1 files changed, 412 insertions, 190 deletions

diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c
index 9033c1c7082..acd599a43bf 100644
--- a/kernel/irq/manage.c
+++ b/kernel/irq/manage.c
@@ -17,6 +17,17 @@
 
 #include "internals.h"
 
+#ifdef CONFIG_IRQ_FORCED_THREADING
+__read_mostly bool force_irqthreads;
+
+static int __init setup_forced_irqthreads(char *arg)
+{
+        force_irqthreads = true;
+        return 0;
+}
+early_param("threadirqs", setup_forced_irqthreads);
+#endif
+
 /**
  *      synchronize_irq - wait for pending IRQ handlers (on other CPUs)
  *      @irq: interrupt number to wait for
@@ -30,7 +41,7 @@
 void synchronize_irq(unsigned int irq)
 {
         struct irq_desc *desc = irq_to_desc(irq);
-        unsigned int status;
+        unsigned int state;
 
         if (!desc)
                 return;
@@ -42,16 +53,16 @@ void synchronize_irq(unsigned int irq)
                  * Wait until we're out of the critical section.  This might
                  * give the wrong answer due to the lack of memory barriers.
                  */
-                while (desc->status & IRQ_INPROGRESS)
+                while (desc->istate & IRQS_INPROGRESS)
                         cpu_relax();
 
                 /* Ok, that indicated we're done: double-check carefully. */
                 raw_spin_lock_irqsave(&desc->lock, flags);
-                status = desc->status;
+                state = desc->istate;
                 raw_spin_unlock_irqrestore(&desc->lock, flags);
 
                 /* Oops, that failed? */
-        } while (status & IRQ_INPROGRESS);
+        } while (state & IRQS_INPROGRESS);
 
         /*
          * We made sure that no hardirq handler is running. Now verify
@@ -73,8 +84,8 @@ int irq_can_set_affinity(unsigned int irq)
 {
         struct irq_desc *desc = irq_to_desc(irq);
 
-        if (CHECK_IRQ_PER_CPU(desc->status) || !desc->irq_data.chip ||
-            !desc->irq_data.chip->irq_set_affinity)
+        if (!desc || !irqd_can_balance(&desc->irq_data) ||
+            !desc->irq_data.chip || !desc->irq_data.chip->irq_set_affinity)
                 return 0;
 
         return 1;
@@ -100,67 +111,169 @@ void irq_set_thread_affinity(struct irq_desc *desc)
         }
 }
 
+#ifdef CONFIG_GENERIC_PENDING_IRQ
+static inline bool irq_can_move_pcntxt(struct irq_desc *desc)
+{
+        return irq_settings_can_move_pcntxt(desc);
+}
+static inline bool irq_move_pending(struct irq_desc *desc)
+{
+        return irqd_is_setaffinity_pending(&desc->irq_data);
+}
+static inline void
+irq_copy_pending(struct irq_desc *desc, const struct cpumask *mask)
+{
+        cpumask_copy(desc->pending_mask, mask);
+}
+static inline void
+irq_get_pending(struct cpumask *mask, struct irq_desc *desc)
+{
+        cpumask_copy(mask, desc->pending_mask);
+}
+#else
+static inline bool irq_can_move_pcntxt(struct irq_desc *desc) { return true; }
+static inline bool irq_move_pending(struct irq_desc *desc) { return false; }
+static inline void
+irq_copy_pending(struct irq_desc *desc, const struct cpumask *mask) { }
+static inline void
+irq_get_pending(struct cpumask *mask, struct irq_desc *desc) { }
+#endif
+
 /**
  *      irq_set_affinity - Set the irq affinity of a given irq
  *      @irq:           Interrupt to set affinity
  *      @cpumask:       cpumask
  *
  */
-int irq_set_affinity(unsigned int irq, const struct cpumask *cpumask)
+int irq_set_affinity(unsigned int irq, const struct cpumask *mask)
 {
         struct irq_desc *desc = irq_to_desc(irq);
         struct irq_chip *chip = desc->irq_data.chip;
         unsigned long flags;
+        int ret = 0;
 
         if (!chip->irq_set_affinity)
                 return -EINVAL;
 
         raw_spin_lock_irqsave(&desc->lock, flags);
 
