[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <xuhsdsl2bhlrlaghar3ru7nhlhjsmaiyjxayryogylsordbwcx@e2jxsb2qdbhd>
Date: Fri, 6 Dec 2024 13:20:23 -0800
From: Shakeel Butt <shakeel.butt@...ux.dev>
To: Hugh Dickins <hughd@...gle.com>
Cc: Chen Ridong <chenridong@...weicloud.com>, Yu Zhao <yuzhao@...gle.com>,
akpm@...ux-foundation.org, mhocko@...nel.org, hannes@...xchg.org, yosryahmed@...gle.com,
roman.gushchin@...ux.dev, muchun.song@...ux.dev, davidf@...eo.com, vbabka@...e.cz,
linux-mm@...ck.org, linux-kernel@...r.kernel.org, cgroups@...r.kernel.org,
chenridong@...wei.com, wangweiyang2@...wei.com
Subject: Re: [next -v1 3/5] memcg: simplify the mem_cgroup_update_lru_size
function
On Fri, Dec 06, 2024 at 12:24:54AM -0800, Hugh Dickins wrote:
[...]
> Another thing to understand: it's called before adding folio to list,
> but after removing folio from list: when it can usefully compare whether
> the emptiness of the list correctly matches lru_size 0.
I think one source of confusion might be that this "emptiness" check has
been removed by commit b4536f0c829c because of maintaining the list size
per-zone and actual list is shared between zones of a node.
> It cannot do so
> when adding if you "simplify" it in the way that you did.
>
Powered by blists - more mailing lists