4 files changed, 244 insertions, 25 deletions

diff --git a/kernel/irq/devres.c b/kernel/irq/devres.c
index 38a25b8d8bff..d06df9c41cba 100644
--- a/kernel/irq/devres.c
+++ b/kernel/irq/devres.c
@@ -26,10 +26,12 @@ static int devm_irq_match(struct device *dev, void *res, void *data)
 }
 
 /**
- *      devm_request_irq - allocate an interrupt line for a managed device
+ *      devm_request_threaded_irq - allocate an interrupt line for a managed device
  *      @dev: device to request interrupt for
  *      @irq: Interrupt line to allocate
  *      @handler: Function to be called when the IRQ occurs
+ *      @thread_fn: function to be called in a threaded interrupt context. NULL
+ *                  for devices which handle everything in @handler
  *      @irqflags: Interrupt type flags
  *      @devname: An ascii name for the claiming device
  *      @dev_id: A cookie passed back to the handler function
@@ -42,9 +44,10 @@ static int devm_irq_match(struct device *dev, void *res, void *data)
  *      If an IRQ allocated with this function needs to be freed
  *      separately, dev_free_irq() must be used.
  */
-int devm_request_irq(struct device *dev, unsigned int irq,
-                     irq_handler_t handler, unsigned long irqflags,
-                     const char *devname, void *dev_id)
+int devm_request_threaded_irq(struct device *dev, unsigned int irq,
+                              irq_handler_t handler, irq_handler_t thread_fn,
+                              unsigned long irqflags, const char *devname,
+                              void *dev_id)
 {
         struct irq_devres *dr;
         int rc;
@@ -54,7 +57,8 @@ int devm_request_irq(struct device *dev, unsigned int irq,
         if (!dr)
                 return -ENOMEM;
 
-        rc = request_irq(irq, handler, irqflags, devname, dev_id);
+        rc = request_threaded_irq(irq, handler, thread_fn, irqflags, devname,
+                                  dev_id);
         if (rc) {
                 devres_free(dr);
                 return rc;
@@ -66,7 +70,7 @@ int devm_request_irq(struct device *dev, unsigned int irq,
 
         return 0;
 }
-EXPORT_SYMBOL(devm_request_irq);
+EXPORT_SYMBOL(devm_request_threaded_irq);
 
 /**
  *      devm_free_irq - free an interrupt
diff --git a/kernel/irq/handle.c b/kernel/irq/handle.c
index 9ebf77968871..26e08754744f 100644
--- a/kernel/irq/handle.c
+++ b/kernel/irq/handle.c
@@ -17,6 +17,7 @@
 #include <linux/kernel_stat.h>
 #include <linux/rculist.h>
 #include <linux/hash.h>
+#include <trace/irq.h>
 #include <linux/bootmem.h>
 
 #include "internals.h"
@@ -338,6 +339,18 @@ irqreturn_t no_action(int cpl, void *dev_id)
         return IRQ_NONE;
 }
 
+static void warn_no_thread(unsigned int irq, struct irqaction *action)
+{
+        if (test_and_set_bit(IRQTF_WARNED, &action->thread_flags))
+                return;
+
+        printk(KERN_WARNING "IRQ %d device %s returned IRQ_WAKE_THREAD "
+               "but no thread function available.", irq, action->name);
+}
+
+DEFINE_TRACE(irq_handler_entry);
+DEFINE_TRACE(irq_handler_exit);
+
 /**
  * handle_IRQ_event - irq action chain handler
  * @irq:        the interrupt number
@@ -350,15 +363,54 @@ irqreturn_t handle_IRQ_event(unsigned int irq, struct irqaction *action)
         irqreturn_t ret, retval = IRQ_NONE;
         unsigned int status = 0;
 
-        WARN_ONCE(!in_irq(), "BUG: IRQ handler called from non-hardirq context!");
-
         if (!(action->flags & IRQF_DISABLED))
                 local_irq_enable_in_hardirq();
 
