On Tue, Nov 18, 2025 at 10:13:54AM +0800, Sébastien Bourdeauducq wrote: > On 11/18/25 2:04 AM, Mike Larkin wrote: > > IIRC when I looked at it it just read an EC register and did some biasing, > > so, > > no, nothing special. > > FWIW Linux also struggles with that broken ACPI: > > [ 6.927809] ACPI: \_TZ_.TZ10: _PSL evaluation failure > [ 6.928103] thermal LNXTHERM:00: registered as thermal_zone0 > [ 6.928105] ACPI: thermal: Thermal Zone [TZ10] (17 C) > [ 6.928204] ACPI: \_TZ_.UAD0: _PSL evaluation failure > [ 6.928283] thermal LNXTHERM:01: registered as thermal_zone1 > [ 6.928285] ACPI: thermal: Thermal Zone [UAD0] (17 C) > > from sensors: > acpitz-acpi-0 > Adapter: ACPI interface > temp1: +16.8°C > temp2: +16.8°C > (and those numbers are below ambient temperature and don't change under > load) > > Changing the "smart fan" (note: irony) settings in the BIOS seems to have no > effect. > > However, Linux keeps the CPU fan spinning for some reason, and thus avoids > the emergency hardware shutdowns. It also avoids the "machine does not > turn-off" bug that has been introduced with the BIOS update. > > Sébastien >
you could probably make a local diff for yourself that calls: acpitz_setfan "_ON_" all the time in acpitz_refresh, regardless of the temperature setting. maybe linux doesn't call _OFF, which is why the fan stays on. dunno otherwise. complain to your system vendor.
