Merge branch 'korg-android-tegra-3.1' into after-upstream-android

Conflicts:
	arch/arm/mach-tegra/Kconfig
	arch/arm/mach-tegra/board-ventana.c
	drivers/misc/Kconfig
	drivers/video/tegra/dc/hdmi.c

Signed-off-by: Dan Willemsen <dwillemsen@nvidia.com>

7 files changed, 561 insertions, 149 deletions

diff --git a/arch/arm/kernel/asm-offsets.c b/arch/arm/kernel/asm-offsets.c
index 16baba2e4369..927522cfc12e 100644
--- a/arch/arm/kernel/asm-offsets.c
+++ b/arch/arm/kernel/asm-offsets.c
@@ -59,9 +59,6 @@ int main(void)
   DEFINE(TI_TP_VALUE,		offsetof(struct thread_info, tp_value));
   DEFINE(TI_FPSTATE,		offsetof(struct thread_info, fpstate));
   DEFINE(TI_VFPSTATE,		offsetof(struct thread_info, vfpstate));
-#ifdef CONFIG_SMP
-  DEFINE(VFP_CPU,		offsetof(union vfp_state, hard.cpu));
-#endif
 #ifdef CONFIG_ARM_THUMBEE
   DEFINE(TI_THUMBEE_STATE,	offsetof(struct thread_info, thumbee_state));
 #endif
diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S
index a87cbf889ff4..4f8e30f183b9 100644
--- a/arch/arm/kernel/entry-armv.S
+++ b/arch/arm/kernel/entry-armv.S
@@ -450,7 +450,7 @@ __und_usr:
 	blo	__und_usr_unknown
 3:	ldrht	r0, [r4]
 	add	r2, r2, #2			@ r2 is PC + 2, make it PC + 4
-	orr	r0, r0, r5, lsl #16
+	orr    	r0, r0, r5, lsl #16
 #else
 	b	__und_usr_unknown
 #endif
diff --git a/arch/arm/kernel/etm.c b/arch/arm/kernel/etm.c
index 1bec8b5f22f0..496b8b84e455 100644
--- a/arch/arm/kernel/etm.c
+++ b/arch/arm/kernel/etm.c
@@ -15,6 +15,7 @@
 #include <linux/init.h>
 #include <linux/types.h>
 #include <linux/io.h>
+#include <linux/slab.h>
 #include <linux/sysrq.h>
 #include <linux/device.h>
 #include <linux/clk.h>
@@ -36,26 +37,36 @@ MODULE_AUTHOR("Alexander Shishkin");
 struct tracectx {
 	unsigned int	etb_bufsz;
 	void __iomem	*etb_regs;
-	void __iomem	*etm_regs;
+	void __iomem	**etm_regs;
+	int		etm_regs_count;
 	unsigned long	flags;
 	int		ncmppairs;
 	int		etm_portsz;
+	u32		etb_fc;
+	unsigned long	range_start;
+	unsigned long	range_end;
+	unsigned long	data_range_start;
+	unsigned long	data_range_end;
+	bool		dump_initial_etb;
 	struct device	*dev;
 	struct clk	*emu_clk;
 	struct mutex	mutex;
 };
 
-static struct tracectx tracer;
+static struct tracectx tracer = {
+	.range_start = (unsigned long)_stext,
+	.range_end = (unsigned long)_etext,
+};
 
 static inline bool trace_isrunning(struct tracectx *t)
 {
 	return !!(t->flags & TRACER_RUNNING);
 }
 
-static int etm_setup_address_range(struct tracectx *t, int n,
+static int etm_setup_address_range(struct tracectx *t, int id, int n,
 		unsigned long start, unsigned long end, int exclude, int data)
 {
-	u32 flags = ETMAAT_ARM | ETMAAT_IGNCONTEXTID | ETMAAT_NSONLY | \
+	u32 flags = ETMAAT_ARM | ETMAAT_IGNCONTEXTID | ETMAAT_IGNSECURITY |
 		    ETMAAT_NOVALCMP;
 
 	if (n < 1 || n > t->ncmppairs)
@@ -71,95 +82,155 @@ static int etm_setup_address_range(struct tracectx *t, int n,
 		flags |= ETMAAT_IEXEC;
 
 	/* first comparator for the range */
-	etm_writel(t, flags, ETMR_COMP_ACC_TYPE(n * 2));
-	etm_writel(t, start, ETMR_COMP_VAL(n * 2));
+	etm_writel(t, id, flags, ETMR_COMP_ACC_TYPE(n * 2));
+	etm_writel(t, id, start, ETMR_COMP_VAL(n * 2));
 
 	/* second comparator is right next to it */
-	etm_writel(t, flags, ETMR_COMP_ACC_TYPE(n * 2 + 1));
-	etm_writel(t, end, ETMR_COMP_VAL(n * 2 + 1));
-
-	flags = exclude ? ETMTE_INCLEXCL : 0;
-	etm_writel(t, flags | (1 << n), ETMR_TRACEENCTRL);
+	etm_writel(t, id, flags, ETMR_COMP_ACC_TYPE(n * 2 + 1));
+	etm_writel(t, id, end, ETMR_COMP_VAL(n * 2 + 1));
+
+	if (data) {
+		flags = exclude ? ETMVDC3_EXCLONLY : 0;
+		if (exclude)
+			n += 8;
+		etm_writel(t, id, flags | BIT(n), ETMR_VIEWDATACTRL3);
+	} else {
+		flags = exclude ? ETMTE_INCLEXCL : 0;
+		etm_writel(t, id, flags | (1 << n), ETMR_TRACEENCTRL);
+	}
 
 	return 0;
 }
 
-static int trace_start(struct tracectx *t)
+static int trace_start_etm(struct tracectx *t, int id)
 {
 	u32 v;
 	unsigned long timeout = TRACER_TIMEOUT;
 
