12 files changed, 1128 insertions, 243 deletions

diff --git a/arch/arm/common/Makefile b/arch/arm/common/Makefile
index 8c60f473e97..70b1eff477b 100644
--- a/arch/arm/common/Makefile
+++ b/arch/arm/common/Makefile
@@ -6,7 +6,6 @@ obj-y				+= firmware.o
 
 obj-$(CONFIG_ICST)		+= icst.o
 obj-$(CONFIG_SA1111)		+= sa1111.o
-obj-$(CONFIG_PCI_HOST_VIA82C505) += via82c505.o
 obj-$(CONFIG_DMABOUNCE)		+= dmabounce.o
 obj-$(CONFIG_SHARP_LOCOMO)	+= locomo.o
 obj-$(CONFIG_SHARP_PARAM)	+= sharpsl_param.o
@@ -14,6 +13,9 @@ obj-$(CONFIG_SHARP_SCOOP)	+= scoop.o
 obj-$(CONFIG_PCI_HOST_ITE8152)  += it8152.o
 obj-$(CONFIG_ARM_TIMER_SP804)	+= timer-sp.o
 obj-$(CONFIG_MCPM)		+= mcpm_head.o mcpm_entry.o mcpm_platsmp.o vlock.o
+CFLAGS_REMOVE_mcpm_entry.o	= -pg
 AFLAGS_mcpm_head.o		:= -march=armv7-a
 AFLAGS_vlock.o			:= -march=armv7-a
 obj-$(CONFIG_TI_PRIV_EDMA)	+= edma.o
+obj-$(CONFIG_BL_SWITCHER)	+= bL_switcher.o
+obj-$(CONFIG_BL_SWITCHER_DUMMY_IF) += bL_switcher_dummy_if.o
diff --git a/arch/arm/common/bL_switcher.c b/arch/arm/common/bL_switcher.c
new file mode 100644
index 00000000000..490f3dced74
--- /dev/null
+++ b/arch/arm/common/bL_switcher.c
@@ -0,0 +1,824 @@
+/*
+ * arch/arm/common/bL_switcher.c -- big.LITTLE cluster switcher core driver
+ *
+ * Created by:	Nicolas Pitre, March 2012
+ * Copyright:	(C) 2012-2013  Linaro Limited
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/atomic.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/sched.h>
+#include <linux/interrupt.h>
+#include <linux/cpu_pm.h>
+#include <linux/cpu.h>
+#include <linux/cpumask.h>
+#include <linux/kthread.h>
+#include <linux/wait.h>
+#include <linux/time.h>
+#include <linux/clockchips.h>
+#include <linux/hrtimer.h>
+#include <linux/tick.h>
+#include <linux/notifier.h>
+#include <linux/mm.h>
+#include <linux/mutex.h>
+#include <linux/smp.h>
+#include <linux/spinlock.h>
+#include <linux/string.h>
+#include <linux/sysfs.h>
+#include <linux/irqchip/arm-gic.h>
+#include <linux/moduleparam.h>
+
+#include <asm/smp_plat.h>
+#include <asm/cputype.h>
+#include <asm/suspend.h>
+#include <asm/mcpm.h>
+#include <asm/bL_switcher.h>
+
+#define CREATE_TRACE_POINTS
+#include <trace/events/power_cpu_migrate.h>
+
+
+/*
+ * Use our own MPIDR accessors as the generic ones in asm/cputype.h have
+ * __attribute_const__ and we don't want the compiler to assume any
+ * constness here as the value _does_ change along some code paths.
+ */
+
+static int read_mpidr(void)
+{
+	unsigned int id;
+	asm volatile ("mrc p15, 0, %0, c0, c0, 5" : "=r" (id));
+	return id & MPIDR_HWID_BITMASK;
+}
+
+/*
+ * Get a global nanosecond time stamp for tracing.
+ */
+static s64 get_ns(void)
+{
+	struct timespec ts;
+	getnstimeofday(&ts);
+	return timespec_to_ns(&ts);
+}
+
+/*
+ * bL switcher core code.
+ */
+
+static void bL_do_switch(void *_arg)
+{
+	unsigned ib_mpidr, ib_cpu, ib_cluster;
+	long volatile handshake, **handshake_ptr = _arg;
+
+	pr_debug("%s\n", __func__);
+
+	ib_mpidr = cpu_logical_map(smp_processor_id());
+	ib_cpu = MPIDR_AFFINITY_LEVEL(ib_mpidr, 0);
+	ib_cluster = MPIDR_AFFINITY_LEVEL(ib_mpidr, 1);
+
+	/* Advertise our handshake location */
+	if (handshake_ptr) {
+		handshake = 0;
+		*handshake_ptr = &handshake;
+	} else
+		handshake = -1;
+
+	/*
+	 * Our state has been saved at this point.  Let's release our
+	 * inbound CPU.
+	 */
+	mcpm_set_entry_vector(ib_cpu, ib_cluster, cpu_resume);
+	sev();
+
+	/*
+	 * From this point, we must assume that our counterpart CPU might
+	 * have taken over in its parallel world already, as if execution
+	 * just returned from cpu_suspend().  It is therefore important to
+	 * be very careful not to make any change the other guy is not
+	 * expecting.  This is why we need stack isolation.
+	 *
+	 * Fancy under cover tasks could be performed here.  For now
+	 * we have none.
+	 */
+
+	/*
+	 * Let's wait until our inbound is alive.
+	 */
+	while (!handshake) {
+		wfe();
+		smp_mb();
+	}
+
+	/* Let's put ourself down. */
+	mcpm_cpu_power_down();
+
+	/* should never get here */
+	BUG();
+}
+
+/*
+ * Stack isolation.  To ensure 'current' remains valid, we just use another
+ * piece of our thread's stack space which should be fairly lightly used.
+ * The selected area starts just above the thread_info structure located
+ * at the very bottom of the stack, aligned to a cache line, and indexed
+ * with the cluster number.
+ */
+#define STACK_SIZE 512
+extern void call_with_stack(void (*fn)(void *), void *arg, void *sp);
+static int bL_switchpoint(unsigned long _arg)
+{
+	unsigned int mpidr = read_mpidr();
+	unsigned int clusterid = MPIDR_AFFINITY_LEVEL(mpidr, 1);
+	void *stack = current_thread_info() + 1;
+	stack = PTR_ALIGN(stack, L1_CACHE_BYTES);
+	stack += clusterid * STACK_SIZE + STACK_SIZE;
+	call_with_stack(bL_do_switch, (void *)_arg, stack);
+	BUG();
+}
+
+/*
+ * Generic switcher interface
+ */
+
+static unsigned int bL_gic_id[MAX_CPUS_PER_CLUSTER][MAX_NR_CLUSTERS];
+static int bL_switcher_cpu_pairing[NR_CPUS];
+
+/*
+ * bL_switch_to - Switch to a specific cluster for the current CPU
+ * @new_cluster_id: the ID of the cluster to switch to.
+ *
+ * This function must be called on the CPU to be switched.
+ * Returns 0 on success, else a negative status code.
+ */
+static int bL_switch_to(unsigned int new_cluster_id)
+{
+	unsigned int mpidr, this_cpu, that_cpu;
+	unsigned int ob_mpidr, ob_cpu, ob_cluster, ib_mpidr, ib_cpu, ib_cluster;
+	struct completion inbound_alive;
+	struct tick_device *tdev;
+	enum clock_event_mode tdev_mode;
+	long volatile *handshake_ptr;
+	int ipi_nr, ret;
+
+	this_cpu = smp_processor_id();
+	ob_mpidr = read_mpidr();
+	ob_cpu = MPIDR_AFFINITY_LEVEL(ob_mpidr, 0);
+	ob_cluster = MPIDR_AFFINITY_LEVEL(ob_mpidr, 1);
+	BUG_ON(cpu_logical_map(this_cpu) != ob_mpidr);
+
+	if (new_cluster_id == ob_cluster)
+		return 0;
+
+	that_cpu = bL_switcher_cpu_pairing[this_cpu];
+	ib_mpidr = cpu_logical_map(that_cpu);
+	ib_cpu = MPIDR_AFFINITY_LEVEL(ib_mpidr, 0);
+	ib_cluster = MPIDR_AFFINITY_LEVEL(ib_mpidr, 1);
+
+	pr_debug("before switch: CPU %d MPIDR %#x -> %#x\n",
+		 this_cpu, ob_mpidr, ib_mpidr);
+
+	this_cpu = smp_processor_id();
+
+	/* Close the gate for our entry vectors */
+	mcpm_set_entry_vector(ob_cpu, ob_cluster, NULL);
+	mcpm_set_entry_vector(ib_cpu, ib_cluster, NULL);
+
+	/* Install our "inbound alive" notifier. */
+	init_completion(&inbound_alive);
+	ipi_nr = register_ipi_completion(&inbound_alive, this_cpu);
+	ipi_nr |= ((1 << 16) << bL_gic_id[ob_cpu][ob_cluster]);
+	mcpm_set_early_poke(ib_cpu, ib_cluster, gic_get_sgir_physaddr(), ipi_nr);
+
+	/*
+	 * Let's wake up the inbound CPU now in case it requires some delay
+	 * to come online, but leave it gated in our entry vector code.
+	 */
+	ret = mcpm_cpu_power_up(ib_cpu, ib_cluster);
+	if (ret) {
+		pr_err("%s: mcpm_cpu_power_up() returned %d\n", __func__, ret);
+		return ret;
+	}
+
+	/*
+	 * Raise a SGI on the inbound CPU to make sure it doesn't stall
+	 * in a possible WFI, such as in bL_power_down().
+	 */
+	gic_send_sgi(bL_gic_id[ib_cpu][ib_cluster], 0);
+
+	/*
+	 * Wait for the inbound to come up.  This allows for other
+	 * tasks to be scheduled in the mean time.
+	 */
+	wait_for_completion(&inbound_alive);
+	mcpm_set_early_poke(ib_cpu, ib_cluster, 0, 0);
+
+	/*
+	 * From this point we are entering the switch critical zone
+	 * and can't take any interrupts anymore.
+	 */
+	local_irq_disable();
+	local_fiq_disable();
+	trace_cpu_migrate_begin(get_ns(), ob_mpidr);
+
+	/* redirect GIC's SGIs to our counterpart */
+	gic_migrate_target(bL_gic_id[ib_cpu][ib_cluster]);
+
+	tdev = tick_get_device(this_cpu);
+	if (tdev && !cpumask_equal(tdev->evtdev->cpumask, cpumask_of(this_cpu)))
+		tdev = NULL;
+	if (tdev) {
+		tdev_mode = tdev->evtdev->mode;
+		clockevents_set_mode(tdev->evtdev, CLOCK_EVT_MODE_SHUTDOWN);
+	}
+
+	ret = cpu_pm_enter();
+
+	/* we can not tolerate errors at this point */
+	if (ret)
+		panic("%s: cpu_pm_enter() returned %d\n", __func__, ret);
+
+	/* Swap the physical CPUs in the logical map for this logical CPU. */
+	cpu_logical_map(this_cpu) = ib_mpidr;
+	cpu_logical_map(that_cpu) = ob_mpidr;
+
+	/* Let's do the actual CPU switch. */
+	ret = cpu_suspend((unsigned long)&handshake_ptr, bL_switchpoint);
+	if (ret > 0)
+		panic("%s: cpu_suspend() returned %d\n", __func__, ret);
+
+	/* We are executing on the inbound CPU at this point */
+	mpidr = read_mpidr();
