Open KinzaQamar opened 2 days ago
I agree and think that this isn't actually reachable. My reasoning: the kmac_rom_vld_o
signal basically says "we read something from ROM on the last cycle". If counter_lnt
is true, it means "we've just read the last non-top item from ROM" and this implies that kmac_rom_vld_o
is high.
One option would be a waiver (possibly UNR), but I think it's probably cleaner to just drop the condition from the RTL. Ideally, by running UNR or a manual FPV flow to prove that the change is correct. I'll take this on.
The nightly regression coverage dashboard reports an uncovered conditional coverage inside rom_ctrl_fsm.sv here. The case 110 is uncovered. If
counter_lnt
is true, it means that we've read the last non-top item from ROM which basically implies thatkmac_rom_vld_o
is also high. That is a non-occurring case.