On Wed 11 Mar 16:25 PDT 2020, Stephen Boyd wrote: > Quoting Bjorn Andersson (2020-03-10 14:27:28) > > On Tue 10 Mar 13:29 PDT 2020, Stephen Boyd wrote: > > > > > Why can't we search in DT for the > > > imem node and then find the pil reloc info compatible string on the > > > first call to this library? Then we don't need an API to see if the > > > device has probed yet (qcom_pil_info_available) > > > > I think this sounds reasonable. > > Great! > > > > > > and we can just ioremap > > > some region of memory that's carved out for this reason. Forcing > > > everything through the regmap is mostly adding pain. > > > > > > > My concern here was simply that we'll end up ioremapping various small > > chunks of the imem region 10 (or so) times. But I agree that things > > would be cleaner here. > > Alright. I'd like the ioremap() approach. ioremap() will "do the right > thing" and reuse mappings if they're already there and overlap in the > page. So it's OK that the syscon/simple-mfd exists and makes a device, > etc. etc., but we don't need to care about it. We can just ioremap() the > area and not worry that the regmap users may have a mapping to the same > place. This is a dedicated carveout inside IMEM so we're safe from other > meddling users. Agreed, thanks for the feedback! Regards, Bjorn
