Re: MVME 5500

[Kate Feng <feng1@bnl.gov>]

GOURNAY Jean-Francois wrote:
> Hello,
>
> I am curious to know what is the process to reload the MOTLoad of the MVME5500. I understand that you have to reload the VxWorks boot file, but what about the MOTLoad CPU monitor? Did you have to reload it?
>   
I guess there was some confusion.  I never had to reload MOTLoad.  It 
should be always there.
When I used vxWorks, there were two files.  Each of them had to be 
loaded by following the vendor's
procedure. One is  boot.bin, and the second one is vxWorks.  However, 
the reason why I needed
to load them is because the board did not come with vxWorks.  It was not 
because memory-write
outside the allocated data memory. Memory-write outside the allocated 
data memory should not require
one to reload the vxWorks or MOTLoad if the memory space was mapped 
correctly.  I think it is because
your vxWorks image was configured and/or built incorrectly. It probably 
maps the illegal space (e.g. the space
of MOTLoad or flash memory) into the user programmable memory area. 
Maybe you can test this with
someone who has a different built vxWorks image and see if this happens 
again. I hope this helps.


Cheers,
Kate


> If yes, what kind of download device is required? Do you have the binary image of MOTLoad? Why did you have to go trough this process? (if you remember ...)
> Thanks,
> Jean-François
>
> J.F. Gournay
> CEA Saclay
> IRFU/SIS
>
> -----Message d'origine-----
> De : tech-talk-bounces@aps.anl.gov [mailto:tech-talk-bounces@aps.anl.gov] De la part de Feng, Kate
> Envoyé : vendredi 19 juin 2009 14:26
> À : GOURNAY Jean-Francois; tech-talk@aps.anl.gov
> Objet : RE: MVME 5500
>
> We do not have the problem with the RTEMS-mvme5500 BSP.  In that case,
> I just have to debug and rewirte the software and reboot the CPU.
>  
> However, for vxWorks, I think you should be able to reload the initial bootload and the
> vxWorks boot yourself.  Did'nt they send you the boot files ?  I thought I reload the
> those vxWorks boot files myself serveral times myself many yeras ago. 
>  
> Cheers,
> Kate
>
> ________________________________
>
> From: tech-talk-bounces@aps.anl.gov on behalf of GOURNAY Jean-Francois
> Sent: Fri 6/19/2009 7:25 AM
> To: tech-talk@aps.anl.gov
> Subject: MVME 5500
>
>
>
> Hello All,
>
>  
>
> This is a problem which not directly related to Epics, but as several labs are using the MVME5500 CPU I hope that maybe someone has encountered this problem and has a solution!
>
>  
>
> Whenever a C program has a misbehaved pointer, which causes a memory-write outside the allocated data memory, we have a crash of the CPU which is absolutely understandable, but then, it is impossible to reboot the CPU. Both flash banks containing the initial bootload and the VxWorks boot are corrupted and the only solution is to send back the CPU to the retailer !!! We had experienced this with several CPUs (we are probably bad C programmers) and didn't get so far any explanation from our retailers or after sales services. We have VxWorks 6.5.
>
>  
>
> Thanks in advance
>
>  
>
> J.F. Gournay
>
> CEA Saclay
>
> IRFU/SIS
>
>  
>
>
>