Re: [PATCH 7/13] Drain per-cpu lists when high-order allocations fail

2007-09-11 Thread Mel Gorman
On Tue, 2007-09-11 at 01:05 +1000, Nick Piggin wrote: > On Monday 10 September 2007 21:22, Mel Gorman wrote: > > Per-cpu pages can accidentally cause fragmentation because they are free, > > but pinned pages in an otherwise contiguous block. When this patch is > > applied, the per-cpu caches are d

Re: [PATCH 7/13] Drain per-cpu lists when high-order allocations fail

2007-09-10 Thread Nick Piggin
On Monday 10 September 2007 21:22, Mel Gorman wrote: > Per-cpu pages can accidentally cause fragmentation because they are free, > but pinned pages in an otherwise contiguous block. When this patch is > applied, the per-cpu caches are drained after the direct-reclaim is entered > if the requested

[PATCH 7/13] Drain per-cpu lists when high-order allocations fail

2007-09-10 Thread Mel Gorman
Subject: Drain per-cpu lists when high-order allocations fail Per-cpu pages can accidentally cause fragmentation because they are free, but pinned pages in an otherwise contiguous block. When this patch is applied, the per-cpu caches are drained after the direct-reclaim is entered if the requeste