On Tue, Oct 4, 2016 at 12:25 AM, Jan Beulich <jbeul...@suse.com> wrote: >>>> On 04.10.16 at 00:38, <m...@kylehuey.com> wrote: >> @@ -2701,9 +2706,13 @@ static int vmx_msr_read_intercept(unsigned int msr, >> uint64_t *msr_content) >> break; >> >> case MSR_INTEL_PLATFORM_INFO: >> - if ( rdmsr_safe(MSR_INTEL_PLATFORM_INFO, *msr_content) ) >> - goto gp_fault; >> - *msr_content = 0; >> + if ( is_pvh_vcpu(current) && !cpu_has_cpuid_faulting ) > > Any specific reason to treat PVHv1 differently? Also don't you mean > || instead of && ?
Looking at it again I think there isn't, because although it ends up calling pv_cpuid it still exits the VM via EXIT_REASON_CPUID, so hardware support for cpuid faulting is not required here. - kyle _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
