linux-kernel - Re: [PATCH v4 2/2] vmscan,cgroup: apply mems

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <aAcV7GmTJGbC1R_s@gourry-fedora-PF4VCD3F>
Date: Tue, 22 Apr 2025 00:07:08 -0400
From: Gregory Price <gourry@...rry.net>
To: Waiman Long <llong@...hat.com>
Cc: linux-mm@...ck.org, cgroups@...r.kernel.org,
	linux-kernel@...r.kernel.org, kernel-team@...a.com,
	hannes@...xchg.org, mhocko@...nel.org, roman.gushchin@...ux.dev,
	shakeel.butt@...ux.dev, muchun.song@...ux.dev, tj@...nel.org,
	mkoutny@...e.com, akpm@...ux-foundation.org
Subject: Re: [PATCH v4 2/2] vmscan,cgroup: apply mems_effective to reclaim

On Mon, Apr 21, 2025 at 10:02:22PM -0400, Waiman Long wrote:
> > +bool cpuset_node_allowed(struct cgroup *cgroup, int nid)
> > +{
> > +	struct cgroup_subsys_state *css;
> > +	struct cpuset *cs;
> > +	bool allowed;
> > +
> > +	/*
> > +	 * In v1, mem_cgroup and cpuset are unlikely in the same hierarchy
> > +	 * and mems_allowed is likely to be empty even if we could get to it,
> > +	 * so return true to avoid taking a global lock on the empty check.
> > +	 */
> > +	if (!cpuset_v2())
> > +		return true;
> > +
> > +	css = cgroup_get_e_css(cgroup, &cpuset_cgrp_subsys);
> > +	if (!css)
> > +		return true;
> > +
> > +	cs = container_of(css, struct cpuset, css);
> > +	rcu_read_lock();
> 
> Sorry, I missed the fact that cgroup_get_e_css() will take a reference to
> the css and so it won't go away. In that case, rcu_read_lock() isn't really
> needed. However, I do want a comment to say that accessing effective_mems
> should normally requrie taking either a cpuset_mutex or callback_lock, but
> is skipped in this case to avoid taking a global lock in the reclaim path at
> the expense that the result may be inaccurate in some rare cases.
> 

I'll add a differential patch here.

~Gregory