The commit a377ac1cd9d7b ("x86/entry: Move user return notifier out of loop") moved fire_user_return_notifiers() into the section with IRQs disabled, and it somewhat inadvertantly fixed the underlying issue that was papered over by commit 1650b4ebc99d ("KVM: Disable irq while unregistering user notifier"). Therefore, the comments and code are outdated. Aslo assert that IRQs are disabled in kvm_on_user_return(), as both fire_user_return_notifiers() and kvm_arch_disable_virtualization_cpu() are now in IRQs disabled state. Signed-off-by: Hou Wenlong --- arch/x86/kvm/x86.c | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 33fba801b205..84fc30a99be1 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -568,18 +568,18 @@ static void kvm_on_user_return(struct user_return_notifier *urn) struct kvm_user_return_msrs *msrs = container_of(urn, struct kvm_user_return_msrs, urn); struct kvm_user_return_msr_values *values; - unsigned long flags; /* - * Disabling irqs at this point since the following code could be - * interrupted and executed through kvm_arch_disable_virtualization_cpu() + * Assert that IRQs are disabled. KVM disables virtualization via IPI + * callback on reboot, and this code isn't safe for re-entrancy, e.g. + * receiving the IRQ after checking "registered" would lead to double + * deletion of KVM's notifier. */ - local_irq_save(flags); - if (msrs->registered) { - msrs->registered = false; - user_return_notifier_unregister(urn); - } - local_irq_restore(flags); + lockdep_assert_irqs_disabled(); + + msrs->registered = false; + user_return_notifier_unregister(urn); + for (slot = 0; slot < kvm_nr_uret_msrs; ++slot) { values = &msrs->values[slot]; if (values->host != values->curr) { -- 2.31.1