Messages in this thread | | | From | Tim Harvey <> | Date | Wed, 9 Feb 2022 08:31:23 -0800 | Subject | Re: [PATCH net v3] net: phy: intel-xway: enable integrated led functions |
| |
On Fri, Feb 4, 2022 at 2:54 PM Andrew Lunn <andrew@lunn.ch> wrote: > > > The PHY_INTERRFACE_MODE_NA is a neat trick that I will remember but it > > would only help with the rgmii delay issue and not the LED issue (this > > patch). The GPY111 has some really nasty errata that is going to cause > > me to have a very hackish work-around anyway and I will be disabling > > the PHY driver to stay out of the way of that workaround > > Well, ideally we want the workaround for the erratas in the kernel > drivers. In the long run, you will get less surprises if you add what > you need to Linux, not hide stuff away in the bootloader.
Andrew,
I agree it is best to get workarounds for errata in the kernel where possible but this one is going to be a mess. There are 3 errata for this part and 2 of them require a toggle of the reset pin as a work around. Even if the mii bus can export a function for the phy to call to toggle reset in my case one of my boards has 2 of these PHY's (1 on a RGMII MAC and 1 on a SGMII MAC) on different MDIO busses that share a reset pin so if I trigger a reset I have to re-configure the other phy as well.
The errata can be summarized as: - 1 out of 100 boots or cable plug events RGMII GbE link will end up going down and up 3 to 4 times then resort to a 100m link; workaround has been found to require a pin level reset - 1 out of 100 boots or cable plug events (varies per board) SGMII will fail link between the MAC and PHY; workaround has been found to require a pin level reset - occasionally the phy will come up with a high bit error rate between the MAC and PHY; workaround has been found to require a soft reset or ANEG restart
> > > As far as changing a driver to force a LED configuration with no dt > > binding input (like this patch does) it feels wrong for exactly the > > same reason - LED configuration for this PHY can be done via > > pin-strapping and this driver now undoes that with this patch. > > Is it possible to read the pin strapping pins? If we know it has been > pin strapped, then a patch to detect this and not change the LED > configuration seems very likely to be accepted. >
No, you can read the current LED configuration but you don't know if it was set via strapping or boot firmware.
Best regards,
Tim
| |