At Diamond we have been running Linux on about 70 of our MVME5500 PPC boards.
It is using a patched 2.6.20 kernel, so over 10 years old, but seems to work. Originally it came from Ajit Prem at Emerson, I believe, as mentioned in this 2006 EPICS presentation.
I can’t find a place to download the patch from now, but can supply our copy if useful.
The large (2.5MB) patch file suggests it would take a lot of effort to port to a newer kernel, but I'm no expert.
Our VME crates have very little VME I/O so would need more testing for general use. We have not tried building the drivers for our Hytec I/O cards or other VME cards yet.
Our build includes limited VME support (only A16 addresses so far), rootfs for the file system and the Busybox toolset. Currently we load the image over the network using the MOTLoad environment parameters.
I've not looked at the Denx tools mentioned in the presentation and cannot comment on RTEMS.
The latest kernel I can find is a 2.6.26 patch for Debian Squeeze on this French website.
I’ve not investigated this any further.
Hope this helps,
Diamond Light Source Ltd.
+44 (0)1235 778054
From: Abbott, Michael (DLSLtd,RAL,TEC)
Sent: 20 September 2017 10:51
To: 'Goetz Pfeiffer'; EPICS tech-talk
Cc: Baker, Keith (DLSLtd,RAL,TEC)
Subject: RE: Ideas / Suggestions for the future of VME-CPU Operating Systems
We would love to see an embedded Linux on our MVME boards. We have a perfectly good root file system that we use on a handful of our boards and on quite a
number of other targets (Libera and PandA in particular), but our kernel is *extremely* old and the patch set is gigantic (Keith who I’ve CC’d will remember where we obtained it). We have not had the effort available to get a maintainable and up to
date kernel version (last I looked it seemed to me that the Tornado support was out of tree), and so we just use our VME Linux on a handful of very basic (limited read only) systems to free up enough vxWorks licences to keep us going elsewhere. I’m pretty
sure our long term plan is to abandon VME altogether … in the *long* term. Migrating back to vxWorks sounds like a horrid idea!
at the Helmholtz-Zentrum Berlin (https://www.helmholtz-berlin.de/) we use EPICS for our control system.
We have a growing number of soft IOCs with Linux and VME bus based IOCs mostly running RTEMS and
some vxWorks 5.4 (Tornado 2.02).
Our CPU boards are MVME162 and MVME2100. We have replaced more than half of the old MVME162 boards with
MVME2100 boards, of which we bought a large supply some years ago.
After migrating most VME CPUs to RTEMS 4.9 we have run into some problems:
Newer CPU boards like the MVME5500 require the "beatnik" board support, which only works with RTEMS 4.10
RTEMS 4.10 has some problems regarding the "cexp" shell and doesn't work on some of our IOCs.
cexp, the shell for RTEMS is not compatible with RTEMS 4.11 and 4.12, but we need it for dynamic loading of objects
gesys, the component that is used to create the RTEMS kernel seems to be a bit of a mess
RTEMS 4.11 and 4.12 are not supported by the EPICS base
Debian Packages for RTEMS are no longer maintained
The intersection of the people using RTEMS and the ones using EPICS seems to be small and getting smaller, so with problems we are much on our own
A possibility would be to use vxWorks again. Our current vxWorks version is
very old and has to be updated. Problems here:
MVME2100 boards do not seem to be supported by vxWorks 6
Possibly high costs for CPU licenses for vxWorks
What are your experiences with this ?
Do you still use VME bus systems ?
Is there a future for RTEMS in EPICS control systems ?
Are there alternatives to RTEMS and vxWorks ?
This e-mail and any attachments may contain confidential, copyright and or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail.
Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd.
Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be transmitted in or with the message.
Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom