Tariq Kurd

discussed with @PeterRugg - we're now trying out `default mode` and `capability mode` Opinions?

> > discussed with @PeterRugg - we're now trying out `default mode` and `capability mode` Opinions? > > How about "compatibility mode" and "capability mode"? After all, the mode is...

> Default isn’t necessarily the default true. I think I should put the name of the mode in a variable! What are we going to call it? * `unsafe` mode?...

We need to choose a name before friday! Opinions?

yes - I'm expecting legacy mode to look like a normal RISC-V core (XLEN access to MTVEC etc.) CRE=1 gives access to the new CSRs in legacy mode, but with...

this is intended to fix https://github.com/riscv/riscv-cheri/issues/213

The reason not to use the same encoding scheme for RV64 was to allow more combinations, so if we ban them all requiring a spin of the silicon to allow...

I'd vote for illegal combinations making the cap invalid, so CBLD would reject them. So if those bit patterns do occur it will only be due to a future extension...

https://github.com/riscv/riscv-CMOs/issues/56 We've already asked for FLUSH to be a valid implementation of INVAL which is almost is in the CMO spec already but it's better to clarify it. Clearly the...

presumably this can happen because a revocation sweep hasn't happened? If the revocation has happened then stale caps won't be around. Or... when closing down a compartment it should be...