PRugg-Cap

@tariqkurd-repo pointed out that the table was slightly wrong in big exponent cases (negative-width bit-slices). This fixes it while hopefully not making it too much more confusing!

Clarify whether nonsense permission combinations are reserved in RV64

3

We currently carefully define the allowed permission encoding combinations in RV32, leaving some reserved. There's also a note describing how to treat the reserved combinations: https://github.com/riscv/riscv-cheri/blob/e3c20ffbfeebe4432a8f3994b31eed81374f297b/src/cap-description.adoc?plain=1#L184-L190 For RV64, we don't...

@sorear raised in https://github.com/riscv/riscv-cheri/issues/40 the following: > Under RV32, any capability represented with EF=0 E=0 can also be represented using EF=1 T8=1. Should the redundant case be considered malformed? I...

Fixes https://github.com/riscv/riscv-cheri/issues/344

Representability check description doesn't capture address-wrapping edge cases

2

The description of the representability check is slightly imprecise in cases where the base or top want to wrap the address space. A correction is described in the bounds section:...

Preserve tag if CBLD checks fail

3

https://github.com/riscv/riscv-cheri/issues/186 got fixed before I noticed, but I think we can safely fix it the other way, i.e. the "otherwise" case changes from "copy cs2 to cd and clear cd...