         do {
+                trace_irq_handler_entry(irq, action);
                 ret = action->handler(irq, action->dev_id);
-                if (ret == IRQ_HANDLED)
+                trace_irq_handler_exit(irq, action, ret);
+
+                switch (ret) {
+                case IRQ_WAKE_THREAD:
+                        /*
+                         * Set result to handled so the spurious check
+                         * does not trigger.
+                         */
+                        ret = IRQ_HANDLED;
+
+                        /*
+                         * Catch drivers which return WAKE_THREAD but
+                         * did not set up a thread function
+                         */
+                        if (unlikely(!action->thread_fn)) {
+                                warn_no_thread(irq, action);
+                                break;
+                        }
+
+                        /*
+                         * Wake up the handler thread for this
+                         * action. In case the thread crashed and was
+                         * killed we just pretend that we handled the
+                         * interrupt. The hardirq handler above has
+                         * disabled the device interrupt, so no irq
+                         * storm is lurking.
+                         */
+                        if (likely(!test_bit(IRQTF_DIED,
+                                             &action->thread_flags))) {
+                                set_bit(IRQTF_RUNTHREAD, &action->thread_flags);
+                                wake_up_process(action->thread);
+                        }
+
+                        /* Fall through to add to randomness */
+                case IRQ_HANDLED:
                         status |= action->flags;
+                        break;
+
+                default:
+                        break;
+                }
+
                 retval |= ret;
                 action = action->next;
         } while (action);
diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c
index 1516ab77355c..2734eca59243 100644
--- a/kernel/irq/manage.c
+++ b/kernel/irq/manage.c
@@ -8,16 +8,15 @@
  */
 
 #include <linux/irq.h>
+#include <linux/kthread.h>
 #include <linux/module.h>
 #include <linux/random.h>
 #include <linux/interrupt.h>
 #include <linux/slab.h>
+#include <linux/sched.h>
 
 #include "internals.h"
 
-#if defined(CONFIG_SMP) && defined(CONFIG_GENERIC_HARDIRQS)
-cpumask_var_t irq_default_affinity;
-
 /**
  *      synchronize_irq - wait for pending IRQ handlers (on other CPUs)
  *      @irq: interrupt number to wait for
@@ -53,9 +52,18 @@ void synchronize_irq(unsigned int irq)
 
                 /* Oops, that failed? */
         } while (status & IRQ_INPROGRESS);
+
+        /*
+         * We made sure that no hardirq handler is running. Now verify
+         * that no threaded handlers are active.
+         */
+        wait_event(desc->wait_for_threads, !atomic_read(&desc->threads_active));
 }
 EXPORT_SYMBOL(synchronize_irq);
 
+#ifdef CONFIG_SMP
+cpumask_var_t irq_default_affinity;
+
 /**
  *      irq_can_set_affinity - Check if the affinity of a given irq can be set
  *      @irq:           Interrupt to check
@@ -72,6 +80,18 @@ int irq_can_set_affinity(unsigned int irq)
         return 1;
 }
 
+static void
+irq_set_thread_affinity(struct irq_desc *desc, const struct cpumask *cpumask)
+{
+        struct irqaction *action = desc->action;
+
+        while (action) {
+                if (action->thread)
+                        set_cpus_allowed_ptr(action->thread, cpumask);
+                action = action->next;
+        }
+}
+
 /**
  *      irq_set_affinity - Set the irq affinity of a given irq
  *      @irq:           Interrupt to set affinity
@@ -89,10 +109,9 @@ int irq_set_affinity(unsigned int irq, const struct cpumask *cpumask)
         spin_lock_irqsave(&desc->lock, flags);
 
 #ifdef CONFIG_GENERIC_PENDING_IRQ
-        if (desc->status & IRQ_MOVE_PCNTXT || desc->status & IRQ_DISABLED) {
-                cpumask_copy(desc->affinity, cpumask);
+        if (desc->status & IRQ_MOVE_PCNTXT)
                 desc->chip->set_affinity(irq, cpumask);
-        } else {
+        else {
                 desc->status |= IRQ_MOVE_PENDING;
                 cpumask_copy(desc->pending_mask, cpumask);
         }
@@ -100,6 +119,7 @@ int irq_set_affinity(unsigned int irq, const struct cpumask *cpumask)
         cpumask_copy(desc->affinity, cpumask);
         desc->chip->set_affinity(irq, cpumask);
 #endif
+        irq_set_thread_affinity(desc, cpumask);
         desc->status |= IRQ_AFFINITY_SET;
         spin_unlock_irqrestore(&desc->lock, flags);
         return 0;
@@ -150,6 +170,8 @@ int irq_select_affinity_usr(unsigned int irq)
 
         spin_lock_irqsave(&desc->lock, flags);
         ret = setup_affinity(irq, desc);
+        if (!ret)
+                irq_set_thread_affinity(desc, desc->affinity);
         spin_unlock_irqrestore(&desc->lock, flags);
 
         return ret;
@@ -401,6 +423,90 @@ int __irq_set_trigger(struct irq_desc *desc, unsigned int irq,
         return ret;
 }
 
