11:14 <hrdl> A while ago I've asked Rockchip about upstreaming "ultra suspend" TF-A support for RK3568. Their reply was that they had no plans to implement it since the feature was "hardware-specific and requires specialized hardware solutions"

11:19 <hrdl> They haven't answered my request for the relevant source code of their bl31 or permission to use reverse-engineered outputs. Would some form of clean-room design be an option to include this function in upstream TF-A or is this a losing game in general?

11:21 <hrdl> Most of the registers touched by the vendored bl31 are documented in the TRM and the PMIC's documentation, but it's still a couple of hundred lines of code we're talking about