update return register accordingly for all arch

[ucgJhe]

update return register accordingly, so setReturnAddress could work as expected.

[elicn]

I am wondering why these haven't failed so far.. @ucgJhe, did you encounter any failures that brought you to fix these?

Also, would appreciate another pair of eyes that are more familiar with those arch that I do.
@wtdcode, what do you think?

[ucgJhe]

I am wondering why these haven't failed so far.. @ucgJhe, did you encounter any failures that brought you to fix these?

Also, would appreciate another pair of eyes that are more familiar with those arch that I do. @wtdcode, what do you think?

Basically, I've been working on improving function call feature and happened to found this is not in the right place.

[xwings]

I am wondering why these haven't failed so far.. @ucgJhe, did you encounter any failures that brought you to fix these?
Also, would appreciate another pair of eyes that are more familiar with those arch that I do. @wtdcode, what do you think?

Basically, I've been working on improving function call feature and happened to found this is not in the right place.

interesting. i think you should start to add in some small unit test. i remember u told me the test run too long. but again, incomplete test it ways better better then no test.

[wtdcode]

I am wondering why these haven't failed so far.. @ucgJhe, did you encounter any failures that brought you to fix these?

Also, would appreciate another pair of eyes that are more familiar with those arch that I do. @wtdcode, what do you think?

The comment seems left by me. IIRC, the reason behind the comment is that, depending on the context, the return address may be on stack, say interrupt handlers. Though in most cases for userland code, LR is used instead.