+static int irq_wait_for_interrupt(struct irqaction *action)
+{
+        while (!kthread_should_stop()) {
+                set_current_state(TASK_INTERRUPTIBLE);
+
+                if (test_and_clear_bit(IRQTF_RUNTHREAD,
+                                       &action->thread_flags)) {
+                        __set_current_state(TASK_RUNNING);
+                        return 0;
+                }
+                schedule();
+        }
+        return -1;
+}
+
+/*
+ * Interrupt handler thread
+ */
+static int irq_thread(void *data)
+{
+        struct sched_param param = { .sched_priority = MAX_USER_RT_PRIO/2, };
+        struct irqaction *action = data;
+        struct irq_desc *desc = irq_to_desc(action->irq);
+        int wake;
+
+        sched_setscheduler(current, SCHED_FIFO, &param);
+        current->irqaction = action;
+
+        while (!irq_wait_for_interrupt(action)) {
+
+                atomic_inc(&desc->threads_active);
+
+                spin_lock_irq(&desc->lock);
+                if (unlikely(desc->status & IRQ_DISABLED)) {
+                        /*
+                         * CHECKME: We might need a dedicated
+                         * IRQ_THREAD_PENDING flag here, which
+                         * retriggers the thread in check_irq_resend()
+                         * but AFAICT IRQ_PENDING should be fine as it
+                         * retriggers the interrupt itself --- tglx
+                         */
+                        desc->status |= IRQ_PENDING;
+                        spin_unlock_irq(&desc->lock);
+                } else {
+                        spin_unlock_irq(&desc->lock);
+
+                        action->thread_fn(action->irq, action->dev_id);
+                }
+
+                wake = atomic_dec_and_test(&desc->threads_active);
+
+                if (wake && waitqueue_active(&desc->wait_for_threads))
+                        wake_up(&desc->wait_for_threads);
+        }
+
+        /*
+         * Clear irqaction. Otherwise exit_irq_thread() would make
+         * fuzz about an active irq thread going into nirvana.
+         */
+        current->irqaction = NULL;
+        return 0;
+}
+
+/*
+ * Called from do_exit()
+ */
+void exit_irq_thread(void)
+{
+        struct task_struct *tsk = current;
+
+        if (!tsk->irqaction)
+                return;
+
+        printk(KERN_ERR
+               "exiting task \"%s\" (%d) is an active IRQ thread (irq %d)\n",
+               tsk->comm ? tsk->comm : "", tsk->pid, tsk->irqaction->irq);
+
+        /*
+         * Set the THREAD DIED flag to prevent further wakeups of the
+         * soon to be gone threaded handler.
+         */
+        set_bit(IRQTF_DIED, &tsk->irqaction->flags);
+}
+
 /*
  * Internal function to register an irqaction - typically used to
  * allocate special interrupts that are part of the architecture.
@@ -437,6 +543,26 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
         }
 
         /*
+         * Threaded handler ?
+         */
+        if (new->thread_fn) {
+                struct task_struct *t;
+
+                t = kthread_create(irq_thread, new, "irq/%d-%s", irq,
+                                   new->name);
+                if (IS_ERR(t))
+                        return PTR_ERR(t);
+                /*
+                 * We keep the reference to the task struct even if
+                 * the thread dies to avoid that the interrupt code
+                 * references an already freed task_struct.
+                 */
+                get_task_struct(t);
+                new->thread = t;
+                wake_up_process(t);
+        }
+
+        /*
          * The following block of code has to be executed atomically
          */
         spin_lock_irqsave(&desc->lock, flags);
@@ -473,15 +599,15 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
         if (!shared) {
                 irq_chip_set_defaults(desc->chip);
 
+                init_waitqueue_head(&desc->wait_for_threads);
+
                 /* Setup the type (level, edge polarity) if configured: */
                 if (new->flags & IRQF_TRIGGER_MASK) {
                         ret = __irq_set_trigger(desc, irq,
                                         new->flags & IRQF_TRIGGER_MASK);
 
-                        if (ret) {
-                                spin_unlock_irqrestore(&desc->lock, flags);
-                                return ret;
-                        }
+                        if (ret)
+                                goto out_thread;
                 } else
                         compat_irq_chip_set_default_handler(desc);
 #if defined(CONFIG_IRQ_PER_CPU)
@@ -549,8 +675,19 @@ mismatch:
                 dump_stack();
         }
 #endif
+        ret = -EBUSY;
+
+out_thread:
         spin_unlock_irqrestore(&desc->lock, flags);
-        return -EBUSY;
+        if (new->thread) {
+                struct task_struct *t = new->thread;
+
+                new->thread = NULL;
+                if (likely(!test_bit(IRQTF_DIED, &new->thread_flags)))
+                        kthread_stop(t);
+                put_task_struct(t);
+        }
+        return ret;
 }
 
