FS#49724 - [linux] 4.6.x modprobe nct6775 stuck in uninterruptible sleep

Attached to Project: Arch Linux
Opened by Riri (chicha) - Wednesday, 15 June 2016, 18:49 GMT
Last edited by Eli Schwartz (eschwartz) - Wednesday, 04 October 2017, 00:01 GMT
Task Type Bug Report
Category Kernel
Status Closed
Assigned To Tobias Powalowski (tpowa)
Architecture All
Severity Low
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 1
Private No

Details

Description:

nct6775 is a hardware monitoring module (fan and voltage).
Since kernel 4.6.2-1 modprobe nct6775 is stuck ending in an uninterruptible sleep, forcing to hard reboot the server (Intel NUC 5I3RYK).

Additional info:
* linux 4.6.2-1
* downgrading to 4.5.4-1 fixes the issue
* dmesg for both kernels added in attachment.


Steps to reproduce:
* upgrade ot linux 4.6.2-1
* modprobe nct6775
* modprobe is stuck.
* sensors do not report data from nct6775.
* downgrade to linux 4.5.4-1
* hard reboot (unplug the server, long press to stop button, ...)
* modprobe nct6775
* sensors report data from nct6775 correctly.

Nothing mentioned on the topic on lkml or kernel's bugzilla so far.
Looking at the diff from kernel's arch config I do not see any evident clue for this issue. Maybe someone here can help ?

Thank you very much in advance to all of you.
Cheers,
CH.
   dmesg-4.5.4-1.log (54.2 KiB)
   dmesg-4.6.2-1.log (53.7 KiB)
This task depends upon

Closed by  Eli Schwartz (eschwartz)
Wednesday, 04 October 2017, 00:01 GMT
Reason for closing:  Fixed
Additional comments about closing:  This is fixed in the latest kernel package.
Comment by Tobias Powalowski (tpowa) - Friday, 24 June 2016, 09:13 GMT
Please do a bisection and report this upstream. We cannot do much here.
Comment by Riri (chicha) - Wednesday, 29 June 2016, 17:53 GMT
Hi Tobias,
thank you very much for your reply.
I reproduced the bug with linux 4.6.3 too unfortunately.

As suggested by michi on the forum I added modules coretemp and nct6775 to the MODULES in mkinitcpio.conf and it fixes the issue.
I am definitely not an expert of initramfs/kernel compilation, but with this new fact do you see an issue on Arch kernel side or you still recommend to report the issue upstream ?

Cheers,
CH.
Comment by Peter Dohm (N3rdle) - Friday, 08 July 2016, 05:03 GMT
i too have encountered this problem on 4.6.3.

i'm quite certain this is an upstream issue. if you're unwilling / unable to report upstream, reply, and i'll do so.

Cheers

Peter
Comment by mattia (nTia89) - Tuesday, 03 October 2017, 19:54 GMT
is this issue still valid with newer kernel?
Comment by Riri (chicha) - Tuesday, 03 October 2017, 20:03 GMT
Hi nTia89,

I am really sorry, I totally forgot to test with newer kernels and come back to you all guys.
You did well ping the thread: I just tested with 4.13.3-1-ARCH and it worked like a charm.
Definitely a kernel regression upstream.

Thanks to all of you for your comments and support.
Cheers,
CH.

Loading...