netdev - Re: [patch net-next] devlink: remove reload failed checks in params get/set callbacks

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: <ZK+9Q/5NC7/eNGH8@nanopsycho>
Date: Thu, 13 Jul 2023 11:00:51 +0200
From: Jiri Pirko <jiri@...nulli.us>
To: Jakub Kicinski <kuba@...nel.org>
Cc: Ido Schimmel <idosch@...dia.com>, netdev@...r.kernel.org,
	pabeni@...hat.com, davem@...emloft.net, edumazet@...gle.com,
	moshe@...dia.com
Subject: Re: [patch net-next] devlink: remove reload failed checks in params
 get/set callbacks

Wed, Jul 12, 2023 at 09:21:03PM CEST, kuba@...nel.org wrote:
>On Wed, 12 Jul 2023 17:20:40 +0200 Jiri Pirko wrote:
>> >> Back then, it was a possible fix. Alternative way to fix this was to
>> >> make sure drivers register/unregister params in the code where it is
>> >> ensured that the data accessed by params callbacks are available.
>> >> But that was problematic as the list of params wes static durint  
>> >
>> >s/wes/was/
>> >s/durint/during/  
>> 
>> Maintainers, I will send v2 with these typos fixed tomorrow, if these
>> are not any other comments.
>
>Feel free to toss in
>
>pw-bot: changes-requested

I see, is this documented somewhere?

>
>so we don't have to update the status manually.
>
>The commit message would benefit from a rewrite, TBH I don't understand
>half of it, specially:

Will do.

>
>  Alternative way to fix this was to make sure drivers
>  register/unregister params in the code where it is ensured that 
>  the data accessed by params callbacks are available.
>
>Can't parse.
>
>  list of params [was] static [during] devlink instance being
>  registered.
>
>You mean that list of params can't change after the instance was
>registered?

Yeah, that was a limitation in history IIRC.


>
>  register/unregister params alongside with the data it touches
>
>Meaning params for a sub-object are registered when the sub-object 
>is registered? An example could help clarify the meaning.
>
>> >> devlink instance being registered.
>> >> 
>> >> Eventually this limitation was lifted and also the alternative fix
>> >> (which also fixed another issue) was done for mlxsw by
>> >> commit 74cbc3c03c82 ("mlxsw: spectrum_acl_tcam: Move devlink param to TCAM code").
>> >> 
>> >> The checks are no longer relevant, each driver should make sure to
>> >> register/unregister params alongside with the data it touches. Remove
>> >> the checks.