Re: [PATCH] x86/intel_rdt: use rdmsr_safe() to workaround AWS host issue

2019-01-10 Thread Borislav Petkov
On Thu, Jan 10, 2019 at 11:32:47AM +0100, Vitaly Kuznetsov wrote: > The other thing is: how can we be sure that there's no hypervisor > exposing these feature already? Even if open-source hypervisors like > KVM/Xen don't do it it doesn't prove anything: there are numerous > proprietary hypervisors

Re: [PATCH] x86/intel_rdt: use rdmsr_safe() to workaround AWS host issue

2019-01-10 Thread Vitaly Kuznetsov
Tony Luck writes: > On Wed, Jan 9, 2019 at 5:00 AM Borislav Petkov wrote: >> >> On Wed, Jan 09, 2019 at 01:09:31PM +0100, Vitaly Kuznetsov wrote: >> > Hm, why is that? In theory, hypervisors can pass through or emulate the >> > required MSRs... >> >> ...and when the theory becomes reality we'll

Re: [PATCH] x86/intel_rdt: use rdmsr_safe() to workaround AWS host issue

2019-01-09 Thread Tony Luck
On Wed, Jan 9, 2019 at 5:00 AM Borislav Petkov wrote: > > On Wed, Jan 09, 2019 at 01:09:31PM +0100, Vitaly Kuznetsov wrote: > > Hm, why is that? In theory, hypervisors can pass through or emulate the > > required MSRs... > > ...and when the theory becomes reality we'll remove the check. In practi

Re: [PATCH] x86/intel_rdt: use rdmsr_safe() to workaround AWS host issue

2019-01-09 Thread Borislav Petkov
On Wed, Jan 09, 2019 at 01:09:31PM +0100, Vitaly Kuznetsov wrote: > Hm, why is that? In theory, hypervisors can pass through or emulate the > required MSRs... ...and when the theory becomes reality we'll remove the check. -- Regards/Gruss, Boris. Good mailing practices for 400: avoid top-po

Re: [PATCH] x86/intel_rdt: use rdmsr_safe() to workaround AWS host issue

2019-01-09 Thread Vitaly Kuznetsov
Borislav Petkov writes: > So, assuming RDT is not going to be supported in a guest Hm, why is that? In theory, hypervisors can pass through or emulate the required MSRs... -- Vitaly

Re: [PATCH] x86/intel_rdt: use rdmsr_safe() to workaround AWS host issue

2019-01-09 Thread Borislav Petkov
On Tue, Jan 08, 2019 at 09:39:46PM -0800, Reinette Chatre wrote: > To clarify please the last sentence applies to real Haswell server CPUs > (not virtualized as prompted this work, my apologies for the confusion) > that support Intel RDT but does not have CPUID enumeration for this > feature. With

Re: [PATCH] x86/intel_rdt: use rdmsr_safe() to workaround AWS host issue

2018-12-20 Thread Vitaly Kuznetsov
Borislav Petkov writes: > On Thu, Dec 20, 2018 at 02:40:46PM +0100, Vitaly Kuznetsov wrote: >> It was found that AWS x1 instances (Xen-based) lack xen.git commit >> 1f1d183d4900 (x86/HVM: don't give the wrong impression of WRMSR succeeding) >> and because of that the wrmsr_safe() check in cache_a

Re: [PATCH] x86/intel_rdt: use rdmsr_safe() to workaround AWS host issue

2018-12-20 Thread Borislav Petkov
On Thu, Dec 20, 2018 at 02:40:46PM +0100, Vitaly Kuznetsov wrote: > It was found that AWS x1 instances (Xen-based) lack xen.git commit > 1f1d183d4900 (x86/HVM: don't give the wrong impression of WRMSR succeeding) > and because of that the wrmsr_safe() check in cache_alloc_hsw_probe() > doesn't help

[PATCH] x86/intel_rdt: use rdmsr_safe() to workaround AWS host issue

2018-12-20 Thread Vitaly Kuznetsov
It was found that AWS x1 instances (Xen-based) lack xen.git commit 1f1d183d4900 (x86/HVM: don't give the wrong impression of WRMSR succeeding) and because of that the wrmsr_safe() check in cache_alloc_hsw_probe() doesn't help: the consequent rdmsr() blows up with unchecked MSR access error: RDMSR