Re: [arch-general] Laptop cooling fan does not work

[Martin Kühne via arch-general]

I just set up some other distro on an IBM laptop, and I noticed that
when I set up the x86 variant, I had very considerable heat problems,
while when I swiched to x86_64, which was just after I acknowledged
that I had the lm flag in /proc/cpuinfo, the temperature issue went
away completely. Whether this would be the same with arch's kernels I
cannot say for certain, though. I also blew through the cooling fan in
the "accelerating" direction, which made the thing cough up some dust
at some point and appeared to run more smoothly afterwards.

It could be that, in the end, these things can be tied to a kernel config thing?
Usually, you might get the necessary hints from lspci and/or, as has
been suggested, detect-sensors.

To sum it up:
· check if your pacman's arch actually matches your cpu. if you have
lm in your /proc/cpuinfo, you should use the x86_64 kernel.
· check for hardware obstruction, dust, correct montage of fan or cooling grill

cheers!
mar77i

Re: [arch-general] Laptop cooling fan does not work

[Jürgen Werner]

Am 25.08.2016 um 05:30 schrieb Peter Nabbefeld:
> Am 23.08.2016 um 21:10 schrieb Jayesh Badwaik:
>>> There seems to be sth. wrong with my Nvidia settings, nvidia-settings
>>> shows an error that X is incorrectly configured and my vidieo card
>>> seems not to be used - probably this causes the problems. Maybe this
>>> is caused by some update or sth. else. I'll check that next weekend.
>>
>> Okay, first you must install lm_sensors package and run
>> $ sensors
>>
> I already had installed it. After even running sensors-detect again, I
> got this output (after replacing the nvidia driver by the nouveau one):
>
> acpitz-virtual-0
> Adapter: Virtual device
> temp1:+52.0°C  (crit = +120.0°C)
>
> nouveau-pci-0100
> Adapter: PCI adapter
> GPU core: +0.60 V
> temp1:N/A  (high = +95.0°C, hyst =  +3.0°C)
>(crit = +105.0°C, hyst =  +5.0°C)
>(emerg = +135.0°C, hyst =  +5.0°C)
> power1:   N/A
>
> coretemp-isa-
> Adapter: ISA adapter
> Physical id 0:  +53.0°C  (high = +84.0°C, crit = +100.0°C)
> Core 0: +53.0°C  (high = +84.0°C, crit = +100.0°C)
> Core 1: +48.0°C  (high = +84.0°C, crit = +100.0°C)
> Core 2: +51.0°C  (high = +84.0°C, crit = +100.0°C)
> Core 3: +49.0°C  (high = +84.0°C, crit = +100.0°C)
>
> As my grafics card is of type NVE4/GK104M, I'd expect temperature data
> to be available - but it obviously isn't (fan control should be
> available ...). Probably I've misunderstodd sth.?
>
>> command. This will tell you about your fan speed for the processor.
>> Next, if the CPU processor is running, determine which driver are you
>> using. For this, you need to do lsmod, which lists all the modules
>> listed (this should contain your video driver module)
>>
>> $ lsmod | grep -i nvidia
>>
>> $ lsmod | grep -i nouveau
>>
>
> nouveau  1478656  1
> ttm77824  1 nouveau
> mxm_wmi16384  1 nouveau
> wmi16384  2 mxm_wmi,nouveau
> video  36864  2 i915,nouveau
> button 16384  2 i915,nouveau
> i2c_algo_bit   16384  2 i915,nouveau
> drm_kms_helper118784  2 i915,nouveau
> drm   294912  8 ttm,i915,drm_kms_helper,nouveau
>
>
>> At least one the above should return something. nouveau is the open
>> source driver.
>>
>>

I must agree with Sean. You are probably focusing to much on the
graphics driver. Most Laptop have only one fan nowadays to cool both,
processor and graphics card. When either one gets to hot, the fan should
react.
Please follow Seans advise to downgrade the kernel or better the whole
system to the state it was before your the upgrade, which broke the fan
functionality. First step should be ruling out hardware failures. Once
you know for certain, that it is indeed a software problem you can start
searching for the particular software. For example, by upgrading package
by package (only for critical packages of course).