On Tue, Apr 10, 2012 at 11:38:33AM +0100, Mel Gorman wrote: > I think the bug you are accidentally fixing is related to how high_pfn > is updated inside that loop. The intent is that when free pages are > isolated that the next scan started from the same place as page > migration may have released those pages again. As it gets updated every > time a page is isolated the scanner is moving faster than it should. > > Try this; > That patch is obviously wrong so you're still looking for some other side-effect of your patch that explains why it appears to behave better. -- Mel Gorman SUSE Labs -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>