-#ifdef CONFIG_GENERIC_PENDING_IRQ
-        if (desc->status & IRQ_MOVE_PCNTXT) {
-                if (!chip->irq_set_affinity(&desc->irq_data, cpumask, false)) {
-                        cpumask_copy(desc->irq_data.affinity, cpumask);
+        if (irq_can_move_pcntxt(desc)) {
+                ret = chip->irq_set_affinity(&desc->irq_data, mask, false);
+                switch (ret) {
+                case IRQ_SET_MASK_OK:
+                        cpumask_copy(desc->irq_data.affinity, mask);
+                case IRQ_SET_MASK_OK_NOCOPY:
                         irq_set_thread_affinity(desc);
+                        ret = 0;
                 }
+        } else {
+                irqd_set_move_pending(&desc->irq_data);
+                irq_copy_pending(desc, mask);
         }
-        else {
-                desc->status |= IRQ_MOVE_PENDING;
-                cpumask_copy(desc->pending_mask, cpumask);
-        }
-#else
-        if (!chip->irq_set_affinity(&desc->irq_data, cpumask, false)) {
-                cpumask_copy(desc->irq_data.affinity, cpumask);
-                irq_set_thread_affinity(desc);
+
+        if (desc->affinity_notify) {
+                kref_get(&desc->affinity_notify->kref);
+                schedule_work(&desc->affinity_notify->work);
         }
-#endif
-        desc->status |= IRQ_AFFINITY_SET;
+        irq_compat_set_affinity(desc);
+        irqd_set(&desc->irq_data, IRQD_AFFINITY_SET);
         raw_spin_unlock_irqrestore(&desc->lock, flags);
-        return 0;
+        return ret;
 }
 
 int irq_set_affinity_hint(unsigned int irq, const struct cpumask *m)
 {
+        unsigned long flags;
+        struct irq_desc *desc = irq_get_desc_lock(irq, &flags);
+
+        if (!desc)
+                return -EINVAL;
+        desc->affinity_hint = m;
+        irq_put_desc_unlock(desc, flags);
+        return 0;
+}
+EXPORT_SYMBOL_GPL(irq_set_affinity_hint);
+
+static void irq_affinity_notify(struct work_struct *work)
+{
+        struct irq_affinity_notify *notify =
+                container_of(work, struct irq_affinity_notify, work);
+        struct irq_desc *desc = irq_to_desc(notify->irq);
+        cpumask_var_t cpumask;
+        unsigned long flags;
+
+        if (!desc || !alloc_cpumask_var(&cpumask, GFP_KERNEL))
+                goto out;
+
+        raw_spin_lock_irqsave(&desc->lock, flags);
+        if (irq_move_pending(desc))
+                irq_get_pending(cpumask, desc);
+        else
+                cpumask_copy(cpumask, desc->irq_data.affinity);
+        raw_spin_unlock_irqrestore(&desc->lock, flags);
+
+        notify->notify(notify, cpumask);
+
+        free_cpumask_var(cpumask);
+out:
+        kref_put(&notify->kref, notify->release);
+}
+
+/**
+ *      irq_set_affinity_notifier - control notification of IRQ affinity changes
+ *      @irq:           Interrupt for which to enable/disable notification
+ *      @notify:        Context for notification, or %NULL to disable
+ *                      notification.  Function pointers must be initialised;
+ *                      the other fields will be initialised by this function.
+ *
+ *      Must be called in process context.  Notification may only be enabled
+ *      after the IRQ is allocated and must be disabled before the IRQ is
+ *      freed using free_irq().
+ */
+int
+irq_set_affinity_notifier(unsigned int irq, struct irq_affinity_notify *notify)
+{
         struct irq_desc *desc = irq_to_desc(irq);
+        struct irq_affinity_notify *old_notify;
         unsigned long flags;
 
+        /* The release function is promised process context */
+        might_sleep();
+
         if (!desc)
                 return -EINVAL;
 
+        /* Complete initialisation of *notify */
+        if (notify) {
+                notify->irq = irq;
+                kref_init(&notify->kref);
+                INIT_WORK(&notify->work, irq_affinity_notify);
+        }
+
         raw_spin_lock_irqsave(&desc->lock, flags);
-        desc->affinity_hint = m;
+        old_notify = desc->affinity_notify;
+        desc->affinity_notify = notify;
         raw_spin_unlock_irqrestore(&desc->lock, flags);
 
+        if (old_notify)
+                kref_put(&old_notify->kref, old_notify->release);
+
         return 0;
 }
-EXPORT_SYMBOL_GPL(irq_set_affinity_hint);
+EXPORT_SYMBOL_GPL(irq_set_affinity_notifier);
 
 #ifndef CONFIG_AUTO_IRQ_AFFINITY
 /*
  * Generic version of the affinity autoselector.
  */
-static int setup_affinity(unsigned int irq, struct irq_desc *desc)
+static int
+setup_affinity(unsigned int irq, struct irq_desc *desc, struct cpumask *mask)
 {
+        struct irq_chip *chip = irq_desc_get_chip(desc);
+        struct cpumask *set = irq_default_affinity;
+        int ret;
+
+        /* Excludes PER_CPU and NO_BALANCE interrupts */
         if (!irq_can_set_affinity(irq))
                 return 0;
 
@@ -168,22 +281,29 @@ static int setup_affinity(unsigned int irq, struct irq_desc *desc)
          * Preserve an userspace affinity setup, but make sure that
          * one of the targets is online.
          */
-        if (desc->status & (IRQ_AFFINITY_SET | IRQ_NO_BALANCING)) {
-                if (cpumask_any_and(desc->irq_data.affinity, cpu_online_mask)
-                    < nr_cpu_ids)
-                        goto set_affinity;
-                else
-                        desc->status &= ~IRQ_AFFINITY_SET;
+        if (irqd_has_set(&desc->irq_data, IRQD_AFFINITY_SET)) {
+                if (cpumask_intersects(desc->irq_data.affinity,
+                                       cpu_online_mask))
+                        set = desc->irq_data.affinity;
+                else {
+                        irq_compat_clr_affinity(desc);
+                        irqd_clear(&desc->irq_data, IRQD_AFFINITY_SET);
+                }
         }
 
