Canberk Topal

@GregAC Do you mind if I take this?

I am looking at our coverage implementation and it looks like for the first item, we already define some bins for it, although they are kind of buried deep. Would...

For the second item we have a more general `cp_edit_locked_pmpcfg` for each pmpcfg register. I believe I can turn that into a cross that has two bins: writes to locked...

I haven't looked at the first one yet but here are my overall findings: For the second one, I am not exactly sure what region 90 means but I can...

After having a more detailed look at the reports and testing out Marno's PR for ePMP this one looks like it's going to be a bigger task than I first...

> > Region override: This one is only possible when we are in NA4 mode in PMP > > Why is this the case? A smaller region inside a larger...

Opened up https://github.com/google/riscv-dv/pull/882 for fixing up broken randomization constraints in riscv-dv that I mentioned in the previous comment.

First item is getting hit with the latest work around pmp randomization. Second item will need further investigation. Third item will be complete after ePMP integration is done. Already hitting...

Second item can be solved by adding a "if we are enabling `pmp_randomize`, do not owerride `pmp_cfg0`" clause to the place where we do PMP related instruction generation in riscv-dv....

I've run `riscv_pmp_full_random_test` 12 times with coverage enabled, looks like we are hitting the last point of this task which was `NA4 and NAPOT matching modes and all privileges other...