linux-kernel - Re: [PATCH v3] net: atlantic: always deep reset on pm op, fixing null deref regression

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: <4b99fc01-5ab4-d803-4177-a1402ac98164@leemhuis.info>
Date:   Sat, 7 May 2022 15:25:51 +0200
From:   Thorsten Leemhuis <regressions@...mhuis.info>
To:     Manuel Ullmann <labre@...teo.de>
Cc:     Igor Russkikh <irusskikh@...vell.com>,
        linux-kernel@...r.kernel.org, netdev@...r.kernel.org,
        regressions@...ts.linux.dev, davem@...emloft.net,
        ndanilov@...vell.com, kuba@...nel.org, pabeni@...hat.com,
        edumazet@...gle.com, Jordan Leppert <jordanleppert@...tonmail.com>,
        Holger Hoffstätte <holger@...lied-asynchrony.com>,
        koo5 <kolman.jindrich@...il.com>
Subject: Re: [PATCH v3] net: atlantic: always deep reset on pm op, fixing null
 deref regression

On 07.05.22 15:10, Manuel Ullmann wrote:
> Thorsten Leemhuis <regressions@...mhuis.info> writes:
> 
>> On 06.05.22 00:09, Manuel Ullmann wrote:
>>> >From d24052938345d456946be0e9ccc337e24d771c79 Mon Sep 17 00:00:00 2001
>>> Date: Wed, 4 May 2022 21:30:44 +0200
>>>
>>> The impact of this regression is the same for resume that I saw on
>>> thaw: the kernel hangs and nothing except SysRq rebooting can be done.
>>>
>>> The null deref occurs at the same position as on thaw.
>>> BUG: kernel NULL pointer dereference
>>> RIP: aq_ring_rx_fill+0xcf/0x210 [atlantic]
>>>
>>> Fixes regression in commit cbe6c3a8f8f4 ("net: atlantic: invert deep
>>> par in pm functions, preventing null derefs"), where I disabled deep
>>> pm resets in suspend and resume, trying to make sense of the
>>> atl_resume_common deep parameter in the first place.
>>>
>>> It turns out, that atlantic always has to deep reset on pm operations
>>> and the parameter is useless. Even though I expected that and tested
>>> resume, I screwed up by kexec-rebooting into an unpatched kernel, thus
>>> missing the breakage.
>>>
>>> This fixup obsoletes the deep parameter of atl_resume_common, but I
>>> leave the cleanup for the maintainers to post to mainline.
>>
>> FWIW, this section starting here and...
>>
>>> PS: I'm very sorry for this regression.
>>>
>>> Changes in v2:
>>> Patch formatting fixes
>>> - Fix Fixes tag
>>> – Simplify stable Cc tag
>>> – Fix Signed-off-by tag
>>>
>>> Changes in v3:
>>> – Prefixed commit reference with "commit" aka I managed to use
>>>   checkpatch.pl.
>>> - Added Tested-by tags for the testing reporters.
>>> – People start to get annoyed by my patch revision spamming. Should be
>>>   the last one.
>>
>> ...ending here needs should be below the "---" line you already have
>> below. For details see:
>> https://www.kernel.org/doc/html/latest/process/submitting-patches.html

Sorry, I caused a misunderstanding. I didn't handle the above, I'm not
one of the net subsystem developers. Either you submit a v4 fixing this
or hope the net maintainer take care of that when they look at it -- but
I guess they would highly prefer it if you'd address this.

>> BTW, same goes for any "#regzbot" commands (like you had in
>> cbe6c3a8f8f4), as things otherwise get confused when a patch for example
>> is posted as part of a stable/longterm -rc review.
> Good to know. Maybe I could patch the handling-regressions documentation
> to include this.

Yeah, I have already thought about it, but didn't get down to it yet.
Only so much hours in a day.

> submitting-patches could also link the subsystem
> specific documentation, e.g. the netdev FAQ, since they handle patches
> with their more bot tests. Would have helped me a bit. Might be a nice
> exercise for properly formatted patching ;)

I agree that the docs for submitting patches could need a few
improvements and that is likely one of them.

>> But don't worry, no big deal, I handled that :-D Many thx for actually
>> directly getting regzbot involved and taking care of this regression!
> Thank you for the final cleanup and you’re welcome. :) Where did you
> handle this? I can’t seem to find the fixup anywhere, i.e. net-next,
> net, linux-next or lkml.

See above, I only handled the regzbot issue, not the issue with this
patch. Sorry for not being clear enough in my wording.

> [...]

Ciao, Thorsten