linux-kernel - Re: [PATCH v3] s390/uvdevice: Support longer secret lists

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: <479b6b3e-72d4-4755-acf4-11011b822682@linux.ibm.com>
Date: Wed, 6 Nov 2024 13:18:46 +0100
From: Steffen Eiden <seiden@...ux.ibm.com>
To: Heiko Carstens <hca@...ux.ibm.com>, Janosch Frank <frankja@...ux.ibm.com>
Cc: linux-kernel@...r.kernel.org, linux-s390@...r.kernel.org,
        Ingo Franzki <ifranzki@...ux.ibm.com>,
        Christoph Schlameuss <schlameuss@...ux.ibm.com>,
        Claudio Imbrenda <imbrenda@...ux.ibm.com>, borntraeger@...ux.ibm.com
Subject: Re: [PATCH v3] s390/uvdevice: Support longer secret lists



On 11/6/24 10:13 AM, Heiko Carstens wrote:
> On Wed, Nov 06, 2024 at 09:54:33AM +0100, Janosch Frank wrote:
>> On 11/6/24 9:10 AM, Heiko Carstens wrote:
>>> On Mon, Nov 04, 2024 at 04:36:09PM +0100, Steffen Eiden wrote:
>>>> +		copy_len = sizeof(list->secrets[0]) * list->num_secr_stored;
>>>> +		WARN_ON(copy_len > sizeof(list->secrets));
>>>
>>> Is this really possible? Without checking the documentation I guess
>>> this is not possible and therefore the WARN_ON() should be removed.
>>>
>>
>> This happening requires a FW error, no?
>> list->num_secr_stored is reported by FW and would need to be >85.
>>
>> We could clamp it down to 85 secrets / 4k - sizeof(header) with a
>> WARN_ON_ONCE() to catch FW problems if that suits you more.
> 
> If this would be an *error* why even add this check? We have tons of
> code without doing sanity checks for firmware provided values - where
> should we start or end?
Yes, this would be an error.

> 
> So imho: either remove this check if this would be firmware error,
> unless there is a good reason do keep this check, or if this is not an
> error convert to WARN_ON_ONCE() and limit the copy_to_user().
OK. I'll remove the check. Sending a fix-up reply.