-        cpumask_and(desc->irq_data.affinity, cpu_online_mask, irq_default_affinity);
-set_affinity:
-        desc->irq_data.chip->irq_set_affinity(&desc->irq_data, desc->irq_data.affinity, false);
-
+        cpumask_and(mask, cpu_online_mask, set);
+        ret = chip->irq_set_affinity(&desc->irq_data, mask, false);
+        switch (ret) {
+        case IRQ_SET_MASK_OK:
+                cpumask_copy(desc->irq_data.affinity, mask);
+        case IRQ_SET_MASK_OK_NOCOPY:
+                irq_set_thread_affinity(desc);
+        }
         return 0;
 }
 #else
-static inline int setup_affinity(unsigned int irq, struct irq_desc *d)
+static inline int
+setup_affinity(unsigned int irq, struct irq_desc *d, struct cpumask *mask)
 {
         return irq_select_affinity(irq);
 }
@@ -192,23 +312,21 @@ static inline int setup_affinity(unsigned int irq, struct irq_desc *d)
 /*
  * Called when affinity is set via /proc/irq
  */
-int irq_select_affinity_usr(unsigned int irq)
+int irq_select_affinity_usr(unsigned int irq, struct cpumask *mask)
 {
         struct irq_desc *desc = irq_to_desc(irq);
         unsigned long flags;
         int ret;
 
         raw_spin_lock_irqsave(&desc->lock, flags);
-        ret = setup_affinity(irq, desc);
-        if (!ret)
-                irq_set_thread_affinity(desc);
+        ret = setup_affinity(irq, desc, mask);
         raw_spin_unlock_irqrestore(&desc->lock, flags);
-
         return ret;
 }
 
 #else
-static inline int setup_affinity(unsigned int irq, struct irq_desc *desc)
+static inline int
+setup_affinity(unsigned int irq, struct irq_desc *desc, struct cpumask *mask)
 {
         return 0;
 }
@@ -219,13 +337,23 @@ void __disable_irq(struct irq_desc *desc, unsigned int irq, bool suspend)
         if (suspend) {
                 if (!desc->action || (desc->action->flags & IRQF_NO_SUSPEND))
                         return;
-                desc->status |= IRQ_SUSPENDED;
+                desc->istate |= IRQS_SUSPENDED;
         }
 
-        if (!desc->depth++) {
-                desc->status |= IRQ_DISABLED;
-                desc->irq_data.chip->irq_disable(&desc->irq_data);
-        }
+        if (!desc->depth++)
+                irq_disable(desc);
+}
+
+static int __disable_irq_nosync(unsigned int irq)
+{
+        unsigned long flags;
+        struct irq_desc *desc = irq_get_desc_buslock(irq, &flags);
+
+        if (!desc)
+                return -EINVAL;
+        __disable_irq(desc, irq, false);
+        irq_put_desc_busunlock(desc, flags);
+        return 0;
 }
 
 /**
@@ -241,17 +369,7 @@ void __disable_irq(struct irq_desc *desc, unsigned int irq, bool suspend)
  */
 void disable_irq_nosync(unsigned int irq)
 {
-        struct irq_desc *desc = irq_to_desc(irq);
-        unsigned long flags;
-
-        if (!desc)
-                return;
-
-        chip_bus_lock(desc);
-        raw_spin_lock_irqsave(&desc->lock, flags);
-        __disable_irq(desc, irq, false);
-        raw_spin_unlock_irqrestore(&desc->lock, flags);
-        chip_bus_sync_unlock(desc);
+        __disable_irq_nosync(irq);
 }
 EXPORT_SYMBOL(disable_irq_nosync);
 
@@ -269,21 +387,24 @@ EXPORT_SYMBOL(disable_irq_nosync);
  */
 void disable_irq(unsigned int irq)
 {
-        struct irq_desc *desc = irq_to_desc(irq);
-
-        if (!desc)
-                return;
-
-        disable_irq_nosync(irq);
-        if (desc->action)
+        if (!__disable_irq_nosync(irq))
                 synchronize_irq(irq);
 }
 EXPORT_SYMBOL(disable_irq);
 
 void __enable_irq(struct irq_desc *desc, unsigned int irq, bool resume)
 {
-        if (resume)
-                desc->status &= ~IRQ_SUSPENDED;
+        if (resume) {
+                if (!(desc->istate & IRQS_SUSPENDED)) {
+                        if (!desc->action)
+                                return;
+                        if (!(desc->action->flags & IRQF_FORCE_RESUME))
+                                return;
+                        /* Pretend that it got disabled ! */
+                        desc->depth++;
+                }
+                desc->istate &= ~IRQS_SUSPENDED;
+        }
 
