KVM: Userspace controlled irq routing

Currently KVM has a static routing from GSI numbers to interrupts (namely,
0-15 are mapped 1:1 to both PIC and IOAPIC, and 16:23 are mapped 1:1 to
the IOAPIC).  This is insufficient for several reasons:

- HPET requires non 1:1 mapping for the timer interrupt
- MSIs need a new method to assign interrupt numbers and dispatch them
- ACPI APIC mode needs to be able to reassign the PCI LINK interrupts to the
  ioapics

This patch implements an interrupt routing table (as a linked list, but this
can be easily changed) and a userspace interface to replace the table.  The
routing table is initialized according to the current hardwired mapping.

Signed-off-by: Avi Kivity <avi@redhat.com>

2 files changed, 64 insertions, 0 deletions

diff --git a/include/linux/kvm.h b/include/linux/kvm.h
index 73f348066866..7a5d73a8d4fa 100644
--- a/include/linux/kvm.h
+++ b/include/linux/kvm.h
@@ -399,6 +399,38 @@ struct kvm_trace_rec {
 #if defined(CONFIG_X86)
 #define KVM_CAP_REINJECT_CONTROL 24
 #endif
+#if defined(CONFIG_X86)||defined(CONFIG_IA64)
+#define KVM_CAP_IRQ_ROUTING 25
+#endif
+
+#ifdef KVM_CAP_IRQ_ROUTING
+
+struct kvm_irq_routing_irqchip {
+        __u32 irqchip;
+        __u32 pin;
+};
+
+/* gsi routing entry types */
+#define KVM_IRQ_ROUTING_IRQCHIP 1
+
+struct kvm_irq_routing_entry {
+        __u32 gsi;
+        __u32 type;
+        __u32 flags;
+        __u32 pad;
+        union {
+                struct kvm_irq_routing_irqchip irqchip;
+                __u32 pad[8];
+        } u;
+};
+
+struct kvm_irq_routing {
+        __u32 nr;
+        __u32 flags;
+        struct kvm_irq_routing_entry entries[0];
+};
+
+#endif
 
 /*
  * ioctls for VM fds
@@ -430,6 +462,7 @@ struct kvm_trace_rec {
                         _IOW(KVMIO,  0x68, struct kvm_coalesced_mmio_zone)
 #define KVM_ASSIGN_PCI_DEVICE _IOR(KVMIO, 0x69, \
                                    struct kvm_assigned_pci_dev)
+#define KVM_SET_GSI_ROUTING       _IOW(KVMIO, 0x6a, struct kvm_irq_routing)
 #define KVM_ASSIGN_IRQ _IOR(KVMIO, 0x70, \
                             struct kvm_assigned_irq)
 #define KVM_REINJECT_CONTROL      _IO(KVMIO, 0x71)
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 99963f36a6db..ce285e01bd57 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -107,6 +107,19 @@ struct kvm_memory_slot {
         int user_alloc;
 };
 
+struct kvm_kernel_irq_routing_entry {
+        u32 gsi;
+        void (*set)(struct kvm_kernel_irq_routing_entry *e,
+                    struct kvm *kvm, int level);
+        union {
+                struct {
+                        unsigned irqchip;
+                        unsigned pin;
+                } irqchip;
+        };
+        struct list_head link;
+};
+
 struct kvm {
         struct mutex lock; /* protects the vcpus array and APIC accesses */
         spinlock_t mmu_lock;
@@ -128,6 +141,7 @@ struct kvm {
 #endif
 
 #ifdef CONFIG_HAVE_KVM_IRQCHIP
+        struct list_head irq_routing; /* of kvm_kernel_irq_routing_entry */
         struct hlist_head mask_notifier_list;
 #endif
 
@@ -480,4 +494,21 @@ static inline int mmu_notifier_retry(struct kvm_vcpu *vcpu, unsigned long mmu_se
 }
 #endif
 
+#ifdef CONFIG_HAVE_KVM_IRQCHIP
+
+#define KVM_MAX_IRQ_ROUTES 1024
+
+int kvm_setup_default_irq_routing(struct kvm *kvm);
+int kvm_set_irq_routing(struct kvm *kvm,
+                        const struct kvm_irq_routing_entry *entries,
+                        unsigned nr,
+                        unsigned flags);
+void kvm_free_irq_routing(struct kvm *kvm);
+
+#else
+
+static inline void kvm_free_irq_routing(struct kvm *kvm) {}
+
+#endif
+
 #endif