-	etb_unlock(t);
-
-	etb_writel(t, 0, ETBR_FORMATTERCTRL);
-	etb_writel(t, 1, ETBR_CTRL);
-
-	etb_lock(t);
-
-	/* configure etm */
 	v = ETMCTRL_OPTS | ETMCTRL_PROGRAM | ETMCTRL_PORTSIZE(t->etm_portsz);
 
 	if (t->flags & TRACER_CYCLE_ACC)
 		v |= ETMCTRL_CYCLEACCURATE;
 
-	etm_unlock(t);
+	if (t->flags & TRACER_TRACE_DATA)
+		v |= ETMCTRL_DATA_DO_ADDR;
+
+	etm_unlock(t, id);
 
-	etm_writel(t, v, ETMR_CTRL);
+	etm_writel(t, id, v, ETMR_CTRL);
 
-	while (!(etm_readl(t, ETMR_CTRL) & ETMCTRL_PROGRAM) && --timeout)
+	while (!(etm_readl(t, id, ETMR_CTRL) & ETMCTRL_PROGRAM) && --timeout)
 		;
 	if (!timeout) {
 		dev_dbg(t->dev, "Waiting for progbit to assert timed out\n");
-		etm_lock(t);
+		etm_lock(t, id);
 		return -EFAULT;
 	}
 
-	etm_setup_address_range(t, 1, (unsigned long)_stext,
-			(unsigned long)_etext, 0, 0);
-	etm_writel(t, 0, ETMR_TRACEENCTRL2);
-	etm_writel(t, 0, ETMR_TRACESSCTRL);
-	etm_writel(t, 0x6f, ETMR_TRACEENEVT);
+	if (t->range_start || t->range_end)
+		etm_setup_address_range(t, id, 1,
+					t->range_start, t->range_end, 0, 0);
+	else
+		etm_writel(t, id, ETMTE_INCLEXCL, ETMR_TRACEENCTRL);
+
+	etm_writel(t, id, 0, ETMR_TRACEENCTRL2);
+	etm_writel(t, id, 0, ETMR_TRACESSCTRL);
+	etm_writel(t, id, 0x6f, ETMR_TRACEENEVT);
+
+	etm_writel(t, id, 0, ETMR_VIEWDATACTRL1);
+	etm_writel(t, id, 0, ETMR_VIEWDATACTRL2);
+
+	if (t->data_range_start || t->data_range_end)
+		etm_setup_address_range(t, id, 2, t->data_range_start,
+					t->data_range_end, 0, 1);
+	else
+		etm_writel(t, id, ETMVDC3_EXCLONLY, ETMR_VIEWDATACTRL3);
+
+	etm_writel(t, id, 0x6f, ETMR_VIEWDATAEVT);
 
 	v &= ~ETMCTRL_PROGRAM;
 	v |= ETMCTRL_PORTSEL;
 
-	etm_writel(t, v, ETMR_CTRL);
+	etm_writel(t, id, v, ETMR_CTRL);
 
 	timeout = TRACER_TIMEOUT;
-	while (etm_readl(t, ETMR_CTRL) & ETMCTRL_PROGRAM && --timeout)
+	while (etm_readl(t, id, ETMR_CTRL) & ETMCTRL_PROGRAM && --timeout)
 		;
 	if (!timeout) {
 		dev_dbg(t->dev, "Waiting for progbit to deassert timed out\n");
-		etm_lock(t);
+		etm_lock(t, id);
 		return -EFAULT;
 	}
 
-	etm_lock(t);
+	etm_lock(t, id);
+	return 0;
+}
+
+static int trace_start(struct tracectx *t)
+{
+	int ret;
+	int id;
+	u32 etb_fc = t->etb_fc;
+
+	etb_unlock(t);
+
+	t->dump_initial_etb = false;
+	etb_writel(t, 0, ETBR_WRITEADDR);
+	etb_writel(t, etb_fc, ETBR_FORMATTERCTRL);
+	etb_writel(t, 1, ETBR_CTRL);
+
+	etb_lock(t);
+
+	/* configure etm(s) */
+	for (id = 0; id < t->etm_regs_count; id++) {
+		ret = trace_start_etm(t, id);
+		if (ret)
+			return ret;
+	}
 
 	t->flags |= TRACER_RUNNING;
 
 	return 0;
 }
 
-static int trace_stop(struct tracectx *t)
+static int trace_stop_etm(struct tracectx *t, int id)
 {
 	unsigned long timeout = TRACER_TIMEOUT;
 
-	etm_unlock(t);
+	etm_unlock(t, id);
 
-	etm_writel(t, 0x440, ETMR_CTRL);
-	while (!(etm_readl(t, ETMR_CTRL) & ETMCTRL_PROGRAM) && --timeout)
+	etm_writel(t, id, 0x441, ETMR_CTRL);
+	while (!(etm_readl(t, id, ETMR_CTRL) & ETMCTRL_PROGRAM) && --timeout)
 		;
 	if (!timeout) {
 		dev_dbg(t->dev, "Waiting for progbit to assert timed out\n");
-		etm_lock(t);
+		etm_lock(t, id);
 		return -EFAULT;
 	}
 