         switch (desc->depth) {
         case 0:
@@ -291,12 +412,11 @@ void __enable_irq(struct irq_desc *desc, unsigned int irq, bool resume)
                 WARN(1, KERN_WARNING "Unbalanced enable for IRQ %d\n", irq);
                 break;
         case 1: {
-                unsigned int status = desc->status & ~IRQ_DISABLED;
-
-                if (desc->status & IRQ_SUSPENDED)
+                if (desc->istate & IRQS_SUSPENDED)
                         goto err_out;
                 /* Prevent probing on this irq: */
-                desc->status = status | IRQ_NOPROBE;
+                irq_settings_set_noprobe(desc);
+                irq_enable(desc);
                 check_irq_resend(desc, irq);
                 /* fall-through */
         }
@@ -318,21 +438,18 @@ void __enable_irq(struct irq_desc *desc, unsigned int irq, bool resume)
  */
 void enable_irq(unsigned int irq)
 {
-        struct irq_desc *desc = irq_to_desc(irq);
         unsigned long flags;
+        struct irq_desc *desc = irq_get_desc_buslock(irq, &flags);
 
         if (!desc)
                 return;
+        if (WARN(!desc->irq_data.chip,
+                 KERN_ERR "enable_irq before setup/request_irq: irq %u\n", irq))
+                goto out;
 
-        if (WARN(!desc->irq_data.chip || !desc->irq_data.chip->irq_enable,
-            KERN_ERR "enable_irq before setup/request_irq: irq %u\n", irq))
-                return;
-
-        chip_bus_lock(desc);
-        raw_spin_lock_irqsave(&desc->lock, flags);
         __enable_irq(desc, irq, false);
-        raw_spin_unlock_irqrestore(&desc->lock, flags);
-        chip_bus_sync_unlock(desc);
+out:
+        irq_put_desc_busunlock(desc, flags);
 }
 EXPORT_SYMBOL(enable_irq);
 
@@ -348,7 +465,7 @@ static int set_irq_wake_real(unsigned int irq, unsigned int on)
 }
 
 /**
- *      set_irq_wake - control irq power management wakeup
+ *      irq_set_irq_wake - control irq power management wakeup
  *      @irq:   interrupt to control
  *      @on:    enable/disable power management wakeup
  *
@@ -359,23 +476,22 @@ static int set_irq_wake_real(unsigned int irq, unsigned int on)
  *      Wakeup mode lets this IRQ wake the system from sleep
  *      states like "suspend to RAM".
  */
-int set_irq_wake(unsigned int irq, unsigned int on)
+int irq_set_irq_wake(unsigned int irq, unsigned int on)
 {
-        struct irq_desc *desc = irq_to_desc(irq);
         unsigned long flags;
+        struct irq_desc *desc = irq_get_desc_buslock(irq, &flags);
         int ret = 0;
 
         /* wakeup-capable irqs can be shared between drivers that
          * don't need to have the same sleep mode behaviors.
          */
-        raw_spin_lock_irqsave(&desc->lock, flags);
         if (on) {
                 if (desc->wake_depth++ == 0) {
                         ret = set_irq_wake_real(irq, on);
                         if (ret)
                                 desc->wake_depth = 0;
                         else
-                                desc->status |= IRQ_WAKEUP;
+                                irqd_set(&desc->irq_data, IRQD_WAKEUP_STATE);
                 }
         } else {
                 if (desc->wake_depth == 0) {
@@ -385,14 +501,13 @@ int set_irq_wake(unsigned int irq, unsigned int on)
                         if (ret)
                                 desc->wake_depth = 1;
                         else
-                                desc->status &= ~IRQ_WAKEUP;
+                                irqd_clear(&desc->irq_data, IRQD_WAKEUP_STATE);
                 }
         }
-
-        raw_spin_unlock_irqrestore(&desc->lock, flags);
+        irq_put_desc_busunlock(desc, flags);
         return ret;
 }
-EXPORT_SYMBOL(set_irq_wake);
+EXPORT_SYMBOL(irq_set_irq_wake);
 
 /*
  * Internal function that tells the architecture code whether a
@@ -401,43 +516,27 @@ EXPORT_SYMBOL(set_irq_wake);
  */
 int can_request_irq(unsigned int irq, unsigned long irqflags)
 {
-        struct irq_desc *desc = irq_to_desc(irq);
-        struct irqaction *action;
         unsigned long flags;
+        struct irq_desc *desc = irq_get_desc_lock(irq, &flags);
+        int canrequest = 0;
 
         if (!desc)
                 return 0;
 