+	pr_debug("after switch: CPU %d MPIDR %#x\n", this_cpu, mpidr);
+	BUG_ON(mpidr != ib_mpidr);
+
+	mcpm_cpu_powered_up();
+
+	ret = cpu_pm_exit();
+
+	if (tdev) {
+		clockevents_set_mode(tdev->evtdev, tdev_mode);
+		clockevents_program_event(tdev->evtdev,
+					  tdev->evtdev->next_event, 1);
+	}
+
+	trace_cpu_migrate_finish(get_ns(), ib_mpidr);
+	local_fiq_enable();
+	local_irq_enable();
+
+	*handshake_ptr = 1;
+	dsb_sev();
+
+	if (ret)
+		pr_err("%s exiting with error %d\n", __func__, ret);
+	return ret;
+}
+
+struct bL_thread {
+	spinlock_t lock;
+	struct task_struct *task;
+	wait_queue_head_t wq;
+	int wanted_cluster;
+	struct completion started;
+	bL_switch_completion_handler completer;
+	void *completer_cookie;
+};
+
+static struct bL_thread bL_threads[NR_CPUS];
+
+static int bL_switcher_thread(void *arg)
+{
+	struct bL_thread *t = arg;
+	struct sched_param param = { .sched_priority = 1 };
+	int cluster;
+	bL_switch_completion_handler completer;
+	void *completer_cookie;
+
+	sched_setscheduler_nocheck(current, SCHED_FIFO, &param);
+	complete(&t->started);
+
+	do {
+		if (signal_pending(current))
+			flush_signals(current);
+		wait_event_interruptible(t->wq,
+				t->wanted_cluster != -1 ||
+				kthread_should_stop());
+
+		spin_lock(&t->lock);
+		cluster = t->wanted_cluster;
+		completer = t->completer;
+		completer_cookie = t->completer_cookie;
+		t->wanted_cluster = -1;
+		t->completer = NULL;
+		spin_unlock(&t->lock);
+
+		if (cluster != -1) {
+			bL_switch_to(cluster);
+
+			if (completer)
+				completer(completer_cookie);
+		}
+	} while (!kthread_should_stop());
+
+	return 0;
+}
+
+static struct task_struct *bL_switcher_thread_create(int cpu, void *arg)
+{
+	struct task_struct *task;
+
+	task = kthread_create_on_node(bL_switcher_thread, arg,
+				      cpu_to_node(cpu), "kswitcher_%d", cpu);
+	if (!IS_ERR(task)) {
+		kthread_bind(task, cpu);
+		wake_up_process(task);
+	} else
+		pr_err("%s failed for CPU %d\n", __func__, cpu);
+	return task;
+}
+
+/*
+ * bL_switch_request_cb - Switch to a specific cluster for the given CPU,
+ *      with completion notification via a callback
+ *
+ * @cpu: the CPU to switch
+ * @new_cluster_id: the ID of the cluster to switch to.
+ * @completer: switch completion callback.  if non-NULL,
+ *	@completer(@completer_cookie) will be called on completion of
+ *	the switch, in non-atomic context.
+ * @completer_cookie: opaque context argument for @completer.
+ *
+ * This function causes a cluster switch on the given CPU by waking up
+ * the appropriate switcher thread.  This function may or may not return
+ * before the switch has occurred.
+ *
+ * If a @completer callback function is supplied, it will be called when
+ * the switch is complete.  This can be used to determine asynchronously
+ * when the switch is complete, regardless of when bL_switch_request()
+ * returns.  When @completer is supplied, no new switch request is permitted
+ * for the affected CPU until after the switch is complete, and @completer
+ * has returned.
+ */
+int bL_switch_request_cb(unsigned int cpu, unsigned int new_cluster_id,
+			 bL_switch_completion_handler completer,
+			 void *completer_cookie)
+{
+	struct bL_thread *t;
+
+	if (cpu >= ARRAY_SIZE(bL_threads)) {
+		pr_err("%s: cpu %d out of bounds\n", __func__, cpu);
+		return -EINVAL;
+	}
+
+	t = &bL_threads[cpu];
+
+	if (IS_ERR(t->task))
+		return PTR_ERR(t->task);
+	if (!t->task)
+		return -ESRCH;
+
+	spin_lock(&t->lock);
+	if (t->completer) {
+		spin_unlock(&t->lock);
+		return -EBUSY;
+	}
+	t->completer = completer;
+	t->completer_cookie = completer_cookie;
+	t->wanted_cluster = new_cluster_id;
+	spin_unlock(&t->lock);
+	wake_up(&t->wq);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(bL_switch_request_cb);
+
+/*
+ * Activation and configuration code.
+ */
+
+static DEFINE_MUTEX(bL_switcher_activation_lock);
+static BLOCKING_NOTIFIER_HEAD(bL_activation_notifier);
+static unsigned int bL_switcher_active;
+static unsigned int bL_switcher_cpu_original_cluster[NR_CPUS];
+static cpumask_t bL_switcher_removed_logical_cpus;
+
+int bL_switcher_register_notifier(struct notifier_block *nb)
+{
+	return blocking_notifier_chain_register(&bL_activation_notifier, nb);
+}
+EXPORT_SYMBOL_GPL(bL_switcher_register_notifier);
+
+int bL_switcher_unregister_notifier(struct notifier_block *nb)
+{
+	return blocking_notifier_chain_unregister(&bL_activation_notifier, nb);
+}
+EXPORT_SYMBOL_GPL(bL_switcher_unregister_notifier);
+
+static int bL_activation_notify(unsigned long val)
+{
+	int ret;
+
+	ret = blocking_notifier_call_chain(&bL_activation_notifier, val, NULL);
+	if (ret & NOTIFY_STOP_MASK)
+		pr_err("%s: notifier chain failed with status 0x%x\n",
+			__func__, ret);
+	return notifier_to_errno(ret);
+}
+
+static void bL_switcher_restore_cpus(void)
+{
+	int i;
+
+	for_each_cpu(i, &bL_switcher_removed_logical_cpus) {
+		struct device *cpu_dev = get_cpu_device(i);
+		int ret = device_online(cpu_dev);
+		if (ret)
+			dev_err(cpu_dev, "switcher: unable to restore CPU\n");
+	}
+}
+
+static int bL_switcher_halve_cpus(void)
+{
+	int i, j, cluster_0, gic_id, ret;
+	unsigned int cpu, cluster, mask;
+	cpumask_t available_cpus;
+
+	/* First pass to validate what we have */
+	mask = 0;
+	for_each_online_cpu(i) {
+		cpu = MPIDR_AFFINITY_LEVEL(cpu_logical_map(i), 0);
+		cluster = MPIDR_AFFINITY_LEVEL(cpu_logical_map(i), 1);
+		if (cluster >= 2) {
+			pr_err("%s: only dual cluster systems are supported\n", __func__);
+			return -EINVAL;
+		}
+		if (WARN_ON(cpu >= MAX_CPUS_PER_CLUSTER))
+			return -EINVAL;
+		mask |= (1 << cluster);
+	}
+	if (mask != 3) {
+		pr_err("%s: no CPU pairing possible\n", __func__);
+		return -EINVAL;
+	}
+
+	/*
+	 * Now let's do the pairing.  We match each CPU with another CPU
+	 * from a different cluster.  To get a uniform scheduling behavior
+	 * without fiddling with CPU topology and compute capacity data,
+	 * we'll use logical CPUs initially belonging to the same cluster.
+	 */
+	memset(bL_switcher_cpu_pairing, -1, sizeof(bL_switcher_cpu_pairing));
+	cpumask_copy(&available_cpus, cpu_online_mask);
+	cluster_0 = -1;
+	for_each_cpu(i, &available_cpus) {
+		int match = -1;
+		cluster = MPIDR_AFFINITY_LEVEL(cpu_logical_map(i), 1);
+		if (cluster_0 == -1)
+			cluster_0 = cluster;
+		if (cluster != cluster_0)
+			continue;
+		cpumask_clear_cpu(i, &available_cpus);
+		for_each_cpu(j, &available_cpus) {
+			cluster = MPIDR_AFFINITY_LEVEL(cpu_logical_map(j), 1);
+			/*
+			 * Let's remember the last match to create "odd"
+			 * pairings on purpose in order for other code not
+			 * to assume any relation between physical and
+			 * logical CPU numbers.
+			 */
+			if (cluster != cluster_0)
+				match = j;
+		}
+		if (match != -1) {
+			bL_switcher_cpu_pairing[i] = match;
+			cpumask_clear_cpu(match, &available_cpus);
+			pr_info("CPU%d paired with CPU%d\n", i, match);
+		}
+	}
+
+	/*
+	 * Now we disable the unwanted CPUs i.e. everything that has no
+	 * pairing information (that includes the pairing counterparts).
+	 */
+	cpumask_clear(&bL_switcher_removed_logical_cpus);
+	for_each_online_cpu(i) {
+		cpu = MPIDR_AFFINITY_LEVEL(cpu_logical_map(i), 0);
+		cluster = MPIDR_AFFINITY_LEVEL(cpu_logical_map(i), 1);
+
+		/* Let's take note of the GIC ID for this CPU */
+		gic_id = gic_get_cpu_id(i);
+		if (gic_id < 0) {
+			pr_err("%s: bad GIC ID for CPU %d\n", __func__, i);
+			bL_switcher_restore_cpus();
+			return -EINVAL;
+		}
+		bL_gic_id[cpu][cluster] = gic_id;
+		pr_info("GIC ID for CPU %u cluster %u is %u\n",
+			cpu, cluster, gic_id);
+
+		if (bL_switcher_cpu_pairing[i] != -1) {
+			bL_switcher_cpu_original_cluster[i] = cluster;
+			continue;
+		}
+
+		ret = device_offline(get_cpu_device(i));
+		if (ret) {
+			bL_switcher_restore_cpus();
+			return ret;
+		}
+		cpumask_set_cpu(i, &bL_switcher_removed_logical_cpus);
+	}
+
+	return 0;
+}
+
+/* Determine the logical CPU a given physical CPU is grouped on. */
+int bL_switcher_get_logical_index(u32 mpidr)
+{
+	int cpu;
+
+	if (!bL_switcher_active)
+		return -EUNATCH;
+
+	mpidr &= MPIDR_HWID_BITMASK;
+	for_each_online_cpu(cpu) {
+		int pairing = bL_switcher_cpu_pairing[cpu];
+		if (pairing == -1)
+			continue;
+		if ((mpidr == cpu_logical_map(cpu)) ||
+		    (mpidr == cpu_logical_map(pairing)))
+			return cpu;
+	}
+	return -EINVAL;
+}
+
+static void bL_switcher_trace_trigger_cpu(void *__always_unused info)
+{