-	etm_lock(t);
+	etm_lock(t, id);
+	return 0;
+}
+
+static int trace_stop(struct tracectx *t)
+{
+	int id;
+	int ret;
+	unsigned long timeout = TRACER_TIMEOUT;
+	u32 etb_fc = t->etb_fc;
+
+	for (id = 0; id < t->etm_regs_count; id++) {
+		ret = trace_stop_etm(t, id);
+		if (ret)
+			return ret;
+	}
 
 	etb_unlock(t);
-	etb_writel(t, ETBFF_MANUAL_FLUSH, ETBR_FORMATTERCTRL);
+	if (etb_fc) {
+		etb_fc |= ETBFF_STOPFL;
+		etb_writel(t, t->etb_fc, ETBR_FORMATTERCTRL);
+	}
+	etb_writel(t, etb_fc | ETBFF_MANUAL_FLUSH, ETBR_FORMATTERCTRL);
 
 	timeout = TRACER_TIMEOUT;
 	while (etb_readl(t, ETBR_FORMATTERCTRL) &
@@ -184,24 +255,15 @@ static int trace_stop(struct tracectx *t)
 static int etb_getdatalen(struct tracectx *t)
 {
 	u32 v;
-	int rp, wp;
+	int wp;
 
 	v = etb_readl(t, ETBR_STATUS);
 
 	if (v & 1)
 		return t->etb_bufsz;
 
-	rp = etb_readl(t, ETBR_READADDR);
 	wp = etb_readl(t, ETBR_WRITEADDR);
-
-	if (rp > wp) {
-		etb_writel(t, 0, ETBR_READADDR);
-		etb_writel(t, 0, ETBR_WRITEADDR);
-
-		return 0;
-	}
-
-	return wp - rp;
+	return wp;
 }
 
 /* sysrq+v will always stop the running trace and leave it at that */
@@ -234,21 +296,18 @@ static void etm_dump(void)
 		printk("%08x", cpu_to_be32(etb_readl(t, ETBR_READMEM)));
 	printk(KERN_INFO "\n--- ETB buffer end ---\n");
 
-	/* deassert the overflow bit */
-	etb_writel(t, 1, ETBR_CTRL);
-	etb_writel(t, 0, ETBR_CTRL);
-
-	etb_writel(t, 0, ETBR_TRIGGERCOUNT);
-	etb_writel(t, 0, ETBR_READADDR);
-	etb_writel(t, 0, ETBR_WRITEADDR);
-
 	etb_lock(t);
 }
 
 static void sysrq_etm_dump(int key)
 {
+	if (!mutex_trylock(&tracer.mutex)) {
+		printk(KERN_INFO "Tracing hardware busy\n");
+		return;
+	}
 	dev_dbg(tracer.dev, "Dumping ETB buffer\n");
 	etm_dump();
+	mutex_unlock(&tracer.mutex);
 }
 
 static struct sysrq_key_op sysrq_etm_op = {
@@ -275,6 +334,10 @@ static ssize_t etb_read(struct file *file, char __user *data,
 	struct tracectx *t = file->private_data;
 	u32 first = 0;
 	u32 *buf;
+	int wpos;
+	int skip;
+	long wlength;
+	loff_t pos = *ppos;
 
 	mutex_lock(&t->mutex);
 
@@ -286,31 +349,39 @@ static ssize_t etb_read(struct file *file, char __user *data,
 	etb_unlock(t);
 
 	total = etb_getdatalen(t);
+	if (total == 0 && t->dump_initial_etb)
+		total = t->etb_bufsz;
 	if (total == t->etb_bufsz)
 		first = etb_readl(t, ETBR_WRITEADDR);
 
+	if (pos > total * 4) {
+		skip = 0;
+		wpos = total;
+	} else {
+		skip = (int)pos % 4;
+		wpos = (int)pos / 4;
+	}
+	total -= wpos;
+	first = (first + wpos) % t->etb_bufsz;
+
 	etb_writel(t, first, ETBR_READADDR);
 
-	length = min(total * 4, (int)len);
-	buf = vmalloc(length);
+	wlength = min(total, DIV_ROUND_UP(skip + (int)len, 4));
+	length = min(total * 4 - skip, (int)len);
+	buf = vmalloc(wlength * 4);
 
-	dev_dbg(t->dev, "ETB buffer length: %d\n", total);
+	dev_dbg(t->dev, "ETB read %ld bytes to %lld from %ld words at %d\n",
+		length, pos, wlength, first);
+	dev_dbg(t->dev, "ETB buffer length: %d\n", total + wpos);
 	dev_dbg(t->dev, "ETB status reg: %x\n", etb_readl(t, ETBR_STATUS));
-	for (i = 0; i < length / 4; i++)
+	for (i = 0; i < wlength; i++)
 		buf[i] = etb_readl(t, ETBR_READMEM);
 
-	/* the only way to deassert overflow bit in ETB status is this */
-	etb_writel(t, 1, ETBR_CTRL);
-	etb_writel(t, 0, ETBR_CTRL);
-
-	etb_writel(t, 0, ETBR_WRITEADDR);
-	etb_writel(t, 0, ETBR_READADDR);
-	etb_writel(t, 0, ETBR_TRIGGERCOUNT);
-
 	etb_lock(t);
 
-	length -= copy_to_user(data, buf, length);
+	length -= copy_to_user(data, (u8 *)buf + skip, length);
 	vfree(buf);
+	*ppos = pos + length;
 
 out:
 	mutex_unlock(&t->mutex);
@@ -347,28 +418,17 @@ static int __devinit etb_probe(struct amba_device *dev, const struct amba_id *id
 	if (ret)
 		goto out;
 
+	mutex_lock(&t->mutex);
 	t->etb_regs = ioremap_nocache(dev->res.start, resource_size(&dev->res));
 	if (!t->etb_regs) {
 		ret = -ENOMEM;
 		goto out_release;
 	}
 
+	t->dev = &dev->dev;
+	t->dump_initial_etb = true;
 	amba_set_drvdata(dev, t);
 
