linux-kernel - Re: [PATCH] mm/hugetlb: fix kernel NULL pointer dereference when replacing free hugetlb folios

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: <299A777D-DC94-4724-BAD7-DD02F7CDBAFA@linux.dev>
Date: Fri, 23 May 2025 11:56:11 +0800
From: Muchun Song <muchun.song@...ux.dev>
To: Ge Yang <yangge1116@....com>
Cc: Oscar Salvador <osalvador@...e.de>,
 akpm@...ux-foundation.org,
 linux-mm@...ck.org,
 linux-kernel@...r.kernel.org,
 stable@...r.kernel.org,
 21cnbao@...il.com,
 david@...hat.com,
 baolin.wang@...ux.alibaba.com,
 liuzixing@...on.cn
Subject: Re: [PATCH] mm/hugetlb: fix kernel NULL pointer dereference when
 replacing free hugetlb folios



> On May 23, 2025, at 11:46, Ge Yang <yangge1116@....com> wrote:
> 
> 
> 
> 在 2025/5/23 11:27, Muchun Song 写道:
>>> On May 23, 2025, at 03:32, Oscar Salvador <osalvador@...e.de> wrote:
>>> 
>>> On Thu, May 22, 2025 at 08:39:39PM +0800, Muchun Song wrote:
>>>> But I think we could use "folio_order() > MAX_PAGE_ORDER" to replace the check
>>>> of hstate_is_gigantic(), right? Then ee could remove the first parameter of hstate
>>>> from alloc_and_dissolve_hugetlb_folio() and obtain hstate in it.
>>> 
>>> Yes, I think we can do that.
>>> So something like the following (compily-tested only) maybe?
>>> 
>>> From d7199339e905f83b54d22849e8f21f631916ce94 Mon Sep 17 00:00:00 2001
>>> From: Oscar Salvador <osalvador@...e.de>
>>> Date: Thu, 22 May 2025 19:51:04 +0200
>>> Subject: [PATCH] TMP
>>> 
>>> ---
>>>  mm/hugetlb.c | 38 +++++++++-----------------------------
>>>  1 file changed, 9 insertions(+), 29 deletions(-)
>> Pretty simple. The code LGTM.
>> Thanks.
> 
> Thanks.
> 
> The implementation of alloc_and_dissolve_hugetlb_folio differs between kernel 6.6 and kernel 6.15. To facilitate backporting, I'm planning to submit another patch based on Oscar Salvador's suggestion.

A separate improving patch LGTM.

>