> I vote for having the exact SoC revision in the binding documentation > rather than wildcards or references to the list of i.MX SoCs. Otherwise > only the driver code gives a clue that the i2c driver matches imx1-i2c, > imx21-i2c and vf610-i2c, but not imx31-i2c.
Dunno if I got all right, so adding my 2 cents: Yes to adding each SoC to the binding docs. No to adding each SoC to the driver as a seperate 'compatible' entry if not really needed to distinguish IP versions. I mean imx31 should have two compatible entries anyhow, one for imx31 and one for imx21 as fallback, no? That all being said: Unless somebody objects, I'll pick the most recent VF610 series today and leave the doc fixup for later. Thanks, Wolfram
signature.asc
Description: Digital signature
