KVM: VMX: Re-enter guest in fastpath for "spurious" preemption timer exits

Re-enter the guest in the fast path if VMX preeemption timer VM-Exit was
"spurious", i.e. if KVM "soft disabled" the timer by writing -1u and by
some miracle the timer expired before any other VM-Exit occurred.  This is
just an intermediate step to cleaning up the preemption timer handling,
optimizing these types of spurious VM-Exits is not interesting as they are
extremely rare/infrequent.

Link: https://lore.kernel.org/r/20240110012705.506918-3-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>

diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
index e3653d11f00156d3685d2282f611dad580a0ab00..a44b09e35fd6ec02fd629b131d176aa9783f9e8d 100644 (file)
--- a/arch/x86/kvm/vmx/vmx.c
+++ b/arch/x86/kvm/vmx/vmx.c
@@ -5992,8 +5992,15 @@ static fastpath_t handle_fastpath_preemption_timer(struct kvm_vcpu *vcpu)
 {
        struct vcpu_vmx *vmx = to_vmx(vcpu);
 
-       if (!vmx->req_immediate_exit &&
-           !unlikely(vmx->loaded_vmcs->hv_timer_soft_disabled)) {
+       /*
+        * In the *extremely* unlikely scenario that this is a spurious VM-Exit
+        * due to the timer expiring while it was "soft" disabled, just eat the
+        * exit and re-enter the guest.
+        */
+       if (unlikely(vmx->loaded_vmcs->hv_timer_soft_disabled))
+               return EXIT_FASTPATH_REENTER_GUEST;
+
+       if (!vmx->req_immediate_exit) {
                kvm_lapic_expired_hv_timer(vcpu);
                return EXIT_FASTPATH_REENTER_GUEST;
        }