[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1058dae6-c89b-45ad-83c1-d903fd9532ae@suse.cz>
Date: Fri, 14 Mar 2025 11:09:55 +0100
From: Vlastimil Babka <vbabka@...e.cz>
To: Shakeel Butt <shakeel.butt@...ux.dev>,
Andrew Morton <akpm@...ux-foundation.org>
Cc: Johannes Weiner <hannes@...xchg.org>, Michal Hocko <mhocko@...nel.org>,
Roman Gushchin <roman.gushchin@...ux.dev>,
Muchun Song <muchun.song@...ux.dev>,
Sebastian Andrzej Siewior <bigeasy@...utronix.de>, linux-mm@...ck.org,
cgroups@...r.kernel.org, linux-kernel@...r.kernel.org,
Meta kernel team <kernel-team@...a.com>
Subject: Re: [RFC PATCH 05/10] memcg: no refilling stock from
obj_cgroup_release
On 3/14/25 07:15, Shakeel Butt wrote:
> obj_cgroup_release is called when all the references to the objcg has
> been released i.e. no more memory objects are pointing to it. Most
> probably objcg->memcg will be pointing to some ancestor memcg and at the
> moment, in obj_cgroup_release, the kernel call
> obj_cgroup_uncharge_pages() to uncharge last remaining memory.
>
> However obj_cgroup_uncharge_pages() refills the local stock. There is
> no need to refill the local stock with some ancestor memcg and flush the
> local stock. In addition this removes the requirement to only call
> obj_cgroup_put() outside of local_lock.
>
> Signed-off-by: Shakeel Butt <shakeel.butt@...ux.dev>
Acked-by: Vlastimil Babka <vbabka@...e.cz>
> ---
> mm/memcontrol.c | 15 +++++++++++----
> 1 file changed, 11 insertions(+), 4 deletions(-)
>
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index 7054b0ebd207..83db180455a1 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -129,8 +129,7 @@ bool mem_cgroup_kmem_disabled(void)
> return cgroup_memory_nokmem;
> }
>
> -static void obj_cgroup_uncharge_pages(struct obj_cgroup *objcg,
> - unsigned int nr_pages);
> +static void memcg_uncharge(struct mem_cgroup *memcg, unsigned int nr_pages);
>
> static void obj_cgroup_release(struct percpu_ref *ref)
> {
> @@ -163,8 +162,16 @@ static void obj_cgroup_release(struct percpu_ref *ref)
> WARN_ON_ONCE(nr_bytes & (PAGE_SIZE - 1));
> nr_pages = nr_bytes >> PAGE_SHIFT;
>
> - if (nr_pages)
> - obj_cgroup_uncharge_pages(objcg, nr_pages);
> + if (nr_pages) {
> + struct mem_cgroup *memcg;
> +
> + memcg = get_mem_cgroup_from_objcg(objcg);
> + mod_memcg_state(memcg, MEMCG_KMEM, -nr_pages);
> + memcg1_account_kmem(memcg, -nr_pages);
> + if (!mem_cgroup_is_root(memcg))
> + memcg_uncharge(memcg, nr_pages);
> + css_put(&memcg->css);
> + }
>
> spin_lock_irqsave(&objcg_lock, flags);
> list_del(&objcg->list);
Powered by blists - more mailing lists