Re: Can Fedora enable CONFIG_DETECT_HUNG_TASK ?

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 5/18/19 11:48 AM, Alan Jenkins wrote:
Hi

I was surprised to realize that Fedora kernels do not enable CONFIG_DETECT_HUNG_TASK.  Is there any reason why it should not enable it?

Fedora kernels appear to be an outlier here, v.s. RHEL, Debian, Ubuntu and SUSE.

I'm not taking a postion on whether ABRT needs to report hung tasks.  If ABRT needs explicit support to report them, that could be added later. We only need to be confident that ABRT isn't going to something we don't want it to.

When the kernel detects all of userspace is hung, this could provide useful console logs on Fedora Server, other non-GUI installs, or expert users who resort to serial or network console logging.  When the hang is more specific, seeing the hung task in the kernel log is a nice and quick pointer towards the problem.

I recently wrote some advice which mentioned the hung task detector.  I observer that it is one of a set of handy crash/hang messages in the kernel.[1]

[1] https://unix.stackexchange.com/questions/518554/debian-stretch-vm-becomes-quasi-unresponsive-every-few-days/

It seems undesirable for Fedora to disable this type of message. Especially considering Fedora's quick turnover of kernels, and exposure to exciting new bugs :-).  I haven't checked if it is enabled in Rawhide, but I feel like it should be enabled for all Fedora kernels.

Thanks for the kernels :-)
Alan


 > Say Y here to enable the kernel to detect "hung tasks", which are bugs that cause the task to be stuck in uninterruptible "D" state indefinitely.
 >
 > When a hung task is detected, the kernel will print the current stack trace (which you should report), but the task will stay in uninterruptible state. If lockdep is enabled then all held locks will also be reported. This feature has negligible overhead.

(This option is not a sophisticated lock cycle detector.  It is just a simple timeout.  The upstream default timeout is 2 minutes.  Once a (default) 10 hung task messages have been logged, the kernel stops logging them).

We do have this on in the debug version of the kernel but not the regular
version. I think this is something we set a while ago but never really
thought about again. My one hesitation about enabling this would be
that certain devices might be particularly prone to hanging and this
could create a lot of noise for us to sort through. That said, I
think this could be a good candidate for turning on in rawhide so it
gets enabled in F31.

If nobody raises any serious objections, I'll see about turning it
on in rawhide.

Thanks,
Laura
_______________________________________________
kernel mailing list -- kernel@xxxxxxxxxxxxxxxxxxxxxxx
To unsubscribe send an email to kernel-leave@xxxxxxxxxxxxxxxxxxxxxxx
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: https://lists.fedoraproject.org/archives/list/kernel@xxxxxxxxxxxxxxxxxxxxxxx




[Index of Archives]     [Fedora General Discussion]     [Older Fedora Users Archive]     [Fedora Advisory Board]     [Fedora Security]     [Fedora Devel Java]     [Fedora Legacy]     [Fedora Desktop]     [ATA RAID]     [Fedora Marketing]     [Fedora Mentors]     [Fedora Package Announce]     [Fedora Package Review]     [Fedora Music]     [Fedora Packaging]     [Centos]     [Fedora SELinux]     [Coolkey]     [Yum Users]     [Tux]     [Yosemite News]     [KDE Users]     [Fedora Art]     [Fedora Docs]     [USB]     [Asterisk PBX]

  Powered by Linux