Re: [PATCH v3] is_system_domain: replace open-coded instances

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

On 12/21/21 4:20 AM, Jan Beulich wrote:
> On 20.12.2021 17:28, Daniel P. Smith wrote:
> > From: Christopher Clark <christopher.w.clark@xxxxxxxxx>
> >
> > This is a split out of the hyperlaunch dom0 series.
> >
> > There were several instances of open-coded domid range checking. This commit
> > replaces those with the is_system_domain or is_system_domid inline function.
> >
> > Signed-off-by: Christopher Clark <christopher.w.clark@xxxxxxxxx>
> > Signed-off-by: Daniel P. Smith <dpsmith@xxxxxxxxxxxxxxxxxxxx>
> > Acked-by: Dario Faggioli <dfaggioli@xxxxxxxx>
> While I'm not outright opposed, I'd still like to raise the question whether
> we really want to intermix "is system domain" and "is in-range domain ID"
> predicates. Personally I'd prefer the latter to remain open-coded range
> checks.
>
> > --- a/xen/include/xen/sched.h
> > +++ b/xen/include/xen/sched.h
> > @@ -613,9 +613,14 @@ extern struct vcpu *idle_vcpu[NR_CPUS];
> >   #define is_idle_domain(d) ((d)->domain_id == DOMID_IDLE)
> >   #define is_idle_vcpu(v)   (is_idle_domain((v)->domain))
> >   
> > +static inline bool is_system_domid(domid_t id)
> > +{
> > +    return (id >= DOMID_FIRST_RESERVED);
> Nit: Generally we omit parentheses in cases like this, ...
ack

> > +}
> > +
> >   static inline bool is_system_domain(const struct domain *d)
> >   {
> > -    return d->domain_id >= DOMID_FIRST_RESERVED;
> ... just like was the case here.
>
> Jan