Commit 27bd5fdc authored by Nikunj A Dadhania's avatar Nikunj A Dadhania Committed by Paolo Bonzini
Browse files

KVM: SEV-ES: Prevent MSR access post VMSA encryption 



KVM currently allows userspace to read/write MSRs even after the VMSA is
encrypted. This can cause unintentional issues if MSR access has side-
effects. For ex, while migrating a guest, userspace could attempt to
migrate MSR_IA32_DEBUGCTLMSR and end up unintentionally disabling LBRV on
the target. Fix this by preventing access to those MSRs which are context
switched via the VMSA, once the VMSA is encrypted.

Suggested-by: default avatarSean Christopherson <seanjc@google.com>
Signed-off-by: default avatarNikunj A Dadhania <nikunj@amd.com>
Signed-off-by: default avatarRavi Bangoria <ravi.bangoria@amd.com>
Message-ID: <20240531044644.768-2-ravi.bangoria@amd.com>
Signed-off-by: default avatarPaolo Bonzini <pbonzini@redhat.com>
parent b4bd5564

arch/x86/kvm/svm/svm.c

+18 −0
Original line number Diff line number Diff line
@@ -2822,10 +2822,24 @@ static int svm_get_msr_feature(struct kvm_msr_entry *msr) 
	return 0;

}



static bool

sev_es_prevent_msr_access(struct kvm_vcpu *vcpu, struct msr_data *msr_info)

{

	return sev_es_guest(vcpu->kvm) &&

	       vcpu->arch.guest_state_protected &&

	       svm_msrpm_offset(msr_info->index) != MSR_INVALID &&

	       !msr_write_intercepted(vcpu, msr_info->index);

}



static int svm_get_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)

{

	struct vcpu_svm *svm = to_svm(vcpu);



	if (sev_es_prevent_msr_access(vcpu, msr_info)) {

		msr_info->data = 0;

		return -EINVAL;

	}



	switch (msr_info->index) {

	case MSR_AMD64_TSC_RATIO:

		if (!msr_info->host_initiated &&
@@ -2976,6 +2990,10 @@ static int svm_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr) 


	u32 ecx = msr->index;

	u64 data = msr->data;



	if (sev_es_prevent_msr_access(vcpu, msr))

		return -EINVAL;



	switch (ecx) {

	case MSR_AMD64_TSC_RATIO: