Messages in this thread | | | Date | Tue, 17 Jan 2023 14:43:03 +0100 | From | Andrew Lunn <> | Subject | Re: [PATCH] net: mdio: force deassert MDIO reset signal |
| |
On Mon, Jan 16, 2023 at 08:39:31AM +0000, Pierluigi Passaro wrote: > On Mon, Jan 16, 2023 at 12:55 AM Andrew Lunn <andrew@lunn.ch> wrote: > > > Specifying the ID as part of the compatible string works for clause 22 PHYs, > > > but for clause 45 PHYs it does not work. The code always wants to read the > > > ID from the PHY itself. But I do not understand things well enough to tell > > > whether that's a fundamental restriction of C45 or just our implementation > > > and the implementation can be changed to fix it. > > > > > > Do you have some thoughts on this? > > > > Do you have more details about what goes wrong? Which PHY driver is > > it? What compatibles do you put into DT for the PHY? > > > We use both AR8033 and ADIN1300: these are 10/100/1000 PHYs. > They are both probed after the MDIO bus, but skipped if the reset was > asserted while probing the MDIO bus. > However, for iMX6UL and iMX7 we use C22, not C45.
I never said it did. Please read the actual emails, then you would of noticed we have go off on tangent and are trying to fix another issue.
Andrew
| |