-        if (desc->status & IRQ_NOREQUEST)
-                return 0;
-
-        raw_spin_lock_irqsave(&desc->lock, flags);
-        action = desc->action;
-        if (action)
-                if (irqflags & action->flags & IRQF_SHARED)
-                        action = NULL;
-
-        raw_spin_unlock_irqrestore(&desc->lock, flags);
-
-        return !action;
-}
-
-void compat_irq_chip_set_default_handler(struct irq_desc *desc)
-{
-        /*
-         * If the architecture still has not overriden
-         * the flow handler then zap the default. This
-         * should catch incorrect flow-type setting.
-         */
-        if (desc->handle_irq == &handle_bad_irq)
-                desc->handle_irq = NULL;
+        if (irq_settings_can_request(desc)) {
+                if (desc->action)
+                        if (irqflags & desc->action->flags & IRQF_SHARED)
+                                canrequest =1;
+        }
+        irq_put_desc_unlock(desc, flags);
+        return canrequest;
 }
 
 int __irq_set_trigger(struct irq_desc *desc, unsigned int irq,
                       unsigned long flags)
 {
-        int ret;
         struct irq_chip *chip = desc->irq_data.chip;
+        int ret, unmask = 0;
 
         if (!chip || !chip->irq_set_type) {
                 /*
@@ -449,23 +548,43 @@ int __irq_set_trigger(struct irq_desc *desc, unsigned int irq,
                 return 0;
         }
 
+        flags &= IRQ_TYPE_SENSE_MASK;
+
+        if (chip->flags & IRQCHIP_SET_TYPE_MASKED) {
+                if (!(desc->istate & IRQS_MASKED))
+                        mask_irq(desc);
+                if (!(desc->istate & IRQS_DISABLED))
+                        unmask = 1;
+        }
+
         /* caller masked out all except trigger mode flags */
         ret = chip->irq_set_type(&desc->irq_data, flags);
 
-        if (ret)
-                pr_err("setting trigger mode %lu for irq %u failed (%pF)\n",
-                       flags, irq, chip->irq_set_type);
-        else {
-                if (flags & (IRQ_TYPE_LEVEL_LOW | IRQ_TYPE_LEVEL_HIGH))
-                        flags |= IRQ_LEVEL;
-                /* note that IRQF_TRIGGER_MASK == IRQ_TYPE_SENSE_MASK */
-                desc->status &= ~(IRQ_LEVEL | IRQ_TYPE_SENSE_MASK);
-                desc->status |= flags;
+        switch (ret) {
+        case IRQ_SET_MASK_OK:
+                irqd_clear(&desc->irq_data, IRQD_TRIGGER_MASK);
+                irqd_set(&desc->irq_data, flags);
+
+        case IRQ_SET_MASK_OK_NOCOPY:
+                flags = irqd_get_trigger_type(&desc->irq_data);
+                irq_settings_set_trigger_mask(desc, flags);
+                irqd_clear(&desc->irq_data, IRQD_LEVEL);
+                irq_settings_clr_level(desc);
+                if (flags & IRQ_TYPE_LEVEL_MASK) {
+                        irq_settings_set_level(desc);
+                        irqd_set(&desc->irq_data, IRQD_LEVEL);
+                }
 
                 if (chip != desc->irq_data.chip)
                         irq_chip_set_defaults(desc->irq_data.chip);
+                ret = 0;
+                break;
+        default:
+                pr_err("setting trigger mode %lu for irq %u failed (%pF)\n",
+                       flags, irq, chip->irq_set_type);
         }
-
+        if (unmask)
+                unmask_irq(desc);
         return ret;
 }
 
@@ -509,8 +628,11 @@ static int irq_wait_for_interrupt(struct irqaction *action)
  * handler finished. unmask if the interrupt has not been disabled and
  * is marked MASKED.
  */
-static void irq_finalize_oneshot(unsigned int irq, struct irq_desc *desc)
+static void irq_finalize_oneshot(struct irq_desc *desc,
+                                 struct irqaction *action, bool force)
 {
+        if (!(desc->istate & IRQS_ONESHOT))
+                return;
 again:
         chip_bus_lock(desc);
         raw_spin_lock_irq(&desc->lock);
@@ -522,26 +644,44 @@ again:
          * The thread is faster done than the hard interrupt handler
          * on the other CPU. If we unmask the irq line then the
          * interrupt can come in again and masks the line, leaves due
-         * to IRQ_INPROGRESS and the irq line is masked forever.
+         * to IRQS_INPROGRESS and the irq line is masked forever.
+         *
+         * This also serializes the state of shared oneshot handlers
+         * versus "desc->threads_onehsot |= action->thread_mask;" in
+         * irq_wake_thread(). See the comment there which explains the
+         * serialization.
          */
-        if (unlikely(desc->status & IRQ_INPROGRESS)) {
+        if (unlikely(desc->istate & IRQS_INPROGRESS)) {
                 raw_spin_unlock_irq(&desc->lock);
                 chip_bus_sync_unlock(desc);
                 cpu_relax();
                 goto again;
         }
 
-        if (!(desc->status & IRQ_DISABLED) && (desc->status & IRQ_MASKED)) {
-                desc->status &= ~IRQ_MASKED;
+        /*
+         * Now check again, whether the thread should run. Otherwise
+         * we would clear the threads_oneshot bit of this thread which
+         * was just set.
+         */
+        if (!force && test_bit(IRQTF_RUNTHREAD, &action->thread_flags))
+                goto out_unlock;
+
+        desc->threads_oneshot &= ~action->thread_mask;
+
+        if (!desc->threads_oneshot && !(desc->istate & IRQS_DISABLED) &&
+            (desc->istate & IRQS_MASKED)) {
+                irq_compat_clr_masked(desc);
+                desc->istate &= ~IRQS_MASKED;
                 desc->irq_data.chip->irq_unmask(&desc->irq_data);
         }
+out_unlock:
         raw_spin_unlock_irq(&desc->lock);
         chip_bus_sync_unlock(desc);
 }
 
