Hi Matt, Pushed it, split in two:
https://review.coreboot.org/c/coreboot/+/93839 (drivers/usb/acpi): when a Bluetooth device has no reset GPIO, _RST now calls a new weak hook instead of staying empty. https://review.coreboot.org/c/coreboot/+/93840 (soc/intel/cnvi): the SoC side fills that hook with the CNVi sideband PLDR — the same abort-PLDR the Wi-Fi CNVP._RST already uses. One thing I did differently from the acpigen.c idea: I kept the PLDR emission in the cnvi common block and reach it from the generic driver via the weak hook (same shape as get_wifi_sar_limits). My reasoning was that acpigen.c is SoC-agnostic while PCRR/PCRO are Intel-PCH-specific, so it felt cleaner to keep the Intel bits in intelblocks/cnvi. Happy to move it if you'd prefer. Tested on an Alder Lake-N board where the BT is CNVi-over-USB: the generated _RST comes out identical to the Wi-Fi CNVi reset on the same board, _PRR points at it, and btintel no longer logs "No support for _PRR ACPI method" (hci0 loads firmware and boots fine). Thanks for the help! Sebastian _______________________________________________ coreboot mailing list -- [email protected] To unsubscribe send an email to [email protected]
