diff options
| author | Sean Christopherson <seanjc@google.com> | 2025-06-11 14:35:40 -0700 |
|---|---|---|
| committer | Sean Christopherson <seanjc@google.com> | 2025-06-20 13:52:41 -0700 |
| commit | e295d2e7fbe69ddec772c951c466dfbfc1c96818 (patch) | |
| tree | 87319d701d3fb957d93e0d43aafe33fb7dfb0668 | |
| parent | irqbypass: Require producers to pass in Linux IRQ number during registration (diff) | |
| download | linux-e295d2e7fbe69ddec772c951c466dfbfc1c96818.tar.gz linux-e295d2e7fbe69ddec772c951c466dfbfc1c96818.zip | |
KVM: x86: Trigger I/O APIC route rescan in kvm_arch_irq_routing_update()
Trigger the I/O APIC route rescan that's performed for a split IRQ chip
after userspace updates IRQ routes in kvm_arch_irq_routing_update(), i.e.
before dropping kvm->irq_lock. Calling kvm_make_all_cpus_request() under
a mutex is perfectly safe, and the smp_wmb()+smp_mb__after_atomic() pair
in __kvm_make_request()+kvm_check_request() ensures the new routing is
visible to vCPUs prior to the request being visible to vCPUs.
In all likelihood, commit b053b2aef25d ("KVM: x86: Add EOI exit bitmap
inference") somewhat arbitrarily made the request outside of irq_lock to
avoid holding irq_lock any longer than is strictly necessary. And then
commit abdb080f7ac8 ("kvm/irqchip: kvm_arch_irq_routing_update renaming
split") took the easy route of adding another arch hook instead of risking
a functional change.
Note, the call to synchronize_srcu_expedited() does NOT provide ordering
guarantees with respect to vCPUs scanning the new routing; as above, the
request infrastructure provides the necessary ordering. I.e. there's no
need to wait for kvm_scan_ioapic_routes() to complete if it's actively
running, because regardless of whether it grabs the old or new table, the
vCPU will have another KVM_REQ_SCAN_IOAPIC pending, i.e. will rescan again
and see the new mappings.
Acked-by: Kai Huang <kai.huang@intel.com>
Link: https://lore.kernel.org/r/20250611213557.294358-2-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
| -rw-r--r-- | arch/x86/kvm/irq_comm.c | 10 | ||||
| -rw-r--r-- | include/linux/kvm_host.h | 4 | ||||
| -rw-r--r-- | virt/kvm/irqchip.c | 2 |
3 files changed, 3 insertions, 13 deletions
diff --git a/arch/x86/kvm/irq_comm.c b/arch/x86/kvm/irq_comm.c index d6d792b5d1bd..e2ae62ff9cc2 100644 --- a/arch/x86/kvm/irq_comm.c +++ b/arch/x86/kvm/irq_comm.c @@ -395,13 +395,6 @@ int kvm_setup_default_irq_routing(struct kvm *kvm) ARRAY_SIZE(default_routing), 0); } -void kvm_arch_post_irq_routing_update(struct kvm *kvm) -{ - if (!irqchip_split(kvm)) - return; - kvm_make_scan_ioapic_request(kvm); -} - void kvm_scan_ioapic_irq(struct kvm_vcpu *vcpu, u32 dest_id, u16 dest_mode, u8 vector, unsigned long *ioapic_handled_vectors) { @@ -466,4 +459,7 @@ void kvm_arch_irq_routing_update(struct kvm *kvm) #ifdef CONFIG_KVM_HYPERV kvm_hv_irq_routing_update(kvm); #endif + + if (irqchip_split(kvm)) + kvm_make_scan_ioapic_request(kvm); } diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h index 3bde4fb5c6aa..9461517b4e62 100644 --- a/include/linux/kvm_host.h +++ b/include/linux/kvm_host.h @@ -1024,14 +1024,10 @@ void vcpu_put(struct kvm_vcpu *vcpu); #ifdef __KVM_HAVE_IOAPIC void kvm_arch_post_irq_ack_notifier_list_update(struct kvm *kvm); -void kvm_arch_post_irq_routing_update(struct kvm *kvm); #else static inline void kvm_arch_post_irq_ack_notifier_list_update(struct kvm *kvm) { } -static inline void kvm_arch_post_irq_routing_update(struct kvm *kvm) -{ -} #endif #ifdef CONFIG_HAVE_KVM_IRQCHIP diff --git a/virt/kvm/irqchip.c b/virt/kvm/irqchip.c index 162d8ed889f2..6ccabfd32287 100644 --- a/virt/kvm/irqchip.c +++ b/virt/kvm/irqchip.c @@ -222,8 +222,6 @@ int kvm_set_irq_routing(struct kvm *kvm, kvm_arch_irq_routing_update(kvm); mutex_unlock(&kvm->irq_lock); - kvm_arch_post_irq_routing_update(kvm); - synchronize_srcu_expedited(&kvm->irq_srcu); new = old; |