KVM: VMX: Don't unblock vCPU w/ Posted IRQ if IRQs are disabled in guest
authorPaolo Bonzini <pbonzini@redhat.com>
Tue, 16 Nov 2021 14:32:47 +0000 (09:32 -0500)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Tue, 8 Mar 2022 18:12:34 +0000 (19:12 +0100)
[ Upstream commit 1831fa44df743a7cdffdf1c12c799bf6f3c12b8c ]

Don't configure the wakeup handler when a vCPU is blocking with IRQs
disabled, in which case any IRQ, posted or otherwise, should not be
recognized and thus should not wake the vCPU.

Fixes: bf9f6ac8d749 ("KVM: Update Posted-Interrupts Descriptor when vCPU is blocked")
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20211009021236.4122790-2-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
arch/x86/kvm/vmx/posted_intr.c

index 21ea58d..696ad48 100644 (file)
@@ -147,7 +147,8 @@ int pi_pre_block(struct kvm_vcpu *vcpu)
        struct pi_desc old, new;
        struct pi_desc *pi_desc = vcpu_to_pi_desc(vcpu);
 
-       if (!vmx_can_use_vtd_pi(vcpu->kvm))
+       if (!vmx_can_use_vtd_pi(vcpu->kvm) ||
+           vmx_interrupt_blocked(vcpu))
                return 0;
 
        WARN_ON(irqs_disabled());