Bug 6635

Summary: /proc/acpi/fan empty, fan always at full power after initial startup
Product: ACPI Reporter: Timo Lindemann (lindem)
Component: Power-FanAssignee: Konstantin Karasyov (konstantin.karasyov)
Status: REJECTED INVALID    
Severity: normal CC: acpi-bugzilla
Priority: P2    
Hardware: i386   
OS: Linux   
Kernel Version: 2.6.16.18 Subsystem:
Regression: --- Bisected commit-id:
Attachments: DSDT as provided by the vendor
dmesg, full version

Description Timo Lindemann 2006-06-01 10:12:05 UTC
Fan will turn on when notebook reaches a temperature of 45 degrees celsius
(according to thermal zone readout). From there, fan blows at full power (which
is really noisy!) and will not stop unless machine is shut down. 

Odd: in /proc/acpi/thermal_zone/THRM/trip_points:

critical (S5):           155 C
passive:                 100 C: tc1=2 tc2=5 tsp=0 devices=0xdff5b6e8

echoing other values into that file (x:x:x:x:x) does not cause errors but doesnt
do anything.

The directory /proc/acpi/fan is empty.

dmesg shows multiple messages such as these:
acpi_thermal-0417 [08] thermal_get_trip_point: Invalid active threshold [0]
acpi_thermal-0417 [11] thermal_get_trip_point: Invalid active threshold [0]

Will attach DSDT as provided by the vendor later.

Thanks for thinking about this :-)
Comment 1 Timo Lindemann 2006-06-01 10:15:18 UTC
Created attachment 8237 [details]
DSDT as provided by the vendor
Comment 2 Timo Lindemann 2006-06-01 10:21:25 UTC
Created attachment 8238 [details]
dmesg, full version

should have mentioned: this is an AMD64 system; chipset is nvidia nforce 410.
Comment 3 Luming Yu 2006-06-07 08:04:02 UTC
Does windows work, and NO Fan problem?
Comment 4 Timo Lindemann 2006-06-07 12:22:03 UTC
Ummm.... I never tested that. I will try to get a windows installed for testing,
and report back. 

Didnt use that system for years. 
Comment 5 Konstantin Karasyov 2006-06-17 06:27:42 UTC
Timo,

No fan device defined in your DSDT, so no fan control is possible though ACPI 
fuctionality.
The reason of 'Invalid active threshold [0]' messages is that '_AC0' (active 
trip point) object was defined without '_AL0' (device list for active cooling) 
in thermal zone 'THRM'.
So, I'm rejecting this bug as it has nothing to do with ACPI.