netdev - Re: [PATCH iwl-next v1] ice: Add support for devlink loopback param.

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <340900d4-b30a-4387-9ce2-1971e8d8024c@intel.com>
Date: Wed, 6 Dec 2023 18:26:13 -0600
From: "Kaminski, Pawel" <pawel.kaminski@...el.com>
To: Jakub Kicinski <kuba@...nel.org>
Cc: intel-wired-lan@...osl.org, netdev@...r.kernel.org,
 Michal Wilczynski <michal.wilczynski@...el.com>,
 Przemek Kitszel <przemyslaw.kitszel@...el.com>
Subject: Re: [PATCH iwl-next v1] ice: Add support for devlink loopback param.

On 2023-12-01 20:37, Jakub Kicinski wrote:
> On Fri,  1 Dec 2023 15:59:49 -0800 Pawel Kaminski wrote:
>> Add support for devlink loopback param. Supported values are "enabled",
>> "disabled" and "prioritized". Default configuration is set to "enabled.
>>
>> By default loopback traffic BW is locked to PF configured BW.
> 
> First off - hairpin-bandwidth or some such would be a much better name.
> Second - you must explain every devlink param in Documentation/
> 
> Also admission ctrl vs prioritizing sounds like different knobs.

While at certain abstraction level I agree, in my opinion it is not 
worth here to divide this to separate knobs, since underlying logic (FW) 
doesn't follow that anyways. It is driver specific and extremely 
unlikely to change in the future. Hopefully next gen card will not need 
this knob at all.

>> HW is
>> capable of higher speeds on loopback traffic. Loopback param set to
>> "prioritized" enables HW BW prioritization for VF to VF traffic,
>> effectively increasing BW between VFs. Applicable to 8x10G and 4x25G
>> cards.
> 
> Not very clear what this means...
> So the VFs are Tx bandwidth limited to link speed.
> How does the device know it can admit extra traffic?
> Presumably this doesn't affect rates set by devlink rate?

I will rewrite the description and explanation in v2 and include 
documentation change.

Thank you,
PK