Re: [PATCH RFC] mm,vmscan: only evict file pages when we have plenty

2012-10-31 Thread Rik van Riel
On 10/31/2012 11:39 AM, Johannes Weiner wrote: On Tue, Oct 30, 2012 at 02:42:04PM -0400, Rik van Riel wrote: If we have more inactive file pages than active file pages, we skip scanning the active file pages alltogether, with the idea that we do not want to evict the working set when there is pl

Re: [PATCH RFC] mm,vmscan: only evict file pages when we have plenty

2012-10-31 Thread Johannes Weiner
On Tue, Oct 30, 2012 at 02:42:04PM -0400, Rik van Riel wrote: > If we have more inactive file pages than active file pages, we > skip scanning the active file pages alltogether, with the idea > that we do not want to evict the working set when there is > plenty of streaming IO in the cache. > > Ho

Re: [PATCH RFC] mm,vmscan: only evict file pages when we have plenty

2012-10-30 Thread Rik van Riel
On 10/30/2012 02:54 PM, Andrew Morton wrote: On Tue, 30 Oct 2012 14:42:04 -0400 Rik van Riel wrote: If we have more inactive file pages than active file pages, we skip scanning the active file pages alltogether, with the idea that we do not want to evict the working set when there is plenty of

Re: [PATCH RFC] mm,vmscan: only evict file pages when we have plenty

2012-10-30 Thread Andrew Morton
On Tue, 30 Oct 2012 14:42:04 -0400 Rik van Riel wrote: > If we have more inactive file pages than active file pages, we > skip scanning the active file pages alltogether, with the idea > that we do not want to evict the working set when there is > plenty of streaming IO in the cache. Yes, I've n

[PATCH RFC] mm,vmscan: only evict file pages when we have plenty

2012-10-30 Thread Rik van Riel
If we have more inactive file pages than active file pages, we skip scanning the active file pages alltogether, with the idea that we do not want to evict the working set when there is plenty of streaming IO in the cache. However, the code forgot to also skip scanning anonymous pages in that situa