-	etb_miscdev.parent = &dev->dev;
-
-	ret = misc_register(&etb_miscdev);
-	if (ret)
-		goto out_unmap;
-
-	t->emu_clk = clk_get(&dev->dev, "emu_src_ck");
-	if (IS_ERR(t->emu_clk)) {
-		dev_dbg(&dev->dev, "Failed to obtain emu_src_ck.\n");
-		return -EFAULT;
-	}
-
-	clk_enable(t->emu_clk);
-
 	etb_unlock(t);
 	t->etb_bufsz = etb_readl(t, ETBR_DEPTH);
 	dev_dbg(&dev->dev, "Size: %x\n", t->etb_bufsz);
@@ -377,6 +437,20 @@ static int __devinit etb_probe(struct amba_device *dev, const struct amba_id *id
 	etb_writel(t, 0, ETBR_CTRL);
 	etb_writel(t, 0x1000, ETBR_FORMATTERCTRL);
 	etb_lock(t);
+	mutex_unlock(&t->mutex);
+
+	etb_miscdev.parent = &dev->dev;
+
+	ret = misc_register(&etb_miscdev);
+	if (ret)
+		goto out_unmap;
+
+	/* Get optional clock. Currently used to select clock source on omap3 */
+	t->emu_clk = clk_get(&dev->dev, "emu_src_ck");
+	if (IS_ERR(t->emu_clk))
+		dev_dbg(&dev->dev, "Failed to obtain emu_src_ck.\n");
+	else
+		clk_enable(t->emu_clk);
 
 	dev_dbg(&dev->dev, "ETB AMBA driver initialized.\n");
 
@@ -384,10 +458,13 @@ out:
 	return ret;
 
 out_unmap:
+	mutex_lock(&t->mutex);
 	amba_set_drvdata(dev, NULL);
 	iounmap(t->etb_regs);
+	t->etb_regs = NULL;
 
 out_release:
+	mutex_unlock(&t->mutex);
 	amba_release_regions(dev);
 
 	return ret;
@@ -402,8 +479,10 @@ static int etb_remove(struct amba_device *dev)
 	iounmap(t->etb_regs);
 	t->etb_regs = NULL;
 
-	clk_disable(t->emu_clk);
-	clk_put(t->emu_clk);
+	if (!IS_ERR(t->emu_clk)) {
+		clk_disable(t->emu_clk);
+		clk_put(t->emu_clk);
+	}
 
 	amba_release_regions(dev);
 
@@ -447,7 +526,10 @@ static ssize_t trace_running_store(struct kobject *kobj,
 		return -EINVAL;
 
 	mutex_lock(&tracer.mutex);
-	ret = value ? trace_start(&tracer) : trace_stop(&tracer);
+	if (!tracer.etb_regs)
+		ret = -ENODEV;
+	else
+		ret = value ? trace_start(&tracer) : trace_stop(&tracer);
 	mutex_unlock(&tracer.mutex);
 
 	return ret ? : n;
@@ -462,36 +544,50 @@ static ssize_t trace_info_show(struct kobject *kobj,
 {
 	u32 etb_wa, etb_ra, etb_st, etb_fc, etm_ctrl, etm_st;
 	int datalen;
+	int id;
+	int ret;
 
-	etb_unlock(&tracer);
-	datalen = etb_getdatalen(&tracer);
-	etb_wa = etb_readl(&tracer, ETBR_WRITEADDR);
-	etb_ra = etb_readl(&tracer, ETBR_READADDR);
-	etb_st = etb_readl(&tracer, ETBR_STATUS);
-	etb_fc = etb_readl(&tracer, ETBR_FORMATTERCTRL);
-	etb_lock(&tracer);
-
-	etm_unlock(&tracer);
-	etm_ctrl = etm_readl(&tracer, ETMR_CTRL);
-	etm_st = etm_readl(&tracer, ETMR_STATUS);
-	etm_lock(&tracer);
+	mutex_lock(&tracer.mutex);
+	if (tracer.etb_regs) {
+		etb_unlock(&tracer);
+		datalen = etb_getdatalen(&tracer);
+		etb_wa = etb_readl(&tracer, ETBR_WRITEADDR);
+		etb_ra = etb_readl(&tracer, ETBR_READADDR);
+		etb_st = etb_readl(&tracer, ETBR_STATUS);
+		etb_fc = etb_readl(&tracer, ETBR_FORMATTERCTRL);
+		etb_lock(&tracer);
+	} else {
+		etb_wa = etb_ra = etb_st = etb_fc = ~0;
+		datalen = -1;
+	}
 
-	return sprintf(buf, "Trace buffer len: %d\nComparator pairs: %d\n"
+	ret = sprintf(buf, "Trace buffer len: %d\nComparator pairs: %d\n"
 			"ETBR_WRITEADDR:\t%08x\n"
 			"ETBR_READADDR:\t%08x\n"
 			"ETBR_STATUS:\t%08x\n"
-			"ETBR_FORMATTERCTRL:\t%08x\n"
-			"ETMR_CTRL:\t%08x\n"
-			"ETMR_STATUS:\t%08x\n",
+			"ETBR_FORMATTERCTRL:\t%08x\n",
 			datalen,
 			tracer.ncmppairs,
 			etb_wa,
 			etb_ra,
 			etb_st,
-			etb_fc,
+			etb_fc
+			);
+
+	for (id = 0; id < tracer.etm_regs_count; id++) {
+		etm_unlock(&tracer, id);
+		etm_ctrl = etm_readl(&tracer, id, ETMR_CTRL);
+		etm_st = etm_readl(&tracer, id, ETMR_STATUS);
+		etm_lock(&tracer, id);
+		ret += sprintf(buf + ret, "ETMR_CTRL:\t%08x\n"
+			"ETMR_STATUS:\t%08x\n",
 			etm_ctrl,
 			etm_st
 			);
+	}
+	mutex_unlock(&tracer.mutex);
+
+	return ret;
 }
 
 static struct kobj_attribute trace_info_attr =
@@ -530,42 +626,121 @@ static ssize_t trace_mode_store(struct kobject *kobj,
 static struct kobj_attribute trace_mode_attr =
 	__ATTR(trace_mode, 0644, trace_mode_show, trace_mode_store);
 