+	trace_cpu_migrate_current(get_ns(), read_mpidr());
+}
+
+int bL_switcher_trace_trigger(void)
+{
+	int ret;
+
+	preempt_disable();
+
+	bL_switcher_trace_trigger_cpu(NULL);
+	ret = smp_call_function(bL_switcher_trace_trigger_cpu, NULL, true);
+
+	preempt_enable();
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(bL_switcher_trace_trigger);
+
+static int bL_switcher_enable(void)
+{
+	int cpu, ret;
+
+	mutex_lock(&bL_switcher_activation_lock);
+	lock_device_hotplug();
+	if (bL_switcher_active) {
+		unlock_device_hotplug();
+		mutex_unlock(&bL_switcher_activation_lock);
+		return 0;
+	}
+
+	pr_info("big.LITTLE switcher initializing\n");
+
+	ret = bL_activation_notify(BL_NOTIFY_PRE_ENABLE);
+	if (ret)
+		goto error;
+
+	ret = bL_switcher_halve_cpus();
+	if (ret)
+		goto error;
+
+	bL_switcher_trace_trigger();
+
+	for_each_online_cpu(cpu) {
+		struct bL_thread *t = &bL_threads[cpu];
+		spin_lock_init(&t->lock);
+		init_waitqueue_head(&t->wq);
+		init_completion(&t->started);
+		t->wanted_cluster = -1;
+		t->task = bL_switcher_thread_create(cpu, t);
+	}
+
+	bL_switcher_active = 1;
+	bL_activation_notify(BL_NOTIFY_POST_ENABLE);
+	pr_info("big.LITTLE switcher initialized\n");
+	goto out;
+
+error:
+	pr_warn("big.LITTLE switcher initialization failed\n");
+	bL_activation_notify(BL_NOTIFY_POST_DISABLE);
+
+out:
+	unlock_device_hotplug();
+	mutex_unlock(&bL_switcher_activation_lock);
+	return ret;
+}
+
+#ifdef CONFIG_SYSFS
+
+static void bL_switcher_disable(void)
+{
+	unsigned int cpu, cluster;
+	struct bL_thread *t;
+	struct task_struct *task;
+
+	mutex_lock(&bL_switcher_activation_lock);
+	lock_device_hotplug();
+
+	if (!bL_switcher_active)
+		goto out;
+
+	if (bL_activation_notify(BL_NOTIFY_PRE_DISABLE) != 0) {
+		bL_activation_notify(BL_NOTIFY_POST_ENABLE);
+		goto out;
+	}
+
+	bL_switcher_active = 0;
+
+	/*
+	 * To deactivate the switcher, we must shut down the switcher
+	 * threads to prevent any other requests from being accepted.
+	 * Then, if the final cluster for given logical CPU is not the
+	 * same as the original one, we'll recreate a switcher thread
+	 * just for the purpose of switching the CPU back without any
+	 * possibility for interference from external requests.
+	 */
+	for_each_online_cpu(cpu) {
+		t = &bL_threads[cpu];
+		task = t->task;
+		t->task = NULL;
+		if (!task || IS_ERR(task))
+			continue;
+		kthread_stop(task);
+		/* no more switch may happen on this CPU at this point */
+		cluster = MPIDR_AFFINITY_LEVEL(cpu_logical_map(cpu), 1);
+		if (cluster == bL_switcher_cpu_original_cluster[cpu])
+			continue;
+		init_completion(&t->started);
+		t->wanted_cluster = bL_switcher_cpu_original_cluster[cpu];
+		task = bL_switcher_thread_create(cpu, t);
+		if (!IS_ERR(task)) {
+			wait_for_completion(&t->started);
+			kthread_stop(task);
+			cluster = MPIDR_AFFINITY_LEVEL(cpu_logical_map(cpu), 1);
+			if (cluster == bL_switcher_cpu_original_cluster[cpu])
+				continue;
+		}
+		/* If execution gets here, we're in trouble. */
+		pr_crit("%s: unable to restore original cluster for CPU %d\n",
+			__func__, cpu);
+		pr_crit("%s: CPU %d can't be restored\n",
+			__func__, bL_switcher_cpu_pairing[cpu]);
+		cpumask_clear_cpu(bL_switcher_cpu_pairing[cpu],
+				  &bL_switcher_removed_logical_cpus);
+	}
+
+	bL_switcher_restore_cpus();
+	bL_switcher_trace_trigger();
+
+	bL_activation_notify(BL_NOTIFY_POST_DISABLE);
+
+out:
+	unlock_device_hotplug();
+	mutex_unlock(&bL_switcher_activation_lock);
+}
+
+static ssize_t bL_switcher_active_show(struct kobject *kobj,
+		struct kobj_attribute *attr, char *buf)
+{
+	return sprintf(buf, "%u\n", bL_switcher_active);
+}
+
+static ssize_t bL_switcher_active_store(struct kobject *kobj,
+		struct kobj_attribute *attr, const char *buf, size_t count)
+{
+	int ret;
+
+	switch (buf[0]) {
+	case '0':
+		bL_switcher_disable();
+		ret = 0;
+		break;
+	case '1':
+		ret = bL_switcher_enable();
+		break;
+	default:
+		ret = -EINVAL;
+	}
+
+	return (ret >= 0) ? count : ret;
+}
+
+static ssize_t bL_switcher_trace_trigger_store(struct kobject *kobj,
+		struct kobj_attribute *attr, const char *buf, size_t count)
+{
+	int ret = bL_switcher_trace_trigger();
+
+	return ret ? ret : count;
+}
+
+static struct kobj_attribute bL_switcher_active_attr =
+	__ATTR(active, 0644, bL_switcher_active_show, bL_switcher_active_store);
+
+static struct kobj_attribute bL_switcher_trace_trigger_attr =
+	__ATTR(trace_trigger, 0200, NULL, bL_switcher_trace_trigger_store);
+
+static struct attribute *bL_switcher_attrs[] = {
+	&bL_switcher_active_attr.attr,
+	&bL_switcher_trace_trigger_attr.attr,
+	NULL,
+};
+
+static struct attribute_group bL_switcher_attr_group = {
+	.attrs = bL_switcher_attrs,
+};
+
+static struct kobject *bL_switcher_kobj;
+
+static int __init bL_switcher_sysfs_init(void)
+{
+	int ret;
+
+	bL_switcher_kobj = kobject_create_and_add("bL_switcher", kernel_kobj);
+	if (!bL_switcher_kobj)
+		return -ENOMEM;
+	ret = sysfs_create_group(bL_switcher_kobj, &bL_switcher_attr_group);
+	if (ret)
+		kobject_put(bL_switcher_kobj);
+	return ret;
+}
+
+#endif  /* CONFIG_SYSFS */
+
+bool bL_switcher_get_enabled(void)
+{
+	mutex_lock(&bL_switcher_activation_lock);
+
+	return bL_switcher_active;
+}
+EXPORT_SYMBOL_GPL(bL_switcher_get_enabled);
+
+void bL_switcher_put_enabled(void)
+{
+	mutex_unlock(&bL_switcher_activation_lock);
+}
+EXPORT_SYMBOL_GPL(bL_switcher_put_enabled);
+
+/*
+ * Veto any CPU hotplug operation on those CPUs we've removed
+ * while the switcher is active.
+ * We're just not ready to deal with that given the trickery involved.
+ */
+static int bL_switcher_hotplug_callback(struct notifier_block *nfb,
+					unsigned long action, void *hcpu)
+{
+	if (bL_switcher_active) {
+		int pairing = bL_switcher_cpu_pairing[(unsigned long)hcpu];
+		switch (action & 0xf) {
+		case CPU_UP_PREPARE:
+		case CPU_DOWN_PREPARE:
+			if (pairing == -1)
+				return NOTIFY_BAD;
+		}
+	}
+	return NOTIFY_DONE;
+}
+
+static bool no_bL_switcher;
+core_param(no_bL_switcher, no_bL_switcher, bool, 0644);
+
+static int __init bL_switcher_init(void)
+{
+	int ret;
+
+	if (!mcpm_is_available())
+		return -ENODEV;
+
+	cpu_notifier(bL_switcher_hotplug_callback, 0);
+
+	if (!no_bL_switcher) {
+		ret = bL_switcher_enable();
+		if (ret)
+			return ret;
+	}
+
+#ifdef CONFIG_SYSFS
+	ret = bL_switcher_sysfs_init();
+	if (ret)
+		pr_err("%s: unable to create sysfs entry\n", __func__);
+#endif
+
+	return 0;
+}
+
+late_initcall(bL_switcher_init);
diff --git a/arch/arm/common/bL_switcher_dummy_if.c b/arch/arm/common/bL_switcher_dummy_if.c
new file mode 100644
index 00000000000..3f47f1203c6
--- /dev/null
+++ b/arch/arm/common/bL_switcher_dummy_if.c
@@ -0,0 +1,71 @@
+/*
+ * arch/arm/common/bL_switcher_dummy_if.c -- b.L switcher dummy interface
+ *
+ * Created by:	Nicolas Pitre, November 2012
+ * Copyright:	(C) 2012-2013  Linaro Limited
+ *
+ * Dummy interface to user space for debugging purpose only.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/fs.h>
+#include <linux/miscdevice.h>
+#include <asm/uaccess.h>
+#include <asm/bL_switcher.h>
+
+static ssize_t bL_switcher_write(struct file *file, const char __user *buf,
+			size_t len, loff_t *pos)
+{
+	unsigned char val[3];
+	unsigned int cpu, cluster;
+	int ret;
+
+	pr_debug("%s\n", __func__);
+
+	if (len < 3)
+		return -EINVAL;
+
+	if (copy_from_user(val, buf, 3))
+		return -EFAULT;
+
+	/* format: <cpu#>,<cluster#> */
+	if (val[0] < '0' || val[0] > '9' ||
+	    val[1] != ',' ||
+	    val[2] < '0' || val[2] > '1')
+		return -EINVAL;
+
+	cpu = val[0] - '0';
+	cluster = val[2] - '0';
+	ret = bL_switch_request(cpu, cluster);
+
+	return ret ? : len;
+}
+
+static const struct file_operations bL_switcher_fops = {
+	.write		= bL_switcher_write,
+	.owner	= THIS_MODULE,
+};
+
+static struct miscdevice bL_switcher_device = {
+	MISC_DYNAMIC_MINOR,
+	"b.L_switcher",
+	&bL_switcher_fops
+};
+
+static int __init bL_switcher_dummy_if_init(void)
+{
+	return misc_register(&bL_switcher_device);
+}
+
+static void __exit bL_switcher_dummy_if_exit(void)
+{
+	misc_deregister(&bL_switcher_device);
+}
+
+module_init(bL_switcher_dummy_if_init);
+module_exit(bL_switcher_dummy_if_exit);
diff --git a/arch/arm/common/edma.c b/arch/arm/common/edma.c
index 117f955a2a0..485be42519b 100644
--- a/arch/arm/common/edma.c
+++ b/arch/arm/common/edma.c
@@ -102,7 +102,13 @@
 #define PARM_OFFSET(param_no)	(EDMA_PARM + ((param_no) << 5))
 
