linux-kernel - Re: [PATCH v2] cpu/hotplug: Do not bail-out in DYING/STARTING sections

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: <e589b71b-9fa2-0ca1-af2e-0c513b9250cd@posteo.net>
Date:   Sun, 29 May 2022 18:01:14 +0000
From:   Derek Dolney <z23@...teo.net>
To:     Peter Zijlstra <peterz@...radead.org>,
        Vincent Donnefort <vdonnefort@...gle.com>
Cc:     tglx@...utronix.de, linux-kernel@...r.kernel.org,
        vschneid@...hat.com, kernel-team@...roid.com
Subject: Re: [PATCH v2] cpu/hotplug: Do not bail-out in DYING/STARTING
 sections

FYI there is also now a patch to fix the driver bug in testing by the
tboot devs at the moment, you could monitor the progress here:

https://sourceforge.net/p/tboot/mailman/message/37659164/

I tested this patch and it works for me.

Derek

On 5/26/22 6:15 AM, Peter Zijlstra wrote:
> On Thu, May 26, 2022 at 09:24:28AM +0100, Vincent Donnefort wrote:
>> On Wed, May 25, 2022 at 06:52:48PM +0200, Peter Zijlstra wrote:
>>> On Mon, May 23, 2022 at 05:05:36PM +0100, Vincent Donnefort wrote:
>>>> The DYING/STARTING callbacks are not expected to fail. However, as reported
>>>> by Derek, drivers such as tboot are still free to return errors within
>>>> those sections. In that case, there's nothing the hotplug machinery can do,
>>>> so let's just proceed and log the failures.
>>>>
>>>
>>> I'm confused. Why isn't this a driver bug?
>>
>> It is a entirely a driver bug which has been reported already. but 453e41085183
>> (cpu/hotplug: Add cpuhp_invoke_callback_range()) changed the behaviour so I
>> thought it would be worth to revert to the original one which is to not break
>> the entire up/down for a single driver error.
> 
> Ah I see. Fair enough I suppose.
>