+static ssize_t trace_range_show(struct kobject *kobj,
+				  struct kobj_attribute *attr,
+				  char *buf)
+{
+	return sprintf(buf, "%08lx %08lx\n",
+			tracer.range_start, tracer.range_end);
+}
+
+static ssize_t trace_range_store(struct kobject *kobj,
+				   struct kobj_attribute *attr,
+				   const char *buf, size_t n)
+{
+	unsigned long range_start, range_end;
+
+	if (sscanf(buf, "%lx %lx", &range_start, &range_end) != 2)
+		return -EINVAL;
+
+	mutex_lock(&tracer.mutex);
+	tracer.range_start = range_start;
+	tracer.range_end = range_end;
+	mutex_unlock(&tracer.mutex);
+
+	return n;
+}
+
+
+static struct kobj_attribute trace_range_attr =
+	__ATTR(trace_range, 0644, trace_range_show, trace_range_store);
+
+static ssize_t trace_data_range_show(struct kobject *kobj,
+				  struct kobj_attribute *attr,
+				  char *buf)
+{
+	unsigned long range_start;
+	u64 range_end;
+	mutex_lock(&tracer.mutex);
+	range_start = tracer.data_range_start;
+	range_end = tracer.data_range_end;
+	if (!range_end && (tracer.flags & TRACER_TRACE_DATA))
+		range_end = 0x100000000ULL;
+	mutex_unlock(&tracer.mutex);
+	return sprintf(buf, "%08lx %08llx\n", range_start, range_end);
+}
+
+static ssize_t trace_data_range_store(struct kobject *kobj,
+				   struct kobj_attribute *attr,
+				   const char *buf, size_t n)
+{
+	unsigned long range_start;
+	u64 range_end;
+
+	if (sscanf(buf, "%lx %llx", &range_start, &range_end) != 2)
+		return -EINVAL;
+
+	mutex_lock(&tracer.mutex);
+	tracer.data_range_start = range_start;
+	tracer.data_range_end = (unsigned long)range_end;
+	if (range_end)
+		tracer.flags |= TRACER_TRACE_DATA;
+	else
+		tracer.flags &= ~TRACER_TRACE_DATA;
+	mutex_unlock(&tracer.mutex);
+
+	return n;
+}
+
+
+static struct kobj_attribute trace_data_range_attr =
+	__ATTR(trace_data_range, 0644,
+		trace_data_range_show, trace_data_range_store);
+
 static int __devinit etm_probe(struct amba_device *dev, const struct amba_id *id)
 {
 	struct tracectx *t = &tracer;
 	int ret = 0;
+	void __iomem **new_regs;
+	int new_count;
 
-	if (t->etm_regs) {
-		dev_dbg(&dev->dev, "ETM already initialized\n");
-		ret = -EBUSY;
+	mutex_lock(&t->mutex);
+	new_count = t->etm_regs_count + 1;
+	new_regs = krealloc(t->etm_regs,
+				sizeof(t->etm_regs[0]) * new_count, GFP_KERNEL);
+
+	if (!new_regs) {
+		dev_dbg(&dev->dev, "Failed to allocate ETM register array\n");
+		ret = -ENOMEM;
 		goto out;
 	}
+	t->etm_regs = new_regs;
 
 	ret = amba_request_regions(dev, NULL);
 	if (ret)
 		goto out;
 
-	t->etm_regs = ioremap_nocache(dev->res.start, resource_size(&dev->res));
-	if (!t->etm_regs) {
+	t->etm_regs[t->etm_regs_count] =
+		ioremap_nocache(dev->res.start, resource_size(&dev->res));
+	if (!t->etm_regs[t->etm_regs_count]) {
 		ret = -ENOMEM;
 		goto out_release;
 	}
 
-	amba_set_drvdata(dev, t);
+	amba_set_drvdata(dev, t->etm_regs[t->etm_regs_count]);
 
-	mutex_init(&t->mutex);
-	t->dev = &dev->dev;
-	t->flags = TRACER_CYCLE_ACC;
+	t->flags = TRACER_CYCLE_ACC | TRACER_TRACE_DATA;
 	t->etm_portsz = 1;
 
-	etm_unlock(t);
-	(void)etm_readl(t, ETMMR_PDSR);
+	etm_unlock(t, t->etm_regs_count);
+	(void)etm_readl(t, t->etm_regs_count, ETMMR_PDSR);
 	/* dummy first read */
-	(void)etm_readl(&tracer, ETMMR_OSSRR);
+	(void)etm_readl(&tracer, t->etm_regs_count, ETMMR_OSSRR);
 
-	t->ncmppairs = etm_readl(t, ETMR_CONFCODE) & 0xf;
-	etm_writel(t, 0x440, ETMR_CTRL);
-	etm_lock(t);
+	t->ncmppairs = etm_readl(t, t->etm_regs_count, ETMR_CONFCODE) & 0xf;
+	etm_writel(t, t->etm_regs_count, 0x441, ETMR_CTRL);
+	etm_writel(t, t->etm_regs_count, new_count, ETMR_TRACEIDR);
+	etm_lock(t, t->etm_regs_count);
 
 	ret = sysfs_create_file(&dev->dev.kobj,
 			&trace_running_attr.attr);
@@ -581,36 +756,68 @@ static int __devinit etm_probe(struct amba_device *dev, const struct amba_id *id
 	if (ret)
 		dev_dbg(&dev->dev, "Failed to create trace_mode in sysfs\n");
 
