linux-kernel - Re: [PATCH 6.15] mm/vma: add give_up_on_oom option on modify/merge, use in uffd release

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: <Z94EnmSpBO2L-Mv1@x1.local>
Date: Fri, 21 Mar 2025 20:30:22 -0400
From: Peter Xu <peterx@...hat.com>
To: Lorenzo Stoakes <lorenzo.stoakes@...cle.com>
Cc: "Liam R. Howlett" <Liam.Howlett@...cle.com>,
	Pedro Falcato <pfalcato@...e.de>,
	Andrew Morton <akpm@...ux-foundation.org>,
	Vlastimil Babka <vbabka@...e.cz>, Jann Horn <jannh@...gle.com>,
	linux-mm@...ck.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 6.15] mm/vma: add give_up_on_oom option on modify/merge,
 use in uffd release

On Fri, Mar 21, 2025 at 05:16:44PM +0000, Lorenzo Stoakes wrote:
> On Fri, Mar 21, 2025 at 11:27:34AM -0400, Liam R. Howlett wrote:
> > +cc Peter due to uffd interests
> 
> Gentle nudge for Peter to make himself uffd maintainer :) I am not a fan of
> this 'happen to know person X often touches Y' stuff, this is what
> MAINTAINERS is for. If you're not there, good chance I won't cc you...
> 
> I also strongly feel we need somebody to take overall responsibility for
> uffd at this point.
> 
> Pints will be bought for this person in Montreal ;)

Thanks for the offer, though I will be absent from lsfmm this year.  I sent
a maintainer file change here, though:

https://lore.kernel.org/linux-mm/20250322002124.131736-1-peterx@redhat.com/

Maybe someone would like to be the 2nd reviewer, and if he/she would be at
the conference maybe the pints will still count? :)

[...]

> > I hate both of them, and I (mostly) blame uffd.  Some blame is on syzbot
> > for exposing me to this unreachable failure. ;-)
> 
> So do I.

I agree with Jann; AFAIU, userfaultfd is innocent, if it used to work.

I don't think I have followed much on the recent vma mgmt changes.  So I am
almost of no use here.. I'll leave that to you experts.  Beers indeed might
help.

Said that, it'll definitely be very appreciated if the old behavior can be
fully recovered, so release() will never fail at such -ENOMEM.

Thanks,

-- 
Peter Xu