[clang] [llvm] [RISCV] Make zvknha a subset of zvknhb (PR #178680)

[Ramkumar Ramachandra via cfe-commits]

artagnon wrote:

> > > zvknha and zvknhb are mutually exclusive [riscv/riscv-crypto#364 (comment)](https://github.com/riscv/riscv-crypto/issues/364#issuecomment-1726782096)
> > 
> > 
> > I don't understand the comment: the table is quite clear?
> 
> The table just says both zvknha and zvknhb implement sha256, it doesn’t mean zvknhb implies zvknha unless there’s wording saying that as @topperc mentioned.

Moreover, the table clearly shows that zvknha is a subset of zvknhb. In hardware, would you have two different circuits implementing SHA256 if you support zvkhna and zvknhb? If not, then I don't see the semantic difference. Perhaps the missing wording is an oversight?

https://github.com/llvm/llvm-project/pull/178680