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: <20241127111126.71fc31e0@kmaincent-XPS-13-7390>
Date: Wed, 27 Nov 2024 11:11:26 +0100
From: Kory Maincent <kory.maincent@...tlin.com>
To: Oleksij Rempel <o.rempel@...gutronix.de>
Cc: Andrew Lunn <andrew@...n.ch>, "David S. Miller" <davem@...emloft.net>,
 Eric Dumazet <edumazet@...gle.com>, Jakub Kicinski <kuba@...nel.org>, Paolo
 Abeni <pabeni@...hat.com>, Jonathan Corbet <corbet@....net>, Donald Hunter
 <donald.hunter@...il.com>, Rob Herring <robh@...nel.org>, Andrew Lunn
 <andrew+netdev@...n.ch>, Simon Horman <horms@...nel.org>, Heiner Kallweit
 <hkallweit1@...il.com>, Russell King <linux@...linux.org.uk>, Liam Girdwood
 <lgirdwood@...il.com>, Mark Brown <broonie@...nel.org>, Thomas Petazzoni
 <thomas.petazzoni@...tlin.com>, linux-kernel@...r.kernel.org,
 netdev@...r.kernel.org, linux-doc@...r.kernel.org, Kyle Swenson
 <kyle.swenson@....tech>, Dent Project <dentproject@...uxfoundation.org>,
 kernel@...gutronix.de, Maxime Chevallier <maxime.chevallier@...tlin.com>
Subject: Re: [PATCH RFC net-next v3 21/27] net: pse-pd: Add support for
 getting and setting port priority

On Wed, 27 Nov 2024 10:30:43 +0100
Oleksij Rempel <o.rempel@...gutronix.de> wrote:

> On Tue, Nov 26, 2024 at 04:52:28PM +0100, Kory Maincent wrote:
> > On Tue, 26 Nov 2024 16:31:55 +0100
> > Kory Maincent <kory.maincent@...tlin.com> wrote:
> >   
> > > Hello Oleksij,
> > > 
> > > Thanks for your quick reviews!
> > > 
> > > On Tue, 26 Nov 2024 09:38:27 +0100
> > > Oleksij Rempel <o.rempel@...gutronix.de> wrote:
> > >   
>  [...]  
>  [...]  
> > 
> > We already talked about it but a policies per port seems irrelevant to me.
> > https://lore.kernel.org/netdev/ZySR75i3BEzNbjnv@pengutronix.de/
> > How do we compare the priority value of ports that use different budget
> > strategy? How do we manage in the same power domain two ports with
> > different budget strategies or disconnection policies?  
> 
> Good question :)
> 
> > We indeed may need a separate interface to configure the PSE power domain
> > budget strategies and disconnection policies.  
> 
> And a way to upload everything in atomic way, but I see it as
> optimization and can be done separately
> 
> > I think not being able to set the budget evaluation strategy is not relevant
> > for now as we don't have PSE which could support both,  
> 
> Both can be implemented for TI. By constantly polling the channel
> current register, it should be possible to implement dynamic strategy.
> 
> > but being able to set the disconnection policies may be relevant.
> > If we don't add this support to this series how do we decide which is the
> > default disconnection policy supported?  
> 
> Use hard coded one ¯\_(ツ)_/¯

I think we could start with disabled disconnection policy for now.
The user cans still play with the priority value which is really reasonable as
there is as many priority values as PSE ports in the static strategy.

Should we still report it in the status as there is no disconnection policy?
Maybe we could add it at the time we will support several disconnection
policies.

> In terms of user configuration:
> 
> Users only need to set the top allowed priority for each port. For example, if
> a port is set to LRC, it will always be considered first for disconnection
> during a budget violation. The connection order of all LRC ports should be
> preserved.
> 
> If a port is set to Index, it will be preserved until all LRC ports are
> disconnected.
> 
> Setting a port to RR will make it the last in line for disconnection, thus
> ensuring the fairest distribution when other more prioritized policies have
> already been applied. However, in practice, it may never be executed if all
> ports have higher priority policies.

That's a nice brainstorm! With that we will have a first idea when we would
like to really implement the disconnection policies.

Regards,
-- 
Köry Maincent, Bootlin
Embedded Linux and kernel engineering
https://bootlin.com

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