Re: [RFC 2/6] roles: provide abstraction for the possible domain roles

["Daniel P. Smith" <dpsmith@xxxxxxxxxxxxxxxxxxxx>]

On 8/8/23 11:15, Jan Beulich wrote:
> On 01.08.2023 22:20, Daniel P. Smith wrote:
> > @@ -1076,7 +1076,8 @@ static always_inline bool is_hardware_domain(const struct 
> > domain *d)
> >       if ( IS_ENABLED(CONFIG_PV_SHIM_EXCLUSIVE) )
> >           return false;
> >   
> > -    return evaluate_nospec(d == hardware_domain);
> > +    return evaluate_nospec(((d->role & ROLE_HARDWARE_DOMAIN) ||
> > +        is_initial_domain(d)) && (d == hardware_domain));
> >   }
> >   
> >   /* This check is for functionality specific to a control domain */
> > @@ -1085,7 +1086,8 @@ static always_inline bool is_control_domain(const struct 
> > domain *d)
> >       if ( IS_ENABLED(CONFIG_PV_SHIM_EXCLUSIVE) )
> >           return false;
> >   
> > -    return evaluate_nospec(d->is_privileged);
> > +    return evaluate_nospec((d->role & ROLE_CONTROL_DOMAIN) ||
> > +        is_initial_domain(d));
> >   }
> Why these complicated conditions, and not just the check of the single
> relevant role bit? (Also note that indentation used here doesn't really
> match our expectations, but there are other style issues in the patch
> as well, which I assume is okay for an RFC.)
They go away with suggestion to move ROLE_UNBOUNDED a concatenation of 
all the other roles.
Ack on the style.

v/r,
dps