Experimental Physics and
Industrial Control System

Andrew Johnson <[email protected]> · Fri, 12 Jan 2007 11:39:05 -0600

Hi Bob,

Dalesio, Leo `Bob` wrote:

What is your position on this? It would be nice to have a drvTs that
worked properly in the release. Are there issues about putting this
in that are not being handled on TImo's end?

I haven't had too close a look at what Timo has done yet, although I

have asked Eric to look at the code he sent out today; the one thing

that worries me slightly was a comment from Timo that his code may not

be completely compatible with "the APS version", although at first

glance Eric doesn't see anything obviously wrong. My guess is that

won't be a big problem, but I will be keeping an eye on that - we have

to be able to talk to the APS timing/event system as well as the newer

derivatives, and we now have interfaces to our system on RTEMS as well

as the original vxWorks on VME cards.

My position is that Base should as standard include general-purpose

timing facilities to allow IOCs on all architectures to use OS time

(where provided) or EPICS soft time, and to permit plug-ins to supply

additional time sources such as an Event system or a Bancomm GPS time

card. There are some situations where we might want an IOC running on

Linux to be able to use soft time instead of OS time, although I'm not

sure whether that's possible with Timo+Dave's code yet (it ought to be

though).

Base should not include any hardware-specific code, thus the interface

to event system and/or GPS time cards must be a plug-in provided in a

support module. This should be how the new design works, but I haven't

looked at the code yet to confirm that.

The decision to not accept Timo's original modifications to drvTS but to

wait for the rewrite was Marty's, and I think the result will be to get

a better system sooner than we would have by patching the original at

the time, but in the meanwhile we have suffered some pain in waiting.

Both Eric and I are currently preparing for the EPICS course at USPAS in

Houston next week, so right now is not the best time for us to discuss

this in detail. This code will not make it into R3.14.9 because that

has already been delayed too much as it is, but after that is released

we can look at it more closely.

- Andrew
--
There is considerable overlap between the intelligence of the smartest
bears and the dumbest tourists -- Yosemite National Park Ranger

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System