ISACOV: Sequential sampling incorrect

Created by: silabs-robin

Type

• Functionally incorrect behavior

Steps to Reproduce

Run any test/regression and look at rev32_seq_cg.

Additional context

It seems like the sampling of sequences of instructions is non-functional.

When I look at coverage metrics for a "full" regression, it seems only "A->A" transitions are covered (e.g. "ADDI->ADDI"). (See screenshot below.) This indicates that something about the coverage definition or its sampling is not working as intended. @AyoubJalali are you seeing the same results as I do?

I don't know what is needed to get this working, but the principle is simple so it could potentially be easy.

---

ps. For "instruction group" coverage, it seems to be working. Except that we need to exclude "UNKNOWN_GROUP".