 #define EDMA_DCHMAP	0x0100  /* 64 registers */
-#define CHMAP_EXIST	BIT(24)
+
+/* CCCFG register */
+#define GET_NUM_DMACH(x)	(x & 0x7) /* bits 0-2 */
+#define GET_NUM_PAENTRY(x)	((x & 0x7000) >> 12) /* bits 12-14 */
+#define GET_NUM_EVQUE(x)	((x & 0x70000) >> 16) /* bits 16-18 */
+#define GET_NUM_REGN(x)		((x & 0x300000) >> 20) /* bits 20-21 */
+#define CHMAP_EXIST		BIT(24)
 
 #define EDMA_MAX_DMACH           64
 #define EDMA_MAX_PARAMENTRY     512
@@ -233,7 +239,6 @@ struct edma {
 	unsigned	num_region;
 	unsigned	num_slots;
 	unsigned	num_tc;
-	unsigned	num_cc;
 	enum dma_event_q 	default_queue;
 
 	/* list of channels with no even trigger; terminated by "-1" */
@@ -269,6 +274,11 @@ static const struct edmacc_param dummy_paramset = {
 	.ccnt = 1,
 };
 
+static const struct of_device_id edma_of_ids[] = {
+	{ .compatible = "ti,edma3", },
+	{}
+};
+
 /*****************************************************************************/
 
 static void map_dmach_queue(unsigned ctlr, unsigned ch_no,
@@ -285,12 +295,6 @@ static void map_dmach_queue(unsigned ctlr, unsigned ch_no,
 			~(0x7 << bit), queue_no << bit);
 }
 
