On Mon, 7 Sep 2026 16:25:31 -0700 Shakeel Butt <[email protected]> wrote:
Hi Shakeel, Thanks for your review on the series! > On Mon, Aug 31, 2026 at 09:37:49AM -0700, Joshua Hahn wrote: > > The existing percpu memcg stock drainer schedules a stock drain worker > > per-cpu, one for every CPU containing the target memcg's stock. > > But you are introducing remote draining. Why do you need async drainer at all? > Just always drain synchronously. I was trying to match parity with the existing code as much as possible. But I see now that this series is a bit complicated because I'm trying to do two things at once and mixing up the page_counter migration and the scalable draining together. In the next version I'll keep parity and only migrate to page_counter and in a follow-up I'll address the drainers separately. Thanks again! Joshua
