]> git.karo-electronics.de Git - linux-beck.git/commit
mm: compaction: partially revert capture of suitable high-order page
authorMel Gorman <mgorman@suse.de>
Fri, 11 Jan 2013 22:32:16 +0000 (14:32 -0800)
committerLinus Torvalds <torvalds@linux-foundation.org>
Fri, 11 Jan 2013 22:54:56 +0000 (14:54 -0800)
commit8fb74b9fb2b182d54beee592350d9ea1f325917a
treeccdcf36cfedb0cf54268226ebde75330e4882539
parentc0a3a20b6c4b5229ef5d26fd9b1c4b1957632aa7
mm: compaction: partially revert capture of suitable high-order page

Eric Wong reported on 3.7 and 3.8-rc2 that ppoll() got stuck when
waiting for POLLIN on a local TCP socket.  It was easier to trigger if
there was disk IO and dirty pages at the same time and he bisected it to
commit 1fb3f8ca0e92 ("mm: compaction: capture a suitable high-order page
immediately when it is made available").

The intention of that patch was to improve high-order allocations under
memory pressure after changes made to reclaim in 3.6 drastically hurt
THP allocations but the approach was flawed.  For Eric, the problem was
that page->pfmemalloc was not being cleared for captured pages leading
to a poor interaction with swap-over-NFS support causing the packets to
be dropped.  However, I identified a few more problems with the patch
including the fact that it can increase contention on zone->lock in some
cases which could result in async direct compaction being aborted early.

In retrospect the capture patch took the wrong approach.  What it should
have done is mark the pageblock being migrated as MIGRATE_ISOLATE if it
was allocating for THP and avoided races that way.  While the patch was
showing to improve allocation success rates at the time, the benefit is
marginal given the relative complexity and it should be revisited from
scratch in the context of the other reclaim-related changes that have
taken place since the patch was first written and tested.  This patch
partially reverts commit 1fb3f8ca0e92 ("mm: compaction: capture a
suitable high-order page immediately when it is made available").

Reported-and-tested-by: Eric Wong <normalperson@yhbt.net>
Tested-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Cc: David Miller <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
include/linux/compaction.h
include/linux/mm.h
mm/compaction.c
mm/internal.h
mm/page_alloc.c