Messages in this thread | | | Date | Mon, 11 Jun 2012 14:06:12 +0100 | From | Mel Gorman <> | Subject | Re: [PATCH v9] mm: compaction: handle incorrect MIGRATE_UNMOVABLE type pageblocks |
| |
On Tue, Jun 05, 2012 at 03:05:40PM +0900, Minchan Kim wrote: > > Let's throw it away until the author send us data. > > > > I guess it's hard to make such workload to prove it's useful normally. > But we can't make sure there isn't such workload in the world. > So I hope listen VOC. At least, Mel might require it. >
I'm playing a lot of catch-up at the moment after being out for a few days so sorry for my silence on this and other threads.
My initial support for this patch was based on an artifical load but one I felt was plausible to trigger if CMA was being used. In a normal workload I thought it might be possible to hit if a large process exited freeing a lot of pagetable pages from MIGRATE_UNMOVABLE blocks at the same time but that is a little unlikely and a test case would also look very artifical.
Hence, I believe that if you require a real workload to demonstrate the benefit of the patch that it will be very difficult to find. The primary decision is if CMA needs this or not. I was under the impression that it was a help for CMA allocation success rates but I may be mistaken.
-- Mel Gorman SUSE Labs
| |