linux-kernel - Re: [PATCH] arm64: mm: fix direct map use over accounting

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: <aR8-7N4_ICuJW3pM@arm.com>
Date: Thu, 20 Nov 2025 16:16:44 +0000
From: Catalin Marinas <catalin.marinas@....com>
To: Ryan Roberts <ryan.roberts@....com>
Cc: Yang Shi <yang@...amperecomputing.com>, will@...nel.org, cl@...two.org,
	linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] arm64: mm: fix direct map use over accounting

On Thu, Nov 20, 2025 at 03:11:21PM +0000, Ryan Roberts wrote:
> On 19/11/2025 23:57, Yang Shi wrote:
> > The create_pgd_mapping() families may be used in other paths other than
> > creating the linear map. And spliting kernel page table may be used for
> > non-linear map either. So we need check whether the address is linear
> > map address or not before accounting to direct map use. Encapsulate the
> > linear map address check in the new add/sub helpers. The new helpers are
> > nop if CONFIG_PROC_FS is not enabled.
[...]
> Overall, my opinion is that this patch is not quite cooked yet. Personally I'd
> prefer to see the first half reverted then we can wait another cycle until this
> is ready and apply it all in one go.

Thanks for the feedback Ryan. I'll revert and queue it again later once
you acked it.

-- 
Catalin