[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <a967118d-ee79-4486-b6ad-c0ecb173e2ea@redhat.com>
Date: Thu, 17 Jul 2025 11:25:20 +0200
From: David Hildenbrand <david@...hat.com>
To: Lorenzo Stoakes <lorenzo.stoakes@...cle.com>
Cc: Baolin Wang <baolin.wang@...ux.alibaba.com>,
Hugh Dickins <hughd@...gle.com>, Andrew Morton <akpm@...ux-foundation.org>,
Matthew Wilcox <willy@...radead.org>, ziy@...dia.com,
Liam.Howlett@...cle.com, npache@...hat.com, ryan.roberts@....com,
dev.jain@....com, baohua@...nel.org, vbabka@...e.cz, rppt@...nel.org,
surenb@...gle.com, mhocko@...e.com, linux-mm@...ck.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] mm: fault in complete folios instead of individual
pages for tmpfs
On 17.07.25 11:20, Lorenzo Stoakes wrote:
> On Thu, Jul 17, 2025 at 10:22:47AM +0200, David Hildenbrand wrote:
>> I recall [1]. But that would, of course, only affect the RSS of a program
>> and not the actual memory consumption (the large folio resides in memory
>> ...).
>>
>> The comments in the code spells that out: "inflating the RSS of the
>> process."
>
> RSS is a pretty sucky measure though, as covered in depth by Vlastimil in
> his KR 2024 talk [2]. So maybe we don't need to worry _so_ much about this
> :)
Well, I pointed an an actual issue where people *did* worry about ;)
I also think that it's probably fine, but some apps/environments
apparently don't enjoy suddenly seeing spikes in RSS. Maybe these are so
specific that only distributions like in the report have to worry about
that (good luck once the toggles are gone, lol).
--
Cheers,
David / dhildenb
Powered by blists - more mailing lists