-static void __init map_queue_tc(unsigned ctlr, int queue_no, int tc_no)
-{
-	int bit = queue_no * 4;
-	edma_modify(ctlr, EDMA_QUETCMAP, ~(0x7 << bit), ((tc_no & 0x7) << bit));
-}
-
 static void __init assign_priority_to_queue(unsigned ctlr, int queue_no,
 		int priority)
 {
@@ -399,7 +403,7 @@ static irqreturn_t dma_irq_handler(int irq, void *data)
 					BIT(slot));
 			if (edma_cc[ctlr]->intr_data[channel].callback)
 				edma_cc[ctlr]->intr_data[channel].callback(
-					channel, DMA_COMPLETE,
+					channel, EDMA_DMA_COMPLETE,
 					edma_cc[ctlr]->intr_data[channel].data);
 		}
 	} while (sh_ipr);
@@ -454,7 +458,7 @@ static irqreturn_t dma_ccerr_handler(int irq, void *data)
 								callback) {
 						edma_cc[ctlr]->intr_data[k].
 						callback(k,
-						DMA_CC_ERROR,
+						EDMA_DMA_CC_ERROR,
 						edma_cc[ctlr]->intr_data
 						[k].data);
 					}
@@ -560,14 +564,38 @@ static int reserve_contiguous_slots(int ctlr, unsigned int id,
 static int prepare_unused_channel_list(struct device *dev, void *data)
 {
 	struct platform_device *pdev = to_platform_device(dev);
-	int i, ctlr;
+	int i, count, ctlr;
+	struct of_phandle_args  dma_spec;
+
+	if (dev->of_node) {
+		count = of_property_count_strings(dev->of_node, "dma-names");
+		if (count < 0)
+			return 0;
+		for (i = 0; i < count; i++) {
+			if (of_parse_phandle_with_args(dev->of_node, "dmas",
+						       "#dma-cells", i,
+						       &dma_spec))
+				continue;
+
+			if (!of_match_node(edma_of_ids, dma_spec.np)) {
+				of_node_put(dma_spec.np);
+				continue;
+			}
 
+			clear_bit(EDMA_CHAN_SLOT(dma_spec.args[0]),
+				  edma_cc[0]->edma_unused);
+			of_node_put(dma_spec.np);
+		}
+		return 0;
+	}
+
+	/* For non-OF case */
 	for (i = 0; i < pdev->num_resources; i++) {
 		if ((pdev->resource[i].flags & IORESOURCE_DMA) &&
 				(int)pdev->resource[i].start >= 0) {
 			ctlr = EDMA_CTLR(pdev->resource[i].start);
 			clear_bit(EDMA_CHAN_SLOT(pdev->resource[i].start),
-					edma_cc[ctlr]->edma_unused);
+				  edma_cc[ctlr]->edma_unused);
 		}
 	}
 
@@ -965,29 +993,23 @@ void edma_set_dest(unsigned slot, dma_addr_t dest_port,
 EXPORT_SYMBOL(edma_set_dest);
 
 /**
- * edma_get_position - returns the current transfer points
+ * edma_get_position - returns the current transfer point
  * @slot: parameter RAM slot being examined
- * @src: pointer to source port position
- * @dst: pointer to destination port position
+ * @dst:  true selects the dest position, false the source
  *
- * Returns current source and destination addresses for a particular
- * parameter RAM slot.  Its channel should not be active when this is called.
+ * Returns the position of the current active slot
  */
-void edma_get_position(unsigned slot, dma_addr_t *src, dma_addr_t *dst)
+dma_addr_t edma_get_position(unsigned slot, bool dst)
 {
-	struct edmacc_param temp;
-	unsigned ctlr;
+	u32 offs, ctlr = EDMA_CTLR(slot);
 
-	ctlr = EDMA_CTLR(slot);
 	slot = EDMA_CHAN_SLOT(slot);
 
-	edma_read_slot(EDMA_CTLR_CHAN(ctlr, slot), &temp);
-	if (src != NULL)
-		*src = temp.src;
-	if (dst != NULL)
-		*dst = temp.dst;
+	offs = PARM_OFFSET(slot);
+	offs += dst ? PARM_DST : PARM_SRC;
+
+	return edma_read(ctlr, offs);
 }
-EXPORT_SYMBOL(edma_get_position);
 
 /**
  * edma_set_src_index - configure DMA source address indexing
@@ -1392,57 +1414,101 @@ void edma_clear_event(unsigned channel)
 }
 EXPORT_SYMBOL(edma_clear_event);
 
-#if IS_ENABLED(CONFIG_OF) && IS_ENABLED(CONFIG_DMADEVICES)
-
-static int edma_of_read_u32_to_s16_array(const struct device_node *np,
-					 const char *propname, s16 *out_values,
-					 size_t sz)
+static int edma_setup_from_hw(struct device *dev, struct edma_soc_info *pdata,
+			      struct edma *edma_cc)
 {
-	int ret;
+	int i;
+	u32 value, cccfg;
+	s8 (*queue_priority_map)[2];
 
-	ret = of_property_read_u16_array(np, propname, out_values, sz);
-	if (ret)
-		return ret;
+	/* Decode the eDMA3 configuration from CCCFG register */
+	cccfg = edma_read(0, EDMA_CCCFG);
 
-	/* Terminate it */
-	*out_values++ = -1;
-	*out_values++ = -1;
+	value = GET_NUM_REGN(cccfg);
+	edma_cc->num_region = BIT(value);
+
+	value = GET_NUM_DMACH(cccfg);
+	edma_cc->num_channels = BIT(value + 1);
+
+	value = GET_NUM_PAENTRY(cccfg);
+	edma_cc->num_slots = BIT(value + 4);
+
+	value = GET_NUM_EVQUE(cccfg);
+	edma_cc->num_tc = value + 1;
+
+	dev_dbg(dev, "eDMA3 HW configuration (cccfg: 0x%08x):\n", cccfg);
+	dev_dbg(dev, "num_region: %u\n", edma_cc->num_region);
+	dev_dbg(dev, "num_channel: %u\n", edma_cc->num_channels);
+	dev_dbg(dev, "num_slot: %u\n", edma_cc->num_slots);
+	dev_dbg(dev, "num_tc: %u\n", edma_cc->num_tc);
+
+	/* Nothing need to be done if queue priority is provided */
+	if (pdata->queue_priority_mapping)
+		return 0;
+
+	/*
+	 * Configure TC/queue priority as follows:
+	 * Q0 - priority 0
+	 * Q1 - priority 1
+	 * Q2 - priority 2
+	 * ...
+	 * The meaning of priority numbers: 0 highest priority, 7 lowest
+	 * priority. So Q0 is the highest priority queue and the last queue has
+	 * the lowest priority.
+	 */
+	queue_priority_map = devm_kzalloc(dev,
+					  (edma_cc->num_tc + 1) * sizeof(s8),
+					  GFP_KERNEL);
+	if (!queue_priority_map)
+		return -ENOMEM;
+
+	for (i = 0; i < edma_cc->num_tc; i++) {
+		queue_priority_map[i][0] = i;
+		queue_priority_map[i][1] = i;
+	}
+	queue_priority_map[i][0] = -1;
+	queue_priority_map[i][1] = -1;
+
+	pdata->queue_priority_mapping = queue_priority_map;
+	pdata->default_queue = 0;
 
 	return 0;
 }
 
-static int edma_xbar_event_map(struct device *dev,
-			       struct device_node *node,
-			       struct edma_soc_info *pdata, int len)
+#if IS_ENABLED(CONFIG_OF) && IS_ENABLED(CONFIG_DMADEVICES)
+
+static int edma_xbar_event_map(struct device *dev, struct device_node *node,
+			       struct edma_soc_info *pdata, size_t sz)
 {
-	int ret, i;
+	const char pname[] = "ti,edma-xbar-event-map";
 	struct resource res;
 	void __iomem *xbar;
-	const s16 (*xbar_chans)[2];
+	s16 (*xbar_chans)[2];
+	size_t nelm = sz / sizeof(s16);
 	u32 shift, offset, mux;
+	int ret, i;
 
-	xbar_chans = devm_kzalloc(dev,
-				  len/sizeof(s16) + 2*sizeof(s16),
-				  GFP_KERNEL);
+	xbar_chans = devm_kzalloc(dev, (nelm + 2) * sizeof(s16), GFP_KERNEL);
 	if (!xbar_chans)
 		return -ENOMEM;
 
 	ret = of_address_to_resource(node, 1, &res);
 	if (ret)
-		return -EIO;
+		return -ENOMEM;
 
 	xbar = devm_ioremap(dev, res.start, resource_size(&res));
 	if (!xbar)
 		return -ENOMEM;
 
-	ret = edma_of_read_u32_to_s16_array(node,
-					    "ti,edma-xbar-event-map",
-					    (s16 *)xbar_chans,
-					    len/sizeof(u32));
+	ret = of_property_read_u16_array(node, pname, (u16 *)xbar_chans, nelm);
 	if (ret)
 		return -EIO;
 
