netdev - Re: [RFC] Questionable RCU/BH usage in cgw_create

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] [day] [month] [year] [list]

Message-ID: <0338581d-89f0-4cca-a360-7693a27eba37@hartkopp.net>
Date: Thu, 21 Dec 2023 13:43:19 +0100
From: Oliver Hartkopp <socketcan@...tkopp.net>
To: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
Cc: Marc Kleine-Budde <mkl@...gutronix.de>, linux-can@...r.kernel.org,
 netdev@...r.kernel.org, Thomas Gleixner <tglx@...utronix.de>
Subject: Re: [RFC] Questionable RCU/BH usage in cgw_create_job().

Hi Sebastian,

I've sent a RFC patch to be reviewed here:

https://lore.kernel.org/linux-can/20231221123703.8170-1-socketcan@hartkopp.net/T/#u

I hope your suggestion to use rcu_replace_pointer()
in the implemented way fits the requirements.

Best regards,
Oliver

On 2023-11-30 20:56, Oliver Hartkopp wrote:
> Hi Sebastian,
> 
> On 30.11.23 17:43, Sebastian Andrzej Siewior wrote:
>> On 2023-10-31 17:52:47 [+0100], To Oliver Hartkopp wrote:
> 
>>> The point is to replace/ update cf_mod at runtime while following RCU
>>> rules so always either new or the old object is observed. Never an
>>> intermediate step.
>>
>> Do you want me to take care of it?
> 
> Yes, sorry.
> 
> In fact I've searched some time what would fit best without getting a 
> clear picture.
> 
> As the changes triggered by the netlink update should come into action 
> with the next processed CAN frame I have thought about adding a shadow 
> 'mod' structure which is written instantly.
> And then a flag could be set, that is switched by the next incoming CAN 
> frame.
> 
> I just would have a problem with performing some memory allocation for 
> the 'mod' updates which might take some unpredictable time.
> 
> If you have some cool ideas please let me know. I'm unsure what is the 
> most effective and performant approach for this use case.
> 
> Many thanks,
> Oliver
>