On Thu, 31 Aug 2017 16:30:59 +0200 David Hildenbrand <da...@redhat.com> wrote:
> On 31.08.2017 16:29, Cornelia Huck wrote: > > On Thu, 31 Aug 2017 15:11:28 +0200 > > David Hildenbrand <da...@redhat.com> wrote: > > > >>>> +S390CPU *s390_cpu_addr2state(uint16_t cpu_addr) > >>>> +{ > >>>> + S390CcwMachineState *ms = S390_CCW_MACHINE(qdev_get_machine()); > >>>> + > >>>> + if (cpu_addr >= max_cpus) { > >>>> + return NULL; > >>>> + } > >>>> + > >>>> + /* Fast lookup via CPU ID */ > >>>> + return ms->cpus[cpu_addr]; > >>>> +} > >>> > >>> I wonder whether that function should rather go into a file in > >>> target/s390x/ instead, since it is also used there and its prototype is > >>> in cpu.h ? > >> > >> I thought about the same thing, but as it works directly on the machine, > >> like ri_allowed() and friends. So I decided to keep it here for now. > >> > >> I'll think about moving the definition also into > >> include/hw/s390x/s390-virtio-ccw.h > > > > It would be a bit nicer. > > > > Adding patches right now to move everything out of cpu.h that lies under > the "/* outside of target/s390x/ */" section. :) > Ah, you really care about your patch count, don't you? :) (I think it's a good idea.)