Experimental Physics and
Industrial Control System

Ron MacKenzie <[email protected]> · Thu, 21 Jul 2005 16:22:03 -0700 (PDT)

Hi Carl,

The issue is that CMLOG comes packaged with an older version of ACE
(ACE 4.0).  That version did not support RTEMS.

So, I took some of the obvious approaches to that problem and they proved
difficult.  It looks like the API has changed across the last couple of
major releases of ACE.

Ron

--
/-------------------------------------------------------------------------
|  Ron MacKenzie
|  http://www.slac.stanford.edu/~ronm/
|  mailto:[email protected]
|
|  This message is from me personally, and does not necessarily
|  reflect the opinion of Stanford University, SLAC, or the US DOE
\-------------------------------------------------------------------------

On Thu, 21 Jul 2005, Carl Cork wrote:

> Have you actually tried to build ACE on RTEMS? The ACE site
> claims to support RTEMS and has comprehensive build/install
> information:
>
>      http://www.cs.wustl.edu/~schmidt/ACE_wrappers/ACE-INSTALL.html
>
> The RTEMS archives also have reported success building ACE
> on i386 platforms:
>
>      http://www.rtems.com/ml/rtems-users/2001/april/msg00030.html
>      http://www.rtems.com/ml/rtems-users/2003/june/msg00118.html
>
> The CMLOG Client and ClientD code don't appear to have too many VxWorks
> dependencies.
>
> Maybe porting to RTEMS would be easier than changing iocLogServer?
>
>
> Allison, Stephanie wrote:
> > Hi All,
> >
> > We're looking at upgrading iocLogServer so that it either writes
> > messages to a file, like it does now, or forwards the messages to the
> > cmlogServer (which would make iocLogServer be a cmlog client).  We
> > think it's easier to do this than to port the CMLOG client and daemon,
> > and particularly ACE, to our RTEMS IOCs.  Part of this effort would
> > also be to add support for messages that contain both ASCII text and
> > other ASCII "tags" like facility, status, severity, and task name.
> >
> > Any comments?  Dumb idea?  Anybody interested in something like this?  Should it be done in such a way that it can go into some future EPICS base version with the new features enabled via some configure file?  Or left unbundled?
> >
> > Thanks,
> >
> > Stephanie Allison
> > SLAC/SSRL, SLAC/LCLS
>
> --
> Carl Cork                                 E-Mail:  [email protected]
> Lawrence Berkeley National Lab               TEL: +1 510 486 4295
> One Cyclotron Road, MS 6R2100                FAX: +1 510 486 5664
> Berkeley, CA, USA 94720-8226                 WEB:  www.lbl.gov
>
>

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System