linux-kernel - Re: [PATCH] Revert "hugetlb: make hugetlb depends on SYSFS or SYSCTL"

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: <a770f778-81c2-3325-b44d-43a95134c189@huawei.com>
Date: Thu, 28 Aug 2025 11:29:13 +0800
From: Miaohe Lin <linmiaohe@...wei.com>
To: Andrew Morton <akpm@...ux-foundation.org>
CC: <muchun.song@...ux.dev>, <osalvador@...e.de>, <david@...hat.com>,
	<linux-mm@...ck.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] Revert "hugetlb: make hugetlb depends on SYSFS or SYSCTL"

On 2025/8/28 10:49, Andrew Morton wrote:
> On Thu, 28 Aug 2025 10:31:51 +0800 Miaohe Lin <linmiaohe@...wei.com> wrote:
> 
>> On 2025/8/27 11:35, Andrew Morton wrote:
>>> On Tue, 26 Aug 2025 11:09:55 +0800 Miaohe Lin <linmiaohe@...wei.com> wrote:
>>>
>>>> Commit f8142cf94d47 ("hugetlb: make hugetlb depends on SYSFS or SYSCTL")
>>>> added dependency on SYSFS or SYSCTL but hugetlb can be used without SYSFS
>>>> or SYSCTL. So this dependency is wrong and should be removed.
>>>>
>>>> This reverts commit f8142cf94d4737ea0c3baffb3b9bad8addcb9b6b.
>>>
>>> f8142cf94d47 said:
>>>
>>>     If CONFIG_SYSFS and CONFIG_SYSCTL are both undefined, hugetlb
>>>     doesn't work now as there's no way to set max huge pages.  Make
>>>     sure at least one of the above configs is defined to make hugetlb
>>>     works as expected.
>>>
>>> So there is now a way to set max huge pages?  A reference tot he
>>> commit which made f8142cf94d47 unneeded might be helpful?
>>
>> The commit is just wrong. It overlooked the scenario of using hugetlb through boot parameters
>> when it was submitted.
>>
> 
> OK.  Could we please have a description of the user-visible effect and
> a decision on whether we should backport the fix?

For users with CONFIG_SYSFS or CONFIG_SYSCTL on, there should be no difference.
For users have CONFIG_SYSFS and CONFIG_SYSCTL both undefined, hugetlbfs can still
works perfectly well through cmdline except a possible kismet warning[1] when select
CONFIG_HUGETLBFS. IMHO, it might not worth a backport.


[1]: https://lore.kernel.org/all/5c99458f-4a91-485f-8a35-3618a992e2e4@csgroup.eu/

Thanks.
.