linux-kernel - Re: [PATCH v3 3/5] zsmalloc: Add a LRU to zs_pool to keep track of zspages in LRU order

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]

Date:   Wed, 16 Nov 2022 15:49:29 -0800
From:   Minchan Kim <minchan@...nel.org>
To:     Johannes Weiner <hannes@...xchg.org>
Cc:     Nhat Pham <nphamcs@...il.com>, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org, ngupta@...are.org,
        senozhatsky@...omium.org, akpm@...ux-foundation.org,
        sjenning@...hat.com, ddstreet@...e.org, vitaly.wool@...sulko.com
Subject: Re: [PATCH v3 3/5] zsmalloc: Add a LRU to zs_pool to keep track of
 zspages in LRU order

On Wed, Nov 16, 2022 at 05:02:57PM -0500, Johannes Weiner wrote:
> On Thu, Nov 10, 2022 at 02:48:32PM -0800, Minchan Kim wrote:
> > On Thu, Nov 10, 2022 at 09:18:31AM -0800, Nhat Pham wrote:
> > > > Please put the LRU logic under config ZSMALLOC_LRU since we don't need the
> > > > additional logic to others.
> > > 
> > > I think the existing CONFIG_ZPOOL would be a good option for this purpose. It
> > > should disable the LRU behavior for non-zswap use case (zram for e.g). The
> > > eviction logic is also currently defined under this. What do you think,
> > > Minchan?
> > 
> > That sounds good.
> > 
> > Sergey and I are working to change zsmalloc zspage size.
> > https://lore.kernel.org/linux-mm/20221031054108.541190-1-senozhatsky@chromium.org/
> > 
> > Could you send a new version once we settle those change down
> > in Andrew's tree to minimize conflict?
> > (Feel free to join the review/discussion if you are also interested ;-))
> 
> I've been reading through that thread, and it doesn't look like it'll
> be ready for the upcoming merge window. (I've tried to contribute

Depending on the discussion status :)

> something useful to it, but it's a fairly difficult tuning problem,
> and I don't know if a sysfs knob is the best answer, either...)

That's the point.

> 
> Would you have any objections to putting Nhat's patches here into 6.2?

I don't want to block due to other issues so no objection from my side.

> 
> It doesn't sound like there was any more feedback (except the trivial
> ifdef around the LRU), and the patches are otherwise ready to go.

In fact, I didn't start the review yet so please post it unless
Sergey objects it.

Thank you.