Re: [RFC 3/6] target/riscv: rvk: add flag support for Zk/Zkn/Zknd/Zknd/Zkne/Zknh/Zks/Zksed/Zksh/Zkr

[liweiwei]

在 2021/11/3 上午9:21, Richard Henderson 写道:
> On 11/2/21 9:06 PM, liweiwei wrote:
> > 在 2021/11/3 上午1:56, Richard Henderson 写道:
> > > On 11/1/21 11:11 PM, liweiwei wrote:
> > > > +        if (cpu->cfg.ext_zk) {
> > > > +            cpu->cfg.ext_zbkb = true;
> > > > +            cpu->cfg.ext_zbkc = true;
> > > > +            cpu->cfg.ext_zbkx = true;
> > > > +            cpu->cfg.ext_zknd = true;
> > > > +            cpu->cfg.ext_zkne = true;
> > > > +            cpu->cfg.ext_zknh = true;
> > > > +            cpu->cfg.ext_zkr = true;
> > > > +        }
> > >
> > > Section 2.12 lists instead the larger Zkn, Zks, Zkt extensions.
> > > I think it's better to follow that.
> > OK. I'll replace this with setting ext_zkn, ext_zks, ext_zkt true. By 
> > the way, Zkt only includes partial RVI/M/C instructions. Is it 
> > neccessary to distinguish them?
>
> I don't know if Zkt needs to be called out; it probably depends on 
> whether it is exposed to the guest via some MISA csr extension.
It seems not  exposed to the guest via some MISA csr extension. And it 
also doesn't define a set of instructions available in the core. It just 
restricts the set of instructions to have 'data-independent execution time'.
> r~