It would be better if you could tell us why you think this unconventional method of handling the end of the interrupt is so necessary.

Tery, I have programmed microcontrollers of all types for well past 20 years and have never needed to do what you are trying to do. So it makes me suspicious when you say you need to handle it this way.

Abhushek showed you one method of how this can be done. But Peters method that he described is much better. That is, in the interrupt routine set a bit variable that says "interrupt has happened". In the main line code master loop periodically poll this same bit variable and when it is seen set then clear it again and transfer program flow to the necessary logic.

It that seems too unwieldly to you then I would ask, "is the job you are trying to do at the target jump some short activity?" If so then just put this additional program logic right in the interrupt routine.

In conclusion I'll just say, what you are trying to do IS unconventional. The need for the supposed transfer like you want is due to the improper design of the mainline program code. So if you can share more of what you are tring to achieve I am certain that myself and others here can help suggest how to re-organize things to get you to a solution that is smooth and does not have to bugger the stack contents at the end of the interrupt.

Michael Karas