Re: [PATCH v4 3/3] mm/compaction: enhance compaction finish condition

2015-02-24 Thread Joonsoo Kim
On Wed, Feb 18, 2015 at 04:04:05PM -0800, Andrew Morton wrote: > On Thu, 12 Feb 2015 16:15:05 +0900 Joonsoo Kim wrote: > > > Compaction has anti fragmentation algorithm. It is that freepage > > should be more than pageblock order to finish the compaction if we don't > > find any freepage in reque

Re: [PATCH v4 3/3] mm/compaction: enhance compaction finish condition

2015-02-24 Thread Joonsoo Kim
On Tue, Feb 17, 2015 at 10:46:04AM +0100, Vlastimil Babka wrote: > On 02/12/2015 08:15 AM, Joonsoo Kim wrote: > >Compaction has anti fragmentation algorithm. It is that freepage > >should be more than pageblock order to finish the compaction if we don't > >find any freepage in requested migratetype

Re: [PATCH v4 3/3] mm/compaction: enhance compaction finish condition

2015-02-18 Thread Andrew Morton
On Thu, 12 Feb 2015 16:15:05 +0900 Joonsoo Kim wrote: > Compaction has anti fragmentation algorithm. It is that freepage > should be more than pageblock order to finish the compaction if we don't > find any freepage in requested migratetype buddy list. This is for > mitigating fragmentation, but,

Re: [PATCH v4 3/3] mm/compaction: enhance compaction finish condition

2015-02-17 Thread Vlastimil Babka
On 02/12/2015 08:15 AM, Joonsoo Kim wrote: Compaction has anti fragmentation algorithm. It is that freepage should be more than pageblock order to finish the compaction if we don't find any freepage in requested migratetype buddy list. This is for mitigating fragmentation, but, there is a lack of

[PATCH v4 3/3] mm/compaction: enhance compaction finish condition

2015-02-11 Thread Joonsoo Kim
Compaction has anti fragmentation algorithm. It is that freepage should be more than pageblock order to finish the compaction if we don't find any freepage in requested migratetype buddy list. This is for mitigating fragmentation, but, there is a lack of migratetype consideration and it is too exce