Manolo de Medici, le mar. 24 mars 2026 21:11:05 +0000, a ecrit: > Creating a processor set reserve physical limited resources, > only root should be allowed to do that.
I agree with that and commited it, but > Further, allowing users to inspect the any processor sets that is not > their own shouldn't be allowed. If they are not allowed to inspect some host properties, they shouldn't have obtained the host_t from the start? Samuel > This modification is even more important if we consider that > processor sets and processor affinities will clash in several > aspects. Principally I plan to set the rules that removing > any processor from a processor set will nullify any affinities > of all processes belonging to that processor set. > --- > include/mach/mach_host.defs | 6 +++--- > 1 file changed, 3 insertions(+), 3 deletions(-) > > diff --git a/include/mach/mach_host.defs b/include/mach/mach_host.defs > index 3b9b0c58..372f3859 100644 > --- a/include/mach/mach_host.defs > +++ b/include/mach/mach_host.defs > @@ -82,7 +82,7 @@ skip; /* old yyy_processor_control */ > * Get default processor set for host. > */ > routine processor_set_default( > - host : host_t; > + host_priv : host_priv_t; > out default_set : processor_set_name_t); > > skip; /* old xxx_processor_set_default_priv */ > @@ -92,7 +92,7 @@ skip; /* old xxx_processor_set_default_priv */ > * and name port for obtaining information. > */ > routine processor_set_create( > - host : host_t; > + host_priv : host_priv_t; > out new_set : processor_set_t; > out new_name : processor_set_name_t); > > @@ -245,7 +245,7 @@ routine processor_set_threads( > * List all processor sets on host. > */ > routine host_processor_sets( > - host : host_t; > + host_priv : host_priv_t; > out processor_sets : processor_set_name_array_t); > > /* > -- > 2.53.0 > >
