From: Oliver Upton Date: Fri, 23 May 2025 19:47:21 +0000 (-0700) Subject: KVM: arm64: Unmap vLPIs affected by changes to GSI routing information X-Git-Tag: v6.16-rc1~78^2^2~2 X-Git-Url: https://git.kernel.dk/?a=commitdiff_plain;h=4bf3693d36af9768c9bcc1df3a12d00ad6ea8083;p=linux-block.git KVM: arm64: Unmap vLPIs affected by changes to GSI routing information KVM's interrupt infrastructure is dodgy at best, allowing for some ugly 'off label' usage of the various UAPIs. In one example, userspace can change the routing entry of a particular "GSI" after configuring irqbypass with KVM_IRQFD. KVM/arm64 is oblivious to this, and winds up preserving the stale translation in cases where vLPIs are configured. Honor userspace's intentions and tear down the vLPI mapping if affected by a "GSI" routing change. Make no attempt to reconstruct vLPIs if the new target is an MSI and just fall back to software injection. Tested-by: Sweet Tea Dorminy Signed-off-by: Oliver Upton Link: https://lore.kernel.org/r/20250523194722.4066715-5-oliver.upton@linux.dev Signed-off-by: Marc Zyngier --- diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c index 1de49b48e35e..505d504b52b5 100644 --- a/arch/arm64/kvm/arm.c +++ b/arch/arm64/kvm/arm.c @@ -2790,6 +2790,7 @@ int kvm_arch_irq_bypass_add_producer(struct irq_bypass_consumer *cons, return kvm_vgic_v4_set_forwarding(irqfd->kvm, prod->irq, &irqfd->irq_entry); } + void kvm_arch_irq_bypass_del_producer(struct irq_bypass_consumer *cons, struct irq_bypass_producer *prod) { @@ -2803,6 +2804,28 @@ void kvm_arch_irq_bypass_del_producer(struct irq_bypass_consumer *cons, kvm_vgic_v4_unset_forwarding(irqfd->kvm, prod->irq); } +bool kvm_arch_irqfd_route_changed(struct kvm_kernel_irq_routing_entry *old, + struct kvm_kernel_irq_routing_entry *new) +{ + if (new->type != KVM_IRQ_ROUTING_MSI) + return true; + + return memcmp(&old->msi, &new->msi, sizeof(new->msi)); +} + +int kvm_arch_update_irqfd_routing(struct kvm *kvm, unsigned int host_irq, + uint32_t guest_irq, bool set) +{ + /* + * Remapping the vLPI requires taking the its_lock mutex to resolve + * the new translation. We're in spinlock land at this point, so no + * chance of resolving the translation. + * + * Unmap the vLPI and fall back to software LPI injection. + */ + return kvm_vgic_v4_unset_forwarding(kvm, host_irq); +} + void kvm_arch_irq_bypass_stop(struct irq_bypass_consumer *cons) { struct kvm_kernel_irqfd *irqfd =