Merge branch 'irq/threaded' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip

* 'irq/threaded' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  genirq: fix devres.o build for GENERIC_HARDIRQS=n
  genirq: provide old request_irq() for CONFIG_GENERIC_HARDIRQ=n
  genirq: threaded irq handlers review fixups
  genirq: add support for threaded interrupts to devres
  genirq: add threaded interrupt handler support

9 files changed, 320 insertions, 26 deletions

diff --git a/include/linux/hardirq.h b/include/linux/hardirq.h
index faa1cf848bcd..45257475623c 100644
--- a/include/linux/hardirq.h
+++ b/include/linux/hardirq.h
@@ -116,7 +116,7 @@
 # define IRQ_EXIT_OFFSET HARDIRQ_OFFSET
 #endif
 
-#ifdef CONFIG_SMP
+#if defined(CONFIG_SMP) || defined(CONFIG_GENERIC_HARDIRQS)
 extern void synchronize_irq(unsigned int irq);
 #else
 # define synchronize_irq(irq)	barrier()
diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h
index 8a9613d0c674..91bb76f44f14 100644
--- a/include/linux/interrupt.h
+++ b/include/linux/interrupt.h
@@ -59,6 +59,18 @@
 #define IRQF_NOBALANCING	0x00000800
 #define IRQF_IRQPOLL		0x00001000
 
+/*
+ * Bits used by threaded handlers:
+ * IRQTF_RUNTHREAD - signals that the interrupt handler thread should run
+ * IRQTF_DIED      - handler thread died
+ * IRQTF_WARNED    - warning "IRQ_WAKE_THREAD w/o thread_fn" has been printed
+ */
+enum {
+	IRQTF_RUNTHREAD,
+	IRQTF_DIED,
+	IRQTF_WARNED,
+};
+
 typedef irqreturn_t (*irq_handler_t)(int, void *);
 
 /**
@@ -71,6 +83,9 @@ typedef irqreturn_t (*irq_handler_t)(int, void *);
  * @next:	pointer to the next irqaction for shared interrupts
  * @irq:	interrupt number
  * @dir:	pointer to the proc/irq/NN/name entry
+ * @thread_fn:	interupt handler function for threaded interrupts
+ * @thread:	thread pointer for threaded interrupts
+ * @thread_flags:	flags related to @thread
  */
 struct irqaction {
 	irq_handler_t handler;
@@ -81,18 +96,68 @@ struct irqaction {
 	struct irqaction *next;
 	int irq;
 	struct proc_dir_entry *dir;
+	irq_handler_t thread_fn;
+	struct task_struct *thread;
+	unsigned long thread_flags;
 };
 
 extern irqreturn_t no_action(int cpl, void *dev_id);
-extern int __must_check request_irq(unsigned int, irq_handler_t handler,
-		       unsigned long, const char *, void *);
+
+#ifdef CONFIG_GENERIC_HARDIRQS
+extern int __must_check
+request_threaded_irq(unsigned int irq, irq_handler_t handler,
+		     irq_handler_t thread_fn,
+		     unsigned long flags, const char *name, void *dev);
+
+static inline int __must_check
+request_irq(unsigned int irq, irq_handler_t handler, unsigned long flags,
+	    const char *name, void *dev)
+{
+	return request_threaded_irq(irq, handler, NULL, flags, name, dev);
+}
+
+extern void exit_irq_thread(void);
+#else
+
+extern int __must_check
+request_irq(unsigned int irq, irq_handler_t handler, unsigned long flags,
+	    const char *name, void *dev);
+
+/*
+ * Special function to avoid ifdeffery in kernel/irq/devres.c which
+ * gets magically built by GENERIC_HARDIRQS=n architectures (sparc,
+ * m68k). I really love these $@%#!* obvious Makefile references:
+ * ../../../kernel/irq/devres.o
+ */
+static inline int __must_check
+request_threaded_irq(unsigned int irq, irq_handler_t handler,
+		     irq_handler_t thread_fn,
+		     unsigned long flags, const char *name, void *dev)
+{
+	return request_irq(irq, handler, flags, name, dev);
+}
+
+static inline void exit_irq_thread(void) { }
+#endif
+
 extern void free_irq(unsigned int, void *);
 
 struct device;
 
