When the kernel has no mapping for the vmcs02 virtual APIC page,
userspace MMIO completion is necessary to process nested posted
interrupts. This is not a configuration that KVM supports. Rather than
silently ignoring the problem, try to exit to userspace with
KVM_INTERNAL_ERROR.
Note that the event that triggers this error is consumed as a
side-effect of a call to kvm_check_nested_events. On some paths
(notably through kvm_vcpu_check_block), the error is dropped. In any
case, this is an incremental improvement over always ignoring the
error.
Signed-off-by: Jim Mattson <jmattson@xxxxxxxxxx>
---
arch/x86/kvm/vmx/nested.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/arch/x86/kvm/vmx/nested.c b/arch/x86/kvm/vmx/nested.c
index 7646e6e561ad..706c31821362 100644
--- a/arch/x86/kvm/vmx/nested.c
+++ b/arch/x86/kvm/vmx/nested.c
@@ -3700,7 +3700,7 @@ static int vmx_complete_nested_posted_interrupt(struct kvm_vcpu *vcpu)
if (max_irr != 256) {
vapic_page = vmx->nested.virtual_apic_map.hva;
if (!vapic_page)
- return 0;
+ goto mmio_needed;
__kvm_apic_update_irr(vmx->nested.pi_desc->pir,
vapic_page, &max_irr);
@@ -3714,6 +3714,10 @@ static int vmx_complete_nested_posted_interrupt(struct kvm_vcpu *vcpu)
nested_mark_vmcs12_pages_dirty(vcpu);
return 0;
+
+mmio_needed:
+ kvm_handle_memory_failure(vcpu, X86EMUL_IO_NEEDED, NULL);
+ return -ENXIO;
}
static void nested_vmx_inject_exception_vmexit(struct kvm_vcpu *vcpu,
--
2.31.1.818.g46aad6cb9e-goog