-	for (i = 0; xbar_chans[i][0] != -1; i++) {
+	/* Invalidate last entry for the other user of this mess */
+	nelm >>= 1;
+	xbar_chans[nelm][0] = xbar_chans[nelm][1] = -1;
+
+	for (i = 0; i < nelm; i++) {
 		shift = (xbar_chans[i][1] & 0x03) << 3;
 		offset = xbar_chans[i][1] & 0xfffffffc;
 		mux = readl(xbar + offset);
@@ -1451,8 +1517,7 @@ static int edma_xbar_event_map(struct device *dev,
 		writel(mux, (xbar + offset));
 	}
 
-	pdata->xbar_chans = xbar_chans;
-
+	pdata->xbar_chans = (const s16 (*)[2]) xbar_chans;
 	return 0;
 }
 
@@ -1460,65 +1525,16 @@ static int edma_of_parse_dt(struct device *dev,
 			    struct device_node *node,
 			    struct edma_soc_info *pdata)
 {
-	int ret = 0, i;
-	u32 value;
+	int ret = 0;
 	struct property *prop;
 	size_t sz;
 	struct edma_rsv_info *rsv_info;
-	s8 (*queue_tc_map)[2], (*queue_priority_map)[2];
-
-	memset(pdata, 0, sizeof(struct edma_soc_info));
-
-	ret = of_property_read_u32(node, "dma-channels", &value);
-	if (ret < 0)
-		return ret;
-	pdata->n_channel = value;
-
-	ret = of_property_read_u32(node, "ti,edma-regions", &value);
-	if (ret < 0)
-		return ret;
-	pdata->n_region = value;
-
-	ret = of_property_read_u32(node, "ti,edma-slots", &value);
-	if (ret < 0)
-		return ret;
-	pdata->n_slot = value;
-
-	pdata->n_cc = 1;
 
 	rsv_info = devm_kzalloc(dev, sizeof(struct edma_rsv_info), GFP_KERNEL);
 	if (!rsv_info)
 		return -ENOMEM;
 	pdata->rsv = rsv_info;
 
-	queue_tc_map = devm_kzalloc(dev, 8*sizeof(s8), GFP_KERNEL);
-	if (!queue_tc_map)
-		return -ENOMEM;
-
-	for (i = 0; i < 3; i++) {
-		queue_tc_map[i][0] = i;
-		queue_tc_map[i][1] = i;
-	}
-	queue_tc_map[i][0] = -1;
-	queue_tc_map[i][1] = -1;
-
-	pdata->queue_tc_mapping = queue_tc_map;
-
-	queue_priority_map = devm_kzalloc(dev, 8*sizeof(s8), GFP_KERNEL);
-	if (!queue_priority_map)
-		return -ENOMEM;
-
-	for (i = 0; i < 3; i++) {
-		queue_priority_map[i][0] = i;
-		queue_priority_map[i][1] = i;
-	}
-	queue_priority_map[i][0] = -1;
-	queue_priority_map[i][1] = -1;
-
-	pdata->queue_priority_mapping = queue_priority_map;
-
-	pdata->default_queue = 0;
-
 	prop = of_find_property(node, "ti,edma-xbar-event-map", &sz);
 	if (prop)
 		ret = edma_xbar_event_map(dev, node, pdata, sz);
@@ -1545,6 +1561,7 @@ static struct edma_soc_info *edma_setup_info_from_dt(struct device *dev,
 		return ERR_PTR(ret);
 
 	dma_cap_set(DMA_SLAVE, edma_filter_info.dma_cap);
+	dma_cap_set(DMA_CYCLIC, edma_filter_info.dma_cap);
 	of_dma_controller_register(dev->of_node, of_dma_simple_xlate,
 				   &edma_filter_info);
 
@@ -1563,7 +1580,6 @@ static int edma_probe(struct platform_device *pdev)
 	struct edma_soc_info	**info = pdev->dev.platform_data;
 	struct edma_soc_info    *ninfo[EDMA_MAX_CC] = {NULL};
 	s8		(*queue_priority_mapping)[2];
-	s8		(*queue_tc_mapping)[2];
 	int			i, j, off, ln, found = 0;
 	int			status = -1;
 	const s16		(*rsv_chans)[2];
@@ -1574,7 +1590,6 @@ static int edma_probe(struct platform_device *pdev)
 	struct resource		*r[EDMA_MAX_CC] = {NULL};
 	struct resource		res[EDMA_MAX_CC];
 	char			res_name[10];
-	char			irq_name[10];
 	struct device_node	*node = pdev->dev.of_node;
 	struct device		*dev = &pdev->dev;
 	int			ret;
@@ -1639,12 +1654,10 @@ static int edma_probe(struct platform_device *pdev)
 		if (!edma_cc[j])
 			return -ENOMEM;
 
-		edma_cc[j]->num_channels = min_t(unsigned, info[j]->n_channel,
-							EDMA_MAX_DMACH);
-		edma_cc[j]->num_slots = min_t(unsigned, info[j]->n_slot,
-							EDMA_MAX_PARAMENTRY);
-		edma_cc[j]->num_cc = min_t(unsigned, info[j]->n_cc,
-							EDMA_MAX_CC);
+		/* Get eDMA3 configuration from IP */
+		ret = edma_setup_from_hw(dev, info[j], edma_cc[j]);
+		if (ret)
+			return ret;
 
 		edma_cc[j]->default_queue = info[j]->default_queue;
 
@@ -1696,14 +1709,21 @@ static int edma_probe(struct platform_device *pdev)
 
 		if (node) {
 			irq[j] = irq_of_parse_and_map(node, 0);
+			err_irq[j] = irq_of_parse_and_map(node, 2);
 		} else {
+			char irq_name[10];
+
 			sprintf(irq_name, "edma%d", j);
 			irq[j] = platform_get_irq_byname(pdev, irq_name);
+
+			sprintf(irq_name, "edma%d_err", j);
+			err_irq[j] = platform_get_irq_byname(pdev, irq_name);
 		}
 		edma_cc[j]->irq_res_start = irq[j];
-		status = devm_request_irq(&pdev->dev, irq[j],
-					  dma_irq_handler, 0, "edma",
-					  &pdev->dev);
+		edma_cc[j]->irq_res_end = err_irq[j];
+
+		status = devm_request_irq(dev, irq[j], dma_irq_handler, 0,
+					  "edma", dev);
 		if (status < 0) {
 			dev_dbg(&pdev->dev,
 				"devm_request_irq %d failed --> %d\n",
@@ -1711,16 +1731,8 @@ static int edma_probe(struct platform_device *pdev)
 			return status;
 		}
 
-		if (node) {
-			err_irq[j] = irq_of_parse_and_map(node, 2);
-		} else {
-			sprintf(irq_name, "edma%d_err", j);
-			err_irq[j] = platform_get_irq_byname(pdev, irq_name);
-		}
-		edma_cc[j]->irq_res_end = err_irq[j];
-		status = devm_request_irq(&pdev->dev, err_irq[j],
-					  dma_ccerr_handler, 0,
-					  "edma_error", &pdev->dev);
+		status = devm_request_irq(dev, err_irq[j], dma_ccerr_handler, 0,
+					  "edma_error", dev);
 		if (status < 0) {
 			dev_dbg(&pdev->dev,
 				"devm_request_irq %d failed --> %d\n",
@@ -1731,14 +1743,8 @@ static int edma_probe(struct platform_device *pdev)
 		for (i = 0; i < edma_cc[j]->num_channels; i++)
 			map_dmach_queue(j, i, info[j]->default_queue);
 
-		queue_tc_mapping = info[j]->queue_tc_mapping;
 		queue_priority_mapping = info[j]->queue_priority_mapping;
 
-		/* Event queue to TC mapping */
-		for (i = 0; queue_tc_mapping[i][0] != -1; i++)
-			map_queue_tc(j, queue_tc_mapping[i][0],
-					queue_tc_mapping[i][1]);
-
 		/* Event queue priority mapping */
 		for (i = 0; queue_priority_mapping[i][0] != -1; i++)
 			assign_priority_to_queue(j,
@@ -1751,7 +1757,7 @@ static int edma_probe(struct platform_device *pdev)
 		if (edma_read(j, EDMA_CCCFG) & CHMAP_EXIST)
 			map_dmach_param(j);
 
-		for (i = 0; i < info[j]->n_region; i++) {
+		for (i = 0; i < edma_cc[j]->num_region; i++) {
 			edma_write_array2(j, EDMA_DRAE, i, 0, 0x0);
 			edma_write_array2(j, EDMA_DRAE, i, 1, 0x0);
 			edma_write_array(j, EDMA_QRAE, i, 0x0);
@@ -1762,11 +1768,6 @@ static int edma_probe(struct platform_device *pdev)
 	return 0;
 }
 
