KVM: SVM: Don't intercept #GP for SEV guests

Never intercept #GP for SEV guests as reading SEV guest private memory
will return cyphertext, i.e. emulating on #GP can't work as intended.

Cc: stable@vger.kernel.org
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Cc: Brijesh Singh <brijesh.singh@amd.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Liam Merwick <liam.merwick@oracle.com>
Message-Id: <20220120010719.711476-4-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c
index 37eb316..defc91a 100644 (file)
--- a/arch/x86/kvm/svm/svm.c
+++ b/arch/x86/kvm/svm/svm.c
@@ -312,7 +312,11 @@ int svm_set_efer(struct kvm_vcpu *vcpu, u64 efer)
                                return ret;
                        }
 
-                       if (svm_gp_erratum_intercept)
+                       /*
+                        * Never intercept #GP for SEV guests, KVM can't
+                        * decrypt guest memory to workaround the erratum.
+                        */
+                       if (svm_gp_erratum_intercept && !sev_guest(vcpu->kvm))
                                set_exception_intercept(svm, GP_VECTOR);
                }
        }
@@ -1010,9 +1014,10 @@ static void init_vmcb(struct kvm_vcpu *vcpu)
         * Guest access to VMware backdoor ports could legitimately
         * trigger #GP because of TSS I/O permission bitmap.
         * We intercept those #GP and allow access to them anyway
-        * as VMware does.
+        * as VMware does.  Don't intercept #GP for SEV guests as KVM can't
+        * decrypt guest memory to decode the faulting instruction.
         */
-       if (enable_vmware_backdoor)
+       if (enable_vmware_backdoor && !sev_guest(vcpu->kvm))
                set_exception_intercept(svm, GP_VECTOR);
 
        svm_set_intercept(svm, INTERCEPT_INTR);