On Fri, Oct 09, 2020 at 04:04:56AM +0300, Laurent Pinchart wrote: > On Mon, Aug 17, 2020 at 07:07:26PM +0300, Andy Shevchenko wrote: > > It's quite unlikely that another page size will be supported, > > but in any case there is still an inconsistency between custom > > page size definition and generic macros used in the driver. > > > > Switch over to the generic PAGE_SIZE for sake of the consistency. > > Is this conceptually correct though ? Does the CIO2 have an intrinsic > page size, or do pages here always refer to system memory pages ? In the > latter case the change is good, otherwise a separate macro seems best. I don't think the hardware is going to change these defaults. But of course we may repeat all macros and constants over the code, which makes a little sense to me, because hardware is using same settings as CPU MMU. In principle we may do that, but in reality the page size change will bring a hell out of it with all code being at least rechecked and I believe partially rewritten (you can imagine how to map for example CIO2 MMU with 11-bit per page to CPU page which is 12-bit). -- With Best Regards, Andy Shevchenko
