Re: [PATCH net] net: dsa: microchip: keep compatibility with device tree blobs with no phy-mode

From: Vladimir Oltean <olteanv@gmail.com>
To: "Alvin Šipraga" <ALSI@bang-olufsen.dk>
Cc: Vladimir Oltean <vladimir.oltean@nxp.com>,
	"netdev@vger.kernel.org" <netdev@vger.kernel.org>,
	Woojung Huh <woojung.huh@microchip.com>,
	"UNGLinuxDriver@microchip.com" <UNGLinuxDriver@microchip.com>,
	Andrew Lunn <andrew@lunn.ch>,
	Vivien Didelot <vivien.didelot@gmail.com>,
	Florian Fainelli <f.fainelli@gmail.com>,
	"David S. Miller" <davem@davemloft.net>,
	Eric Dumazet <edumazet@google.com>,
	Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
	Russell King <linux@armlinux.org.uk>,
	Michael Grzeschik <m.grzeschik@pengutronix.de>,
	Oleksij Rempel <linux@rempel-privat.de>,
	Thorsten Leemhuis <regressions@leemhuis.info>,
	Rasmus Villemoes <linux@rasmusvillemoes.dk>,
	Craig McQueen <craig@mcqueen.id.au>
Subject: Re: [PATCH net] net: dsa: microchip: keep compatibility with device tree blobs with no phy-mode
Date: Thu, 18 Aug 2022 18:18:06 +0300	[thread overview]
Message-ID: <20220818151806.yv5crkdab6tpkvph@skbuf> (raw)
In-Reply-To: <20220818150644.ic6pf5arbh34lr5z@bang-olufsen.dk>

On Thu, Aug 18, 2022 at 03:06:44PM +0000, Alvin Šipraga wrote:
> Reviewed-by: Alvin Šipraga <alsi@bang-olufsen.dk>

Thanks.

> Some quick grepping shows at least a few other drivers which do not set
> PHY_INTERFACE_MODE_GMII for their ports with internal PHY:
> 
>     bcm_sf2
>     ar9331 (*)
>     lantiq_gswip
> 
> Should these be "fixed" too? Or only if somebody reports a regression?
> 
> (*) I note that ar9331 ought not to rely on DSA workarounds, per your
> other patchset, so I there is actually no need to "fix" that one, since
> the new validation you are introducing will require a phy-mode to be
> specified for those switches' ports anyway.

Note that my patch set which you are talking about
https://patchwork.kernel.org/project/netdevbpf/cover/20220818115500.2592578-1-vladimir.oltean@nxp.com/
only enforces phy-mode presence for CPU and DSA ports. Whereas internal
PHYs are more typical on user ports. So the problem is equally
applicable to ar9331, as long as it won't specify a phy-mode but rely on
DSA's assumption that user ports with lacking DT descriptions want to
connect to an internal PHY.

IDK, if I had infinite time I'd go fix every driver. I don't know why
some have GMII and some don't, and I'm not exactly keen to go study
every driver's subtleties unless a problem is reported.