 #ifdef CONFIG_SMP
 /*
- * Check whether we need to change the affinity of the interrupt thread.
+ * Check whether we need to chasnge the affinity of the interrupt thread.
  */
 static void
 irq_thread_check_affinity(struct irq_desc *desc, struct irqaction *action)
@@ -573,6 +713,32 @@ irq_thread_check_affinity(struct irq_desc *desc, struct irqaction *action) { }
 #endif
 
 /*
+ * Interrupts which are not explicitely requested as threaded
+ * interrupts rely on the implicit bh/preempt disable of the hard irq
+ * context. So we need to disable bh here to avoid deadlocks and other
+ * side effects.
+ */
+static void
+irq_forced_thread_fn(struct irq_desc *desc, struct irqaction *action)
+{
+        local_bh_disable();
+        action->thread_fn(action->irq, action->dev_id);
+        irq_finalize_oneshot(desc, action, false);
+        local_bh_enable();
+}
+
+/*
+ * Interrupts explicitely requested as threaded interupts want to be
+ * preemtible - many of them need to sleep and wait for slow busses to
+ * complete.
+ */
+static void irq_thread_fn(struct irq_desc *desc, struct irqaction *action)
+{
+        action->thread_fn(action->irq, action->dev_id);
+        irq_finalize_oneshot(desc, action, false);
+}
+
+/*
  * Interrupt handler thread
  */
 static int irq_thread(void *data)
@@ -582,7 +748,14 @@ static int irq_thread(void *data)
         };
         struct irqaction *action = data;
         struct irq_desc *desc = irq_to_desc(action->irq);
-        int wake, oneshot = desc->status & IRQ_ONESHOT;
+        void (*handler_fn)(struct irq_desc *desc, struct irqaction *action);
+        int wake;
+
+        if (force_irqthreads & test_bit(IRQTF_FORCED_THREAD,
+                                        &action->thread_flags))
+                handler_fn = irq_forced_thread_fn;
+        else
+                handler_fn = irq_thread_fn;
 
         sched_setscheduler(current, SCHED_FIFO, &param);
         current->irqaction = action;
@@ -594,23 +767,20 @@ static int irq_thread(void *data)
                 atomic_inc(&desc->threads_active);
 
                 raw_spin_lock_irq(&desc->lock);
-                if (unlikely(desc->status & IRQ_DISABLED)) {
+                if (unlikely(desc->istate & IRQS_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
+                         * but AFAICT IRQS_PENDING should be fine as it
                          * retriggers the interrupt itself --- tglx
                          */
-                        desc->status |= IRQ_PENDING;
+                        irq_compat_set_pending(desc);
+                        desc->istate |= IRQS_PENDING;
                         raw_spin_unlock_irq(&desc->lock);
                 } else {
                         raw_spin_unlock_irq(&desc->lock);
-
-                        action->thread_fn(action->irq, action->dev_id);
-
-                        if (oneshot)
-                                irq_finalize_oneshot(action->irq, desc);
+                        handler_fn(desc, action);
                 }
 
                 wake = atomic_dec_and_test(&desc->threads_active);
@@ -619,6 +789,9 @@ static int irq_thread(void *data)
                         wake_up(&desc->wait_for_threads);
         }
 
+        /* Prevent a stale desc->threads_oneshot */
+        irq_finalize_oneshot(desc, action, true);
+
         /*
          * Clear irqaction. Otherwise exit_irq_thread() would make
          * fuzz about an active irq thread going into nirvana.
@@ -633,6 +806,7 @@ static int irq_thread(void *data)
 void exit_irq_thread(void)
 {
         struct task_struct *tsk = current;
+        struct irq_desc *desc;
 
         if (!tsk->irqaction)
                 return;
@@ -641,6 +815,14 @@ void exit_irq_thread(void)
                "exiting task \"%s\" (%d) is an active IRQ thread (irq %d)\n",
                tsk->comm ? tsk->comm : "", tsk->pid, tsk->irqaction->irq);
 
+        desc = irq_to_desc(tsk->irqaction->irq);
+
+        /*
+         * Prevent a stale desc->threads_oneshot. Must be called
+         * before setting the IRQTF_DIED flag.
+         */
+        irq_finalize_oneshot(desc, tsk->irqaction, true);
+
         /*
          * Set the THREAD DIED flag to prevent further wakeups of the
          * soon to be gone threaded handler.
@@ -648,6 +830,22 @@ void exit_irq_thread(void)
         set_bit(IRQTF_DIED, &tsk->irqaction->flags);
 }
 