 /**
@@ -576,6 +713,7 @@ static struct irqaction *__free_irq(unsigned int irq, void *dev_id)
 {
         struct irq_desc *desc = irq_to_desc(irq);
         struct irqaction *action, **action_ptr;
+        struct task_struct *irqthread;
         unsigned long flags;
 
         WARN(in_interrupt(), "Trying to free IRQ %d from IRQ context!\n", irq);
@@ -622,6 +760,10 @@ static struct irqaction *__free_irq(unsigned int irq, void *dev_id)
                 else
                         desc->chip->disable(irq);
         }
+
+        irqthread = action->thread;
+        action->thread = NULL;
+
         spin_unlock_irqrestore(&desc->lock, flags);
 
         unregister_handler_proc(irq, action);
@@ -629,6 +771,12 @@ static struct irqaction *__free_irq(unsigned int irq, void *dev_id)
         /* Make sure it's not being used on another CPU: */
         synchronize_irq(irq);
 
+        if (irqthread) {
+                if (!test_bit(IRQTF_DIED, &action->thread_flags))
+                        kthread_stop(irqthread);
+                put_task_struct(irqthread);
+        }
+
 #ifdef CONFIG_DEBUG_SHIRQ
         /*
          * It's a shared IRQ -- the driver ought to be prepared for an IRQ
@@ -681,9 +829,12 @@ void free_irq(unsigned int irq, void *dev_id)
 EXPORT_SYMBOL(free_irq);
 
 /**
- *      request_irq - allocate an interrupt line
+ *      request_threaded_irq - allocate an interrupt line
  *      @irq: Interrupt line to allocate
- *      @handler: Function to be called when the IRQ occurs
+ *      @handler: Function to be called when the IRQ occurs.
+ *                Primary handler for threaded interrupts
+ *      @thread_fn: Function called from the irq handler thread
+ *                  If NULL, no irq thread is created
  *      @irqflags: Interrupt type flags
  *      @devname: An ascii name for the claiming device
  *      @dev_id: A cookie passed back to the handler function
@@ -695,6 +846,15 @@ EXPORT_SYMBOL(free_irq);
  *      raises, you must take care both to initialise your hardware
  *      and to set up the interrupt handler in the right order.
  *
+ *      If you want to set up a threaded irq handler for your device
+ *      then you need to supply @handler and @thread_fn. @handler ist
+ *      still called in hard interrupt context and has to check
+ *      whether the interrupt originates from the device. If yes it
+ *      needs to disable the interrupt on the device and return
+ *      IRQ_THREAD_WAKE which will wake up the handler thread and run
+ *      @thread_fn. This split handler design is necessary to support
+ *      shared interrupts.
+ *
  *      Dev_id must be globally unique. Normally the address of the
  *      device data structure is used as the cookie. Since the handler
  *      receives this value it makes sense to use it.
@@ -710,8 +870,9 @@ EXPORT_SYMBOL(free_irq);
  *      IRQF_TRIGGER_*          Specify active edge(s) or level
  *
  */
-int request_irq(unsigned int irq, irq_handler_t handler,
-                unsigned long irqflags, const char *devname, void *dev_id)
+int request_threaded_irq(unsigned int irq, irq_handler_t handler,
+                         irq_handler_t thread_fn, unsigned long irqflags,
+                         const char *devname, void *dev_id)
 {
         struct irqaction *action;
         struct irq_desc *desc;
@@ -759,6 +920,7 @@ int request_irq(unsigned int irq, irq_handler_t handler,
                 return -ENOMEM;
 
         action->handler = handler;
+        action->thread_fn = thread_fn;
         action->flags = irqflags;
         action->name = devname;
         action->dev_id = dev_id;
@@ -788,4 +950,4 @@ int request_irq(unsigned int irq, irq_handler_t handler,
 #endif
         return retval;
 }
-EXPORT_SYMBOL(request_irq);
+EXPORT_SYMBOL(request_threaded_irq);
diff --git a/kernel/irq/numa_migrate.c b/kernel/irq/numa_migrate.c
index 243d6121e50e..44bbdcbaf8d2 100644
--- a/kernel/irq/numa_migrate.c
+++ b/kernel/irq/numa_migrate.c
@@ -54,6 +54,7 @@ static bool init_copy_one_irq_desc(int irq, struct irq_desc *old_desc,
 static void free_one_irq_desc(struct irq_desc *old_desc, struct irq_desc *desc)
 {
         free_kstat_irqs(old_desc, desc);
+        free_desc_masks(old_desc, desc);
         arch_free_chip_data(old_desc, desc);
 }