Re: [PATCH] mm/page_alloc: use ac->high_zoneidx for classzone_idx

2018-05-16 Thread Mel Gorman
On Wed, May 16, 2018 at 11:35:55AM +0200, Vlastimil Babka wrote: > On 05/08/2018 03:00 AM, Joonsoo Kim wrote: > >> classzone predates my involvement with Linux but I would be less concerneed > >> about what the original intent was and instead ensure that classzone index > >> is consistent, sane and

Re: [PATCH] mm/page_alloc: use ac->high_zoneidx for classzone_idx

2018-05-16 Thread Vlastimil Babka
On 05/08/2018 03:00 AM, Joonsoo Kim wrote: >> classzone predates my involvement with Linux but I would be less concerneed >> about what the original intent was and instead ensure that classzone index >> is consistent, sane and potentially renamed while preserving the intent of >> "reserve pages in

Re: [PATCH] mm/page_alloc: use ac->high_zoneidx for classzone_idx

2018-05-08 Thread Andrew Morton
On Fri, 4 May 2018 11:33:22 +0100 Mel Gorman wrote: > On Fri, May 04, 2018 at 09:03:02AM +0200, Vlastimil Babka wrote: > > > min watermark for NORMAL zone on node 0 > > > allocation initiated on node 0: 750 + 4096 = 4846 > > > allocation initiated on node 1: 750 + 0 = 750 > > > > > > This waterm

Re: [PATCH] mm/page_alloc: use ac->high_zoneidx for classzone_idx

2018-05-07 Thread Joonsoo Kim
Hello, Mel. Thanks for precious input! 2018-05-04 19:33 GMT+09:00 Mel Gorman : > On Fri, May 04, 2018 at 09:03:02AM +0200, Vlastimil Babka wrote: >> > min watermark for NORMAL zone on node 0 >> > allocation initiated on node 0: 750 + 4096 = 4846 >> > allocation initiated on node 1: 750 + 0 = 750

Re: [PATCH] mm/page_alloc: use ac->high_zoneidx for classzone_idx

2018-05-04 Thread Mel Gorman
On Fri, May 04, 2018 at 09:03:02AM +0200, Vlastimil Babka wrote: > > min watermark for NORMAL zone on node 0 > > allocation initiated on node 0: 750 + 4096 = 4846 > > allocation initiated on node 1: 750 + 0 = 750 > > > > This watermark difference could cause too many numa_miss allocation > > in so

Re: [PATCH] mm/page_alloc: use ac->high_zoneidx for classzone_idx

2018-05-04 Thread Joonsoo Kim
2018-05-04 16:03 GMT+09:00 Vlastimil Babka : > On 05/04/2018 06:30 AM, js1...@gmail.com wrote: >> From: Joonsoo Kim >> >> Currently, we use the zone index of preferred_zone which represents >> the best matching zone for allocation, as classzone_idx. It has a problem >> on NUMA system with ZONE_MOV

Re: [PATCH] mm/page_alloc: use ac->high_zoneidx for classzone_idx

2018-05-04 Thread Vlastimil Babka
On 05/04/2018 06:30 AM, js1...@gmail.com wrote: > From: Joonsoo Kim > > Currently, we use the zone index of preferred_zone which represents > the best matching zone for allocation, as classzone_idx. It has a problem > on NUMA system with ZONE_MOVABLE. > > In NUMA system, it can be possible that

[PATCH] mm/page_alloc: use ac->high_zoneidx for classzone_idx

2018-05-03 Thread js1304
From: Joonsoo Kim Currently, we use the zone index of preferred_zone which represents the best matching zone for allocation, as classzone_idx. It has a problem on NUMA system with ZONE_MOVABLE. In NUMA system, it can be possible that each node has different populated zones. For example, node 0 c