On Fri, Jul 01, 2022 at 03:19:25PM -0600, Rob Herring wrote: > On Fri, Jul 01, 2022 at 05:04:59PM +0200, Thierry Reding wrote: > > From: Thierry Reding <treding@xxxxxxxxxx> > > > > Tegra194 has two separate instances of the pin controller, one called > > AON (in the always-on domain) and another called "main". Instead of > > treating them as a single pin controller, split them up into two > > separate controllers. Doing so allows the mapping between the pinmux and > > GPIO controllers to be trivial identity mappings and more cleanly > > separates the AON from the main IP blocks. > > Doesn't this break compatibility? I had added this paragraph to the cover letter: Note that while this changes the DT node in an incompatible way, this doesn't have any practical implications for backwards-compatibility. The reason for this is that device trees have only reconfigured a very narrow subset of pins of the main controller, so the new driver will remain backwards-compatible with old device trees. I've now added that to the commit message of this patch in v2 to hopefully clarify that for posterity. Thanks, Thierry
Attachment:
signature.asc
Description: PGP signature
