Since we recognize invalid structions in the regression tests, we get a failure in regression test
for stm8. There are other such new failures, but most of the failing tests call some variant of printf() or use function pointers. This test doesn't as is quite short, which rules out bugs in printf() and function pointers.
The invalid instructions might be due to stack corruption in code generation. But if they aren't, then we have a bug in the assembler or simulator: The assembler wrongly generating an incorrect instruction or the simulator incorrectly reporting an error.