Experimental Physics and
Industrial Control System

Andrew Johnson <[email protected]> · Thu, 1 Oct 2015 11:50:15 -0500

Hi Mark,

On 10/01/2015 09:53 AM, Mark Rivers wrote:
> I have another question.  What does EPICS base do when the process()
> routine for an input record returns an error?  Is there a specific
> error code it should return, or is any non-zero value OK?

The return value should be a status code which gets passed back to
whatever triggered the processing. Unfortunately it is not uniformly
handled inside the IOC at the moment; in some places it gets ignored
completely (callback.c), but in others it gets turned into a message
string and printed (e.g. by the iocsh dbtr command) and may even be
passed back to the CA client that triggered the processing as a CA
exception.

The status codes are the S_* macros defined in various header files such
as devSup.h drvSup.h and recSup.h. The message string is the comment
that immediately follows the macro definition which gets extracted from
the header file at IOC build-time. Support modules can add their own
status codes but the interface for doing that has not been revised for
many years and might not work too well (I have looked at it a little but
haven't done much work to improve it yet).

HTH,

- Andrew

-- 
Light thinks it travels faster than anything but it is wrong.
No matter how fast light travels, it finds the darkness has
always got there first, and is waiting for it.
    -- Terry Pratchett, Reaper Man

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System