Commit 73ce02e9 authored by KOSAKI Motohiro's avatar KOSAKI Motohiro Committed by Linus Torvalds

mm: stop kswapd's infinite loop at high order allocation

Wassim Dagash reported following kswapd infinite loop problem.

  kswapd runs in some infinite loop trying to swap until order 10 of zone
  highmem is OK.... kswapd will continue to try to balance order 10 of zone
  highmem forever (or until someone release a very large chunk of highmem).

For non order-0 allocations, the system may never be balanced due to
fragmentation but kswapd should not infinitely loop as a result.

Instead, recheck all watermarks at order-0 as they are the most important.
If watermarks are ok, kswapd will go back to sleep.

[akpm@linux-foundation.org: fix comment]
Reported-by: default avatarwassim dagash <wassim.dagash@gmail.com>
Signed-off-by: default avatarKOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Reviewed-by: default avatarNick Piggin <npiggin@suse.de>
Signed-off-by: default avatarMel Gorman <mel@csn.ul.ie>
Signed-off-by: default avatarAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
parent 594fe1a0
...@@ -1867,6 +1867,23 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order) ...@@ -1867,6 +1867,23 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order)
try_to_freeze(); try_to_freeze();
/*
* Fragmentation may mean that the system cannot be
* rebalanced for high-order allocations in all zones.
* At this point, if nr_reclaimed < SWAP_CLUSTER_MAX,
* it means the zones have been fully scanned and are still
* not balanced. For high-order allocations, there is
* little point trying all over again as kswapd may
* infinite loop.
*
* Instead, recheck all watermarks at order-0 as they
* are the most important. If watermarks are ok, kswapd will go
* back to sleep. High-order users can still perform direct
* reclaim if they wish.
*/
if (sc.nr_reclaimed < SWAP_CLUSTER_MAX)
order = sc.order = 0;
goto loop_again; goto loop_again;
} }
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment