linux-kernel - Re: [PATCH v5.10] arm64: fix a concurrency issue in emulation_proc

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: <78883f84-dceb-9b24-b1f3-321d84091a85@huawei.com>
Date:   Tue, 31 Jan 2023 16:23:19 +0800
From:   Ruan Jinjie <ruanjinjie@...wei.com>
To:     Greg KH <gregkh@...uxfoundation.org>
CC:     <catalin.marinas@....com>, <will@...nel.org>,
        <haibinzhang@...cent.com>, <hewenliang4@...wei.com>,
        <linux-arm-kernel@...ts.infradead.org>,
        <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v5.10] arm64: fix a concurrency issue in
 emulation_proc_handler()



On 2023/1/31 15:27, Greg KH wrote:
> On Tue, Jan 31, 2023 at 02:52:11PM +0800, ruanjinjie wrote:
>> This patch is addressing an issue in stable linux-5.10 only.
>>
>> In linux-6.1, the related code is refactored in 124c49b1b("arm64:
>> armv8_deprecated: rework deprected instruction handling") and
>> this issue was incidentally fixed. However, the patch changes a lot and
>> is not specific to this issue.
> 
> Then what about 5.15.y?  You can not upgrade to that kernel and have a
> regression, right?

This patch has a pre-dependency af483947d ("arm64: fix oops in
concurrently setting insn_emulation sysctls"), which has not merged into
branches except 5.10.y, so the other branches don't apply.

> 
> And nit, you need a ' ' before the '(' character.
> 
> But why can we just not take the original commit that fixed this issue?
> That way almost always is the best (prevents regressions, makes
> backports easier, is actually tested, etc.) ?

Thank you! It is ok to take the original commit to fix this issue.

> 
> thanks,
> 
> greg k-h