Argonne National Laboratory

Experimental Physics and
Industrial Control System

1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  2014  <20152016  2017  Index 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  2014  <20152016  2017 
<== Date ==> <== Thread ==>

Subject: Re: VxWorks IOC machine check exception
From: Ron Sluiter <sluiter@aps.anl.gov>
To: "J. Lewis Muir" <jlmuir@imca-cat.org>, EPICS Tech-Talk <tech-talk@aps.anl.gov>
Date: Thu, 15 Oct 2015 16:03:12 -0500
Hello Lewis,

It is unlikely that you can determine the cause just from this VxWorks task crash message. The tNetTask may not be the root cause of the crash; it may just be the current running task when the machine check exception was taken.

If the "Exception next instruction address: 0x001722a0" message is consistent, you might try to dump instructions around that address ("l 0x001722a0") and get lucking enough to find an external reference that would get you some idea of where execution was when the error occurred. Without a source-code debugger there is not much chance of determining what is wrong and, unfortunately, you are running an old version of VxWorks that we (APS) no longer have source-code debugging tools for.

My suggestion is upgrade to VxWorks 6.9.4.1. We have source-code debugging for that via the network and if that does not work (because of a hard crash), we have a Wind River ICE2.

Ron

On 10/15/2015 3:15 PM, J. Lewis Muir wrote:
Hello, Everyone!

I had a VxWorks IOC crash on me.  The only information I have is the
following from the console:

   machine check
   Exception next instruction address: 0x001722a0
   Machine Status Register: 0x0008b030
   Condition Register: 0x20000043
   Task: 0x1ef70f10 "tNetTask"^G
   Link Register: 0x0017228c

Is there any way to figure out the cause?  I gather that tNetTask has to
do with the VxWorks networking stack?  Does that tell me where the issue
was, or is that just where the issue happened to show up?  Any way to
use addr2line with this information?

Environment:

* VME processor module: Motorola MVME-5100 (vxWorks-ppc604_long)
   - CPU: MPC 750
   - Memory: 512 MB
* OS: VxWorks 5.5.2
* Kernel: WIND 2.6
* BSP: mv5100-anj-no-nfsv3 (from Andrew Johnson)
* Build toolchain: Tornado 2.2.2
* EPICS 3.14.12.3 (+ Known Problems patches except for the patch from
   2013-12-04)

Thank you!

Lewis


References:
VxWorks IOC machine check exception J. Lewis Muir

Navigate by Date:
Prev: VxWorks IOC machine check exception J. Lewis Muir
Next: Re: Modbus Binary Output Readbacks Wallace, Alex
Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  2014  <20152016  2017 
Navigate by Thread:
Prev: VxWorks IOC machine check exception J. Lewis Muir
Next: multiple pollgroup with s7nodave David Michel
Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  2014  <20152016  2017 
ANJ, 16 Dec 2015 Valid HTML 4.01! · Home · News · About · Base · Modules · Extensions · Distributions · Download ·
· EPICS V4 · IRMIS · Talk · Bugs · Documents · Links · Licensing ·