+static void irq_setup_forced_threading(struct irqaction *new)
+{
+        if (!force_irqthreads)
+                return;
+        if (new->flags & (IRQF_NO_THREAD | IRQF_PERCPU | IRQF_ONESHOT))
+                return;
+
+        new->flags |= IRQF_ONESHOT;
+
+        if (!new->thread_fn) {
+                set_bit(IRQTF_FORCED_THREAD, &new->thread_flags);
+                new->thread_fn = new->handler;
+                new->handler = irq_default_primary_handler;
+        }
+}
+
 /*
  * Internal function to register an irqaction - typically used to
  * allocate special interrupts that are part of the architecture.
@@ -657,9 +855,9 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
 {
         struct irqaction *old, **old_ptr;
         const char *old_name = NULL;
-        unsigned long flags;
-        int nested, shared = 0;
-        int ret;
+        unsigned long flags, thread_mask = 0;
+        int ret, nested, shared = 0;
+        cpumask_var_t mask;
 
         if (!desc)
                 return -EINVAL;
@@ -683,15 +881,11 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
                 rand_initialize_irq(irq);
         }
 
-        /* Oneshot interrupts are not allowed with shared */
-        if ((new->flags & IRQF_ONESHOT) && (new->flags & IRQF_SHARED))
-                return -EINVAL;
-
         /*
          * Check whether the interrupt nests into another interrupt
          * thread.
          */
-        nested = desc->status & IRQ_NESTED_THREAD;
+        nested = irq_settings_is_nested_thread(desc);
         if (nested) {
                 if (!new->thread_fn)
                         return -EINVAL;
@@ -701,6 +895,8 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
                  * dummy function which warns when called.
                  */
                 new->handler = irq_nested_primary_handler;
+        } else {
+                irq_setup_forced_threading(new);
         }
 
         /*
@@ -724,6 +920,11 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
                 new->thread = t;
         }
 
+        if (!alloc_cpumask_var(&mask, GFP_KERNEL)) {
+                ret = -ENOMEM;
+                goto out_thread;
+        }
+
         /*
          * The following block of code has to be executed atomically
          */
@@ -735,29 +936,40 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
                  * Can't share interrupts unless both agree to and are
                  * the same type (level, edge, polarity). So both flag
                  * fields must have IRQF_SHARED set and the bits which
-                 * set the trigger type must match.
+                 * set the trigger type must match. Also all must
+                 * agree on ONESHOT.
                  */
                 if (!((old->flags & new->flags) & IRQF_SHARED) ||
-                    ((old->flags ^ new->flags) & IRQF_TRIGGER_MASK)) {
+                    ((old->flags ^ new->flags) & IRQF_TRIGGER_MASK) ||
+                    ((old->flags ^ new->flags) & IRQF_ONESHOT)) {
                         old_name = old->name;
                         goto mismatch;
                 }
 
-#if defined(CONFIG_IRQ_PER_CPU)
                 /* All handlers must agree on per-cpuness */
                 if ((old->flags & IRQF_PERCPU) !=
                     (new->flags & IRQF_PERCPU))
                         goto mismatch;
-#endif
 
                 /* add new interrupt at end of irq queue */
                 do {
+                        thread_mask |= old->thread_mask;
                         old_ptr = &old->next;
                         old = *old_ptr;
                 } while (old);
                 shared = 1;
         }
 
+        /*
+         * Setup the thread mask for this irqaction. Unlikely to have
+         * 32 resp 64 irqs sharing one line, but who knows.
+         */
+        if (new->flags & IRQF_ONESHOT && thread_mask == ~0UL) {
+                ret = -EBUSY;
+                goto out_mask;
+        }
+        new->thread_mask = 1 << ffz(thread_mask);
+
         if (!shared) {
                 irq_chip_set_defaults(desc->irq_data.chip);
 
@@ -769,42 +981,44 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
                                         new->flags & IRQF_TRIGGER_MASK);
 
                         if (ret)
-                                goto out_thread;
-                } else
-                        compat_irq_chip_set_default_handler(desc);
-#if defined(CONFIG_IRQ_PER_CPU)
-                if (new->flags & IRQF_PERCPU)
-                        desc->status |= IRQ_PER_CPU;
-#endif
+                                goto out_mask;
+                }
 
-                desc->status &= ~(IRQ_AUTODETECT | IRQ_WAITING | IRQ_ONESHOT |
-                                  IRQ_INPROGRESS | IRQ_SPURIOUS_DISABLED);
+                desc->istate &= ~(IRQS_AUTODETECT | IRQS_SPURIOUS_DISABLED | \
+                                  IRQS_INPROGRESS | IRQS_ONESHOT | \
+                                  IRQS_WAITING);
+
+                if (new->flags & IRQF_PERCPU) {
+                        irqd_set(&desc->irq_data, IRQD_PER_CPU);
+                        irq_settings_set_per_cpu(desc);
+                }
 
                 if (new->flags & IRQF_ONESHOT)
-                        desc->status |= IRQ_ONESHOT;
+                        desc->istate |= IRQS_ONESHOT;
 
