Major Fault ONLY after Download for JSR or RET

[theColonel26]

On site and another Programmer is getting a Major Fault when downloading code that I gave him to interface withour machines.

I have JSRs that pass UDTs tags in and return UDTs tags. it only throws a on program start. If it is AFIed, started, and then the AFI is removed the Routine call works fine.

Processor is V30 5580

 

[dmroeder]

Is there some sort of infinite loop where the scan is trapped in that routine? Was the major fault related to the watchdog timer? Edit: what is the major fault?

 

[plvlce]

it only throws a on program start. If it is AFIed, started, and then the AFI is removed the Routine call works fine.

By 'program start' do you mean entering run mode? At a guess something used in the routine is set (initialized?) elsewhere, but the logic for that is after the JSR in the scan cycle. Or perhaps something using the first scan bit is causing the issue.

If, after getting it running via add/removal of AFI, it is flipped to program and then back to run does it fault out? If not that would suggest to me some specific tag values in what you are downloading is the issue.

+1 to wanting the actual fault.

 

[Operaghost]

Check the number of input parameters on the JSR matches the input parameters on your SBR instruction. Likewise check the RET and verify it is returning the same number of return parameters as the JSR is expecting.

If the JSR sends 1 input parameter, but in the subroutine there is no SBR to match that, no fault occurs. But if you pass 1 input and it is expecting 2, then a fault occurs.

Likewise with the RET. If you return 1 parameter but the JSR is expecting 2, then you have a fault.

Too many parameters being passed isn't an issue. The extras are discarded. But too few causes a fault.

OG