On 01/13/2015 08:18 AM, Joonsoo Kim wrote: > On Mon, Jan 12, 2015 at 05:35:47PM +0100, Vlastimil Babka wrote: >> Hm what if we avoided dirtying the cache line in the non-deferred case? Would be >> simpler, too? >> >> if (zone->compact_considered + 1 >= defer_limit) >> return false; >> >> zone->compact_considered++; >> >> trace_mm_compaction_defer_compaction(zone, order); >> >> return true; > > Okay. I will include this minor optimization in next version of this > patch. Hm, on second thought, the "+ 1" part would break compaction_restarting() and it's ugly anyway. Removing "+ 1" would increase the number of compaction_deferred() attempts until success by one. Which should be negligible, but maybe not good to hide it in a tracepoint patch. Sorry for the noise. > Thanks. > -- 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/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>