KVM: x86: Print CPU of last attempted VM-entry when dumping VMCS/VMCB
authorJim Mattson <jmattson@google.com>
Mon, 21 Jun 2021 22:16:48 +0000 (15:16 -0700)
committerPaolo Bonzini <pbonzini@redhat.com>
Thu, 24 Jun 2021 08:31:13 +0000 (04:31 -0400)
Failed VM-entry is often due to a faulty core. To help identify bad
cores, print the id of the last logical processor that attempted
VM-entry whenever dumping a VMCS or VMCB.

Signed-off-by: Jim Mattson <jmattson@google.com>
Message-Id: <20210621221648.1833148-1-jmattson@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
arch/x86/kvm/svm/svm.c
arch/x86/kvm/vmx/vmx.c

index 4cee285..8834822 100644 (file)
@@ -3132,6 +3132,8 @@ static void dump_vmcb(struct kvm_vcpu *vcpu)
                return;
        }
 
+       pr_err("VMCB %p, last attempted VMRUN on CPU %d\n",
+              svm->current_vmcb->ptr, vcpu->arch.last_vmentry_cpu);
        pr_err("VMCB Control Area:\n");
        pr_err("%-20s%04x\n", "cr_read:", control->intercepts[INTERCEPT_CR] & 0xffff);
        pr_err("%-20s%04x\n", "cr_write:", control->intercepts[INTERCEPT_CR] >> 16);
index ab6f682..684daa3 100644 (file)
@@ -5724,6 +5724,8 @@ void dump_vmcs(struct kvm_vcpu *vcpu)
        if (cpu_has_secondary_exec_ctrls())
                secondary_exec_control = vmcs_read32(SECONDARY_VM_EXEC_CONTROL);
 
+       pr_err("VMCS %p, last attempted VM-entry on CPU %d\n",
+              vmx->loaded_vmcs->vmcs, vcpu->arch.last_vmentry_cpu);
        pr_err("*** Guest State ***\n");
        pr_err("CR0: actual=0x%016lx, shadow=0x%016lx, gh_mask=%016lx\n",
               vmcs_readl(GUEST_CR0), vmcs_readl(CR0_READ_SHADOW),