Experimental Physics and
Industrial Control System

Bill Brown <[email protected]> · Fri, 26 Jun 1998 09:29:15 -0700

At 09:21 AM 6/26/98 -0500, you wrote:
>I've been using an mvme177 on the most heavily loaded bpm ioc for over a
year.
>I've never done an actual timing measurement, but with a 25MHz 167 this ioc
>used to go to 0% idle time. A very bad thing. With a 60MHz 177 it's idle time
>is  now typically 50% to 60%.
>
>You do need a BSP as Ned pointed out. The 177 will crash and burn with the 
>167's vxWorks. *.o files built for the "167" run fine on the 177.
>
>We've not had any luck booting a 177 from Solaris. Admittedly,  I haven't
>tried booting from Solaris in the past year and don't remember what
version we
>were running at the time. 
>
>I need to upgrade a number of ioc's to faster processors. The 177 is nice,
>but a little pricey. I'm planning on buying 172's when we can support them
>at APS. My understanding is we've got the 172 BSP working under vxWorks 5.3.
>That's the good news. The bad news is we're still running vxWorks 5.2 on our
>production systems :(
>
>Frank
>

What are the major problems with moving to 5.3?  Will apps compiled with
5.2 headers not run?  We're on the edge of making this transition, and
if we absolutely gotta convert everything in one swell foop, it probably
aint' gonna happen real soon.  And the longer we put it off, the worse
the pain is gonna be!

I guess the good news is that it looks to me like the structure of 3.13.0
makes it easier, altho perhaps a little spendy in terms of disk space,
to move forward one system (ioc) at a time.  Don't wanna think about
what happens when ca is no longer rearward compatible.  Maybe someone
will write a "bridge" between versions.  Shuddddder!

Different subject: is anyone using EPICS built for Heurikon hkn260 cpus
yet?  Or is everyone still running the '167 apps on them.  Is there
enough speed gain to make the effort worth while?  IOW - is there a
noticable difference between running '040-optomised code on a '060
and running '060-optomised code on a '060?

Disclaimer:  Any opinions are my own and have	    | -bill
    nothing to do with the official policy or the   |  [email protected]
    management of L.B.N.L, who probably couldn't    |  Berkeley, CA
    care less about employees who play with trains.    aka [email protected]

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System