On Wed, Jan 04, 2017 at 03:41:59PM -0800, David Rientjes wrote: > There is no thp defrag option that currently allows MADV_HUGEPAGE regions > to do direct compaction and reclaim while all other thp allocations simply > trigger kswapd and kcompactd in the background and fail immediately. > > The "defer" setting simply triggers background reclaim and compaction for > all regions, regardless of MADV_HUGEPAGE, which makes it unusable for our > userspace where MADV_HUGEPAGE is being used to indicate the application is > willing to wait for work for thp memory to be available. > > The "madvise" setting will do direct compaction and reclaim for these > MADV_HUGEPAGE regions, but does not trigger kswapd and kcompactd in the > background for anybody else. > > For reasonable usage, there needs to be a mesh between the two options. > This patch introduces a fifth mode, "background", that will do direct > reclaim and compaction for MADV_HUGEPAGE regions and trigger background > reclaim and compaction for everybody else so that hugepages may be > available in the near future. > > A proposal to allow direct reclaim and compaction for MADV_HUGEPAGE > regions as part of the "defer" mode, making it a very powerful setting and > avoids breaking userspace, was offered: > http://marc.info/?t=148236612700003. This additional mode is a > compromise. > > This patch also cleans up the helper function for storing to "enabled" > and "defrag" since the former supports three modes while the latter > supports five and triple_flag_store() was getting unnecessarily messy. > > Signed-off-by: David Rientjes <rientjes@xxxxxxxxxx> > --- > I don't understand Mel's suggestion of "defer-fault" as option naming. > defer-fault was intended to reflect "defer faults but not anything else" with the only sensible alternative being madvise requests. While not a major fan of the background name, I don't have a better suggestion either other than defer-fault. There are likely to be objections based on how this should be specified and investigating alternative proposals such as fine-grained control of how background compaction should be done but I hadn't proposed them and hadn't intended to work on such patches. This patch appears to give the semantics you want and I said I would ack such a configuration option so; Acked-by: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx> -- 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/ . Don't email: <a href=mailto:"dont@xxxxxxxxx"> email@xxxxxxxxx </a>