> > IMO it would be better to use id->driver_data unconditionally and save a > > call to of_device_get_match_data() > > What would you suggest to move ahead? > > I'd suggest using id->driver_data unconditionally, but I'd like Wolfram's > opinion on this (CC'ed).
IIUC: I'd suggest to leave it as it is. There are people working on removing the quirky fallback to i2c_device_id in case of_compatible is not provided. If that gets mainline, you don't have i2c_device_id anymore with DT matching. So all binding methods should be self-contained. -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: not available URL: <https://lists.freedesktop.org/archives/dri-devel/attachments/20161020/f332f660/attachment.sig>
