Re: spin lock sleepq chain held by 0xc0fc3540 for > 5 seconds

From: Ian FREISLICH <if_at_hetzner.co.za>
Date: Sat, 08 May 2004 10:45:40 +0200
John Baldwin wrote:
> On Friday 07 May 2004 09:34 am, Ian FREISLICH wrote:
> > Hi
> >
> > Had this twice on my dual PII server runnning current.  It was doing
> > a make world -j8 both times.
> >
> > FreeBSD 5.2-CURRENT #4: Thu May  6 15:58:56 SAST 2004
> 
> :(  This is a known reversal that can cause deadlock (it did in this case).  
> It is being worked on but completely fixing it is going to take some time.  
> Just to verify, you had processes swapped to disk when this happened?

More than likely.  By this point in the make world, the load average
is normally over 40 and it normally uses a small amount of swap.
I'll make sure it used swap by regularly outputting swapinfo to the
console.  This well and truely wedges the system, so there is no
way to get into the debugger after it happens.

Ian

--
Ian Freislich
Received on Fri May 07 2004 - 23:45:42 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:53 UTC