KVM: Move IO APIC to its own lock

The allows removal of irq_lock from the injection path.

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

3 files changed, 79 insertions, 28 deletions

diff --git a/virt/kvm/ioapic.c b/virt/kvm/ioapic.c
index 9fe140bb38ec..38a2d20b89de 100644
--- a/virt/kvm/ioapic.c
+++ b/virt/kvm/ioapic.c
@@ -182,6 +182,7 @@ int kvm_ioapic_set_irq(struct kvm_ioapic *ioapic, int irq, int level)
         union kvm_ioapic_redirect_entry entry;
         int ret = 1;
 
+        mutex_lock(&ioapic->lock);
         if (irq >= 0 && irq < IOAPIC_NUM_PINS) {
                 entry = ioapic->redirtbl[irq];
                 level ^= entry.fields.polarity;
@@ -198,34 +199,51 @@ int kvm_ioapic_set_irq(struct kvm_ioapic *ioapic, int irq, int level)
                 }
                 trace_kvm_ioapic_set_irq(entry.bits, irq, ret == 0);
         }
+        mutex_unlock(&ioapic->lock);
+
         return ret;
 }
 
-static void __kvm_ioapic_update_eoi(struct kvm_ioapic *ioapic, int pin,
-                                    int trigger_mode)
+static void __kvm_ioapic_update_eoi(struct kvm_ioapic *ioapic, int vector,
+                                     int trigger_mode)
 {
-        union kvm_ioapic_redirect_entry *ent;
+        int i;
+
+        for (i = 0; i < IOAPIC_NUM_PINS; i++) {
+                union kvm_ioapic_redirect_entry *ent = &ioapic->redirtbl[i];
 
-        ent = &ioapic->redirtbl[pin];
+                if (ent->fields.vector != vector)
+                        continue;
 
-        kvm_notify_acked_irq(ioapic->kvm, KVM_IRQCHIP_IOAPIC, pin);
+                /*
+                 * We are dropping lock while calling ack notifiers because ack
+                 * notifier callbacks for assigned devices call into IOAPIC
+                 * recursively. Since remote_irr is cleared only after call
+                 * to notifiers if the same vector will be delivered while lock
+                 * is dropped it will be put into irr and will be delivered
+                 * after ack notifier returns.
+                 */
+                mutex_unlock(&ioapic->lock);
+                kvm_notify_acked_irq(ioapic->kvm, KVM_IRQCHIP_IOAPIC, i);
+                mutex_lock(&ioapic->lock);
+
+                if (trigger_mode != IOAPIC_LEVEL_TRIG)
+                        continue;
 
-        if (trigger_mode == IOAPIC_LEVEL_TRIG) {
                 ASSERT(ent->fields.trig_mode == IOAPIC_LEVEL_TRIG);
                 ent->fields.remote_irr = 0;
-                if (!ent->fields.mask && (ioapic->irr & (1 << pin)))
-                        ioapic_service(ioapic, pin);
+                if (!ent->fields.mask && (ioapic->irr & (1 << i)))
+                        ioapic_service(ioapic, i);
         }
 }
 
 void kvm_ioapic_update_eoi(struct kvm *kvm, int vector, int trigger_mode)
 {
         struct kvm_ioapic *ioapic = kvm->arch.vioapic;
-        int i;
 
-        for (i = 0; i < IOAPIC_NUM_PINS; i++)
-                if (ioapic->redirtbl[i].fields.vector == vector)
-                        __kvm_ioapic_update_eoi(ioapic, i, trigger_mode);
+        mutex_lock(&ioapic->lock);
+        __kvm_ioapic_update_eoi(ioapic, vector, trigger_mode);
+        mutex_unlock(&ioapic->lock);
 }
 
 static inline struct kvm_ioapic *to_ioapic(struct kvm_io_device *dev)
@@ -250,8 +268,8 @@ static int ioapic_mmio_read(struct kvm_io_device *this, gpa_t addr, int len,
         ioapic_debug("addr %lx\n", (unsigned long)addr);
         ASSERT(!(addr & 0xf));  /* check alignment */
 
-        mutex_lock(&ioapic->kvm->irq_lock);
         addr &= 0xff;
+        mutex_lock(&ioapic->lock);
         switch (addr) {
         case IOAPIC_REG_SELECT:
                 result = ioapic->ioregsel;
@@ -265,6 +283,8 @@ static int ioapic_mmio_read(struct kvm_io_device *this, gpa_t addr, int len,
                 result = 0;
                 break;
         }
+        mutex_unlock(&ioapic->lock);
+
         switch (len) {
         case 8:
                 *(u64 *) val = result;
@@ -277,7 +297,6 @@ static int ioapic_mmio_read(struct kvm_io_device *this, gpa_t addr, int len,
         default:
                 printk(KERN_WARNING "ioapic: wrong length %d\n", len);
         }
-        mutex_unlock(&ioapic->kvm->irq_lock);
         return 0;
 }
 
@@ -293,15 +312,15 @@ static int ioapic_mmio_write(struct kvm_io_device *this, gpa_t addr, int len,
                      (void*)addr, len, val);
         ASSERT(!(addr & 0xf));  /* check alignment */
 
