Attention is currently required from: Furquan Shaikh, Martin Roth, Tim Wawrzynczak, Rob Barnes. Raul Rangel has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/54134 )
Change subject: mb/google/guybrush: Add SoC thermal zone ......................................................................
Patch Set 1:
(1 comment)
File src/mainboard/google/guybrush/variants/baseboard/devicetree.cb:
https://review.coreboot.org/c/coreboot/+/54134/comment/648d688c_fb7f7db3 PS1, Line 55: chip drivers/acpi/thermal_zone : register "name" = ""SOC"" : : register "temperature_sensor_id" = "0" : : register "polling_period" = "10000" : : # EC is configured to power off the system at 92C, so add one degree of buffer : # so the OS can gracefully shutdown : register "critical_temperature" = "91" : : # EC is configured to assert PROCHOT at 90C. That drastically lowers : # performance. Instead we will tell the OS to start throttling the CPUs at : # 85C in hopes that we don't hit the PROCHOT limit. : register "passive_config" = "{ : .temperature = 85, : .time_constant_1 = 2, : .time_constant_2 = 5, : .time_sampling_period = 2000, : }" : : device generic 0 on end : end
Another suggestion: The new alias syntax allows you to add a pointer to another `struct device` like […]
I think that's a good idea. If we go that route, I think I want to move the acpigen_write_TMP method into `struct device_operations`. This way we can cleanly support multiple temperature controllers. The ec_lpc.c code would set .acpi_write_TMP in its ops. This would all be guarded with DRIVERS_ACPI_THERMAL_ZONE so we don't increase the device tree size for everyone.
What do you think?