-static const struct of_device_id edma_of_ids[] = {
-	{ .compatible = "ti,edma3", },
-	{}
-};
-
 static struct platform_driver edma_driver = {
 	.driver = {
 		.name	= "edma",
diff --git a/arch/arm/common/it8152.c b/arch/arm/common/it8152.c
index 001f4913799..5114b68e99d 100644
--- a/arch/arm/common/it8152.c
+++ b/arch/arm/common/it8152.c
@@ -257,7 +257,7 @@ static int it8152_needs_bounce(struct device *dev, dma_addr_t dma_addr, size_t s
  */
 static int it8152_pci_platform_notify(struct device *dev)
 {
-	if (dev->bus == &pci_bus_type) {
+	if (dev_is_pci(dev)) {
 		if (dev->dma_mask)
 			*dev->dma_mask = (SZ_64M - 1) | PHYS_OFFSET;
 		dev->coherent_dma_mask = (SZ_64M - 1) | PHYS_OFFSET;
@@ -268,7 +268,7 @@ static int it8152_pci_platform_notify(struct device *dev)
 
 static int it8152_pci_platform_notify_remove(struct device *dev)
 {
-	if (dev->bus == &pci_bus_type)
+	if (dev_is_pci(dev))
 		dmabounce_unregister_dev(dev);
 
 	return 0;
diff --git a/arch/arm/common/mcpm_entry.c b/arch/arm/common/mcpm_entry.c
index 370236dd1a0..f91136ab447 100644
--- a/arch/arm/common/mcpm_entry.c
+++ b/arch/arm/common/mcpm_entry.c
@@ -27,6 +27,17 @@ void mcpm_set_entry_vector(unsigned cpu, unsigned cluster, void *ptr)
 	sync_cache_w(&mcpm_entry_vectors[cluster][cpu]);
 }
 
+extern unsigned long mcpm_entry_early_pokes[MAX_NR_CLUSTERS][MAX_CPUS_PER_CLUSTER][2];
+
+void mcpm_set_early_poke(unsigned cpu, unsigned cluster,
+			 unsigned long poke_phys_addr, unsigned long poke_val)
+{
+	unsigned long *poke = &mcpm_entry_early_pokes[cluster][cpu][0];
+	poke[0] = poke_phys_addr;
+	poke[1] = poke_val;
+	__sync_cache_range_w(poke, 2 * sizeof(*poke));
+}
+
 static const struct mcpm_platform_ops *platform_ops;
 
 int __init mcpm_platform_register(const struct mcpm_platform_ops *ops)
@@ -37,6 +48,11 @@ int __init mcpm_platform_register(const struct mcpm_platform_ops *ops)
 	return 0;
 }
 
+bool mcpm_is_available(void)
+{
+	return (platform_ops) ? true : false;
+}
+
 int mcpm_cpu_power_up(unsigned int cpu, unsigned int cluster)
 {
 	if (!platform_ops)
@@ -51,7 +67,8 @@ void mcpm_cpu_power_down(void)
 {
 	phys_reset_t phys_reset;
 
-	BUG_ON(!platform_ops);
+	if (WARN_ON_ONCE(!platform_ops || !platform_ops->power_down))
+		return;
 	BUG_ON(!irqs_disabled());
 
 	/*
@@ -89,11 +106,27 @@ void mcpm_cpu_power_down(void)
 	BUG();
 }
 
+int mcpm_wait_for_cpu_powerdown(unsigned int cpu, unsigned int cluster)
+{
+	int ret;
+
+	if (WARN_ON_ONCE(!platform_ops || !platform_ops->wait_for_powerdown))
+		return -EUNATCH;
+
+	ret = platform_ops->wait_for_powerdown(cpu, cluster);
+	if (ret)
+		pr_warn("%s: cpu %u, cluster %u failed to power down (%d)\n",
+			__func__, cpu, cluster, ret);
+
+	return ret;
+}
+
 void mcpm_cpu_suspend(u64 expected_residency)
 {
 	phys_reset_t phys_reset;
 
-	BUG_ON(!platform_ops);
+	if (WARN_ON_ONCE(!platform_ops || !platform_ops->suspend))
+		return;
 	BUG_ON(!irqs_disabled());
 
 	/* Very similar to mcpm_cpu_power_down() */
@@ -138,7 +171,7 @@ void __mcpm_cpu_down(unsigned int cpu, unsigned int cluster)
 	dmb();
 	mcpm_sync.clusters[cluster].cpus[cpu].cpu = CPU_DOWN;
 	sync_cache_w(&mcpm_sync.clusters[cluster].cpus[cpu].cpu);
-	dsb_sev();
+	sev();
 }
 
 /*
@@ -154,7 +187,7 @@ void __mcpm_outbound_leave_critical(unsigned int cluster, int state)
 	dmb();
 	mcpm_sync.clusters[cluster].cluster = state;
 	sync_cache_w(&mcpm_sync.clusters[cluster].cluster);
-	dsb_sev();
+	sev();
 }
 
 /*
diff --git a/arch/arm/common/mcpm_head.S b/arch/arm/common/mcpm_head.S
index 39c96df3477..e02db4b81a6 100644
--- a/arch/arm/common/mcpm_head.S
+++ b/arch/arm/common/mcpm_head.S
@@ -15,6 +15,7 @@
 
 #include <linux/linkage.h>
 #include <asm/mcpm.h>
+#include <asm/assembler.h>
 
 #include "vlock.h"
 
@@ -47,6 +48,7 @@
 
 ENTRY(mcpm_entry_point)
 
+ ARM_BE8(setend        be)
  THUMB(	adr	r12, BSYM(1f)	)
  THUMB(	bx	r12		)
  THUMB(	.thumb			)
@@ -71,12 +73,19 @@ ENTRY(mcpm_entry_point)
 	 * position independent way.
 	 */
 	adr	r5, 3f
-	ldmia	r5, {r6, r7, r8, r11}
+	ldmia	r5, {r0, r6, r7, r8, r11}
+	add	r0, r5, r0			@ r0 = mcpm_entry_early_pokes
 	add	r6, r5, r6			@ r6 = mcpm_entry_vectors
 	ldr	r7, [r5, r7]			@ r7 = mcpm_power_up_setup_phys
 	add	r8, r5, r8			@ r8 = mcpm_sync
 	add	r11, r5, r11			@ r11 = first_man_locks
 
+	@ Perform an early poke, if any
+	add	r0, r0, r4, lsl #3
+	ldmia	r0, {r0, r1}
+	teq	r0, #0
+	strne	r1, [r0]
+
 	mov	r0, #MCPM_SYNC_CLUSTER_SIZE
 	mla	r8, r0, r10, r8			@ r8 = sync cluster base
 
@@ -195,7 +204,8 @@ mcpm_entry_gated:
 
 	.align	2
 
-3:	.word	mcpm_entry_vectors - .
+3:	.word	mcpm_entry_early_pokes - .
+	.word	mcpm_entry_vectors - 3b
 	.word	mcpm_power_up_setup_phys - 3b
 	.word	mcpm_sync - 3b
 	.word	first_man_locks - 3b
@@ -214,6 +224,10 @@ first_man_locks:
 ENTRY(mcpm_entry_vectors)
 	.space	4 * MAX_NR_CLUSTERS * MAX_CPUS_PER_CLUSTER
 
+	.type	mcpm_entry_early_pokes, #object
+ENTRY(mcpm_entry_early_pokes)
+	.space	8 * MAX_NR_CLUSTERS * MAX_CPUS_PER_CLUSTER
+
 	.type	mcpm_power_up_setup_phys, #object
 ENTRY(mcpm_power_up_setup_phys)
 	.space  4		@ set by mcpm_sync_init()
diff --git a/arch/arm/common/mcpm_platsmp.c b/arch/arm/common/mcpm_platsmp.c
index 1bc34c7567f..92e54d7c6f4 100644
--- a/arch/arm/common/mcpm_platsmp.c
+++ b/arch/arm/common/mcpm_platsmp.c
@@ -19,14 +19,23 @@
 #include <asm/smp.h>
 #include <asm/smp_plat.h>
 
+static void cpu_to_pcpu(unsigned int cpu,
+			unsigned int *pcpu, unsigned int *pcluster)
+{
+	unsigned int mpidr;
+
+	mpidr = cpu_logical_map(cpu);
+	*pcpu = MPIDR_AFFINITY_LEVEL(mpidr, 0);
+	*pcluster = MPIDR_AFFINITY_LEVEL(mpidr, 1);
+}
+
 static int mcpm_boot_secondary(unsigned int cpu, struct task_struct *idle)
 {
-	unsigned int mpidr, pcpu, pcluster, ret;
+	unsigned int pcpu, pcluster, ret;
 	extern void secondary_startup(void);
 
