Experimental Physics and
Industrial Control System

Andrew Johnson <[email protected]> · Fri, 3 Jun 2011 17:03:09 -0500

Hi Mathias,

On 2011-06-03 Steiner, Mathias wrote:
> I have a softIOC with a shared-resource problem.
> A couple dozen 'sub' records need access to the same file.
> I would like to put a mutex around the file access, but I'm stumped as to
> where the corresoponding variable should be defined.
> 
> It seems to me that it has to go one level "up" from the subroutine.
> So my question is:  Where to park the '->lock' ?

Doing file access from a subroutine record or (device support for that matter) 
is really not a good idea; you're breaking the rule that record processing 
code should never block for any significant period of time.

What are these records doing with the file?  The usual way we do file I/O is 
to write a sequence program that handles all the file operations, and have the 
sequence interact with any necessary records inside the database.  Can you 
rework your design to follow that model?  We don't have a solution to what 
you're asking because it's not compatible with the IOC's processing model.

- Andrew
-- 
Optimization is the process of taking something that works and
replacing it with something that almost works, but costs less.
-- Roger Needham

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System