-        mutex_lock(&ioapic->kvm->irq_lock);
         if (len == 4 || len == 8)
                 data = *(u32 *) val;
         else {
                 printk(KERN_WARNING "ioapic: Unsupported size %d\n", len);
-                goto unlock;
+                return 0;
         }
 
         addr &= 0xff;
+        mutex_lock(&ioapic->lock);
         switch (addr) {
         case IOAPIC_REG_SELECT:
                 ioapic->ioregsel = data;
@@ -312,15 +331,14 @@ static int ioapic_mmio_write(struct kvm_io_device *this, gpa_t addr, int len,
                 break;
 #ifdef  CONFIG_IA64
         case IOAPIC_REG_EOI:
-                kvm_ioapic_update_eoi(ioapic->kvm, data, IOAPIC_LEVEL_TRIG);
+                __kvm_ioapic_update_eoi(ioapic, data, IOAPIC_LEVEL_TRIG);
                 break;
 #endif
 
         default:
                 break;
         }
-unlock:
-        mutex_unlock(&ioapic->kvm->irq_lock);
+        mutex_unlock(&ioapic->lock);
         return 0;
 }
 
@@ -349,6 +367,7 @@ int kvm_ioapic_init(struct kvm *kvm)
         ioapic = kzalloc(sizeof(struct kvm_ioapic), GFP_KERNEL);
         if (!ioapic)
                 return -ENOMEM;
+        mutex_init(&ioapic->lock);
         kvm->arch.vioapic = ioapic;
         kvm_ioapic_reset(ioapic);
         kvm_iodevice_init(&ioapic->dev, &ioapic_mmio_ops);
@@ -360,3 +379,26 @@ int kvm_ioapic_init(struct kvm *kvm)
         return ret;
 }
 
+int kvm_get_ioapic(struct kvm *kvm, struct kvm_ioapic_state *state)
+{
+        struct kvm_ioapic *ioapic = ioapic_irqchip(kvm);
+        if (!ioapic)
+                return -EINVAL;
+
+        mutex_lock(&ioapic->lock);
+        memcpy(state, ioapic, sizeof(struct kvm_ioapic_state));
+        mutex_unlock(&ioapic->lock);
+        return 0;
+}
+
+int kvm_set_ioapic(struct kvm *kvm, struct kvm_ioapic_state *state)
+{
+        struct kvm_ioapic *ioapic = ioapic_irqchip(kvm);
+        if (!ioapic)
+                return -EINVAL;
+
+        mutex_lock(&ioapic->lock);
+        memcpy(ioapic, state, sizeof(struct kvm_ioapic_state));
+        mutex_unlock(&ioapic->lock);
+        return 0;
+}
diff --git a/virt/kvm/ioapic.h b/virt/kvm/ioapic.h
index 6e461ade6365..419c43b667ab 100644
--- a/virt/kvm/ioapic.h
+++ b/virt/kvm/ioapic.h
@@ -45,6 +45,7 @@ struct kvm_ioapic {
         struct kvm_io_device dev;
         struct kvm *kvm;
         void (*ack_notifier)(void *opaque, int irq);
+        struct mutex lock;
 };
 
 #ifdef DEBUG
@@ -74,4 +75,7 @@ int kvm_ioapic_set_irq(struct kvm_ioapic *ioapic, int irq, int level);
 void kvm_ioapic_reset(struct kvm_ioapic *ioapic);
 int kvm_irq_delivery_to_apic(struct kvm *kvm, struct kvm_lapic *src,
                 struct kvm_lapic_irq *irq);
+int kvm_get_ioapic(struct kvm *kvm, struct kvm_ioapic_state *state);
+int kvm_set_ioapic(struct kvm *kvm, struct kvm_ioapic_state *state);
+
 #endif
diff --git a/virt/kvm/irq_comm.c b/virt/kvm/irq_comm.c
index 6c946141dbcc..fadf4408a820 100644
--- a/virt/kvm/irq_comm.c
+++ b/virt/kvm/irq_comm.c
@@ -146,8 +146,8 @@ static int kvm_set_msi(struct kvm_kernel_irq_routing_entry *e,
  */
 int kvm_set_irq(struct kvm *kvm, int irq_source_id, u32 irq, int level)
 {
-        struct kvm_kernel_irq_routing_entry *e;
-        int ret = -1;
+        struct kvm_kernel_irq_routing_entry *e, irq_set[KVM_NR_IRQCHIPS];
+        int ret = -1, i = 0;
         struct kvm_irq_routing_table *irq_rt;
         struct hlist_node *n;
 
@@ -162,14 +162,19 @@ int kvm_set_irq(struct kvm *kvm, int irq_source_id, u32 irq, int level)
         rcu_read_lock();
         irq_rt = rcu_dereference(kvm->irq_routing);
         if (irq < irq_rt->nr_rt_entries)
-                hlist_for_each_entry(e, n, &irq_rt->map[irq], link) {
-                        int r = e->set(e, kvm, irq_source_id, level);
-                        if (r < 0)
-                                continue;
-
-                        ret = r + ((ret < 0) ? 0 : ret);
-                }
+                hlist_for_each_entry(e, n, &irq_rt->map[irq], link)
+                        irq_set[i++] = *e;
         rcu_read_unlock();
+
+        while(i--) {
+                int r;
+                r = irq_set[i].set(&irq_set[i], kvm, irq_source_id, level);
+                if (r < 0)
+                        continue;
+
+                ret = r + ((ret < 0) ? 0 : ret);
+        }
+
         return ret;
 }