Hi Niklas,
On Sat, Jul 6, 2024 at 10:39 AM Niklas Söderlund
<niklas.soderlund+renesas@xxxxxxxxxxxx> wrote:
> On 2024-07-05 23:49:56 +0200, Marek Vasut wrote:
> > On 7/3/24 11:36 AM, Niklas Söderlund wrote:
> > > On 2024-07-03 10:24:26 +0200, Geert Uytterhoeven wrote:
> > > > Niklas: commit 54bf0c27380b95a2 ("arm64: dts: renesas: r8a779g0: Use
> > > > MDIO node for all AVB devices") did keep the reset-gpios property in
> > > > the PHY node. I guess it should be moved one level up?
> > >
> > > It's possible to have a rest-gpios property both in the mdio node and
> > > the phy node. The former resets the whole bus while the later a single
> > > PHY, at least that's my understanding.
> >
> > My understanding of reset GPIO in the MDIO node is that it is used in case
> > there might be multiple PHYs with shared reset GPIO on the same MDIO bus.
> > Like on the NXP iMX28 .
>
> There is a use-case for a single PHY on the MDIO bus too, at least in
> Linux as I understand it. If the boot process leave the PHY in a bad
> state which prevents it from being probed. A GPIO reset in the MDIO node
> is used when the MDIO bus is registered thus resetting all (in this
> use-case the one) PHYs which later allows them to be probed. A GPIO
> reset on the PHY node is only used after a PHY have been probed, at
> least that is my understanding.
The reset is also asserted on driver unbind.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
