stuck sensor and fan at full speed

One · 2013-05-29 20:50:22

Hello,

I have been having some issues lately with my HP Probook 4530s.

I was a bit annoyed that even at rest (laptop on, no programs running besides dwm) the fan seemed to always be at full speed. I seemed to have managed to track the issue down to the radeon module that was loaded even though X was running on the integrated Intel graphics. So I blacklisted that module as follows and things improved for some time.

# echo "blacklist radeon" > /etc/modprobe.d/radeon.conf

However, after some weeks (and probably after some kernel updates) I am now having even worse issues.

I have Windows 7, Ubuntu and Arch linux on the same machine, and as far as I understand, there seems to be a sensor that does not get updated under Linux, only under Windows.

$ acpi -t
Thermal 0: ok, 0.0 degrees C
Thermal 1: ok, 0.0 degrees C
Thermal 2: ok, 0.0 degrees C
Thermal 3: ok, 20.0 degrees C
Thermal 4: ok, 46.0 degrees C
Thermal 5: ok, 33.0 degrees C
Thermal 6: ok, 89.0 degrees C
Thermal 7: ok, 54.0 degrees C

$ sensors
acpitz-virtual-0
Adapter: Virtual device
temp1:        +46.0°C  (crit = +128.0°C)
temp2:        +89.0°C  (crit = +128.0°C)
temp3:        +27.0°C  (crit = +128.0°C)
temp4:        +37.0°C  (crit = +128.0°C)
temp5:        +20.0°C  (crit = +128.0°C)
temp6:         +0.0°C  (crit = +128.0°C)
temp7:         +0.0°C  (crit = +128.0°C)
temp8:         +0.0°C  (crit = +128.0°C)

coretemp-isa-0000
Adapter: ISA adapter
Physical id 0:  +49.0°C  (high = +86.0°C, crit = +100.0°C)
Core 0:         +49.0°C  (high = +86.0°C, crit = +100.0°C)
Core 1:         +44.0°C  (high = +86.0°C, crit = +100.0°C)

$ uptime 
 21:29:23 up 11 min,  1 user,  load average: 0.10, 0.22, 0.24

I am referring to the one at 89°C, despite the practically null load on the system (dwm with Firefox and urxvt on).

Because of this sensor, the fan is at full speed constantly, and whereas other values change (see below for a comparison I got while writing this) that one remains constant and high. I am running a custom version of kernel 3.9.3-1, but I get the very same behaviour with the -ARCH one. I get the same problem with Ubuntu.

$ sensors
acpitz-virtual-0
Adapter: Virtual device
temp1:        +54.0°C  (crit = +128.0°C)
temp2:        +89.0°C  (crit = +128.0°C)
temp3:        +34.0°C  (crit = +128.0°C)
temp4:        +46.0°C  (crit = +128.0°C)
temp5:        +20.0°C  (crit = +128.0°C)
temp6:         +0.0°C  (crit = +128.0°C)
temp7:         +0.0°C  (crit = +128.0°C)
temp8:         +0.0°C  (crit = +128.0°C)

coretemp-isa-0000
Adapter: ISA adapter
Physical id 0:  +57.0°C  (high = +86.0°C, crit = +100.0°C)
Core 0:         +57.0°C  (high = +86.0°C, crit = +100.0°C)
Core 1:         +51.0°C  (high = +86.0°C, crit = +100.0°C)

I am not sure how to tackle this issue, whether the sensor is really stuck, whether it is not recognised properly, what it corresponds to.

edit: I should point out that cpu load under Windows was high the last time I used it, which explains the high value for that sensor. However the computer was shut down for at least 1 hour between the intensive usage under Windows and the regular usage with Arch Linux.

I would be grateful for any insight in this matter.

Thanks!

Last edited by One (2013-05-29 20:55:51)

One · 2013-05-29 21:33:51

I've seen the intel_pstate issue, and tried the new kernel 3.9.4-1-ARCH. The sensor is still up there even with intel_pstate=distable. At least, this change at the kernel line brought back to life the cpu governors (I was left with performance and powersave )

WonderWoofy · 2013-05-29 22:19:08

One wrote:

At least, this change at the kernel line brought back to life the cpu governors (I was left with performance and powersave

Did you do absolutely no research into what this new intel_pstate actually is beore deciding that you should disable it? I can see disabling it for the sake of trying to fix that sensor, but disabling it because you want ondemand back is a poor reason if you don't understand what it does.

That is way OT in regard to your issue though. I think that this is a hardware issue that may be beyond the scope of the Arch Linux forums. By that, I mean you may have to contact HP on this one. Does the temp on that sensor actually change at all, or does it stay exactly at 89 degrees? Does it report this same thing in windows? Or at if that info is not available (I have no idea about windows) does using windows result in the same contant fan usage?

Seeing that you have indicated that it is consistent between operating systems, I think that you are in need of actual service on this machine...

One · 2013-05-30 07:51:55

Thank you, WonderWoofy.

I did some research into intel_pstate, and having multiple laptop issues at the same time I decided to disable it, and focus on this sensor, which causes the fan to go berserk.

I don't think it's a hardware issue because the sensor seems to work fine under Windows. I turned the laptop on again after the night, and it was still on 89 in Arch, but after booting up Windows and going back to Arch it is now 46°C. That is why I believe that it gets set correctly in Windows. Also, the fan in Windows works as one would expect.

Arch Linux

#1 2013-05-29 20:50:22

stuck sensor and fan at full speed

#2 2013-05-29 21:33:51

Re: stuck sensor and fan at full speed

#3 2013-05-29 22:19:08

Re: stuck sensor and fan at full speed

#4 2013-05-30 07:51:55

Re: stuck sensor and fan at full speed

Board footer