netdev - Re: [RFC PATCH] net: introduce HW Rate Limiting Driver API

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: <b6512bf1-5343-4130-a962-db2617a85fda@redhat.com>
Date: Tue, 30 Jul 2024 15:34:24 +0200
From: Paolo Abeni <pabeni@...hat.com>
To: Jiri Pirko <jiri@...nulli.us>
Cc: netdev@...r.kernel.org, Jakub Kicinski <kuba@...nel.org>,
 Madhu Chittim <madhu.chittim@...el.com>,
 Sridhar Samudrala <sridhar.samudrala@...el.com>,
 Simon Horman <horms@...nel.org>, John Fastabend <john.fastabend@...il.com>,
 Sunil Kovvuri Goutham <sgoutham@...vell.com>,
 Jamal Hadi Salim <jhs@...atatu.com>
Subject: Re: [RFC PATCH] net: introduce HW Rate Limiting Driver API

Hi,

On 7/30/24 14:18, Jiri Pirko wrote:
> Wed, May 08, 2024 at 10:20:51PM CEST, pabeni@...hat.com wrote:
> 
>> + * NET_SHAPER_SCOPE_QUEUE are available on both PFs and VFs devices.
> 
> This is interesting. Do you mean you can put a shaper on a specific VF
> queue from hypervisor? I was thinking about it recently, I have some
> concerns.
> 
> In general a nic user expects all queues to behave in the same way,
> unless he does some sort of configuration (dcb for example).
> VF (the VM side) is not different, it's also a nic.
> 
> If you allow the hypervisor to configure shapers on specifig VF queues,
> you are breaking VM's user expectation. He did not configure any
> different queue treating, yet they are treated differently.
> 
> Is that okay? What do you think?

I'm unsure why you are looking to this old version...

The idea to allow configuring the VF's queues from the hypervisor has 
been removed from the most recent version, for roughly the same reasons 
you mention above.

Cheers,

Paolo