Hi Michael, thank you very much for debugging and reporting this!
Sorry for the major inconvenience.
You're welcome and the only inconvenience was doing the bisect. ;-)
We have some known issues with saving and restoring ASPM state on
suspend/resume, in particular with ASPM L1 Substates, which are
enabled on this device.
David Box has a patch in the works that should fix one of those
issues:
https://lore.kernel.org/r/20231221011250.191599-1-david.e.box@xxxxxxxxxxxxxxx
It's not merged yet, but it's possible it might fix or at least be
related to this. If you try it out, please let us know what happens.
I gave the proposed patch a quick test on top of kernel 6.6.8 and it
didn't affect this issue (the reboot still happens on resume attempt).
