Re: [Xen-devel] [PATCH 6/6] x86/msr: Blacklist various MSRs which guests definitely shouldn't be using

2018-03-01 Thread Jan Beulich
>>> On 26.02.18 at 18:35, wrote: > @@ -183,6 +187,15 @@ int guest_rdmsr(const struct vcpu *v, uint32_t msr, > uint64_t *val) > ret = guest_rdmsr_x2apic(v, msr, val); > goto out; > > +case 0xc80: > +/* Silicon Debug Inferface not advertised to guests. */ > +

Re: [Xen-devel] [PATCH 6/6] x86/msr: Blacklist various MSRs which guests definitely shouldn't be using

2018-02-27 Thread Roger Pau Monné
On Mon, Feb 26, 2018 at 05:35:19PM +, Andrew Cooper wrote: > The main purpose is to blacklist the Intel Resource Director Technology MSRs. > We do not yet virtualise support for guests, but Linux has been observed to > probe for these MSRs without checking CPUID first. > > The architecturally

[Xen-devel] [PATCH 6/6] x86/msr: Blacklist various MSRs which guests definitely shouldn't be using

2018-02-26 Thread Andrew Cooper
The main purpose is to blacklist the Intel Resource Director Technology MSRs. We do not yet virtualise support for guests, but Linux has been observed to probe for these MSRs without checking CPUID first. The architecturally inaccessable ranges don't need to fall back into the legacy ranges, becau