Re: [RFC PATCH] memcg: use root_mem_cgroup when css is inherited
From: Johannes Weiner <hidden>
Date: 2022-08-25 13:35:17
Also in:
linux-mm, lkml
From: Johannes Weiner <hidden>
Date: 2022-08-25 13:35:17
Also in:
linux-mm, lkml
On Thu, Aug 25, 2022 at 06:11:09PM +0800, Zhaoyang Huang wrote:
Sorry for any misunderstanding among the discussion. My purpose is real and simple as I have stated from the very beginning that I would like to have per-app cgroup hierarchy to charge memory to root if it is not enabled explicitly for memory. The reason has also been stated like reclaim and workingset regression in suren's report. I don't think my proposal will do any harm to current v2's mechanism besides asking for the admin echo "+memory" to their desire group.
It would permit children to escape parental control, completely breaking the hierarchy and delegation guarantees currently in place. That just isn't going to happen. Nacked-by: Johannes Weiner [off-list ref] I would suggest focusing on finding the root cause for the reclaim problem you are describing. It's possible all we need is a fix to reclaim or the workingset code.