-                if (!(desc->status & IRQ_NOAUTOEN)) {
-                        desc->depth = 0;
-                        desc->status &= ~IRQ_DISABLED;
-                        desc->irq_data.chip->irq_startup(&desc->irq_data);
-                } else
+                if (irq_settings_can_autoenable(desc))
+                        irq_startup(desc);
+                else
                         /* Undo nested disables: */
                         desc->depth = 1;
 
                 /* Exclude IRQ from balancing if requested */
-                if (new->flags & IRQF_NOBALANCING)
-                        desc->status |= IRQ_NO_BALANCING;
+                if (new->flags & IRQF_NOBALANCING) {
+                        irq_settings_set_no_balancing(desc);
+                        irqd_set(&desc->irq_data, IRQD_NO_BALANCING);
+                }
 
                 /* Set default affinity mask once everything is setup */
-                setup_affinity(irq, desc);
-
-        } else if ((new->flags & IRQF_TRIGGER_MASK)
-                        && (new->flags & IRQF_TRIGGER_MASK)
-                                != (desc->status & IRQ_TYPE_SENSE_MASK)) {
-                /* hope the handler works with the actual trigger mode... */
-                pr_warning("IRQ %d uses trigger mode %d; requested %d\n",
-                                irq, (int)(desc->status & IRQ_TYPE_SENSE_MASK),
-                                (int)(new->flags & IRQF_TRIGGER_MASK));
+                setup_affinity(irq, desc, mask);
+
+        } else if (new->flags & IRQF_TRIGGER_MASK) {
+                unsigned int nmsk = new->flags & IRQF_TRIGGER_MASK;
+                unsigned int omsk = irq_settings_get_trigger_mask(desc);
+
+                if (nmsk != omsk)
+                        /* hope the handler works with current  trigger mode */
+                        pr_warning("IRQ %d uses trigger mode %u; requested %u\n",
+                                   irq, nmsk, omsk);
         }
 
         new->irq = irq;
@@ -818,8 +1032,8 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
          * Check whether we disabled the irq via the spurious handler
          * before. Reenable it and give it another chance.
          */
-        if (shared && (desc->status & IRQ_SPURIOUS_DISABLED)) {
-                desc->status &= ~IRQ_SPURIOUS_DISABLED;
+        if (shared && (desc->istate & IRQS_SPURIOUS_DISABLED)) {
+                desc->istate &= ~IRQS_SPURIOUS_DISABLED;
                 __enable_irq(desc, irq, false);
         }
 
@@ -849,6 +1063,9 @@ mismatch:
 #endif
         ret = -EBUSY;
 
+out_mask:
+        free_cpumask_var(mask);
+
 out_thread:
         raw_spin_unlock_irqrestore(&desc->lock, flags);
         if (new->thread) {
@@ -871,9 +1088,14 @@ out_thread:
  */
 int setup_irq(unsigned int irq, struct irqaction *act)
 {
+        int retval;
         struct irq_desc *desc = irq_to_desc(irq);
 
-        return __setup_irq(irq, desc, act);
+        chip_bus_lock(desc);
+        retval = __setup_irq(irq, desc, act);
+        chip_bus_sync_unlock(desc);
+
+        return retval;
 }
 EXPORT_SYMBOL_GPL(setup_irq);
 
@@ -924,13 +1146,8 @@ static struct irqaction *__free_irq(unsigned int irq, void *dev_id)
 #endif
 
         /* If this was the last handler, shut down the IRQ line: */
-        if (!desc->action) {
-                desc->status |= IRQ_DISABLED;
-                if (desc->irq_data.chip->irq_shutdown)
-                        desc->irq_data.chip->irq_shutdown(&desc->irq_data);
-                else
-                        desc->irq_data.chip->irq_disable(&desc->irq_data);
-        }
+        if (!desc->action)
+                irq_shutdown(desc);
 
 #ifdef CONFIG_SMP
         /* make sure affinity_hint is cleaned up */
@@ -1004,6 +1221,11 @@ void free_irq(unsigned int irq, void *dev_id)
         if (!desc)
                 return;
 
+#ifdef CONFIG_SMP
+        if (WARN_ON(desc->affinity_notify))
+                desc->affinity_notify = NULL;
+#endif
+
         chip_bus_lock(desc);
         kfree(__free_irq(irq, dev_id));
         chip_bus_sync_unlock(desc);
@@ -1074,7 +1296,7 @@ int request_threaded_irq(unsigned int irq, irq_handler_t handler,
         if (!desc)
                 return -EINVAL;
 
-        if (desc->status & IRQ_NOREQUEST)
+        if (!irq_settings_can_request(desc))
                 return -EINVAL;
 
         if (!handler) {
@@ -1149,7 +1371,7 @@ int request_any_context_irq(unsigned int irq, irq_handler_t handler,
         if (!desc)
                 return -EINVAL;
 
-        if (desc->status & IRQ_NESTED_THREAD) {
+        if (irq_settings_is_nested_thread(desc)) {
                 ret = request_threaded_irq(irq, NULL, handler,
                                            flags, name, dev_id);
                 return !ret ? IRQC_IS_NESTED : ret;