-	mpidr = cpu_logical_map(cpu);
-	pcpu = MPIDR_AFFINITY_LEVEL(mpidr, 0);
-	pcluster = MPIDR_AFFINITY_LEVEL(mpidr, 1);
+	cpu_to_pcpu(cpu, &pcpu, &pcluster);
+
 	pr_debug("%s: logical CPU %d is physical CPU %d cluster %d\n",
 		 __func__, cpu, pcpu, pcluster);
 
@@ -47,6 +56,15 @@ static void mcpm_secondary_init(unsigned int cpu)
 
 #ifdef CONFIG_HOTPLUG_CPU
 
+static int mcpm_cpu_kill(unsigned int cpu)
+{
+	unsigned int pcpu, pcluster;
+
+	cpu_to_pcpu(cpu, &pcpu, &pcluster);
+
+	return !mcpm_wait_for_cpu_powerdown(pcpu, pcluster);
+}
+
 static int mcpm_cpu_disable(unsigned int cpu)
 {
 	/*
@@ -73,6 +91,7 @@ static struct smp_operations __initdata mcpm_smp_ops = {
 	.smp_boot_secondary	= mcpm_boot_secondary,
 	.smp_secondary_init	= mcpm_secondary_init,
 #ifdef CONFIG_HOTPLUG_CPU
+	.cpu_kill		= mcpm_cpu_kill,
 	.cpu_disable		= mcpm_cpu_disable,
 	.cpu_die		= mcpm_cpu_die,
 #endif
diff --git a/arch/arm/common/scoop.c b/arch/arm/common/scoop.c
index a5c3dc38aa1..a20fa80776d 100644
--- a/arch/arm/common/scoop.c
+++ b/arch/arm/common/scoop.c
@@ -182,7 +182,6 @@ static int scoop_probe(struct platform_device *pdev)
 	struct scoop_config *inf;
 	struct resource *mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	int ret;
-	int temp;
 
 	if (!mem)
 		return -EINVAL;
@@ -232,8 +231,6 @@ static int scoop_probe(struct platform_device *pdev)
 
 	return 0;
 
-	if (devptr->gpio.base != -1)
-		temp = gpiochip_remove(&devptr->gpio);
 err_gpio:
 	platform_set_drvdata(pdev, NULL);
 err_ioremap:
diff --git a/arch/arm/common/sharpsl_param.c b/arch/arm/common/sharpsl_param.c
index d56c932580e..025f6ce3859 100644
--- a/arch/arm/common/sharpsl_param.c
+++ b/arch/arm/common/sharpsl_param.c
@@ -15,6 +15,7 @@
 #include <linux/module.h>
 #include <linux/string.h>
 #include <asm/mach/sharpsl_param.h>
+#include <asm/memory.h>
 
 /*
  * Certain hardware parameters determined at the time of device manufacture,
@@ -25,8 +26,10 @@
  */
 #ifdef CONFIG_ARCH_SA1100
 #define PARAM_BASE	0xe8ffc000
+#define param_start(x)	(void *)(x)
 #else
 #define PARAM_BASE	0xa0000a00
+#define param_start(x)	__va(x)
 #endif
 #define MAGIC_CHG(a,b,c,d) ( ( d << 24 ) | ( c << 16 )  | ( b << 8 ) | a )
 
@@ -41,7 +44,7 @@ EXPORT_SYMBOL(sharpsl_param);
 
 void sharpsl_save_param(void)
 {
-	memcpy(&sharpsl_param, (void *)PARAM_BASE, sizeof(struct sharpsl_param_info));
+	memcpy(&sharpsl_param, param_start(PARAM_BASE), sizeof(struct sharpsl_param_info));
 
 	if (sharpsl_param.comadj_keyword != COMADJ_MAGIC)
 		sharpsl_param.comadj=-1;
diff --git a/arch/arm/common/timer-sp.c b/arch/arm/common/timer-sp.c
index e901d0f3e0b..fd6bff0c5b9 100644
--- a/arch/arm/common/timer-sp.c
+++ b/arch/arm/common/timer-sp.c
@@ -66,7 +66,7 @@ static long __init sp804_get_clock_rate(struct clk *clk)
 
 static void __iomem *sched_clock_base;
 
-static u32 sp804_read(void)
+static u64 notrace sp804_read(void)
 {
 	return ~readl_relaxed(sched_clock_base + TIMER_VALUE);
 }
@@ -104,7 +104,7 @@ void __init __sp804_clocksource_and_sched_clock_init(void __iomem *base,
 
 	if (use_sched_clock) {
 		sched_clock_base = base;
-		setup_sched_clock(sp804_read, 32, rate);
+		sched_clock_register(sp804_read, 32, rate);
 	}
 }
 
@@ -175,7 +175,7 @@ static struct clock_event_device sp804_clockevent = {
 
 static struct irqaction sp804_timer_irq = {
 	.name		= "timer",
-	.flags		= IRQF_DISABLED | IRQF_TIMER | IRQF_IRQPOLL,
+	.flags		= IRQF_TIMER | IRQF_IRQPOLL,
 	.handler	= sp804_timer_interrupt,
 	.dev_id		= &sp804_clockevent,
 };
@@ -271,10 +271,14 @@ static void __init integrator_cp_of_init(struct device_node *np)
 	void __iomem *base;
 	int irq;
 	const char *name = of_get_property(np, "compatible", NULL);
+	struct clk *clk;
 
 	base = of_iomap(np, 0);
 	if (WARN_ON(!base))
 		return;
+	clk = of_clk_get(np, 0);
+	if (WARN_ON(IS_ERR(clk)))
+		return;
 
 	/* Ensure timer is disabled */
 	writel(0, base + TIMER_CTRL);
@@ -283,13 +287,13 @@ static void __init integrator_cp_of_init(struct device_node *np)
 		goto err;
 
 	if (!init_count)
-		sp804_clocksource_init(base, name);
+		__sp804_clocksource_and_sched_clock_init(base, name, clk, 0);
 	else {
 		irq = irq_of_parse_and_map(np, 0);
 		if (irq <= 0)
 			goto err;
 
-		sp804_clockevents_init(base, irq, name);
+		__sp804_clockevents_init(base, irq, clk, name);
 	}
 
 	init_count++;
diff --git a/arch/arm/common/via82c505.c b/arch/arm/common/via82c505.c
deleted file mode 100644
index 6cb362e56d2..00000000000
--- a/arch/arm/common/via82c505.c
+++ /dev/null
@@ -1,83 +0,0 @@
-#include <linux/kernel.h>
-#include <linux/pci.h>
-#include <linux/interrupt.h>
-#include <linux/mm.h>
-#include <linux/init.h>
-#include <linux/ioport.h>
-#include <linux/io.h>
-
-
-#include <asm/mach/pci.h>
-
-#define MAX_SLOTS		7
-
-#define CONFIG_CMD(bus, devfn, where)   (0x80000000 | (bus->number << 16) | (devfn << 8) | (where & ~3))
-
-static int
-via82c505_read_config(struct pci_bus *bus, unsigned int devfn, int where,
-		      int size, u32 *value)
-{
-	outl(CONFIG_CMD(bus,devfn,where),0xCF8);
-	switch (size) {
-	case 1:
-		*value=inb(0xCFC + (where&3));
-		break;
-	case 2:
-		*value=inw(0xCFC + (where&2));
-		break;
-	case 4:
-		*value=inl(0xCFC);
-		break;
-	}
-	return PCIBIOS_SUCCESSFUL;
-}
-
-static int
-via82c505_write_config(struct pci_bus *bus, unsigned int devfn, int where,
-		       int size, u32 value)
-{
-	outl(CONFIG_CMD(bus,devfn,where),0xCF8);
-	switch (size) {
-	case 1:
-		outb(value, 0xCFC + (where&3));
-		break;
-	case 2:
-		outw(value, 0xCFC + (where&2));
-		break;
-	case 4:
-		outl(value, 0xCFC);
-		break;
-	}
-	return PCIBIOS_SUCCESSFUL;
-}
-
-struct pci_ops via82c505_ops = {
-	.read	= via82c505_read_config,
-	.write	= via82c505_write_config,
-};
-
-void __init via82c505_preinit(void)
-{
-	printk(KERN_DEBUG "PCI: VIA 82c505\n");
-	if (!request_region(0xA8,2,"via config")) {
-		printk(KERN_WARNING"VIA 82c505: Unable to request region 0xA8\n");
-		return;
-	}
-	if (!request_region(0xCF8,8,"pci config")) {
-		printk(KERN_WARNING"VIA 82c505: Unable to request region 0xCF8\n");
-		release_region(0xA8, 2);
-		return;
-	}
-
-	/* Enable compatible Mode */
-	outb(0x96,0xA8);
-	outb(0x18,0xA9);
-	outb(0x93,0xA8);
-	outb(0xd0,0xA9);
-
-}
-
-int __init via82c505_setup(int nr, struct pci_sys_data *sys)
-{
-	return (nr == 0);
-}