[PATCH 3/4] ARM: mvebu: NETGEAR ReadyNAS 104 .dts file cleanup

[Jason Gunthorpe]

On Fri, Nov 22, 2013 at 09:07:01PM +0100, Arnaud Ebalard wrote:
>  			mdio {
>  				phy0: ethernet-phy at 0 {
> +					compatible = "marvell,88e1318s";
>  					reg = <0>;
>  				};
>  
>  				phy1: ethernet-phy at 1 {
> +					compatible = "marvell,88e1318s";
>  					reg = <1>;
>  				};
>  			};

Just an inquiry, are these compatible strings OK?

Documentation/devicetree/bindings/net/phy.txt doesn't really give
guidance.

For self-discoverable busses I would expect to see a structured
compatible string, or property, that encodes the phy-id - such as what
the PCI binding does with vid/did:

"compatible" Construct a list of names in most-specific to
             least-specific order. The names shall be derived from
	     values of the Vendor ID, Device ID, Subsystem Vendor ID,
	     Subsystem ID, Revision ID and Class Code bytes, and shall
             have the following form, and be placed in the list in the
             following order:

  pciVVVV,DDDD.SSSS.ssss.RR (1)

So, I would think something like this:

compatible = "ethernet-phy-id0141,0e90"; /* Marvell 88E1318 */

?

A big rational for this is that the kernel already has complete
infrastructure for matching drivers to the numeric phy-id.

I have a small patch to make the OF mdio bus scanner get the phy-id
from DT. I didn't use compatible string in this patch, but it would be
easy to change, and this demonstrates what we could do with a
structured compatible string (attached).

Grant? Rob? I could make some patches for this if you agree.

Regards,
Jason

>From 1aab2047b6dfb8b814673718dfa7ef0ea0c7a0ef Mon Sep 17 00:00:00 2001
From: Jason Gunthorpe <jgunthorpe at obsidianresearch.com>
Date: Thu, 12 Sep 2013 17:36:13 -0600
Subject: [PATCH] of_mdio: Allow the DT to specify the phy ID and avoid
 autoprobing

This makes the generic of_mdiobus_register parse the DT property
'phy-id'. If present it should be the numeric constant that matches the
phy-id register normally readable through MDIO.

When the ID is given the phy autoprobing is defeated and the phy is
created directly.

This is necessary to support phy's that cannot be autoprobed when
of_mdiobus_register is called. Specifically, my case has the phy in
reset at that time, the reset is only released once the ethernet
driver starts.

Signed-off-by: Jason Gunthorpe <jgunthorpe at obsidianresearch.com>
---
 drivers/of/of_mdio.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/of/of_mdio.c b/drivers/of/of_mdio.c
index d5a57a9..dbe6932 100644
--- a/drivers/of/of_mdio.c
+++ b/drivers/of/of_mdio.c
@@ -81,7 +81,12 @@ int of_mdiobus_register(struct mii_bus *mdio, struct device_node *np)
 
 		is_c45 = of_device_is_compatible(child,
 						 "ethernet-phy-ieee802.3-c45");
-		phy = get_phy_device(mdio, addr, is_c45);
+		paddr = of_get_property(child, "phy-id", &len);
+		if (paddr && len <= sizeof(*paddr) && !is_c45)
+			phy = phy_device_create(mdio, addr, be32_to_cpup(paddr),
+						0, NULL);
+		else
+			phy = get_phy_device(mdio, addr, is_c45);
 
 		if (!phy || IS_ERR(phy)) {
 			dev_err(&mdio->dev,
-- 
1.8.1.2