linux-kernel - Re: [PATCH] zswap: improve memory.zswap.writeback inheritance

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] [day] [month] [year] [list]

Message-ID: <CAKEwX=MPtsX0OexWfL3j-4TcPuAdoDWP4E2BzehrkFs67bTv9Q@mail.gmail.com>
Date: Fri, 27 Sep 2024 09:40:56 -0700
From: Nhat Pham <nphamcs@...il.com>
To: Michal Koutný <mkoutny@...e.com>
Cc: Ivan Shapovalov <intelfx@...elfx.name>, linux-kernel@...r.kernel.org, 
	Mike Yuan <me@...dnzj.com>, Tejun Heo <tj@...nel.org>, Zefan Li <lizefan.x@...edance.com>, 
	Johannes Weiner <hannes@...xchg.org>, Jonathan Corbet <corbet@....net>, Yosry Ahmed <yosryahmed@...gle.com>, 
	Chengming Zhou <chengming.zhou@...ux.dev>, Michal Hocko <mhocko@...nel.org>, 
	Roman Gushchin <roman.gushchin@...ux.dev>, Shakeel Butt <shakeel.butt@...ux.dev>, 
	Muchun Song <muchun.song@...ux.dev>, Andrew Morton <akpm@...ux-foundation.org>, 
	Chris Li <chrisl@...nel.org>, cgroups@...r.kernel.org, linux-doc@...r.kernel.org, 
	linux-mm@...ck.org
Subject: Re: [PATCH] zswap: improve memory.zswap.writeback inheritance

On Fri, Sep 27, 2024 at 8:01 AM Michal Koutný <mkoutny@...e.com> wrote:
>
>
> What about assigning this semantic to an empty string ("")?
> That would be the default behavior and also the value shown when reading
> the file (to distinguish this for explicitly configured values).

Yeah that's better than -1, I agree. Still a bit confusing, but at
least the semantic is "we are not making a choice at the memcg".

>
> (The weirdness of 0, 1, -1, -1, -1  would remain. Maybe switching this
> via the mount option could satisfy any user. Admittedly, I tend to
> confuse this knob with swap.max.)

Yeah a mount option, or in general some sort of global knob (with
proper documentation) would be preferable.

And yeah, I hear you with the swap.max confusion. That's why I tried
to make it explicit in the documentation, because the difference is
subtle and can trip up users. Hopefully, when zswap and swap are
decoupled, users can conceptualize them as two separate tiers, and the
confusion will lessen...