../

Meeting with Kevin

Why HARM and BPS differ

  • The template restricts some of the signals to a or c

    • This choice was made so that we avoid redundant rules
    • But this also restricts the kind of kinds of relations that we are mining
    • The reasonable assumptions is that the outputs signals don’t dictate the behavior of the previous stages
    • But should we make this assumptions
    • We know that some signals loop back
    • What would we do if didn’t have access to the RTL? What if this was just a bunch of signals? We know the directionality of the signals by interpreting the _o or _i suffix. But what if the signals weren’t named like this?

  • How does the program breakdown for HARM look like?

  • Can I just permute all the signal? This should be all possible globally implies relationship. We can then measure our coverage.