-extern int __must_check devm_request_irq(struct device *dev, unsigned int irq,
-			    irq_handler_t handler, unsigned long irqflags,
-			    const char *devname, void *dev_id);
+extern int __must_check
+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);
+
+static inline int __must_check
+devm_request_irq(struct device *dev, unsigned int irq, irq_handler_t handler,
+		 unsigned long irqflags, const char *devname, void *dev_id)
+{
+	return devm_request_threaded_irq(dev, irq, handler, NULL, irqflags,
+					 devname, dev_id);
+}
+
 extern void devm_free_irq(struct device *dev, unsigned int irq, void *dev_id);
 
 /*
diff --git a/include/linux/irq.h b/include/linux/irq.h
index 974890b3c52f..ca507c9426b0 100644
--- a/include/linux/irq.h
+++ b/include/linux/irq.h
@@ -22,6 +22,7 @@
 #include <linux/irqnr.h>
 #include <linux/errno.h>
 #include <linux/topology.h>
+#include <linux/wait.h>
 
 #include <asm/irq.h>
 #include <asm/ptrace.h>
@@ -158,6 +159,8 @@ struct irq_2_iommu;
  * @affinity:		IRQ affinity on SMP
  * @cpu:		cpu index useful for balancing
  * @pending_mask:	pending rebalanced interrupts
+ * @threads_active:	number of irqaction threads currently running
+ * @wait_for_threads:	wait queue for sync_irq to wait for threaded handlers
  * @dir:		/proc/irq/ procfs entry
  * @name:		flow handler name for /proc/interrupts output
  */
@@ -189,6 +192,8 @@ struct irq_desc {
 	cpumask_var_t		pending_mask;
 #endif
 #endif
+	atomic_t		threads_active;
+	wait_queue_head_t       wait_for_threads;
 #ifdef CONFIG_PROC_FS
 	struct proc_dir_entry	*dir;
 #endif
diff --git a/include/linux/irqreturn.h b/include/linux/irqreturn.h
index c5584ca5b8c9..819acaaac3f5 100644
--- a/include/linux/irqreturn.h
+++ b/include/linux/irqreturn.h
@@ -5,10 +5,12 @@
  * enum irqreturn
  * @IRQ_NONE		interrupt was not from this device
  * @IRQ_HANDLED		interrupt was handled by this device
+ * @IRQ_WAKE_THREAD	handler requests to wake the handler thread
  */
 enum irqreturn {
 	IRQ_NONE,
 	IRQ_HANDLED,
+	IRQ_WAKE_THREAD,
 };
 
 typedef enum irqreturn irqreturn_t;
diff --git a/include/linux/sched.h b/include/linux/sched.h
index b94f3541f67b..c96140210d1c 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1294,6 +1294,11 @@ struct task_struct {
 /* Protection of (de-)allocation: mm, files, fs, tty, keyrings */
 	spinlock_t alloc_lock;
 
+#ifdef CONFIG_GENERIC_HARDIRQS
+	/* IRQ handler threads */
+	struct irqaction *irqaction;
+#endif
+
 	/* Protection of the PI data structures: */
 	spinlock_t pi_lock;
 
diff --git a/kernel/exit.c b/kernel/exit.c
index 32cbf2607cb0..abf9cf3b95c6 100644
--- a/kernel/exit.c
+++ b/kernel/exit.c
@@ -923,6 +923,8 @@ NORET_TYPE void do_exit(long code)
 		schedule();
 	}
 
+	exit_irq_thread();
+
 	exit_signals(tsk);  /* sets PF_EXITING */
 	/*
 	 * tsk->flags are checked in the futex code to protect against
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 343acecae629..d82142be8dd2 100644
--- a/kernel/irq/handle.c
+++ b/kernel/irq/handle.c
@@ -339,6 +339,15 @@ 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);
 
@@ -363,8 +372,47 @@ irqreturn_t handle_IRQ_event(unsigned int irq, struct irqaction *action)
 		trace_irq_handler_entry(irq, action);
 		ret = action->handler(irq, action->dev_id);
 		trace_irq_handler_exit(irq, action, ret);
-		if (ret == IRQ_HANDLED)
+
+		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..7e2e7dd4cd2f 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
@@ -100,6 +120,7 @@ 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 +171,8 @@ 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 +424,90 @@ 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 +544,26 @@ __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 +600,15 @@ __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 +676,19 @@ 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 +714,7 @@ 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 +761,10 @@ 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 +772,12 @@ 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 +830,12 @@ 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 +847,15 @@ 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 +871,9 @@ 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 +921,7 @@ 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 +951,4 @@ int request_irq(unsigned int irq, irq_handler_t handler,
 #endif
 	return retval;
 }
-EXPORT_SYMBOL(request_irq);
+EXPORT_SYMBOL(request_threaded_irq);