-	dev_dbg(t->dev, "ETM AMBA driver initialized.\n");
+	ret = sysfs_create_file(&dev->dev.kobj, &trace_range_attr.attr);
+	if (ret)
+		dev_dbg(&dev->dev, "Failed to create trace_range in sysfs\n");
+
+	ret = sysfs_create_file(&dev->dev.kobj, &trace_data_range_attr.attr);
+	if (ret)
+		dev_dbg(&dev->dev,
+			"Failed to create trace_data_range in sysfs\n");
+
+	dev_dbg(&dev->dev, "ETM AMBA driver initialized.\n");
+
+	/* Enable formatter if there are multiple trace sources */
+	if (new_count > 1)
+		t->etb_fc = ETBFF_ENFCONT | ETBFF_ENFTC;
+
+	t->etm_regs_count = new_count;
 
 out:
+	mutex_unlock(&t->mutex);
 	return ret;
 
 out_unmap:
 	amba_set_drvdata(dev, NULL);
-	iounmap(t->etm_regs);
+	iounmap(t->etm_regs[t->etm_regs_count]);
 
 out_release:
 	amba_release_regions(dev);
 
+	mutex_unlock(&t->mutex);
 	return ret;
 }
 
 static int etm_remove(struct amba_device *dev)
 {
-	struct tracectx *t = amba_get_drvdata(dev);
+	int i;
+	struct tracectx *t = &tracer;
+	void __iomem	*etm_regs = amba_get_drvdata(dev);
+
+	sysfs_remove_file(&dev->dev.kobj, &trace_running_attr.attr);
+	sysfs_remove_file(&dev->dev.kobj, &trace_info_attr.attr);
+	sysfs_remove_file(&dev->dev.kobj, &trace_mode_attr.attr);
+	sysfs_remove_file(&dev->dev.kobj, &trace_range_attr.attr);
+	sysfs_remove_file(&dev->dev.kobj, &trace_data_range_attr.attr);
 
 	amba_set_drvdata(dev, NULL);
 
-	iounmap(t->etm_regs);
-	t->etm_regs = NULL;
+	mutex_lock(&t->mutex);
+	for (i = 0; i < t->etm_regs_count; i++)
+		if (t->etm_regs[i] == etm_regs)
+			break;
+	for (; i < t->etm_regs_count - 1; i++)
+		t->etm_regs[i] = t->etm_regs[i + 1];
+	t->etm_regs_count--;
+	if (!t->etm_regs_count) {
+		kfree(t->etm_regs);
+		t->etm_regs = NULL;
+	}
+	mutex_unlock(&t->mutex);
 
+	iounmap(etm_regs);
 	amba_release_regions(dev);
 
-	sysfs_remove_file(&dev->dev.kobj, &trace_running_attr.attr);
-	sysfs_remove_file(&dev->dev.kobj, &trace_info_attr.attr);
-	sysfs_remove_file(&dev->dev.kobj, &trace_mode_attr.attr);
-
 	return 0;
 }
 
@@ -619,6 +826,10 @@ static struct amba_id etm_ids[] = {
 		.id	= 0x0003b921,
 		.mask	= 0x0007ffff,
 	},
+	{
+		.id	= 0x0003b950,
+		.mask	= 0x0007ffff,
+	},
 	{ 0, 0 },
 };
 
@@ -636,6 +847,8 @@ static int __init etm_init(void)
 {
 	int retval;
 
+	mutex_init(&tracer.mutex);
+
 	retval = amba_driver_register(&etb_driver);
 	if (retval) {
 		printk(KERN_ERR "Failed to register etb\n");
diff --git a/arch/arm/kernel/leds.c b/arch/arm/kernel/leds.c
index 0f107dcb0347..136e8376a3eb 100644
--- a/arch/arm/kernel/leds.c
+++ b/arch/arm/kernel/leds.c
@@ -9,6 +9,8 @@
  */
 #include <linux/module.h>
 #include <linux/init.h>
+#include <linux/notifier.h>
+#include <linux/cpu.h>
 #include <linux/sysdev.h>
 #include <linux/syscore_ops.h>
 
@@ -101,6 +103,25 @@ static struct syscore_ops leds_syscore_ops = {
 	.resume		= leds_resume,
 };
 
+static int leds_idle_notifier(struct notifier_block *nb, unsigned long val,
+                                void *data)
+{
+	switch (val) {
+	case IDLE_START:
+		leds_event(led_idle_start);
+		break;
+	case IDLE_END:
+		leds_event(led_idle_end);
+		break;
+	}
+
+	return 0;
+}
+
+static struct notifier_block leds_idle_nb = {
+	.notifier_call = leds_idle_notifier,
+};
+
 static int __init leds_init(void)
 {
 	int ret;
@@ -109,8 +130,12 @@ static int __init leds_init(void)
 		ret = sysdev_register(&leds_device);
 	if (ret == 0)
 		ret = sysdev_create_file(&leds_device, &attr_event);
-	if (ret == 0)
+
+	if (ret == 0) {
 		register_syscore_ops(&leds_syscore_ops);
+		idle_notifier_register(&leds_idle_nb);
+	}
+
 	return ret;
 }
 
diff --git a/arch/arm/kernel/process.c b/arch/arm/kernel/process.c
index 1a347f481e5e..ea7ec069fe88 100644
--- a/arch/arm/kernel/process.c
+++ b/arch/arm/kernel/process.c
@@ -31,9 +31,9 @@
 #include <linux/random.h>
 #include <linux/hw_breakpoint.h>
 #include <linux/cpuidle.h>
+#include <linux/console.h>
 
