netdev - Re: [PATCH net-next 0/2] ice: support FEC automatic disable

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: <2602a3a2-0bb4-8da7-0691-aaa5566278dd@nvidia.com>
Date:   Thu, 25 Aug 2022 10:08:32 +0300
From:   Gal Pressman <gal@...dia.com>
To:     Jakub Kicinski <kuba@...nel.org>
Cc:     Jacob Keller <jacob.e.keller@...el.com>,
        Saeed Mahameed <saeedm@...dia.com>, netdev@...r.kernel.org
Subject: Re: [PATCH net-next 0/2] ice: support FEC automatic disable

On 24/08/2022 19:25, Jakub Kicinski wrote:
> On Wed, 24 Aug 2022 16:35:41 +0300 Gal Pressman wrote:
>> On 23/08/2022 18:04, Jacob Keller wrote:
>>> 2) always treat ETHTOOL_FEC_AUTO as "automatic + allow disable"
>>>
>>>   This could work, but it means that behavior will differ depending on the
>>>   firmware version. Users have no way to know that and might be surprised to
>>>   find the behavior differ across devices which have different firmware
>>>   which do or don't support this variation of automatic selection.  
>> Hi Jacob,
>> This is exactly how it's already implemented in mlx5, and I don't really
>> understand how firmware version is related? Is it specific to your
>> device firmware?
>> Maybe you can workaround that in the driver?
>>
>> I feel like we're going the wrong way here having different flags
>> interpretations by different drivers.
> Hm, according to my notes the drivers supporting a single bit and
> multiple bits were evenly split when I wrote the docs. Either way
> we're going to make someone unhappy?

IMHO, passing both auto and off is confusing.
BTW, I think this is also inconsistent with the direction SONiC took (or
is taking?) with the auto fec mode interpretation?

> While we're talking about mlx5 FEC, what does it report on get?
> I wasn't sure if it reports the supported or configured mode.

I think it's the configured mode and the active mode.