linux-kernel - Re: [PATCH v2 1/9] mm/vmscan: make active/inactive ratio as 1:1 for anon lru

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: <20200312144749.GG29835@cmpxchg.org>
Date:   Thu, 12 Mar 2020 10:47:49 -0400
From:   Johannes Weiner <hannes@...xchg.org>
To:     js1304@...il.com
Cc:     Andrew Morton <akpm@...ux-foundation.org>, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org, Michal Hocko <mhocko@...nel.org>,
        Hugh Dickins <hughd@...gle.com>,
        Minchan Kim <minchan@...nel.org>,
        Vlastimil Babka <vbabka@...e.cz>,
        Mel Gorman <mgorman@...hsingularity.net>, kernel-team@....com,
        Joonsoo Kim <iamjoonsoo.kim@....com>
Subject: Re: [PATCH v2 1/9] mm/vmscan: make active/inactive ratio as 1:1 for
 anon lru

On Thu, Feb 20, 2020 at 02:11:45PM +0900, js1304@...il.com wrote:
> From: Joonsoo Kim <iamjoonsoo.kim@....com>
> 
> Current implementation of LRU management for anonymous page has some
> problems. Most important one is that it doesn't protect the workingset,
> that is, pages on the active LRU list. Although, this problem will be
> fixed in the following patchset, the preparation is required and
> this patch does it.
> 
> What following patchset does is to restore workingset protection. In this
> case, newly created or swap-in pages are started their lifetime on the
> inactive list. If inactive list is too small, there is not enough chance
> to be referenced and the page cannot become the workingset.
> 
> In order to provide enough chance to the newly anonymous pages, this patch
> makes active/inactive LRU ratio as 1:1.

Patch 8/9 is a revert of this patch. I assume you did this for the
series to be bisectable and partially revertable, but I'm not sure
keeping only the first and second patch would be safe: they reduce
workingset protection quite dramatically on their own (on a 10G system
from 90% of RAM to 50% e.g.) and likely cause regressions.

So while patch 2 is probably a lot better with patch 1 than without,
it seems a bit unnecessary since we cannot keep patch 2 on its own. We
need the rest of the series to make these changes.

On the other hand, the patch is small and obviously correct. So no
strong feelings either way.

> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@....com>

Reviewed-by: Johannes Weiner <hannes@...xchg.org>