 #include <asm/cacheflush.h>
-#include <asm/leds.h>
 #include <asm/processor.h>
 #include <asm/system.h>
 #include <asm/thread_notify.h>
@@ -63,6 +63,18 @@ static volatile int hlt_counter;
 
 #include <mach/system.h>
 
+#ifdef CONFIG_SMP
+void arch_trigger_all_cpu_backtrace(void)
+{
+	smp_send_all_cpu_backtrace();
+}
+#else
+void arch_trigger_all_cpu_backtrace(void)
+{
+	dump_stack();
+}
+#endif
+
 void disable_hlt(void)
 {
 	hlt_counter++;
@@ -92,8 +104,37 @@ static int __init hlt_setup(char *__unused)
 __setup("nohlt", nohlt_setup);
 __setup("hlt", hlt_setup);
 
+#ifdef CONFIG_ARM_FLUSH_CONSOLE_ON_RESTART
+void arm_machine_flush_console(void)
+{
+	printk("\n");
+	pr_emerg("Restarting %s\n", linux_banner);
+	if (console_trylock()) {
+		console_unlock();
+		return;
+	}
+
+	mdelay(50);
+
+	local_irq_disable();
+	if (!console_trylock())
+		pr_emerg("arm_restart: Console was locked! Busting\n");
+	else
+		pr_emerg("arm_restart: Console was locked!\n");
+	console_unlock();
+}
+#else
+void arm_machine_flush_console(void)
+{
+}
+#endif
+
 void arm_machine_restart(char mode, const char *cmd)
 {
+	/* Flush the console to make sure all the relevant messages make it
+	 * out to the console drivers */
+	arm_machine_flush_console();
+
 	/* Disable interrupts first */
 	local_irq_disable();
 	local_fiq_disable();
@@ -183,8 +224,8 @@ void cpu_idle(void)
 
 	/* endless idle loop with no priority at all */
 	while (1) {
+		idle_notifier_call_chain(IDLE_START);
 		tick_nohz_stop_sched_tick(1);
-		leds_event(led_idle_start);
 		while (!need_resched()) {
 #ifdef CONFIG_HOTPLUG_CPU
 			if (cpu_is_offline(smp_processor_id()))
@@ -209,8 +250,8 @@ void cpu_idle(void)
 				local_irq_enable();
 			}
 		}
-		leds_event(led_idle_end);
 		tick_nohz_restart_sched_tick();
+		idle_notifier_call_chain(IDLE_END);
 		preempt_enable_no_resched();
 		schedule();
 		preempt_disable();
@@ -253,6 +294,77 @@ void machine_restart(char *cmd)
 	arm_pm_restart(reboot_mode, cmd);
 }
 
+/*
+ * dump a block of kernel memory from around the given address
+ */
+static void show_data(unsigned long addr, int nbytes, const char *name)
+{
+	int	i, j;
+	int	nlines;
+	u32	*p;
+
+	/*
+	 * don't attempt to dump non-kernel addresses or
+	 * values that are probably just small negative numbers
+	 */
+	if (addr < PAGE_OFFSET || addr > -256UL)
+		return;
+
+	printk("\n%s: %#lx:\n", name, addr);
+
+	/*
+	 * round address down to a 32 bit boundary
+	 * and always dump a multiple of 32 bytes
+	 */
+	p = (u32 *)(addr & ~(sizeof(u32) - 1));
+	nbytes += (addr & (sizeof(u32) - 1));
+	nlines = (nbytes + 31) / 32;
+
+
+	for (i = 0; i < nlines; i++) {
+		/*
+		 * just display low 16 bits of address to keep
+		 * each line of the dump < 80 characters
+		 */
+		printk("%04lx ", (unsigned long)p & 0xffff);
+		for (j = 0; j < 8; j++) {
+			u32	data;
+			if (probe_kernel_address(p, data)) {
+				printk(" ********");
+			} else {
+				printk(" %08x", data);
+			}
+			++p;
+		}
+		printk("\n");
+	}
+}
+
+static void show_extra_register_data(struct pt_regs *regs, int nbytes)
+{
+	mm_segment_t fs;
+
+	fs = get_fs();
+	set_fs(KERNEL_DS);
+	show_data(regs->ARM_pc - nbytes, nbytes * 2, "PC");
+	show_data(regs->ARM_lr - nbytes, nbytes * 2, "LR");
+	show_data(regs->ARM_sp - nbytes, nbytes * 2, "SP");
+	show_data(regs->ARM_ip - nbytes, nbytes * 2, "IP");
+	show_data(regs->ARM_fp - nbytes, nbytes * 2, "FP");
+	show_data(regs->ARM_r0 - nbytes, nbytes * 2, "R0");
+	show_data(regs->ARM_r1 - nbytes, nbytes * 2, "R1");
+	show_data(regs->ARM_r2 - nbytes, nbytes * 2, "R2");
+	show_data(regs->ARM_r3 - nbytes, nbytes * 2, "R3");
+	show_data(regs->ARM_r4 - nbytes, nbytes * 2, "R4");
+	show_data(regs->ARM_r5 - nbytes, nbytes * 2, "R5");
+	show_data(regs->ARM_r6 - nbytes, nbytes * 2, "R6");
+	show_data(regs->ARM_r7 - nbytes, nbytes * 2, "R7");
+	show_data(regs->ARM_r8 - nbytes, nbytes * 2, "R8");
+	show_data(regs->ARM_r9 - nbytes, nbytes * 2, "R9");
+	show_data(regs->ARM_r10 - nbytes, nbytes * 2, "R10");
+	set_fs(fs);
+}
+
 void __show_regs(struct pt_regs *regs)
 {
 	unsigned long flags;
@@ -312,6 +424,8 @@ void __show_regs(struct pt_regs *regs)
 		printk("Control: %08x%s\n", ctrl, buf);
 	}
 #endif
+
+	show_extra_register_data(regs, 128);
 }
 
