23:06 <sawyer> Hi all, wasn't able to find anything in the asahi-re logs but I have a 16 inch m1 on its way and was wondering if either someone else is working on vrr/"pro-motion" support or if that would be something where help would be welcome? I don't have a ton of prior reverse engineering experience, but hopefully have enough systems dev background to make some progress

23:51 <jannau> sawyer: that would be dcp. drivers/gpu/drm/apple in the linux tree and proxyclient/hv/trace_dcp.py and proxyclient/m1n1/fw/dcp/ in m1n1. there is little content here since we mostly RE it through HV tracing, discussions are mostly in #asahi-dev since it doesn't involve binary RE

23:52 <jannau> I don't think anyone has looked at VRR/"pro-motion" yet. RE through tracing might be a little bit annoying since the current python based proxy tracing is slow

23:58 <jannau> on the otherhand I'm not sure if there's that much to RE. the swap surface structure has 3 timestamps. ts3 seems to be the current time and is always set, ts1/ts2 are not always set (at least during boot). ts1 and ts2 have a delta less than a frame duration, ts1 is in the future (maybe a deadline), ts2 is less clear. timestamps have a resolution of 24MHz