[Leaplist] BUG: soft lockup - CPU#0 stuck for 4304s!

[Jason Boxman]

Jason Boxman wrote:
> While this comes up in Google quite a bit, no one seems to know what 
> exactly is going on.  It happens for me under
> 
> Linux zimbra 2.6.26-2-686 #1 SMP Mon May 11 19:00:59 UTC 2009
>   i686 GNU/Linux
> 
> But the issue seems to have been backported into recent RHEL kernels as 
> well.  Anyone have any ideas on what I might do to resolve it?

http://docs.xensource.com/XenServer/4.0.1/guest/ch04s08.html#rhel5_limitations
https://bugzilla.redhat.com/show_bug.cgi?id=250994

Sounds like it might be harmless, but not impact free.  Since this is a 
full VM, sounds like there may be no resolution for me with my existing 
host hardware.

I'll have to see what kinds of behaviors are the result of this during 
an actual duty cycle and respond accordingly.  Hopefully it'll all be 
just fine.

> Comment #26 From Chris Lalancette 2008-02-13 08:43:28 EDT 
> This BZ is specifically about PV guests having softlockups.  The problem with
> fully virtualized guests may or may not be solvable....the problem is that fully
> virtualized guests don't actually know when they are being paused, etc, so they
> can't necessarily take steps to correct for it.  When they do wake up, they will
> realize they haven't run for a long time and think they have a softlockup.
> 
> I'm not 100% certain that is what is happening in the fully-virt case, but it
> seems likely.  In any case, this BZ does not address that at all.
> 
> Chris Lalancette  


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.