 void show_regs(struct pt_regs * regs)
diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c
index d88ff0230e82..a63326923e84 100644
--- a/arch/arm/kernel/smp.c
+++ b/arch/arm/kernel/smp.c
@@ -53,6 +53,7 @@ enum ipi_msg_type {
 	IPI_CALL_FUNC,
 	IPI_CALL_FUNC_SINGLE,
 	IPI_CPU_STOP,
+	IPI_CPU_BACKTRACE,
 };
 
 int __cpuinit __cpu_up(unsigned int cpu)
@@ -301,17 +302,7 @@ asmlinkage void __cpuinit secondary_start_kernel(void)
 	 */
 	platform_secondary_init(cpu);
 
-	/*
-	 * Enable local interrupts.
-	 */
 	notify_cpu_starting(cpu);
-	local_irq_enable();
-	local_fiq_enable();
-
-	/*
-	 * Setup the percpu timer for this CPU.
-	 */
-	percpu_timer_setup();
 
 	calibrate_delay();
 
@@ -323,10 +314,23 @@ asmlinkage void __cpuinit secondary_start_kernel(void)
 	 * before we continue.
 	 */
 	set_cpu_online(cpu, true);
+
+	/*
+	 * Setup the percpu timer for this CPU.
+	 */
+	percpu_timer_setup();
+
 	while (!cpu_active(cpu))
 		cpu_relax();
 
 	/*
+	 * cpu_active bit is set, so it's safe to enable interrupts
+	 * now.
+	 */
+	local_irq_enable();
+	local_fiq_enable();
+
+	/*
 	 * OK, it's off to the idle thread for us
 	 */
 	cpu_idle();
@@ -412,6 +416,7 @@ static const char *ipi_types[NR_IPI] = {
 	S(IPI_CALL_FUNC, "Function call interrupts"),
 	S(IPI_CALL_FUNC_SINGLE, "Single function call interrupts"),
 	S(IPI_CPU_STOP, "CPU stop interrupts"),
+	S(IPI_CPU_BACKTRACE, "CPU backtrace"),
 };
 
 void show_ipi_list(struct seq_file *p, int prec)
@@ -562,6 +567,58 @@ static void ipi_cpu_stop(unsigned int cpu)
 		cpu_relax();
 }
 
+static cpumask_t backtrace_mask;
+static DEFINE_RAW_SPINLOCK(backtrace_lock);
+
+/* "in progress" flag of arch_trigger_all_cpu_backtrace */
+static unsigned long backtrace_flag;
+
+void smp_send_all_cpu_backtrace(void)
+{
+	unsigned int this_cpu = smp_processor_id();
+	int i;
+
+	if (test_and_set_bit(0, &backtrace_flag))
+		/*
+		 * If there is already a trigger_all_cpu_backtrace() in progress
+		 * (backtrace_flag == 1), don't output double cpu dump infos.
+		 */
+		return;
+
+	cpumask_copy(&backtrace_mask, cpu_online_mask);
+	cpu_clear(this_cpu, backtrace_mask);
+
+	pr_info("Backtrace for cpu %d (current):\n", this_cpu);
+	dump_stack();
+
+	pr_info("\nsending IPI to all other CPUs:\n");
+	smp_cross_call(&backtrace_mask, IPI_CPU_BACKTRACE);
+
+	/* Wait for up to 10 seconds for all other CPUs to do the backtrace */
+	for (i = 0; i < 10 * 1000; i++) {
+		if (cpumask_empty(&backtrace_mask))
+			break;
+		mdelay(1);
+	}
+
+	clear_bit(0, &backtrace_flag);
+	smp_mb__after_clear_bit();
+}
+
+/*
+ * ipi_cpu_backtrace - handle IPI from smp_send_all_cpu_backtrace()
+ */
+static void ipi_cpu_backtrace(unsigned int cpu, struct pt_regs *regs)
+{
+	if (cpu_isset(cpu, backtrace_mask)) {
+		raw_spin_lock(&backtrace_lock);
+		pr_warning("IPI backtrace for cpu %d\n", cpu);
+		show_regs(regs);
+		raw_spin_unlock(&backtrace_lock);
+		cpu_clear(cpu, backtrace_mask);
+	}
+}
+
 /*
  * Main handler for inter-processor interrupts
  */
@@ -594,6 +651,10 @@ asmlinkage void __exception_irq_entry do_IPI(int ipinr, struct pt_regs *regs)
 		ipi_cpu_stop(cpu);
 		break;
 
+	case IPI_CPU_BACKTRACE:
+		ipi_cpu_backtrace(cpu, regs);
+		break;
+
 	default:
 		printk(KERN_CRIT "CPU%u: Unknown IPI message 0x%x\n",
 		       cpu, ipinr);
diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c
index bc9f9da782cb..4ef9f0d04e5a 100644
--- a/arch/arm/kernel/traps.c
+++ b/arch/arm/kernel/traps.c
@@ -466,7 +466,9 @@ do_cache_op(unsigned long start, unsigned long end, int flags)
 		if (end > vma->vm_end)
 			end = vma->vm_end;
 
-		flush_cache_user_range(vma, start, end);
+		up_read(&mm->mmap_sem);
+		flush_cache_user_range(start, end);
+		return;
 	}
 	up_read(&mm->mmap_sem);
 }