netdev - Re: [PATCH net-next RFC v4 4/5] net: Let the active time stamping layer be selectable.

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: <ZF1Y3OurlvukFBs1@shell.armlinux.org.uk>
Date: Thu, 11 May 2023 22:06:36 +0100
From: "Russell King (Oracle)" <linux@...linux.org.uk>
To: Jakub Kicinski <kuba@...nel.org>
Cc: Vladimir Oltean <vladimir.oltean@....com>,
	Köry Maincent <kory.maincent@...tlin.com>,
	netdev@...r.kernel.org, glipus@...il.com,
	maxime.chevallier@...tlin.com, vadim.fedorenko@...ux.dev,
	richardcochran@...il.com, gerhard@...leder-embedded.com,
	thomas.petazzoni@...tlin.com, krzysztof.kozlowski+dt@...aro.org,
	robh+dt@...nel.org
Subject: Re: [PATCH net-next RFC v4 4/5] net: Let the active time stamping
 layer be selectable.

On Thu, May 11, 2023 at 09:25:39AM -0700, Jakub Kicinski wrote:
> Yes, we don't want to lose the simplification benefit for the common
> cases. I think we should make the "please call me for PHY requests"
> an opt in.
> 
> Annoyingly the "please call me for PHY/all requests" needs to be
> separate from "this MAC driver supports PHY timestamps". Because in
> your example the switch driver may not in fact implement PHY stamping,
> it just wants to know about the configuration.
> 
> So we need a bit somewhere (in ops? in some other struct? in output 
> of get_ts?) to let the driver declare that it wants to see all TS
> requests. (I've been using bits in ops, IDK if people find that
> repulsive or neat :))

I haven't thought this through fully, but just putting this out there
as a potential suggestion...

Would it help at all if we distilled the entire timestamping interface
into a separate set of ops which are registered independently from the
NDO, and NDO has a call to get the ops for the layer being configured?

That would allow a netdev driver to return the ops appropriate for the
MAC layer or its own PHY layer, or maybe phylib.

In the case of phylib, as there is a raft of drivers that only bind to
their phylib PHY in the NDO open method, we'd need to figure out how
to get the ops for the current mode at that time.

There's probably lots I've missed though...

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 80Mbps down 10Mbps up. Decent connectivity at last!