writeback.c: do not count anon pages as dirtyable memory
authorJohannes Weiner <hannes@cmpxchg.org>
Wed, 29 Jan 2014 22:05:41 +0000 (14:05 -0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 13 Feb 2014 21:48:00 +0000 (13:48 -0800)
commit48526149964e69fc54a06c409e13d36990386464
tree844e34d12c42f34711531573af92243fbf273ef8
parent03381bd28963f97a976d4742468359f12474ea39
mm/page-writeback.c: do not count anon pages as dirtyable memory

commit a1c3bfb2f67ef766de03f1f56bdfff9c8595ab14 upstream.

The VM is currently heavily tuned to avoid swapping.  Whether that is
good or bad is a separate discussion, but as long as the VM won't swap
to make room for dirty cache, we can not consider anonymous pages when
calculating the amount of dirtyable memory, the baseline to which
dirty_background_ratio and dirty_ratio are applied.

A simple workload that occupies a significant size (40+%, depending on
memory layout, storage speeds etc.) of memory with anon/tmpfs pages and
uses the remainder for a streaming writer demonstrates this problem.  In
that case, the actual cache pages are a small fraction of what is
considered dirtyable overall, which results in an relatively large
portion of the cache pages to be dirtied.  As kswapd starts rotating
these, random tasks enter direct reclaim and stall on IO.

Only consider free pages and file pages dirtyable.

Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Reported-by: Tejun Heo <tj@kernel.org>
Tested-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
include/linux/vmstat.h
mm/page-writeback.c
mm/vmscan.c