Stefano Stabellini <sstabell...@kernel.org> writes: > On Mon, 25 Jul 2016, Vitaly Kuznetsov wrote: >> >> While we're not obliged to have the same type for xen_vcpu_id on all >> arches I see no point in diverging without a reason. I can do v3 making >> the mapping uint32 > > I agree that making the mapping uint32_t would be desirable. It would > even make sense from the int types point of view in Linux. > >>and indicating the missing value as U32_MAX-1 if nobody is against the >>idea. > > Why U32_MAX-1? (int)-1 is (unsigned)U32_MAX. Even > XEN_INVALID_MAX_VCPU_ID is defined as (~0U).
Yes, my bad, of course it should be U32_MAX. -- Vitaly _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel