> Is this ready to be queued up for next? It looks good to me.

Ulf, out of your head, do you have an idea why preparing HS400 during
set_ios works but populating host->ops->prepare_hs400_tuning() does not
(according to Simon's tests)?