On Thu, Mar 9, 2017 at 12:28 PM, Huang, Ying <ying.hu...@intel.com> wrote: > Balbir Singh <bsinghar...@gmail.com> writes: > >> On Wed, 2017-03-08 at 15:26 +0800, Huang, Ying wrote: >>> From: Huang Ying <ying.hu...@intel.com> >>> >>> This patch make it possible to charge or uncharge a set of continuous >>> swap entries in the swap cgroup. The number of swap entries is >>> specified via an added parameter. >>> >>> This will be used for the THP (Transparent Huge Page) swap support. >>> Where a swap cluster backing a THP may be allocated and freed as a >>> whole. So a set of (HPAGE_PMD_NR) continuous swap entries backing one >>> THP need to be charged or uncharged together. This will batch the >>> cgroup operations for the THP swap too. >> >> A quick look at the patches makes it look sane. I wonder if we would >> make sense to track THP swapout separately as well >> (from a memory.stat perspective) > > The patchset is just the first step of THP swap optimization. So the > THP will still be split after putting the THP into the swap cache. This > makes it unnecessary to change mem_cgroup_swapout(). I am working on a > following up patchset to further delaying THP splitting after swapping > out the THP to the disk. In that patchset, I will change > mem_cgroup_swapout() too.
Fair enough Thanks, Balbir
