[PATCH 9/9] KVM: x86: Disable intercept for Intel processor trace MSRs

[PATCH 9/9] KVM: x86: Disable intercept for Intel processor trace MSRs

[Luwei Kang]

From: Chao Peng <chao.p.peng@linux.intel.com>

Trap for Intel processor trace is none sense. Pass through to
guest directly.

Signed-off-by: Chao Peng <chao.p.peng@linux.intel.com>
Signed-off-by: Luwei Kang <luwei.kang@intel.com>
---
 arch/x86/kvm/vmx.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 3c9ce3e..58606ce 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -7076,6 +7076,20 @@ static __init int hardware_setup(void)
 	if (pt_mode == PT_MODE_GUEST)
 		pt_register_virt_ops(&pt_virt_ops);
 
+	if (pt_mode == PT_MODE_GUEST || pt_mode == PT_MODE_HOST_GUEST) {
+		u32 i, eax, ebx, ecx, edx;
+
+		cpuid_count(0x14, 1, &eax, &ebx, &ecx, &edx);
+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_CTL, false);
+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_STATUS, false);
+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_OUTPUT_BASE, false);
+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_OUTPUT_MASK, false);
+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_CR3_MATCH, false);
+		for (i = 0; i < (eax & 0x7); i++)
+			vmx_disable_intercept_for_msr(MSR_IA32_RTIT_ADDR0_A + i,
+									false);
+	}
+
 	return alloc_kvm_area();
 
 out:
-- 
1.8.3.1

Re: [PATCH 9/9] KVM: x86: Disable intercept for Intel processor trace MSRs

[Konrad Rzeszutek Wilk]

On October 16, 2017 8:14:11 AM EDT, Luwei Kang <luwei.kang@intel.com> wrote:
>From: Chao Peng <chao.p.peng@linux.intel.com>
>
>Trap for Intel processor trace is none sense. Pass through to
>guest directly.


And none of those MSRs can be subverted by the guest? That is none of these should be filtered / audited first?


>
>Signed-off-by: Chao Peng <chao.p.peng@linux.intel.com>
>Signed-off-by: Luwei Kang <luwei.kang@intel.com>
>---
> arch/x86/kvm/vmx.c | 14 ++++++++++++++
> 1 file changed, 14 insertions(+)
>
>diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
>index 3c9ce3e..58606ce 100644
>--- a/arch/x86/kvm/vmx.c
>+++ b/arch/x86/kvm/vmx.c
>@@ -7076,6 +7076,20 @@ static __init int hardware_setup(void)
> 	if (pt_mode == PT_MODE_GUEST)
> 		pt_register_virt_ops(&pt_virt_ops);
> 
>+	if (pt_mode == PT_MODE_GUEST || pt_mode == PT_MODE_HOST_GUEST) {
>+		u32 i, eax, ebx, ecx, edx;
>+
>+		cpuid_count(0x14, 1, &eax, &ebx, &ecx, &edx);
>+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_CTL, false);
>+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_STATUS, false);
>+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_OUTPUT_BASE, false);
>+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_OUTPUT_MASK, false);
>+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_CR3_MATCH, false);
>+		for (i = 0; i < (eax & 0x7); i++)
>+			vmx_disable_intercept_for_msr(MSR_IA32_RTIT_ADDR0_A + i,
>+									false);
>+	}
>+
> 	return alloc_kvm_area();
> 
> out:


Thanks!

RE: [PATCH 9/9] KVM: x86: Disable intercept for Intel processor trace MSRs

[Kang, Luwei]

> On October 16, 2017 8:14:11 AM EDT, Luwei Kang <luwei.kang@intel.com> wrote:
> >From: Chao Peng <chao.p.peng@linux.intel.com>
> >
> >Trap for Intel processor trace is none sense. Pass through to guest
> >directly.
> 
> 
> And none of those MSRs can be subverted by the guest? That is none of these should be filtered / audited first?
> 
   Currently, I think it is no necessary. For example, host context(MSRs) would be saved and guest context would be restored before VM-entry in host-guest mode. And IA32_RTIT_CTL will be written with the value of the associated Guest State field of the VMCS on VM entry. I am not very clear what your point. Could you make a specific scenes?

Thanks,
Luwei Kang

> 
> >
> >Signed-off-by: Chao Peng <chao.p.peng@linux.intel.com>
> >Signed-off-by: Luwei Kang <luwei.kang@intel.com>
> >---
> > arch/x86/kvm/vmx.c | 14 ++++++++++++++
> > 1 file changed, 14 insertions(+)
> >
> >diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index
> >3c9ce3e..58606ce 100644
> >--- a/arch/x86/kvm/vmx.c
> >+++ b/arch/x86/kvm/vmx.c
> >@@ -7076,6 +7076,20 @@ static __init int hardware_setup(void)
> > 	if (pt_mode == PT_MODE_GUEST)
> > 		pt_register_virt_ops(&pt_virt_ops);
> >
> >+	if (pt_mode == PT_MODE_GUEST || pt_mode == PT_MODE_HOST_GUEST) {
> >+		u32 i, eax, ebx, ecx, edx;
> >+
> >+		cpuid_count(0x14, 1, &eax, &ebx, &ecx, &edx);
> >+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_CTL, false);
> >+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_STATUS, false);
> >+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_OUTPUT_BASE, false);
> >+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_OUTPUT_MASK, false);
> >+		vmx_disable_intercept_for_msr(MSR_IA32_RTIT_CR3_MATCH, false);
> >+		for (i = 0; i < (eax & 0x7); i++)
> >+			vmx_disable_intercept_for_msr(MSR_IA32_RTIT_ADDR0_A + i,
> >+									false);
> >+	}
> >+
> > 	return alloc_kvm_area();
> >
> > out:
> 
> 
> Thanks!